syllabus

Transcription

syllabus

SYLLABUS
PAOKC-cursus Klinische Chemie en
Laboratoriumgeneeskunde
CONSULTVERLENING
Nederlandse Vereniging voor Klinische
Chemie en Laboratoriumgeneeskunde
Donderdag 19 september 2013
De ReeHorst, Ede
PROGRAMMA
___________________________________________________________________________
09.30 – 10.00
Ontvangst met koffie/thee
10.00 – 10.05
Opening en welkom
Dr drs WP Oosterhuis
10.05 – 10.50
Usefulness and necessity of consultation by laboratory specialists
– developments from an international perspective
Dr ID Watson
10.50 – 11.10
Richtlijn NVKC Consultverlening door specialisten laboratoriumgeneeskunde (klinische chemie) – achtergrond, minimum- en streefnormen en vormen van consultverlening
Dr HJ Vermeer
11.10 – 11.30
LESA: het aanvragen van de juiste testen bij de juiste vraagstelling
Dr JLP van Duijnhoven
11.30 – 11.50
Feedback aanvraaggedrag en diagnostisch toetsoverleg
Drs J Trietsch
11.50 – 12.45
Lunch
12.45 – 13.30
Reflecterend testen
Dr drs WP Oosterhuis, Drs AGJM Roos
13.30 – 14.35
Interpreteren en becommentariëren van uitslagen
- Anemie
Dr ing MPG Leers
- Stolling
Drs NCV Péquériaux
- Functietesten Dr JMM Rondeel
14.35 – 14.55
Adviseren over specialistisch onderzoek – hemoglobinopathieën
Dr HJ Adriaansen
14.55 – 15.25
Koffie/theepauze
15.25 – 16.10
Automatisering consultverlening
- Consultregistratie in LIS
Dr M Oostendorp
- RippleDown als tool – in de praktijk
Dr MWM Schellings
16.10 – 16.40
Consultverlening – achteromzien en vooruitblikken
Dr JW Janssen
16.40 – 16.45
Evaluatie en afsluiting
16.45 – 17.15
Borrel
1
SPREKERS
_____________________________________________
Dr HJ Adriaansen
Arts Klinische Chemie
Gelre Ziekenhuizen, Apeldoorn
Laboratoriumspecialist Klinische Chemie
Elkerliek Ziekenhuis, Helmond
Dr JW Janssen
Sint Franciscus Gasthuis, Rotterdam
Dr ing MPG Leers
Atrium Medisch Centrum Parkstad, Heerlen
Dr M Oostendorp
Laboratoriumspecialist Klinische Chemie i.o.
Universitair Medisch Centrum, Utrecht
Jeroen Bosch Ziekenhuis, Den Bosch
Dr JMM Rondeel
Isala Klinieken, Zwolle
Drs AGJM Roos
Huisarts
Brunssum
Dr MWM Schellings
Laboratoriumspecialist Klinische Chemie i.o.
Maxima Medisch Centrum, Veldhoven
Drs JP Trietsch
Huisarts - onderzoeker
Gezondheidscentrum Terwinselen, Kerkrade
Vakgroep Huisartsgeneeskunde, Maastricht Universitair Medisch Centrum
Dr HJ Vermeer
Albert Schweitzer Ziekenhuis, Dordrecht
Dr ID Watson
President EFLM, Consultant Biochemist and Toxicologist
University Hospital Aintree, Liverpool, United Kingdom
2
ORGANISATIE
_____________________________________________
Georganiseerd door de SKMS werkgroep Implementatie Richtlijn Consultverlening:
(projectnr 18358855)
Dr RTP Jansen
voorheen Directeur SKML, Nijmegen
Dr P van ’t Sant
Jeroen Bosch Ziekenhuis, Den Bosch
Dr WPHG Verboeket-van de Venne
Senior Onderzoeker Klinische Chemie
Dr H de Waard
tevens voorzitter PAOKC Commissie
Rijnstate Ziekenhuis, Arnhem
Met ondersteuning en medewerking van:
Mevr E Barmen ’t Loo
NVKC Bureau
Utrecht
3
SPONSOREN
___________________________________________________________________
Deze PAOKC cursus Consultverlening is mede mogelijk gemaakt door bijdragen van:
Beckman Coulter
Mips
Roche
Sysmex
4
INHOUD
___________________________________________________________________________
Pag.
Usefulness and necessity of consultation by specialists in
laboratory medicine
Dr ID Watson
6
Richtlijn NVKC Consultverlening door specialisten laboratoriumgeneeskunde (klinische chemie)
Dr HJ Vermeer
14
LESA, eerste herziening van de Landelijke Eerstelijns SamenwerkingsAfspraak ‘Rationeel aanvragen van laboratoriumdiagnostiek’
40
Feedback aanvraaggedrag en diagnostisch toetsoverleg
Drs J Trietsch
51
Reflecterend testen
Dr drs WP Oosterhuis, drs AGJM Roos
67
Interpreteren van becommentariëren van anemie uitslagen
Dr ing MPG Leers
75
Interpreteren en becommentariëren van stollingsuitslagen
85
Interpreteren en becommentariëren van functietest uitslagen
Dr JMM Rondeel
105
Adviseren over specialistisch onderzoek – hemoglobinopathieën
Dr HJ Adriaansen
113
Automatisering consultverlening: consultregistratie in een LIS
Dr M Oostendorp
128
Automatisering consultverlening: RippleDown als tool – in de
praktijk
Dr MWM Schellings
133
Consultverlening: achteromzien en vooruitblikken
Dr JW Janssen
148
5
USEFULNESS AND NECESSITY OF CONSULTATION BY
SPECIALISTS IN LABORATORY MEDICINE
ID Watson, EFLM President
Advice on the use of the laboratory, interpretation of the results and possible
interventions, is a key element in providing a clinical laboratory service. The selection
of tests to be used to investigate/monitor specific presentations and the application of
tests in individual cases is a basic role; however advising on the management of
patients is also a necessary part of providing a clinical laboratory service. I would
contend that providing the Specialists in Laboratory Medicine that offer such advice
are suitably qualified this should not be contentious. Laboratory based medical staff
would be expected to be better able for such a role, but this does not preclude highly
trained scientific staff. At a basic level they would be expected to be on their National
Register, to be registerable as Eur.Sp.Lab.Med. and qualified to a higher level, in the
UK to have obtained the FRCPath, a qualification awarded by examination to both
medical and scientific staff.
As the Specialist in Laboratory Medicine will only know some elements of the patient
background it is axiomatic that any advice given is in line with best practice and that
there has been discussion with a responsible clinician, the use of agreed guidelines
and protocols is advisable. However there will be times when intervention advice
must be provided urgently and under the individuals own cognisance, therefore they
need to be assured of defence cover from their employer and/or a medical protection
society in case of litigation.
There are recurring scenarios that require effective clinical laboratory intervention: in
secondary care: the investigation, interpretation and treatment of hyponatraemia; in
primary care: the monitoring of thyroxin replacement therapy and less frequent
situations such as investigation of suspected Cushing’s Syndrome where a
laboratory specialist consultation could be helpful to a non-specialist. There are of
course very rare presentations that require collaboration by the whole team e.g.
managing antidotal treatment of methanol poisoned patients on haemodialysis.
6
Most of us can think of individual patients where our intervention has made a clear
difference yet with the expansion of roles in patient care there is less awareness of
the meaning and interventions dependant on laboratory results. There is a need for a
patient focused approach to laboratory medicine, which necessitates pro-active
consultation by Specialists in Laboratory Medicine with healthcare givers, there is
also the prospect that we should extend such knowledge sharing with patients,
perhaps then better laboratory information can be linked to outcome.
7
Measuring the Clinical Impact of Pathologist Reviews of
Blood and Body Fluid Smears
A Laboratory Outcome Study
Linda M. Sandhaus, MD; David N. Wald, MD, PhD; Kenan J. Sauder, MD; Erica L. Steele, DO; Howard J. Meyerson, MD
● Context.—Despite the widespread practice of pathologist review of blood and body fluid smears, little is known
about its impact on improving patient care.
Objective.—To assess the clinical usefulness of pathologist review of blood and body fluid smears.
Design.—Survey study. Pathology residents contacted
the ordering physician after pathologist reviews were reported to assess their clinical impact.
Results.—Ninety-six pathologist reviews met criteria for
study inclusion, and 64 ordering physicians were successfully contacted during the 2-month study period. Of the
64 cases, 19 reviews (30%) had been seen by the physician
within 24 to 48 hours after the report was issued and 33
(51%) had not been seen; in 4 (6%) instances, physicians
did not remember whether they had seen the review. Eight
reviews (13%) were considered urgent enough to warrant
immediate communication by the pathologist. Of the 27
reviews that were seen or directly communicated, 23
(85%) contributed to clinical diagnosis and/or patient
management.
Conclusions.—This study demonstrates the contribution
of pathologist reviews of blood and body fluids to clinical
diagnosis and patient management. The results also highlight the problem of a lack of physician awareness of clinical pathology results.
(Arch Pathol Lab Med. 2007;131:468–472)
P
ments are appended to the leukocyte differential count for
complete blood count and leukocyte differential count
(CBC/DIFF) reports and body fluid cell counts in the laboratory information system (LIS). On average, 220 blood
smears and 170 body fluid smears are reviewed monthly,
which provides a substantial test volume on which to evaluate the clinical usefulness of the test.
hysicians rely on the laboratory for accurate interpretations of peripheral blood and body fluid smears.
The College of American Pathologists laboratory accreditation program requires that laboratories establish specific
criteria for peripheral blood smear review by a pathologist
or a technologist with qualifications in laboratory hematology.1 However, there is no explicit requirement that the
pathologists’ review of blood smear findings be reported.
Pathologist reviews (PRs) of blood and body fluid smears
serve several purposes. First, they fulfill an important
quality assurance function for the hematology laboratory
by providing a secondary review of significant abnormal
findings. Second, they offer interpretive information that
may be clinically useful. Third, the PR process is an essential component of residency training in laboratory hematology.
Despite wide acceptance of the PR process, little is
known about what impact this laboratory practice actually
has on improving patient care. We designed a study to
evaluate the clinical usefulness of PRs of peripheral blood
and body fluid smears in an academic medical center. The
hematology laboratory at the University Hospitals of
Cleveland has had criteria for PR of abnormal blood and
body fluid smears for more than 20 years. The PR com-
MATERIALS AND METHODS
Pathologist Reviews
Accepted for publication August 3, 2006.
From the Department of Pathology, University Hospitals of Cleveland, Cleveland, Ohio.
The authors have no relevant financial interest in the products or
companies described in this article.
Reprints: Linda M. Sandhaus, MD, University Hospitals of Cleveland,
Department of Pathology, 11100 Euclid Ave, Cleveland, OH 44106 (email: [email protected]).
Criteria for PRs are defined for quantitative and qualitative
abnormalities of CBC/DIFF parameters, and include age-related
criteria for anemia, mean cell volume, and absolute lymphocyte
counts. Pathologist reviews are ‘‘reflex-ordered’’ by the LIS
through a bidirectional interface with the automated hematology
analyzer for those samples that meet numeric criteria based on
analyzer results. Pathologist reviews are manually accessioned by
laboratory technologists on smears that meet qualitative criteria
based on actual smear review. The result field, ‘‘PATHOLOGIST
REVIEW: pending,’’ appears in the initial cell count report for
each PR that is ordered. Each weekday morning, clinical pathology residents review the slides and corresponding reports for
each pending PR and obtain relevant clinical information from
the LIS and the anatomic pathology information system. The
slide reviews are then ‘‘signed out’’ with an attending hematopathologist, and interpretive comments are entered into the LIS
in the PATHOLOGIST REVIEW field and are verified electronically by the attending pathologist. For results that are deemed
urgent, the ordering physician (OP) is contacted directly by pager. On weekends, the ‘‘on-call’’ pathology resident reviews urgent
blood and body fluid smears and consults with an attending pathologist, as needed.
468 Arch Pathol Lab Med—Vol 131, March 2007
Pathologist Reviews of Blood and Body Fluid Smears—Sandhaus et al
8
Table 1. Pathologist Review Criteria for
Study Inclusion*
1. Anemia: microcytic anemia (MCV ⬍ 70), macrocytic anemia
(MCV ⬎ 110), or anemia with morphologic features of hemolysis
2. Lymphocytosis suggestive of viral infection or lymphoproliferative disorder
3. Findings suggestive of previously undiagnosed myeloproliferative disorder, myelodysplastic syndrome, or acute leukemia
4. Thrombocytopenia, neutropenia, or combined cytopenias in
a nononcology patient
5. Cerebrospinal fluid or other body fluid with cells suspicious
for malignancy
6. Cerebrospinal fluid with leukocytosis consistent with meningitis
* MCV indicates mean cell volume.
Study Design
The study involved human subjects research, and therefore,
institutional review board approval was obtained with a waiver
of informed consent. We anticipated that the majority of OPs,
who are the subjects of the research, would be housestaff physicians in internal medicine and pediatrics. After obtaining approval from the residency program directors for these 2 departments, the principal investigators met with the housestaff at their
morning reports to explain the study and to request their participation. An information sheet describing the study was distributed at that time, with an explanation that subject and patient
identities would be protected by deidentification of the data.
For practical reasons, it was not possible to obtain clinical follow-up on all PRs. Therefore, we narrowed the PR criteria for
study inclusion to focus on those that were most likely to have
clinical significance (Table 1). Reactive neutrophilias, joint fluid
crystal exams, newborn blood smears, cytopenias on known cancer chemotherapy patients, and negative body fluid smears were
excluded. Pathologist reviews on patients with previously diagnosed hematologic diseases or previously described hematologic
abnormalities were also excluded. During the 2-month study period (September 7 through October 30, 2005), cases that met the
criteria for study inclusion were identified at daily sign-out. Pathology residents contacted the OP by alpha-numeric pager the
day after the PR was reported. A maximum of 3 attempts to reach
the OP were made on successive days, up to 48 hours after the
PR was issued. For PRs performed on Fridays, calls were made
the following Monday. It was not always possible to reach OPs
from the Emergency Department (ED) the next day. In those cases, follow-up was attempted with the inpatient housestaff for
those patients who were admitted to the hospital. Similarly, if a
housestaff physician could not be reached due to ‘‘post-call’’
hours limitations, another housestaff physician on the same clinical service, or the attending physician, was paged. The interview
consisted of the following 5 questions:
1.
2.
3.
4.
Did you see the PR on this patient sample?
Did the PR contribute to the clinical diagnosis?
Did the PR affect patient management?
Did the PR lead to the ordering of additional laboratory tests?
If yes, what tests were ordered?
5. Did the PR result in the consultation of a subspecialty service?
If yes, what service was consulted?
If the OP had not seen the PR, then the interview was stopped
and the physician was informed of the content of the PR. Pathologist reviews that were considered urgent for patient care (eg,
suspicious for leukemia or other serious blood disorder) were
called directly to the OP at the time of sign-out, and the interview
was conducted during the same telephone call.
Arch Pathol Lab Med—Vol 131, March 2007
Figure 1. Flow chart of pathologist reviews (PRs) conducted during
study period. MD indicates physician; DR, do not remember. * Includes
19 PRs seen by the ordering physician and 7 PRs called to the ordering
physician.
RESULTS
There were 603 total PRs during the study period; 96
(16%) of these met the criteria for study inclusion (Figure
1). The distribution of PRs by diagnosis category is shown
in Figure 2 and the distribution of PRs by physician type,
patient type, and patient location is shown in Table 2.
Contact with an OP was made in 64 (67%) of 96 cases.
Nineteen PRs (30%) had been read by an OP prior to being
contacted by a pathology resident, 8 PRs (13%) were considered urgent enough to warrant immediate communication, 33 PRs (51%) had not seen by an OP at the time of
contact, and in 4 instances (6%), the physicians did not
remember whether they had seen the PR (Figure 3).
For the remaining 32 cases (33%), no contact was made
with a physician. These instances involved cases where
Table 2. Distribution of 96 Pathologist Reviews by
Physician and Patient Type
Percentage
Physician type
Housestaff
Attending
Nurse practitioner
50
48
2
Patient type
Adult
Pediatric
72
28
Patient location
Inpatient
Outpatient
55
45
Pathologist Reviews of Blood and Body Fluid Smears—Sandhaus et al 469
9
Figure 2. Distribution of pathologist reviews
by diagnosis category, N ⫽ 96. MPD/MDS,
myeloproliferative disorder/myelodysplastic
syndrome.
Figure 3. Distribution of 64 pathologist reviews for which contact was made with an
ordering physician. In 13% of the cases, the
pathologist reviews were considered urgent
and were called directly to the ordering physician.
the physician did not return the pages after 3 attempts
(10), a hematology/oncology consultation had already
been requested before the PR was issued (8), the physician
was on vacation (2), no OP was listed in the LIS (10), or
the physician listed in the LIS as the OP denied knowledge
of the patient (2).
The distribution of PRs that were seen by the OPs is
shown in Table 3. Since the total number of cases is small,
statistical significance was not evaluated. The highest proportions of PRs that were seen by physicians were for cytopenias (83%) and body fluids with suspicious or malignant cells (50%). A somewhat higher proportion of inpatient PRs was seen compared to outpatient PRs (39% vs
28%). There were 12 PRs performed on patients from the
ED; 5 of these were from patients who were discharged
from the ED and 7 were from patients who were admitted
to the hospital. None of the PRs for the discharged patients
was seen by a physician, whereas 4 of 7 of the PRs on ED
patients who were admitted to the hospital were seen by
a physician, generally from the inpatient service.
The 27 PRs that were read by clinicians (19) or directly
called to them (8) could be evaluated for their clinical effect on patient care (Table 4). In 23 (85%) of these 27 cases,
the OP responded that the PR contributed to the clinical
diagnosis. The diagnoses for these patients included 7 new
cases of acute leukemia, 6 cases of viral or bacterial meningitis, 3 cases of iron-deficiency anemia, 2 cases of microangiopathic hemolytic anemia, 1 malignant effusion, 1
case of bacterial sepsis, and 3 cases for which no specific
diagnosis was made. Direct effects on patient management
prompted by the PR were evident from 6 cases in which
additional laboratory tests including serum iron studies
and bone marrow examination were performed and 3 cases in which consultation by the hematology/oncology ser-
10
Table 3. Distribution of 64 Pathologist Reviews (PRs)
for Which Contact Was Made With an Ordering
Physician*
No. Seen/Total
PRs (%)
No. Called
Directly
Total (%)
Adult
14/40 (35)
6
20/46 (44)
Pediatrics
5/16 (31)
2
7/18 (39)
Attending
8/25 (32)
6
14/31 (45)
Housestaff
10/30 (33)
2
12/32 (38)
NP
1/1 (100)
0
1/1 (100)
A
4/24 (17)
1
5/25 (20)
B
6/12 (50)
1
7/13 (54)
C
5/6 (83)
2
7/8 (88)
L
2/9 (22)
1
3/10 (30)
M
2/5 (40)
3
5/8 (63)
Outpatient
7/25 (28)
4
11/29 (38)
Inpatient
12/31 (39)
16/35 (46)
Total
19/56 (34)
8
27/64 (42)
* NP indicates nurse practitioner; A, anemia; B, body fluid; C, cytopenias; L, lymphocytosis; and M, myelproliferative disorder/myelodysplastic syndrome.
Table 4. Clinical Effects of 27 Pathologist Reviews
Pathologist
Review Type
Affected
Diagnosis
Additional
Tests
Ordered
Clinical
Consultation
Anemia
3/5
2
0
Body fluid
7/7
1
1
MPD/MDS*
5/5
3
2
Cytopenias
6/7
0
0
Lymphocytosis
2/3
0
0
Total
23/27
6
3
* MPD/MDS indicates myelproliferative disorder/myelodysplastic
syndrome.
vice was obtained. Of the 9 cases in which additional laboratory tests or hematology/oncology consultation were
obtained, 5 involved PRs that were considered urgent and
were called directly to the OP.
COMMENT
Although PRs of peripheral blood and body fluid
smears are a well-established and widely accepted practice in hospital laboratories, few references in the published literature address the clinical usefulness of the
test.2–4 Javidian et al2 stated that the PR is an important
aspect of quality assurance and of training pathology residents. However, another study demonstrated that the review of blood smears for anemia did not correlate with
diagnostic accuracy or the number of tests that were subsequently ordered.4 At an April 2005 conference, the Institute for Quality in Laboratory Medicine proposed evaluating clinician follow-up of laboratory test results as a
quality indicator for clinical laboratories. This recommendation was the motivation for this laboratory quality outcome study.
The results of our study demonstrate the ‘‘added value’’
of PRs of abnormal blood and body fluid smears for patient care. A limitation of the study is the small sample
size. Clinical pathology outcome studies are difficult to
design and perform. Physicians are a notoriously difficult
group to survey, and chart reviews are extremely labor
intensive and time consuming. To minimize the imposition on clinicians and to maximize response rate, the study
Arch Pathol Lab Med—Vol 131, March 2007
focused on a subset of PRs, and the study period was
limited to 2 months in duration. In spite of these limitations, the value of the PRs is clear, as 85% of the PRs that
were seen by a clinician were considered clinically useful.
The number of PRs that were useful to clinicians during
this time period might actually have been higher than our
results suggest because we only followed up on a fraction
(16%) of the total PRs performed during the study period.
The large number of PRs that were not included in the
study consisted of severe normocytic anemias, reactive
neutrophilias, cytopenias on oncology chemotherapy patients, neonatal blood smears, joint fluids, negative body
fluids, and repeat PRs. Many of these PRs might also have
been clinically useful to the primary clinical team and
consultants. In some cases, the value of the PR was immediately evident, such as the 7 new cases of acute leukemia and 1 case in which carcinoma cells were first detected in the pleural fluid cytospin. However, some PRs
might be clinically useful even when they do not lead directly to a specific diagnosis because they help to exclude
other possibilities. Although PRs can contribute to more
prompt diagnoses and treatments for patients, it is not
possible to determine from this study whether hospital
length of stay was affected.
The results suggest that PRs for body fluids, unexplained cytopenias, and findings suggestive of myeloproliferative or myelodysplastic disorders may be more closely followed by clinicians than PRs performed for the other
indications. This behavior makes sense because anemia is
a fairly common finding in hospitalized patients, whereas
leukopenia and thrombocytopenia are more worrisome
findings that may signal previously unsuspected bone
marrow disease or iatrogenic complications. Likewise, the
presence of unsuspected blasts on a leukocyte differential
count is a cause for great concern and is likely to alert a
clinician to follow-up on the pending PR. When invasive
procedures such as thoracentesis or spinal tap are done
for the purpose of collecting diagnostic samples, physicians would be expected to follow-up on the final results.
Also, when the differential diagnosis is contingent on specific blood or body fluid findings, the clinicians may look
for the PR to confirm or exclude diagnoses.
It is a matter for concern that as many as 51% of PRs
were not seen by an OP up to 2 days after the report was
available in the hospital information system. It is possible
that our results might underestimate the percentage of
PRs that were actually seen by clinicians, as a member of
the clinical team other than the OP might have seen some
PRs. Several reasons for the OP not seeing the PRs were
offered: (1) physician went off service, (2) ‘‘a nurse checks
the labs,’’ (3) the attending physician only checks laboratory results 2 days a week, (4) the person listed as the OP
denied knowledge of the patient, and (5) ED patients are
not followed up on by the ED physician. Some of these
reasons are especially worrisome because they suggest the
possibility that other laboratory data on patients might
also be missed. These patterns of physician behavior highlight a major issue in pathology and laboratory medicine,
namely a lack of clinician awareness of laboratory and pathology results, and emphasize the importance of calling
in critical laboratory results.
The process of pathologist review contributes to training pathology residents in laboratory hematology, hematopathology, and cytopathology. Daily review of abnormal
blood smears involves residents in the operations of the
Pathologist Reviews of Blood and Body Fluid Smears—Sandhaus et al 471
11
system improves communication with clinicians about laboratory results and encourages medical staff to come to
the laboratory to review smears with the hematopathologists, which they do regularly.
laboratory and requires them to integrate blood smear
findings with automated analyzer outputs, technologist interpretations, and clinical information. A solid foundation
in blood smear morphology is a prerequisite for bone marrow morphology and is essential to hematopathologic diagnosis. Many anatomic pathology–trained cytopathologists are not comfortable reviewing air-dried, Wrightstained cytospin preparations from the hematology laboratory. Pathology residents benefit from correlating
cytospin smears prepared in the hematology laboratory
with Papanicolaou-stained cytopathology smears. Our PR
system also provides an opportunity for graded responsibility for residents, as required by the Accreditation
Council for Graduate Medical Education. Finally, the PR
1. Sarewitz S, ed. Hematology and Coagulation Checklist. Northfield, Ill: College of American Pathologists; October 6, 2005:40.
2. Javidian P, Garshelis L, Peterson P. Pathologist review of the peripheral
smear: a mandatory quality assurance activity? Clin Lab Med. 1993;13:853–861.
3. Peterson P, Blomberg DJ, Rabinovitch A, Cornbleet PJ, for the Hematology
and Clinical Microscopy Resource Committee of the College of American Pathologists. Physician review of the peripheral blood smear: when and why—an
opinion. Lab Hematol. 2001;7:175–179.
4. Simmons JO, Noel GL, Diehl LF. Does review of peripheral blood smears
help in the initial workup of common anemias? J Gen Intern Med. 1989;4:473–
481.
References
12
RICHTLIJN NVKC CONSULTVERLENING DOOR SPECIALISTEN
LABORATORIUMGENEESKUNDE (KLINISCHE CHEMIE)
HJ Vermeer, laboratoriumspecialist klinische chemie
Het is duidelijk een open deur te stellen dat de laboratoriumgeneeskunde momenteel
sterk in beweging is. De huidige discussies rondom de medische laboratoria vanuit
overheid, zorgverzekeraars en samenleving spitsen zich voornamelijk toe op de
beheersing van kosten terwijl de focus binnen de laboratoria zelf sterk ligt op
consolidatie, integratie en de vorming van netwerken. Een vraag die niet altijd de
aandacht krijgt die ze verdient is deze: hoe is de laboratoriumgeneeskunde in staat
extra value toe te voegen aan haar laboratoriumverrichtingen? Het verlenen van
gevraagde en ongevraagde consulten betekent een belangrijke toevoeging van
waarde aan een laboratoriumtest. Redenerend vanuit het perspectief van de patiënt
gaat het immers maar om één ding en wel de juiste vertaling van laboratoriumdata in
zinvolle informatie aangaande zijn of haar ziekteproces.
Het leveren van bewijs dat consultverlening door specialisten klinische chemie
daadwerkelijk bijdraagt aan betere clinical outcomes blijft nog steeds complexe
materie.
Behalve
research
vraagt
dit
ook
om
standaardisering
van
de
consultverlening zelf. Tijdens de NVKC Voorjaarscongressen in 2009 en 2010 is
derhalve besloten tot het opstellen van een richtlijn consultverlening en deze richtlijn
is tijdens de algemene ledenvergadering van het NVKC Voorjaarscongres 2012
aangenomen. Dit is niet zonder slag of stoot gegaan: de leden voelden zeker een
behoefte tot een meer uniforme werkwijze van consultverlening maar ervoeren een
aarzeling ten aanzien van de nog matige evidence rondom de gestelde normen. De
vraag hoe hard er reeds richtlijnen rondom consultverlening kunnen worden
geformuleerd waaraan men zich tegelijkertijd toetsbaar dient op te stellen, is
daarmee nog niet volledig beantwoord.
In deze presentatie wordt de aanloop naar het opstellen van de richtlijn geschetst.
Daarnaast zal er aandacht zijn voor context waarbinnen deze richtlijn is opgesteld en
zullen de in de richtlijn opgenomen minimum- en streefnormen in kort bestek
gepresenteerd worden. Ten slotte zullen de resultaten getoond worden van een mini-
13
enquête inzake de mate van implementatie in de dagelijkse praktijk sedert de
accordering van de richtlijn.
14
Mini review
The Future of Laboratory Medicine:
Understanding the New Pressures
Mauro Panteghini*
*Australasian Association of Clinical Biochemists’ David Curnow Plenary Lecturer, 2004
Laboratorio Analisi Chimico Cliniche 1, Azienda Ospedaliera “Spedali Civili”, 25125 Brescia, Italy
For correspondence: Prof Mauro Panteghini e-mail: [email protected]
Abstract
Since the future role of Laboratory Medicine is strongly and equally challenged by economic and new technological pressures, it
is essential to take a broad view of the discipline and present to the administrators and other decision-makers the full spectrum of
activities and benefits Laboratory Medicine can provide. In particular, the importance and the true impact of Laboratory Medicine
can only be achieved by adding value to laboratory tests, represented by their effectiveness in influencing the management of
patients and related clinical outcomes.
Introduction
Experiencing a Paradigm Shift
Clinical laboratories represent an area of healthcare that has
always undergone major changes because of technological
advances and external economic pressures.1 In the recent
past, many new diagnostic techniques and laboratory tests
have been introduced as a result of both research on the
fundamental pathogenesis of diseases and the development of
new methods in themselves.
Reaction on the part of administrators and decision makers to
decreased availability of funds has begun on several fronts,
and the funding position of clinical laboratories throughout
the world is becoming critical. Laboratories are indeed an
easy target for economic restrictions and limitations due to
their technological characteristics.2 Furthermore, laboratory
testing on hospital inpatients usually is reimbursed under a
diagnostic-related group (DRG). Under this arrangement,
the hospital is paid a fixed rate for a DRG regardless of how
many (or how few) tests actually are performed. Reducing
laboratory costs will therefore improve the profit margin of
the hospital.3
The two Nobel prizes awarded respectively to the inventors
of monoclonal antibodies (G. Koehler and C. Milstein, 1984)
and the polymerase chain reaction (K.B. Mullis, 1993) are
only the more visible tips of a huge iceberg of innovation
in the field. Without these techniques, many immunoassays
and methods of molecular genetic testing that are currently
taken for granted would simply have been impossible. On
the other hand, in recent years, significant changes have
been made to health care systems and care policy, largely
because governments have had to address extremely complex
economic issues.2
In clinical laboratories, cost savings have frequently been
realised by consolidation of laboratory sections with the
creation of central core laboratories. Further economies of
scale have been sought through regionalisation of laboratory
services with the creation of individual laboratories serving
different health care facilities.4 In some situations, supposed
savings have also been achieved by the addition of automated
pre-analytical specimen handling using robotic systems.5
Clin Biochem Rev Vol 25 November 2004 I 207
15
Panteghini M
Unfortunately, this “technological” approach to lowering
costs per assay has frequently been used to undermine the
influence of laboratory professionals and to further isolate
them from clinical problems.1 On the other hand, laboratory
professionals are usually trained to concentrate on the technical
performance and on the achievement and maintenance of the
highest quality test results generated in laboratories. Often
forgotten is the value of clinical information associated with
clinical laboratory testing. But it is clearly not enough to
report the right results if such data are not used for patient
care. From the patient’s point of view the conversion of
data into useful information is the only thing that counts.6
The entire picture requires a general knowledge model
that moves from laboratory data to information, into new
knowledge to facilitate medical decisions by care givers and,
ultimately, the intervention and outcome.7 This integration
and understanding is the real challenge faced by laboratory
pathologists and scientists in an era when the number of
available test parameters have increased enormously and
the available funds have significantly decreased. Thus, the
survival of Laboratory Medicine in such an environment
ultimately depends on the ability to add value to the care of
patients. The key to appreciating the importance and the true
impact of diagnostic testing can only be achieved if the cost
aspects are considered in the wider overall context of health
economics and not within the more blinkered area of pure
laboratory economics where, almost by definition, every test
represents a cost, and its value is outside the scope of the
laboratory practice.8
Measuring the Outcome of Laboratory Practice
How can this thinking be applied in Laboratory Medicine? It is
clear that the “raison d’être” of laboratories should be assessed
only in the context of the impact of their output on clinical
services, and the other benefits from the laboratory service.
In other words, clinical laboratories have to use outcomes
research to be competitive in a changed health-care landscape
that is characterised by financial problems, and in the use
of a wide variety of medical procedures and technologies.9
Laboratory professionals must now think more globally and
perform studies that demonstrate the impact of laboratory
tests on overall patient health, the cost of patient care, and
other less tangible utilitarian measures, such as quality of life
and patient satisfaction.10 Understanding laboratory-related
outcomes enables the clinical laboratory to become involved
with institutional process improvement, including practice
guideline development, redesign of laboratory services,
and application of patient satisfaction measures within the
organisation.11
Assessment of clinical outcomes in relation to clinical
diagnostics is, however, difficult.12 Typical measures in
outcomes include morbidity, mortality, quality of life,
satisfaction with care, and cost of care, but there are many
problems performing outcome studies in Laboratory
Medicine, such as the gap between the outcome measures
and the biochemical testing.1 Frequently, there is a role for
surrogate markers to be used to assess the clinical impact
of laboratory practice (Table 1).10 In fact, it is easier and
quicker to measure changes in utilisation of resources, such
as the length of hospital stay or the number of clinic visits,
than it is to assess the years of life gained. These outcomes
may not be traditional, but they are valuable, and we should
start using them. One of the best examples of a surrogate
outcome is glycated haemoglobin (HbA1c), which can be used
as a surrogate marker of glycaemic control and for assessing
compliance with therapy in diabetic patients.
Three levels of laboratory-related patient outcomes have been
defined.11 The first-order laboratory outcome is simply the
performance of a given test result, in terms of sensitivity and
specificity in actual practice. Thus every test has at least four
sets of outcomes associated with it; namely, the consequences
of a true positive, a true negative, a false positive, and a false
negative result. The second-order laboratory outcome is the
probability of disease in the patient as estimated by the caregiver
receiving the laboratory result; namely, the predictive value of
the test as determined using Bayes’ theorem. The third-order
laboratory outcome is the actual probability of a change in
health status of the patient resulting from any therapeutic
interventions either instituted or foregone based on the test
Table 1. Types of outcome measures.
Clinical outcome
Surrogate outcome
Mortality
Morbidity
Quality of life, e.g. quality-adjusted life year (QALY)
Cost of episode
Cost of treatment
Length of stay
Number of clinic visits
Disease markers, e.g. HbA1c, LDL cholesterol
Complication rate
Readmission rate
208 I Clin Biochem Rev Vol 25 November 2004
16
The Future of Laboratory Medicine
result. In the end, all healthcare measures, including laboratory
tests, should be judged with respect to their ability to maintain
or restore a patient’s health.
Presently, there are good examples of situations where the
judicious choice and use of diagnostic testing can significantly
reduce the overall costs of treating the patient, accompanied
frequently by a better overall clinical outcome for the patient.
In certain clinical situations the introduction of new and more
effective laboratory tests has influenced the management of
patients and related clinical outcomes directly. One example
of this is the introduction of cardiac troponin for the diagnosis
and treatment of patients with diseases in the spectrum of acute
coronary syndrome.13 Cardiac troponin could be the paradigm
of the new role of Laboratory Medicine in many diseases.14 As
yet, no other clinical information or any other diagnostic test
can replace the information provided by the measurement of
troponin. Cardiac troponins are presently regarded as the most
specific and sensitive of the currently available diagnostic
techniques for myocardial damage, and the redefined criteria
used to classify acute coronary syndrome patients presenting
with ischaemic symptoms as myocardial infarction patients
are heavily predicated on an increased concentration of these
markers in blood.15 Troponins also are the only markers
identifying high-risk coronary patients who should be treated
with anti-thrombotic agents, such as glycoprotein IIb/IIIa
antagonists, and referred for invasive evaluation at the
earliest convenience.16 When compared with the traditional
diagnostic approach (elevated CK-MB), troponin is markedly
effective in altering patient management by enabling early
discharge of patients, resulting in significant cost savings
and increasing bed availability. In a British study conducted
over six months, the introduction of troponin led to a saving
of more than £20,000 to the hospital from fewer bed days
and reduced patient episode cost.17 In another study of more
than 850 consecutive patients presenting to the emergency
department with suspected myocardial infarction who were
randomised to receive a standard evaluation with serial
electrocardiograms and CK-MB tests (control group) with or
without a serial cardiac troponin evaluation, the length of stay
was significantly shorter and hospital charges were less for
patients who had troponin measurements, with an impressive
potential annual saving of about US$4 million.18 Collinson et
al. recently showed that 5% of all admissions in their hospital
for suspected acute coronary syndrome were incorrectly
classified as myocardial infarction using the traditional WHO
criteria.19 The potential annual drug cost for treatment of these
patients as infarction patients was approximately £56,000,
with a 10-year estimated cost close to half a million pounds in
wasted resources.19
Another example is represented by the use of B-type
natriuretic peptide (BNP) in screening symptomatic patients
for left ventricular dysfunction. In a recently published
analysis, screening of high risk individuals by BNP before
echocardiogram appeared to be more cost-effective than
referring all subjects for echocardiography, with a reduction
in the cost of screening per detected case of left ventricular
systolic dysfunction by 21%.20
In addition to diagnostic problems, clinical laboratories are
now increasingly becoming involved in assisting physicians
to make therapeutic decisions. For instance, the recently
updated guidelines of the U.S. National Cholesterol Education
Program for treatment of hypercholesterolaemia in adults
are based on well-defined low-density lipoprotein (LDL)
cholesterol values, indicating when drug therapy should
be initiated and what the treatment goals will be.21 Another
example is represented by HbA1c. The clinical use of this
marker as a target for more aggressive therapy in order to
reduce the development and the progression of retinopathy,
nephropathy, and neuropathy in diabetes mellitus patients is
now well recognised. But it has recently been reported that
HbA1c also predicts mortality in non-diabetic men, with an
increasing risk as the concentration increases, even below the
commonly used upper reference limit.22 A last example is a
recently published study, demonstrating that procalcitoninguided treatment of lower respiratory tract infections is
able to significantly reduce antibiotic use in this type of
disease without any compromise in outcome.23 Low serum
procalcitonin concentrations identified patients without
clinically relevant bacterial infections, in whom antimicrobial
therapy can be safely withheld. Thus, in view of the current
overuse of antibiotics in acute respiratory tract infections,
treatment based on procalcitonin measurement may have
important financial and clinical implications. In addition to
lower costs, a reduction of antibiotic use also results in fewer
side effects and, in the long-term, leads to diminishing drug
resistance.
Changing Role for Medical Laboratory Professionals
In order to meet the changing testing needs, the role of
the laboratory in patient management should therefore be
improved by adding value to laboratory tests derived from
appropriate test request and utilisation. This brings us to what
the laboratory scientist actually does within his own laboratory.
Although it is fundamental that he takes responsibility for
how laboratory tests are used for patient care, many people
still emphasise the development of analytical expertise at the
expense of the application of laboratory science to Medicine.
Some reasons can be enumerated to explain this situation:
reluctance by laboratory scientists to involve themselves in
test structuring and requesting and in the inspection of work as
it arrives because it is assumed that all requests are clinically
17
Panteghini M
necessary (it is a fact that once blood has been taken and the
request has reached the laboratory, it is easier to perform the
test than to discuss its suitability with the referrers); poor
communication and integration between wards and laboratory,
due in part to the uncommunicative attitude of some clinicians
to the “service” departments; and, last but not least, the
need for an excellent cultural and scientific background for
implementing outcome research. This requires the laboratory
scientists to have knowledge in a diverse group of medical
specialties and organisational and leadership skills that are
necessary for functioning successfully in inter-departmental
multidisciplinary teams.
Table 2. Practical significance of biological variation:
knowledge of the biological variation for the analyte is
required in order to answer the following questions.
On the other hand, physicians who frequently request
laboratory tests outside of their field of expertise lack the
knowledge base to order the optimal sequence of tests and
to correctly interpret the results.24 Conversely, medical
laboratory professionals, combining clinical knowledge with
experience in the performance of laboratory assays, have the
unique expertise to advise their clinical colleagues in regard to
the appropriate test selection and interpretation of laboratory
results.25 Knowledge of analytical and biological variation
and the influence of physiological status and co-morbidities
are critical in the interpretation of laboratory results, but many
clinicians are unaware of these. For example, the reliability
of information derived from a laboratory test may heavily
depend on the quality of the analytical performance of the
assay being used for the corresponding measurement.
ensuring that consumers of our services actually use these
aids to test interpretation. Recent studies have provided
information on the biological variation of BNP and N-terminal
proBNP, showing broad fluctuations of their concentrations
in the blood of healthy subjects.30 The critical difference for
these markers has been calculated as being approximately 7090%. Therefore, caution should be exercised in interpreting
concentration changes of BNP of less than 80% on average as
being related to medical therapy. Minor changes could simply
be due only to the random fluctuation of the biomarker around
the homeostatic set point of the individual and not to the effect
of a given therapeutic regimen.31
It is well demonstrated that the use of the more sensitive
cardiac troponin instead of the traditional criteria for the
diagnosis of myocardial infarction leads to an average increase
in the number of infarcts diagnosed, from 20 to 30%, in
patients admitted with suspected acute coronary syndrome.26
However, the percentage of patients re-categorised from
angina to myocardial infarction is also critically dependent
on the performance of the troponin assay used.27 Since
experimental data indicates that various commercial methods
have significantly different sensitivities for detection of cardiac
troponin in blood samples with very low concentrations of this
biomarker, the selection of the troponin assay by the clinical
laboratory represents one of the major factors influencing the
clinical performance of this important biomarker.28
Biological variation is frequently the most important source
of variability in laboratory measurements. Knowledge of
the biological variability is critical to understanding the
significance of a laboratory result (Table 2). The importance
of the critical difference, also called “reference change value”,
is to determine whether changes in an individual’s serial
results are really significant. Only by knowing analytical and
biological variability is it possible to calculate this figure.29
Laboratories need to put these tools into everyday practice,
What is the significance of this result?
When should I measure it again?
Has this result changed significantly over time?
Is the performance of the analytical assay
appropriate (imprecision, bias)?
A demonstration of the possible influence of the physiological
situation on the clinical value of laboratory tests can be
derived from the behaviour of pancreatic amylase in infants
and children. Due to the slow development and maturation of
some functions of the exocrine pancreas, pancreatic amylase
reaches adult concentrations only after the fifth year of life.32
As a consequence, the use of this enzyme for the diagnosis
of acute pancreatitis in young children should be avoided,
and be replaced with the measurement of pancreatic lipase.
Nevertheless, some paediatricians are unaware of this and
continue to request an amylase determination in children with
acute abdominal pain and suspected acute pancreatitis.33
Co-morbidities are also critical in test interpretation, as in
the case of the influence of a reduction in the glomerular
filtration rate on blood concentrations of C-telopeptide of
type I collagen (CTx), a biomarker of bone resorption.34
Thus, in patients with impaired renal function, measurement
of serum CTx needs to be interpreted with great caution. In
this type of patient, other serum markers of bone resorption,
such as tartrate-resistant acid phosphatase 5b isoform, which
is not influenced by renal function, should be considered.35
It is clear from my personal experience that physicians are
greatly confused by the amount of information and make
many errors in the selection and interpretation of laboratory
tests. As an example, Figure 1 displays the results of an
18
audit on the reasons for the request of measurement of bone
turnover markers in different clinical departments done in my
hospital some years ago. When asked to explain the reason for
the test request at the time of ordering, orthopaedic surgeons
were unable to formulate sound reasons in all but one case,
and the number of profiles decreased from 49 in the three
month period before the introduction of the specific request,
to only one in the same period the year after the introduction
of the justification process. Clearly, the exercise helped to
identify misconceptions and ignorance on the use of these
types of tests. Other authors have shown that the involvement
of laboratory professionals in test selection and interpretation
can significantly decrease the likelihood of some types of
medical errors.24
Promoting the Laboratory-Clinic Interface
The laboratory-clinic interface is, therefore, of
fundamental importance to ensure that the patient is
given high quality care, because it provides the boundary
for the multidisciplinary activities which result in the
improvement of the appropriateness of test requests
and in the exchange of information on test results.36,37
consultancy in Laboratory Medicine: 1. use of reflex testing
and algorithms; 2. providing interpretative comments; and 3.
organisation of clinical audits.1
Many examples demonstrate the effectiveness of reflex testing
and algorithms for shortening the time of diagnosis and
rationalising the use of laboratory testing. The most common
example where a cascade of tests is performed based on an
abnormal (frequently chance) biochemical finding, is in the case
where monoclonal gammopathy is suspected. In this case, an
abnormal band found on protein electrophoresis might trigger
the performance of immunofixation and monoclonal protein
quantitation to confirm the presence of this abnormality. Figure
2 shows another example related to an algorithm proposed
for the interpretation of hyperamylasaemia.38 This work-up
begins with the measurement of amylase in serum. A high
value leads to reflexive testing for pancreatic lipase, followed
by serum creatinine or isoamylase assays. The algorithm is
able to determine, with a high degree of confidence, if the
underlying pathophysiology is due to the presence of acute
pancreatitis or of other causes of hyperamylasaemia, such as
extra-pancreatic abdominal disorders or renal insufficiency.38
In order to fill the need for better quality health care, avoidance
of medical errors, and cost reduction, three strategies have
been recommended for supporting and promoting clinical
Figure 1. Results of an audit on the reasons of the request for measurement of bone turnover markers. Apr-Jun ’96: number
of profiles before the introduction of the specific request; Apr-Jun ’97: number of profiles after the introduction of the specific
request.
19
Panteghini M
Figure 2. Proposed algorithm for the interpretation of hyperamylasaemia. Adapted from ref. 38. URL, upper reference limit.
The second recommended strategy is to provide a patientspecific comment and, if necessary, graphical interpretation
of complex test results in order to allow a more objective
utilisation of the data.39 Adding an interpretative comment
to the patient’s results and, eventually, giving advice on any
action that should be undertaken represents an essential tool
for adding value to laboratory reports. An audit of this type
of activity in our institution demonstrated the impact of the
availability of laboratory-generated interpretative comments
on clinical decision making.40 Our investigation showed that
comments appeared to be useful to better classify patients with
suspected acute coronary syndrome in 70% of cases out of a
total of 60 requests of cardiac marker tests. Only in less than
15% of these cases were the laboratory comments fully ignored
by the clinicians.40 Similar findings were recently obtained at
the Massachusetts General Hospital in Boston.41 We may pose
questions on the responsibility of and accountability for these
actions, and on potential pitfalls of making a judgement on
clinical issues based on the knowledge of biochemical pattern
recognition, without necessarily having an insight into the
clinical process of patient management.42 However, if we
consider that in many cases laboratory investigations should
aim to identify a pathophysiological process rather to confirm
a diagnosis, I don’t see any problems in a laboratory comment
reporting, for instance, “a significant increase of specific
cardiac markers consistent with the presence of myocardial
necrosis” or “a significant increase of bone resorption markers
consistent with increased osteoclast activity”. As laboratory
specialists, while assuming responsibility to guarantee
reliable laboratory information, we have to educate physicians
to accept laboratory results as information describing a
pathophysiological process, not a morphological diagnosis.43
Using bone disorders as an example, Jabor and Palicka have
well illustrated the issue of the rational and non-rational use
of laboratory tests.44 If the clinical question is to make the
diagnosis of osteoporosis, the correct test is bone densitometry,
which can provide a morphological diagnosis. Conversely,
biochemical markers should be used if clinicians need to
ascertain any modifications in the activity of osteoblasts and
osteoclasts in order to identify alterations of bone turnover,
including the effect of appropriate therapies.44
Although the practice of commenting varies among countries,
audit findings show that still too few laboratories regularly
add interpretative comments to their reports. In a recent
national survey performed in the field of cardiac biomarkers,
only 9% of participants declared that they performed this type
of activity, even when, as in 46% of cases, clinicians required
advice from the laboratory, especially for interpretative
doubts, or when test results were not consistent with clinical
and analytical information.45 The largest barrier to the
wide implementation of a program to generate narrative
20
interpretations in the clinical laboratory is probably the lack of
a sufficient number of specialists in one laboratory to provide
adequate interpretations. A recent report clearly shows the
potential negative consequences of using laboratory staff with
inadequate expertise for commenting.46
The third pillar of the model system of clinical consulting
is clinical audit. Audit in Laboratory Medicine may be
defined as a process of review and assessment of laboratory
performance.47 It is important that laboratories find out
whether they are providing a useful service for the clinicians
they serve, in order to ensure that they provide the optimum
service to the patient. Once again, this activity requires
co-operation with functional areas outside the laboratory,
reflecting the real world of medicine: a co-operative venture
among medical specialty fields.48 As an example, biochemical
protocols for diagnosing and monitoring patients with acute
coronary syndrome in our hospital are subjected to constant
refinement and, if necessary, to changes in parallel with
analytical innovations and new recommendations coming
from expert groups.49 The continuous availability of new tests
in this field is forcing laboratory professionals and clinicians
to revise and compare diagnostic strategies and different
protocols to evaluate whether the new tests are to be used in
addition to, or instead of, other more traditional tests.50 Our
experience shows that the collaboration and co-operation
between those with expertise in Cardiology and Laboratory
Medicine working in the hospital may permit us to achieve a
significant delay reduction through a continuous improvement
of the processes, as well as introduction of changes aimed
at further improving the results, thus ensuring better patient
triage.49
to maximise the influence of the laboratory results on the
management of patients. Advances in science and technology
will continue to result in the introduction of more complex,
expensive, and difficult-to-interpret tests. By integrating
pathophysiologic rationale and preferences of the clinicians
responsible for the care of the patient with valid and upto-date clinical research evidence, Laboratory Medicine,
supported by computerised information and expert systems,
will promote the use of this new knowledge in a timely and
responsible manner, contributing to the provision of better
care more economically. It is undoubtedly impossible to
predict the future, but that does not mean that it is impossible
to prepare for it, keeping the best interest of the patient first
in mind. As laboratory professionals, we will remain viable
only if we build our own future and educate others about the
contribution that Laboratory Medicine can and does make to
health care.
References
1.
Plebani M. Charting the course of medical
laboratories in a changing environment. Clin Chim
Acta 2002;319:87-100.
2.
Pansini N. The national health system: future
possibilities for the clinical laboratory. Clin Chim
Acta 2002;319:101-5.
3.
Young DS, Sachais BS, Jefferies LC. Laboratory costs
in the context of disease. Clin Chem 2000;46:967-75.
4.
Burke MD. Laboratory medicine in the 21th century.
Am J Clin Pathol 2000;114:841-6.
5.
Boyd JC, Felder RA, Savory J. Robotics and the
changing face of the clinical laboratory. Clin Chem
1996;42:1901-10.
6.
Marques MB, McDonald JM. Defining/measuring
the value of clinical information. Clin Leadersh
Manag Rev 2000;14:275-9.
7.
Goldschmidt HMJ. Postanalytical factors and their
influence on analytical quality specifications. Scand
J Clin Lab Invest 1999;59:551-4.
8.
Marshall DA, O’Brien BJ. Economic evaluation
of diagnostic tests. In: Price CP, Christenson RH
editors. Evidence-based laboratory medicine. From
principles to outcomes, AACC Press, Washington;
2003. pp. 159-86.
9.
Lundberg GD. The need for an outcomes research
agenda for clinical laboratory testing. JAMA
1998;280:565-6.
10.
St. John A, Price CP. Measures of outcome. In:
Price CP, Christenson RH editors. Evidence-based
Conclusions
Some years ago, presidents of European Societies of
Laboratory Medicine were asked what they considered to
be the most relevant issues for the future development of
their profession.51 The implementation of request strategies,
the diagnostic validation of tests and knowledge of test
interpretation were indeed ranked as the most important issues.
Today, the complexity of the health-care environment and the
availability of an ever expanding array of laboratory tests
have further increased the need for more integration between
clinical information and laboratory data.6 This is especially
true in genetic testing, because it should be performed as an
adjunct to the management of the individual and must be
used in conjunction with the total information concerning the
patient. The impact of the clinical laboratory on the medical
environment of the future will be not only to maintain the
highest quality generated data and to improve the total quality
of the process of providing laboratory information, but also
21
Panteghini M
blood cholesterol in adults (Adult treatment panel
III). JAMA 2001;285:2486-97.
laboratory medicine. From principles to outcomes,
AACC Press, Washington; 2003. pp. 55-74.
11.
Bissell MG. Introduction: what’s in a laboratory
outcome? In: Bissell MG editor. Laboratory-related
measures of patient outcomes: An introduction,
12.
Bruns DE. Laboratory-related outcomes
healthcare. Clin Chem 2001;47:1547-52.
22.
Khaw KT, Wareham N, Luben R, et al. Glycated
haemoglobin, diabetes, and mortality in men in
Norfolk cohort of European Prospective Investigation
of Cancer and nutrition (EPIC-Norfolk). Br Med J
2001;322;15-8.
23.
Chist-Crain M, Jaccard-Stolz D, Bingisser R, et al.
Effect of procalcitonin-guided treatment on antibiotic
use and outcome in lower respiratory tract infections:
cluster-randomized, single-blinded intervention trial.
Lancet 2004;363:600-7.
in
13.
Panteghini M. Acute coronary syndrome.
Biochemical strategies in the troponin era. Chest
2002;122:1428-35.
14.
Panteghini M. Role and importance of biochemical
markers in clinical cardiology. Eur Heart J
2004;25:1187-96.
24.
Jaffe AS, Ravkilde J, Roberts R, et al. It’s time
for a change to a troponin standard. Circulation
2000;102:1216-20.
Kratz A, Laposata M. Enhanced clinical consulting –
moving toward the core competencies of laboratory
professionals. Clin Chim Acta 2002;319:117-25.
25.
Price CP, Christenson RH. Teaching evidence-based
laboratory medicine: a cultural experience. In:
Price CP, Christenson RH editors. Evidence-based
laboratory medicine. From principles to outcomes,
26.
Koukkunen H, Penttilä K, Kemppainen A, et al.
Differences in the diagnosis of myocardial infarction by
troponin T compared with clinical and epidemiologic
criteria. Am J Cardiol 2001;88:727-31.
27.
Ferguson JL, Beckett GJ, Stoddart M, Walker SW,
Fox KAA. Myocardial infarction redefined: the new
ACC/ESC definition, based on cardiac troponin,
increases the apparent incidence of infarction. Heart
2002;88:343-7.
15.
16.
Braunwald E, Antman EM, Beasley JW, et al. ACC/
AHA guidelines for the management of patients
with unstable angina and non-ST-segment elevation
myocardial infarction: a report of the American
College of Cardiology/American Heart Association
task force on practice guidelines (Committee on the
management of patients with unstable angina). J Am
Coll Cardiol 2000;36:970-1062.
17.
Owen A, Khan W, Griffiths KD. Troponin T: role
in altering patient management and enabling earlier
discharge from a district general hospital. Ann Clin
Biochem 2001;38:135-9.
18.
Zarich S, Bradley K, Seymour J, et al. Impact of
troponin T determinations on hospital resource
utilization and costs in the evaluation of patients
with suspected myocardial ischemia. Am J Cardiol
2001;88:732-6.
28.
Panteghini M, Pagani F, Yeo KTJ, et al. Evaluation
of imprecision for cardiac troponin assays at lowrange concentrations. Clin Chem 2004;50:327-32.
29.
Fraser CG, Hyltoft Petersen P. The importance of
imprecision. Ann Clin Biochem 1991;28:207-11.
19.
Collinson PO, Rao AC, Canepa-Anson R, Joseph
S. Impact of European Society of Cardiology/
American College of Cardiology guidelines on
diagnostic classification of patients with suspected
acute coronary syndromes. Ann Clin Biochem
2003;40:156-60.
30.
Wu AHB, Smith A, Wieczorek S, et al. Biological
variation for N-terminal pro- and B-type natriuretic
peptides and implications for therapeutic monitoring
of patients with congestive heart failure. Am J
Cardiol 2003;92:628-31.
31.
20.
Nielsen OW, McDonagh TA, Robb SD, Dargie HJ.
Retrospective analysis of the cost-effectiveness of
using plasma brain natriuretic peptide in screening
for left ventricular systolic dysfunction in the general
population. J Am Coll Cardiol 2003;41:113-20.
Panteghini M, Clerico A. Understanding the clinical
biochemistry of N-terminal pro-B-type natriuretic
peptide: the prerequisite for its optimal clinical use.
Clin Lab 2004;50:325-31.
32.
Gillard BK, Simbala JA, Goodglick L. Reference
intervals for amylase isoenzymes in serum and
plasma of infants and children. Clin Chem 1983 ;29:
1119-23.
33.
Tikanoja T, Rautiainen P, Lijala M, Svens E, Tikanoja
S. Hyperamylasemia after cardiac surgery in infants
and children. Intensive Care Med 1996;22:959-63.
21.
Expert panel on detection, evaluation, and treatment
of high blood cholesterol in adults. Executive
summary of the third report of the National
Cholesterol Education Program (NCEP). Expert
panel on detection, evaluation, and treatment of high
22
34.
35.
Pagani F, Bonetti G, Stefini F, Panteghini M.
Evaluation of a fully automated assay to measure Ctelopeptide of type I collagen in serum. Clin Chem
Lab Med 2000;38:1111-3.
Pagani F, Boselli C, Panteghini M. Evaluation of an
immunoassay specific for serum tartrate-resistant acid
phosphatase (sTRAP) 5b isoform, a novel marker of
bone turnover. Clin Chem Lab Med 2003;41(suppl):
S87.
36.
Büttner J. Good laboratory practice: the medical aspects.
Eur J Clin Chem Clin Biochem 1997;35:251-6.
37.
Plebani M. The clinical importance of laboratory
reasoning. Clin Chim Acta 1999;280:35-45.
38.
Panteghini M, Pagani F. Clinical evaluation of an
algorithm for the interpretation of hyperamylasemia.
Arch Path Lab Med 1991;115:355-8.
39.
Dighe AS, Soderberg BL, Laposata M. Narrative
interpretations for clinical laboratory evaluations.
Am J Clin Pathol 2001;116:S123-8.
40.
Panteghini M, Cuccia C, Pagani F, Bonetti G. Gli
“enzimi cardiaci” nell’era delle troponine: cosa
salvare. Biochim Clin 1999;23:378-85.
41.
Laposata M. Patient-specific narrative interpretations
of complex clinical laboratory evaluations: who is
competent to provide them? Clin Chem 2004;50:471-2.
42.
Waise A, Plebani M. Which surrogate marker can
be used to assess the effectiveness of the laboratory
and its contribution to clinical outcome? Ann Clin
Biochem 2001;38:589-95.
43.
Plebani M. The changing face of clinical laboratories.
Clin Chem Lab Med 1999;37:711-7.
44.
Jabor A, Palicka V. Rational use of clinical chemistry
investigations: from diagnoses to processes. Ann
Clin Biochem 1998;35:351-3.
45.
Sciacovelli L, Zardo L, Secchiero S, Zaninotto M,
Plebani M. Interpretative comments and reference
ranges in EQA programs as a tool for improving
laboratory appropriateness and effectiveness. Clin
Chim Acta 2003;333:209-19.
46.
Lim EM, Sikaris KA, Gill J, et al. Quality assessment
of interpretative commenting in clinical chemistry.
Clin Chem 2004;50:632-7.
47.
Plebani M, Chiozza ML. Audit in laboratory medicine.
Eur J Clin Chem Clin Biochem 1996;34:655-7.
48.
Lewandrowski K. Managing utilization of new
diagnostic tests. Clin Leadersh Manag Rev
2003;17:318-24.
49.
Panteghini M, Pagani F, Bonetti G, Cuccia C.
Biochemical algorithms in the troponin era: audit
of some care maps one year after their introduction.
Biochim Clin 2000;24:469-75.
50.
Panteghini M. Biochemical markers of cardiac
damage: what is current, what is redundant? Biochim
Clin 2000;24:431-8.
51
Guder WG, Büttner J. Clinical chemistry in
laboratory medicine in Europe – Past, present and
future challenges. Eur J Clin Chem Clin Biochem
1997;35:487-94.
23
Richtlijn NVKC Consultverlening door specialisten laboratoriumgeneeskunde (klinische chemie)
Oosterhuis WP1,2,7,8, Verboeket-van de Venne WPHG2, Kuiper-Kramer PA1,3,
Ulenkate HJLM4,7, Vermeer HJ5,7,8, Jansen RTP1,6
1
Werkgroep Consultfunctie, NVKC
2
3
Isala klinieken, Zwolle
4
ZorgSaam Ziekenhuis, Terneuzen
5
Albert Schweitzer Ziekenhuis, Dordrecht
6
UMC St. Radboud, SKML, Nijmegen
7
Werkgroep Richtlijnen, NVKC
8
Commissie Kwaliteit, NVKC
Correspondentie:
Dr. Drs. W. Oosterhuis
Laboratorium voor Klinische Chemie en Hematologie
Atrium Medisch Centrum Parkstad
Postbus 4446
6401 CX HEERLEN
E: [email protected]
T: 045 5766341
F: 045 5766575
projectnr. 4123579
Evidence based medicine bij laboratoriumdiagnostiek
projectnr. 4123039
Richtlijn feedback eerste lijn
24
Samenvatting Richtlijn NVKC Consultverlening door specialisten laboratoriumgeneeskunde (klinische chemie)
Met deze richtlijn consultverlening willen we bewerkstelligen dat er meer uniformiteit
komt in de consultatieve taken van de laboratoriumspecialist. Dit draagt bij aan een
betere dienstverlening vanuit het laboratorium en een betere patiëntenzorg.
Uniformiteit in de verwerking van afwijkende combinaties van uitslagen kan hieraan
een bijdrage leveren.
In de richtlijn wordt de term ‘laboratoriumspecialist’ gebruikt. Hieronder wordt
verstaan een specialist laboratoriumgeneeskunde (klinische chemie), arts klinische
chemie of andere laboratoriumspecialist, geregistreerd via een formeel register.
In de richtlijn zijn aanbevelingen herkenbaar opgenomen en voorzien van status,
minimumnorm of streefnorm. Minimumnormen geven een harde ondergrens aan en
moeten terugkomen in lokale procedures. Streefnormen geven optimale zorg aan.
Een samenvatting van alle aanbevelingen is onderstaand weergegeven.
Gewenste condities voor consultverlening
Minimumnorm Ö Er is te allen tijde een laboratoriumspecialist beschikbaar –
aanwezig of bereikbaar – voor het verlenen van consulten.
Streefnorm
Ö In opleidingsziekenhuizen met maatschappen die opleidingen
verzorgen worden vaak verplichte patiëntenbesprekingen georganiseerd. De laboratoriumspecialist neemt bij voorkeur deel aan
relevante patiëntenbesprekingen.
Registratie consultverlening
Minimumnorm Ö Consulten dienen te worden geregistreerd.
Streefnorm
Ö Een consult dient altijd door de behandelaar te kunnen worden
ingezien. Dit is mogelijk door middel van registratie van een
consult in een elektronisch patiëntendossier via een aan het
resultaat gekoppelde commentaartekst in het uitslagrapport of per
brief.
25
NZa verrichtingencodes
Streefnorm
Ö Er wordt door de laboratoriumspecialist bij het registreren van
consulten gebruik gemaakt van NZa verrichtingencodes.
Signalering afwijkende uitslagen en becommentariëring
Streefnorm
Ö De laboratoriumspecialist neemt proactief de taak op zich van
signalering en advies met betrekking tot aanvullende diagnostiek.
In sommige gevallen kan vervolgonderzoek meteen ingezet worden (reflex- of reflecterend testen).
Streefnorm
Ö De laboratoriumspecialist voorziet onderzoeken van interpretatief
commentaar, indien dit naar verwachting bij een relevant gedeelte
van de aanvragers zal bijdragen aan een juiste interpretatie van de
uitslag.
Functieproeven
Streefnorm
Ö Als er bij aanvragers behoefte aan is, voorziet de laboratoriumspecialist functieonderzoeken van interpretatief commentaar.
Feedback
Streefnorm
Ö De laboratoriumspecialist geeft feedback aan de aanvragers,
zodat ze hun aanvraaggedrag kunnen vergelijken met andere
aanvragers.
26
Inleiding
In het meerjarenbeleidsplan 2009-2013 van de NVKC getiteld ‘Van meten naar
consult, van chemisch naar medisch’ (1) wordt een versterking van de consultfunctie
van de laboratoriumspecialist bepleit. Niet het laboratoriumonderzoek op zichzelf,
maar de betekenis van het laboratoriumonderzoek voor de patiënt en de
zorgverlening krijgt hier een centrale plaats. De toenemende complexiteit van het
klinisch chemisch areaal vraagt om een andere rol van de moderne laboratoriumspecialist. Meer en meer vraagt de medisch specialist, maar ook de huisarts en
verloskundige, om advisering op het gebied van selectie, interpretatie en follow-up
van diagnostische testen. Om aan deze veranderende rol tegemoet te komen, zal in
de opleiding, in de na- en bijscholing (ook van andere specialismen) en tijdens
symposia en andere settings ruim aandacht moeten worden gegeven aan de
consultatieve rol van de specialist laboratoriumgeneeskunde (klinische chemie).
De CCKL 4e Praktijkrichtlijn (2) geeft eveneens aan dat klinische consultatie een
wezenlijk onderdeel uitmaakt van de dienstverlening door het klinisch chemisch
laboratorium en dus van de competentie van de laboratoriumspecialist. ISO 15189
(3), waarop de praktijkrichtlijn van de CCKL gebaseerd is, stelt expliciet (4.7
Advisory services): “The laboratory shall establish arrangements for communicating
with users on the following: advising on choice of examinations and use of the
services, including required type of sample, clinical indications and limitations of
examination procedures and the frequency of requesting the examination; advising
on individual clinical cases; professional judgments on the interpretation of the
results of examinations”. Consultverlening wordt daarmee gezien als een integrale
taak van de laboratoriumspecialist.
Ten slotte is consultverlening een nieuwe competentie in het opleidingscurriculum
van de specialist laboratoriumgeneeskunde (klinische chemie) geworden (4).
Op 17 april 2009 werd tijdens het NVKC voorjaarcongres samen met de Federatie
Medisch Laboratorium Specialismen (FMLS) een symposium georganiseerd met als
titel: ‘Klinisch chemicus uit de kast, de kliniek in!’(5). Er werd besloten een werkgroep
in te stellen om de versterking van de consultverlening door laboratoriumspecialisten
27
nader uit te werken. De werkgroep Consultfunctie bestaat uit leden uit de
verschillende geledingen van de FMLS.
De werkgroep is tot verschillende aanbevelingen gekomen. Het is wenselijk dat de
harmonisatie tussen laboratoria, die verregaand is doorgevoerd in de analytische
fase, ook wordt toegepast in de pre- en postanalytische fase. Zo is aangetoond, dat
er tussen laboratoria aanzienlijke verschillen bestaan in het serviceniveau wat betreft
anemiediagnostiek en andere min of meer geprotocolleerde diagnostiek (6,7). De
werkgroep heeft drie voorbeelden vastgesteld om nader uit te werken en afspraken
over vast te leggen: anemiediagnostiek, hemochromatose en het doorbellen van
laboratoriumuitslagen (8).
Tijdens het NVKC-voorjaarscongres op 23 april 2010 stond het thema ‘Richtlijnen’
centraal (8). Er werd melding gemaakt van de komst van de huidige richtlijn
‘Consultverlening door specialisten laboratoriumgeneeskunde (klinische chemie)’.
Het uitgangspunt moet zijn dat de laboratoriumspecialist pro-actief actie onderneemt
richting aanvrager over bepaalde uitslagen en initiatie van eventueel vervolgonderzoek. Hierover dienen afspraken vastgelegd te worden. Daarvoor gelden soms
andere regels dan voor ‘evidence based’ aanbevelingen. Met bekrachtigde
afspraken wordt de laboratoriumspecialist zichtbaarder.
Achtergrond
Wat wordt verstaan onder consultverlening?
Consulteren
wordt
over
het
algemeen
gedefinieerd
als
“raadplegen”
of
“beraadslagen met”. In de medische wereld vraagt men een specialist in consult. De
specialist verleent het consult voor het oplossen van klinische problemen die buiten
de competentie liggen van de verwijzende arts die het consult aanvraagt (9,10).
Onder consultverlening door de laboratoriumspecialist wordt hier verstaan: “elke
vorm van informatieverstrekking aan medische hulpverleners op het terrein van de
laboratoriumgeneeskunde die van belang is voor de diagnostiek c.q. behandeling
van een specifieke patiënt”. De informatie kan mondeling en/of schriftelijk gegeven
worden en het advies kan gevraagd of ongevraagd gegeven worden. Onder de
28
consultatieve rol van de laboratoriumspecialist wordt zowel de klinisch consultatieve
(voorbeeld: interpretatie van uitslagen) als de technisch analytische rol (voorbeeld:
mogelijke storende factoren) verstaan (1).
Verschillende vormen van consultatie
In 2009 verscheen een overzichtsartikel met de verschillende mogelijkheden tot
consultverlening en de meerwaarde die dit kan hebben voor de kliniek (11). In de
preanalytische fase kan de consultverlenende functie actief uitgeoefend worden. Zo
zouden bepaalde testen alleen na overleg met de laboratoriumspecialist kunnen
worden aangevraagd. Op deze manier is de laboratoriumspecialist direct betrokken
bij het diagnostische proces. Bovendien kan de laboratoriumspecialist sturing geven
door actief telefonisch contact te zoeken met de aanvragende specialist om de
aanvraag te bespreken. Aanvullend zou de laboratoriumspecialist regelmatig
(bijvoorbeeld jaarlijks) het aanvraagprofiel van een individuele aanvrager of van een
maatschap kunnen bespreken (feedback).
In
de
postanalytische
fase
worden
verschillende
vormen
van
consultatie
onderscheiden. Ten eerste is er de alarmerende rol, waarbij aan de hand van
doorbelgrenzen sterk pathologische uitslagen direct worden doorgegeven aan de
aanvrager (12). Deze alarmfunctie van het laboratorium wordt meestal door analisten uitgevoerd. Hier ligt echter een mogelijkheid voor de laboratoriumspecialist om
zelf met de aanvrager te overleggen en zo invulling te geven aan de consultfunctie.
Een tweede belangrijke vorm van consultverlening in de postanalytische fase is die
van het interpreteren en becommentariëren van testresultaten. In veel klinisch
chemische laboratoria vindt deze vorm van consultverlening plaats bij specialistisch
onderzoek, zoals morfologische en immunofenotypische analyse van beenmerg en
lymfklieren, hemostase-onderzoek bij patiënten met verdenking trombofilie of
bloedingsneiging, liquoronderzoek met betrekking tot de bloedhersenbarrière en de
aanwezigheid van intrathecale productie van immunoglobulines en onderzoek naar
aanwezigheid van M-proteïnen. Het interpreteren en becommentariëren van anemieprotocollen, fertiliteitonderzoek of semenanalyse past ook bij deze vorm van
consultverlening.
29
Ten derde kan de laboratoriumspecialist een adviserende rol hebben. Voor
specialistische onderzoeken mist de aanvrager soms de benodigde kennis. De
laboratoriumspecialist is dan bij uitstek geschikt om de aanvrager te adviseren bij de
beoordeling van uitslagen van aandoeningen die minder vaak voorkomen.
Voorbeelden hiervan zijn de screening en diagnostiek van familiaire hypercholesterolemie en het herkennen en diagnosticeren van hemoglobinopathieën. Het
laboratorium kan hierbij een belangrijke rol spelen om het diagnostische proces zo
goed mogelijk te laten plaatsvinden.
Een vierde belangrijke vorm van consulteren is die waarbij de laboratoriumspecialist
in overleg treedt met de aanvrager over het in te zetten medisch beleid, zoals bij het
aanvragen van bloedproducten. De laboratoriumspecialist kan aan de hand van het
compatibiliteitsonderzoek besluiten om te overleggen of een bloedtransfusie niet
uitgesteld kan worden in verband met nog in te zetten vervolgonderzoek. Aan de
hand van de in veel instellingen gehanteerde 4-5-6 regel (13) kan de
laboratoriumspecialist in overleg treden met de aanvrager over de noodzaak van de
transfusie.
Ten slotte is reflecterend testen bij uitstek een consultfunctie (14-17). De
laboratoriumspecialist voegt op basis van patronen van laboratoriumuitslagen,
zonder tussenkomst van de aanvrager, extra testen toe om zo de aanvrager verder
te ondersteunen in de diagnostiek. Er wordt gerichte diagnostiek uitgevoerd bij
patiënten met voldoende pre-test waarschijnlijkheid op een aandoening. Nadat de
diagnostiek compleet is, voegt de laboratoriumspecialist nog een ondersteunend
commentaar toe aan het uitslagrapport zodat de aanvrager de uitslagen
gemakkelijker kan interpreteren.
Consultatieve taken tijdens de opleiding
De specialist laboratoriumgeneeskunde (klinische chemie) in opleiding kan voor
deelaspecten, zoals preanalyse en point of care testen (POCT), reeds in een vroeg
stadium van de opleiding consultatieve taken uitvoeren: het principe van “training on
the job” (4). Een assistent in opleiding leert het best adviezen te geven als de
opleider en het opleidingsteam dit structureel beoefenen (4). Het begint met
30
patiëntenbesprekingen goed voor te bereiden. Bij de voorbereiding is er afstemming
tussen de assistent in opleiding en een lid van het opleidingsteam.
Conclusie
Ondanks het belang dat gehecht wordt aan het versterken van de consultatieve
functie van het klinisch chemisch laboratorium, is praktisch onderzoek hierover zeer
beperkt. Er lijken grote verschillen te bestaan tussen laboratoria (en laboratoriumspecialisten) wat betreft de mate van interactie met de kliniek c.q. eerste lijn en er
ontbreekt op dit punt een standaard. Hoewel consultverlening is verankerd in de
opleiding tot specialist laboratoriumgeneeskunde (klinische chemie), bestaat er (nog)
geen duidelijk omschreven beleid binnen de beroepsgroep van de NVKC om de
consultatieve functie te structureren en te versterken. Een proactieve benadering lijkt
hierbij aangewezen.
Aanbevelingen consultverlening
Gewenste condities voor consultverlening
Voor
een
goede
consultverlening
is
een
laagdrempelige
bereikbaarheid
noodzakelijk, waarbij de clinicus/aanvrager altijd (eventueel via bereikbaarheidsdiensten) een laboratoriumspecialist kan consulteren. Aangezien een proactieve
instelling van de laboratoriumspecialist van belang is bij het versterken van de
consultatieve taken is het deelnemen aan patiëntenbesprekingen wenselijk.
Klinische informatie, bijvoorbeeld via inzage in het elektronisch patiëntendossier
(EPD), is hierbij – maar bijvoorbeeld ook bij het interpreteren en becommentariëren
van testresultaten – onontbeerlijk (zie Richtlijn NVKC Vrijgave van laboratoriumuitslagen (7)).
Aanbeveling 1:
minimumnorm
Er is te allen tijde een laboratoriumspecialist beschikbaar – aanwezig of bereikbaar –
voor het verlenen van consulten.
31
Aanbeveling 2:
streefnorm
In opleidingsziekenhuizen met maatschappen die opleidingen verzorgen worden
vaak verplichte patiëntenbesprekingen georganiseerd. De laboratoriumspecialist
neemt bij voorkeur deel aan relevante patiëntenbesprekingen.
Registratie consultverlening
Van meerdere kanten is ervoor gepleit consulten te registreren (18,19). In een
recente evaluatie bleek echter een minderheid van de laboratoria dit routinematig uit
te voeren (20). Registratie van een consult van een laboratoriumspecialist is aangewezen als het consult invloed kan hebben op de behandeling van de patiënt.
Systematische registratie maakt bovendien statistische verwerking van de omvang
en aard van de consultverlening mogelijk.
Aanbeveling 3:
minimumnorm
Consulten dienen te worden geregistreerd.
Aanbeveling 4:
streefnorm
Een consult dient altijd door de behandelaar te kunnen worden ingezien. Dit is
mogelijk door middel van registratie van een consult in een elektronisch
patiëntendossier via een aan het resultaat gekoppelde commentaartekst in het
uitslagrapport of per brief.
NZa verrichtingencodes
In een publicatie over verrichtingencodes is in detail de interpretatie beschreven van
de twee codes die er zijn om consulten van laboratoriumspecialisten te registreren
(18). De codering maakt onderscheid naar degene die het consult initieert.
Code 070027: advies op verzoek van een behandelaar over een individuele patiënt.
Het gaat alleen om registratie van werkzaamheden van de laboratoriumspecialist
zelf en dus niet om werkzaamheden van de medewerkers van het laboratorium.
Het betreft vragen van een behandelaar over de interpretatie van een uitslag
(bijvoorbeeld: passen de uitslagen bij het betreffende ziektebeeld); vragen over
de analytische kwaliteit, omdat bij een individuele patiënt het resultaat niet
32
verwacht wordt; mogelijke interferenties van geneesmiddelen of andere verbindingen bij een individuele patiënt; welk vervolgonderzoek bij betreffende patiënt
geïndiceerd is.
Code 070028: advies van de laboratoriumspecialist aan de behandelaar. Het betreft
hier een advies op initiatief van de laboratoriumspecialist aan de behandelaar
over de laboratoriumuitslagen van een individuele patiënt.
Het betreft bijvoorbeeld de volgende adviezen: de betekenis van een (sterk)
afwijkende uitslag of een onverwachte uitslag (daarbij wordt de uitslag niet alleen
doorgebeld, maar ook besproken); schriftelijk interpretatief commentaar bij een
uitslag (automatisch gegenereerd commentaar valt niet onder deze code); een
advies voor vervolgonderzoek; een unieke brief over een patiënt aan de
behandelaar met daarin de uitslagen van laboratoriumonderzoek.
Aanbeveling 5:
streefnorm
Er wordt door de laboratoriumspecialist bij het registreren van consulten gebruik
gemaakt van NZa verrichtingencodes.
Doorgeven van afwijkende uitslagen
Bij sterk afwijkende laboratoriumuitslagen kan de laboratoriumspecialist een bijdrage
leveren aan de diagnostiek en eventueel vervolgonderzoek. Dit is een mogelijkheid
om de consultfunctie te versterken. Het zelf doorbellen van sterk afwijkende
uitslagen maakt het ook mogelijk om te verifiëren of de ontvanger van de uitslagen
deze juist interpreteert. Hierbij moet worden opgemerkt dat dit in het bijzonder geldt
voor aanvragen vanuit de eerste lijn. Buiten kantooruren heeft het doorbellen van
uitslagen door een laboratoriumspecialist naar afdelingen zoals de intensive care of
een acute opname afdeling zelden toegevoegde waarde. Er wordt een regionale lijst
met doorbelgrenswaarden opgesteld, die ook gecommuniceerd is met de aanvragers
en door hen opvraagbaar is (21). Het is wenselijk dat de lijst met een paar
aanvragers tot stand is gekomen c.q. afgestemd. Eventueel bestaat er differentiatie
in de doorbelwaarden voor (poli)klinische- en huisartspatiënten en voor kinderen en
volwassenen (12,22,23)(zie ook Richtlijn NVKC voor vrijgave van laboratoriumuitslagen (7)).
33
Signalering afwijkende uitslagen en becommentariëring
Naast doorbellen kan het klinisch chemisch laboratorium een bijdrage leveren aan
de diagnostiek door herkenning van afwijkende uitslagen of afwijkende patronen. Het
betreft laboratoriumuitslagen, waarvan het niet onwaarschijnlijk is dat de aanvrager
de afwijking onjuist interpreteert (24). Het is onmogelijk hiervan een sluitende
opsomming te geven. Enkele voorbeelden zijn M. Gilbert, hemochromatose, hemoglobinopathieën.
M. Gilbert
Bij een verhoogd bilirubine zonder afwijkende leverenzymwaarden of aanwijzingen
voor hemolyse, moet rekening gehouden worden met het syndroom van Gilbert, een
aangeboren lichte conjugatiestoornis. Het syndroom van Gilbert komt bij 6% van de
bevolking voor, maar vaak wordt het niet als zodanig herkend (25). Herkenning kan
voorkómen dat de patiënt onnodig onderzoek moet ondergaan.
Hemochromatose
Hereditaire hemochromatose is een ziekte die wordt gekarakteriseerd door een
voortschrijdende ijzerstapeling, voornamelijk in de lever, die op termijn leidt tot
orgaanschade. Een hoog ferritine kan ook ontstaan door steatose/NAFLD of door
veelvuldige bloedtransfusies. Bij verhoogd ferritine in combinatie met een verhoogde
ijzerverzadiging (>45%) is DNA-onderzoek naar mutaties geïndiceerd (26). Het
verdient aanbeveling om een protocol beschikbaar te hebben bij verdenking
‘hemochromatose’. De inhoud van het protocol kan per laboratorium verschillen.
Hemoglobinopathie
Hemoglobinopathie is een erfelijke bloedziekte waarbij het lichaam onvoldoende en
afwijkend hemoglobine aanmaakt. Bij anemieonderzoek dient eventueel vervolgonderzoek naar de verdenking hemoglobinopathie plaats te vinden. Parameters als
de Q-index, ret-He en ZPP kunnen hierbij van nut zijn (27).
34
Aanbeveling 6:
streefnorm
De laboratoriumspecialist neemt proactief de taak op zich van signalering en advies
met
betrekking
tot
aanvullende
diagnostiek.
In
sommige
gevallen
kan
vervolgonderzoek meteen ingezet worden (reflex- of reflecterend testen).
Uitslagen worden meestal zonder interpretatie doorgegeven. Sommige combinaties
van uitslagen zijn echter complex en het kan de patiënt ten goede komen als het
laboratorium de uitslagen voorziet van een interpretatief commentaar. Ook dit
commentaar is op te vatten als een consult. Bij onderzoek onder huisartsen bleek
dat gemiddeld 50% van de uitslagen met karakteristieke afwijkingen juist werd
geïnterpreteerd (24). Dit benadrukt dat interpretatief commentaar wel degelijk een
aanvullende waarde kan hebben.
Aanbeveling 7:
streefnorm
De laboratoriumspecialist voorziet onderzoeken van interpretatief commentaar,
indien dit naar verwachting bij een relevant gedeelte van de aanvragers zal bijdragen
aan een juiste interpretatie van de uitslag.
Functieproeven
Interpretatie van analyseresultaten in geval van functieproeven leent zich bij uitstek
voor toevoeging van commentaar door de laboratoriumspecialist. Men kan hierbij
denken aan endocrinologische functietesten (dexamethasonremmingstest, test om
groeihormoondeficiëntie vast te stellen), lactosetoleratietest, etc.
Aanbeveling 8:
streefnorm
Als er bij aanvragers behoefte aan is, voorziet de laboratoriumspecialist functieonderzoeken van interpretatief commentaar.
Feedback
Het geven van feedback op het aanvraaggedrag van aanvragers is een effectieve
methode om dit aanvraaggedrag te beïnvloeden. Het is een vorm van consultatie die
betrekking heeft op de preanalytische fase. Het laboratorium kan de benodigde
gegevens voor de feedbackrapportage aanleveren in grafiek- of tabelvorm en
35
aanbieden aan de aanvragers (28,29). Desgewenst kan het aanvraaggedrag besproken worden; bijvoorbeeld als bijscholing met huisartsen.
Aanbeveling 9:
streefnorm
De laboratoriumspecialist geeft feedback aan de aanvragers, zodat ze hun aanvraaggedrag kunnen vergelijken met andere aanvragers.
36
Referenties
1.
Meerjarenbeleidsplan 2009-2013: ‘Van meten naar consult, van chemisch naar
medisch’. 2009; NVKC, Utrecht.
http://www.nvkc.nl/organisatie/documents/NVKCbeleidsplanboekje.pdf
2.
CCKL praktijkrichtlijn voor een kwaliteitssysteem voor laboratoria in de
gezondheidszorg. Loeber J.G. en Slagter S. 4e gewijzigde druk (2005)
gebaseerd op ISO 15189. ISBN 90.267.2094.7.
3.
International Organization for Standardization (ISO) 15189:2011. Medical
laboratories – Particular requirements for quality and competence.
4.
Bartels PCM, Willems JL. Consultatie en interpretatie van analyseresultaten:
kerncompetenties van de klinisch chemicus. Ned Tijdschr Klin Chem
Labgeneesk 2009; 34: 165-168.
5.
Kuiper-Kramer PA, Jansen RTP, Oosterhuis WP, Buiting M. Consultfunctie
binnen de klinische chemie: ‘Klinisch chemicus, uit de kast de kliniek in!’ Ned
Tijdschr Klin Chem Labgeneesk 2010; 35: 88-90.
6.
Verboeket-van de Venne WPHG, Oosterhuis WP, Keuren JFW, Ulenkate HJLM,
Leers MPG. Richtlijn NVKC Reflexdiagnostiek bij anemie, 2012.
7.
Oosterhuis WP, Ulenkate HJLM, Horst M van der, Vermeer HJ, Wulkan RW,
Thelen
M.
Richtlijn
NVKC
Vrijgave
van
laboratoriumuitslagen
(eerste
herziening), 2012.
8.
Thelen MHM, Wielders JPM, Oosterhuis WP Ulenkate HJLM, Ruiter C, Burgers
J, Jansen RTP. Vrijdagmiddagsessie ‘Richtlijnen’ NVKC-congres 2010. Ned
9.
Burke MD. Clinical laboratory consultation. Clin Chem 1995; 41: 1237-1240.
10. Burke MD. Clinical laboratory consultation: appropriateness to laboratory
medicine. Clin Chim Acta 2003; 333: 125-129.
11. Oosterhuis WP, Raijmakers MTM, Leers MPG, Keuren JFW, Verboeket-van de
Venne WPHG, Munnix ICA, Kleinveld HA. Consultfunctie: van klinisch chemicus
naar laboratoriumspecialist. Ned Tijdschr Klin Chem Labgeneesk 2009; 34: 214218.
12. Ulenkate HJLM, Dongen CAJM van, Oosterhuis WP, Horst M van der, Dols J,
Volmer M, Wulkan RW. Doorbellen van uitslagen: criteria in verschillende
ziekenhuizen. Ned Tijdschr Klin Chem 2003; 28: 76.
13. Richtlijn Bloedtransfusie, CBO, 2011. http://www.diliguide.nl/document/2903
37
14. Darby D, Kelly AM. Reflective testing – what do our service users think? Ann
Clin Biochem 2006; 43: 361-368.
15. Simpson WG, Twomey PJ. Reflective testing. J Clin Pathol 2004; 57: 239-240.
16. Oosterhuis
WP,
Kleinveld
HA.
‘Reflecterend’
testen:
het
laboratorium
ondersteunt de huisarts actief met professioneel vervolgonderzoek. Ned
17. Keuren JFW, Kleinveld HA, Oosterhuis WP. ‘Reflecterend’ testen wordt
gewaardeerd door huisartsen en heeft een positieve invloed op diagnose en
behandeling. Ned Tijdschr Klin Chem Labgeneesk 2008; 33: 182-183.
18. Doelman CJA. De klinisch chemicus en DBC 2003. Ned Tijdschr Klin Chem 2003;
28: 40-42.
19. Ulenkate HJLM. Registratie van de consulten van de klinisch chemicus:
leerzaam en een ‘must’ ter verbetering van de dienstverlening. Ned Tijdschr Klin
Chem Labgeneesk 2005; 30: 55-60.
20. Kortlandt W, Fischer JC, Doelman CJA, Hens JJH, Henskens YMC, Keyzer JJ.
Het gebruik van en wensen voor een electronisch consultregistratie systeem (ECRS). Ned Tijdschr Klin Chem Labgeneesk 2011, 36, 75.
21. NHG/NVKC/SAN/NVMM. Rationeel aanvragen van laboratoriumdiagnostiek –
Eerste herziening. LESA (Landelijke Eerstelijns Samenwerkings Afspraak)
2012.
22. Ulenkate H, Dongen C van, Oosterhuis W, Horst M van der, Dols JLS, Volmer M,
Wulkan R. Telephone reporting to clinicians of extreme values: criteria in several
hospitals. Clin Chem Lab Med 2003; 41: S382.
23. Richtlijn Elektrolytstoornissen, NIV, ISBN 90-8523-080-2, 2005: 1-105.
24. Verboeket-van de Venne WPHG, Oosterhuis WP, Waard H de, Sant P van ‘t,
Kleinveld HA. Beïnvloedt ‘reflecterend testen’ het beoordelen van casuïstiek
door huisartsen? Ned Tijdschr Klin Chem Labgeneesk 2011; 36: 272-274.
25. Keularts IMLW, Meijden BB van der, Wielders JPM. Genotypische bevestiging
van syndroom van Gilbert: een geruststelling van de patiënt. Ned Tijdschr Klin
Chem Labgeneesk 2008; 33: 43-47.
26. Richtlijn Hereditaire Hemochromatose. Diagnostiek en behandeling van
hereditaire hemochromatose. NIV/NVKC-VAL, mei 2007.
38
27. Leers MPG, Keuren JFW, Oosterhuis WP. The value of the Thomas-plot in the
diagnostic work up of anemic patients referred by general practitioners. Int Jnl
Lab Hem 2010; 32: 572-581.
28. Ulenkate H, Versluys C. Terugkoppeling naar aanvragers over aanvraaggedrag
m.b.v. het LIS Labosys. Ned Tijdschr Klin Chem Labgeneesk 2011, 36, 75.
29. Feedback software, ontwikkeld in het kader van SKMS projectnr. 4123039
(Feedback eerste lijn). www.feedbackrapportage.nl
39
LANDELIJKE EERSTELIJNS SAMENWERKINGSAFSPRAAK
(herziening 2012)
Eerste herziening van de Landelijke Eerstelijns Samenwerkingsafspraak ‘Rationeel
aanvragen van laboratoriumdiagnostiek’
Labots-Vogelesang SM, Ten Boekel E, Rutten WPF, Weel JFL, Guldemond FI, Hens
JJH, Klein Ikkink A, Souverijn JHM, Van Balen JAM, Van der Laan JR, Van
Duijnhoven JLP, Walma EP, Woutersen-Koch H.
Ten geleide
In 2006 verscheen de Landelijke Eerstelijns Samenwerkingsafspraak Rationeel
aanvragen van laboratoriumdiagnostiek. In de afgelopen vijf jaar zijn veel van de
richtlijnen (NHG-Standaarden maar ook andere landelijke richtlijnen) waarop de
hoofdstukken van de LESA zijn gebaseerd, herzien. Daarnaast neemt naast het
NHG, de NVKC en de SAN sinds 2009 ook de Nederlandse Vereniging voor
Medische Microbiologie (NVMM) deel aan de LESA. Op grond van deze
ontwikkelingen zijn veel hoofdstukken van de LESA aangepast en leek het ons zinvol
de LESA opnieuw te publiceren.
Een groot aantal hoofdstukken is aangepast op grond van een herziene versie van
de onderliggende richtlijn (atriumfibrilleren, coeliakie, diarree, diep veneuze
trombose,
hartfalen,
hemochromatose,
leveraandoeningen,
nieraandoeningen,
overgevoeligheid, schildklierfunctiestoornissen en subfertiliteit). Enkele hoofdstukken
zijn samengevoegd tot één nieuw hoofdstuk (reumatoïde artritis en jicht tot ‘artritis’
en hypertensie en cholesterol tot ‘cardiovasculair risicomanagement’). De naam van
het hoofdstuk mononucleosis is gewijzigd in ‘acute keelpijn’. Daarnaast zijn, op grond
van de deelname van de NVMM, alle hoofdstukken waarin microbiologisch
onderzoek een rol speelt opnieuw beoordeeld en waar nodig herzien (acute keelpijn,
diarree, leveraandoeningen, maagklachten, SOA, subfertiliteit en urineweginfecties).
Ook het probleemgeoriënteerd formulier is aangepast aan de herziene hoofdstukken
(zie bijlage probleemgeoriënteerd aanvraagformulier).
Deze herziene versie van de LESA zal worden uitgegeven als een makkelijk te
raadplegen boekje en als een zogenaamd ‘levend document’, toegankelijk via
40
internet. Op deze manier kunnen in de toekomst op eenvoudige wijze wijzigingen in
hoofdstukken worden aangebracht en nieuwe hoofdstukken worden toegevoegd,
bijvoorbeeld na het verschijnen van herzieningen of nieuwe NHG-Standaarden. Zo
kan de LESA continu actueel gehouden worden. Om deze reden is tevens besloten
om de nummering van de hoofdstukken en van de bijlagen te laten vervallen.
Ten slotte zijn er de afgelopen jaren nog enkele andere ontwikkelingen geweest die
van belang zijn voor de LESA. Tegenwoordig wordt door de meeste laboratoria ter
bepaling van de nierfunctie een geschatte creatinineklaring (eGFR) gegeven. In de
niet-herziene hoofdstukken wordt vaak nog geadviseerd het serumcreatininegehalte
te bepalen.
Daarnaast was in de vorige versie van de LESA Het Diagnostisch Kompas de
onderlegger voor veel hoofdstukken (met name wat betreft de referentiewaarden).
Het Diagnostisch Kompas is voortgezet in het Handboek medische laboratoriumdiagnostiek. In de nieuwe hoofdstukken wordt derhalve naar dit handboek verwezen.
Utrecht, januari 2012
41
Belangrijkste wijzigingen
De NVMM neemt sinds deze versie deel aan de ontwikkeling van deze LESA en om
die reden zijn de hoofdstukken waarin microbiologisch onderzoek een rol speelt
opnieuw beoordeeld en waar nodig, herzien.
Inleiding
De Landelijke Eerstelijns Samenwerkingsafspraak Rationeel aanvragen van
laboratoriumdiagnostiek is opgesteld door een werkgroep van het Nederlands
Huisartsen Genootschap (NHG), de Nederlandse Vereniging voor Klinische Chemie
(NVKC), de Nederlandse Vereniging voor Medische Microbiologie (NVMM) en de
Centra voor Medische Diagnostiek (SAN). Een Landelijke Eerstelijns Samenwerkingsafspraak (LESA) geeft richtlijnen voor de samenwerking tussen huisartsen
en andere beroepsgroepen die in de eerste lijn werkzaam zijn en houdt daarbij
rekening met de verschillen in taken en verantwoordelijkheden van de verschillende
beroepsgroepen.
Kenmerkend voor een LESA is dat de richtlijnen op een zodanige manier worden
gepresenteerd dat door de betrokken beroepsgroepen, in dit geval huisartsen,
klinisch chemici en medisch microbiologen op regionaal niveau werkafspraken over
de aanbevelingen kunnen worden gemaakt.
De doelstelling van de richtlijnen van de LESA Rationeel aanvragen van
laboratoriumdiagnostiek (verder te noemen ‘de LESA’) is het optimaal gebruikmaken
van laboratoriumdiagnostiek door de juiste diagnostiek bij de juiste indicatie te
bevorderen en onnodige diagnostiek of het aanvragen van diagnostiek op onjuiste
indicatie te voorkómen. In verschillende publicaties is aangetoond dat het
probleemgeoriënteerd aanvragen van laboratoriumdiagnostiek een eenvoudige
manier is om tot verandering in aanvraaggedrag te komen en het invoeren van
richtlijnen voor het laboratoriumonderzoek te bevorderen (1,2). Uit een onderzoek
naar het aanvraaggedrag van huisartsen betreffende negentien laboratoriumbepalingen, blijkt dat actieve betrokkenheid bij het maken van richtlijnen en meer dan
één jaar ervaring met probleemgeoriënteerd aanvragen van laboratoriumdiagnostiek
geassocieerd zijn met respectievelijk 27% en 41% minder aangevraagde bepalingen
(3).
42
De LESA is gebaseerd op wetenschappelijke gegevens en consensusafspraken in
de werkgroep. Bij de bespreking van de wetenschappelijke literatuur en de gemaakte
keuzes is de werkgroep uitgegaan van de NHG-Standaarden. Voor de onderwerpen
waarvoor geen NHG-Standaarden beschikbaar zijn, is uitgegaan van andere
algemeen geaccepteerde richtlijnen, zoals multidisciplinaire richtlijnen en algemene
richtlijnen in het vakgebied klinische chemie, laboratoriumgeneeskunde en
microbiologie. Voor onderwerpen waarvoor geen algemene richtlijnen voorhanden
waren, is literatuuronderzoek gedaan. Verder zijn de ervaringen met het
probleemgeoriënteerde aanvraagformulier en de huidige werkwijze van laboratoria
van belang geweest bij de totstandkoming van de LESA (4).
Inhoud en opbouw LESA
De frequentie van voorkomen en aanbevelingen voor laboratoriumonderzoek in de
richtlijnen vormden de basis voor de selectie van aandoeningen voor opname in de
LESA.
De hoofdstukken hebben een vaste opzet. Bij de achtergrondinformatie over de
aandoening (vooral epidemiologie en pathofysiologie) wordt bij voorkeur verwezen
naar goed toegankelijke publicaties.
In elk hoofdstuk wordt bij elke aanbevolen bepaling aangegeven wat de indicatie
voor het aanvragen van deze bepaling is en worden de achtergronden van de
bepaling en de referentiewaarden besproken. Indien mogelijk wordt aangegeven wat
de laboratoriumbepaling toevoegt aan de anamnese en het lichamelijk onderzoek. Bij
de achtergrondinformatie over de bepaling is zo veel mogelijk informatie over de
sensitiviteit en specificiteit van de bepaling vermeld, over de prevalentie en over de
positief en negatief voorspellende waarden.
Bij het opstellen van de aanbevelingen in de LESA is rekening gehouden met de
ruimtelijke beperkingen van een probleemgeoriënteerd aanvraagformulier op papier.
Het is te verwachten dat deze beperking in de nabije toekomst vervalt als de huisarts
de diagnostiek elektronisch kan aanvragen met een binnen het Huisarts Informatie
Systeem functionerende diagnostiekmodule (zoals bijvoorbeeld via ‘Zorgdomein’).
43
Samenwerking tussen huisartsen, klinisch chemici en microbiologen
Van huisartsen wordt verwacht dat zij laboratoriumdiagnostiek zo veel mogelijk
probleemgeoriënteerd aanvragen. Daarnaast is het voor een goede interpretatie van
de uitslagen van belang dat de huisarts de relevante klinische gegevens op het
aanvraagformulier vermeldt.
Huisartsen en klinisch chemici
Op grond van zijn expertise over onder andere testkarakteristieken en analytische en
biologische variatie van de bepaling, informeert de klinisch chemicus de huisarts over
de indicatie voor een test en de interpretatie van testuitslagen.
De klinisch chemicus dient uitslagen die sterk afwijkend en klinisch relevant zijn, zo
snel mogelijk te rapporteren aan de huisarts (zie verder). De huisarts is ervoor
verantwoordelijk dat bij een sterk afwijkende uitslag de juiste actie wordt
ondernomen.
De klinisch chemicus geeft de huisarts desgevraagd feedback over zijn
aanvraaggedrag, waarbij eventueel een vergelijking wordt gemaakt met het
aanvraaggedrag van andere artsen in de regio.
Samenwerking tussen klinisch chemici en huisartsen geeft ook de mogelijkheid tot
samenwerking op het punt van kwaliteitsbewaking van testapparatuur die in de
huisartsenpraktijk wordt gebruikt.
Huisartsen en microbiologen
Op grond van zijn kennis over de methoden om de aanwezigheid van microorganismen aan te tonen in de verschillende lichaamsvloeistoffen, informeert de
microbioloog de huisarts over de indicatie voor een onderzoek en over de
interpretatie van testuitslagen. Ook de wijze van verzamelen van het in te sturen
materiaal is een belangrijk onderdeel van afspraken. De microbioloog dient de
positieve testuitslagen, met name wanneer deze klinisch relevant zijn, zo snel
mogelijk te rapporteren aan de huisarts. De huisarts is ervoor verantwoordelijk dat bij
een positieve uitslag de juiste actie wordt ondernomen.
De microbioloog en de huisarts kunnen onderling afspraken maken over feedback op
het aanvraaggedrag van de huisarts, waarbij ook de mogelijkheid bestaat een
vergelijking te maken met het aanvraaggedrag van andere artsen in de regio. Op
44
grond van het kwaliteitsbeleid kan de microbioloog eveneens een rol vervullen in het
gebruik van testmateriaal bij onderzoek dat de huisarts in eigen beheer uitvoert.
Het microbiologisch onderzoek is als apart hoofdstuk in de LESA opgenomen.
Daarnaast wordt dit onderzoek ook besproken in de hoofdstukken waar dit van
toepassing
is
(diarree,
SOA,
subfertiliteit,
maagklachten,
acute
keelpijn,
leveraandoeningen en urineweginfecties). Voor een goede interpretatie van de
kweekuitslag moet de huisarts bij het aanvragen van de kweek de volgende
informatie vermelden: de herkomst van het materiaal (bijvoorbeeld catheterurine), het
klinisch beeld (bijvoorbeeld verblijf in het buitenland bij parasitologisch onderzoek
van feces) en antibioticagebruik (bijvoorbeeld bij een urineweginfectie).
Gesprekspunten voor de regio
In de LESA-werkgroep is overeenstemming bereikt over de aanbevelingen
betreffende de aandoeningen die in de inhoudelijke hoofdstukken worden besproken.
Geadviseerd wordt niet af te wijken van deze aanbevelingen. De werkgroep
adviseert om regionaal bijeenkomsten met het regionaal laboratorium te organiseren,
voor overleg en desgewenst het maken van afspraken over de volgende punten:
Het regionaal probleemgeoriënteerd aanvraagformulier
In veel regio’s wordt een aanvraagformulier gebruikt dat niet geheel overeenkomt
met het landelijke model. Tijdens de bijeenkomsten kunnen de volgende
onderwerpen aan de orde komen:
-
Zijn er bepalingen op het formulier die niet in de LESA worden geadviseerd
(bijvoorbeeld FSH en LH, vitamine D)?
-
Zijn er belangrijke aandoeningen/bepalingen weggelaten?
-
Wat betreft het microbiologisch onderzoek: is er voldoende ruimte op het
formulier om klinisch relevante informatie te geven of is er een apart
aanvraagformulier voor microbiologische bepalingen?
-
Is er ruimte om bepalingen aan te kruisen op indicaties die buiten de
genoemde aandoeningen vallen (bijvoorbeeld kinkhoestserologie)?
-
Is er sprake van (een apart formulier met) een alfabetische lijst? Kan deze
vervallen?
45
Bespreking van hoofdstukken uit de LESA
Bespreking van specifieke hoofdstukken geeft vaak aanleiding tot het maken van
aanvullende werkafspraken. Discussiepunten komen vaak voort uit de verschillen in
werkwijze in laboratoria. Sommige laboratoria geven bijvoorbeeld de gelegenheid bij
urineweginfecties een dipslide op te sturen en te laten beoordelen, waarbij tevens
resistentiebepaling plaatsvindt. Het bacteriologisch onderzoek bij diarree kan
plaatsvinden door middel van een kweek, maar sommige laboratoria gebruiken tests
op basis van DNA-onderzoek, die sneller een betrouwbaarder uitslag geven. Een
tripletest is dan niet meer noodzakelijk. Bij mononucleosis infectiosa heeft het de
voorkeur om, indien geïndiceerd, de specifieke antistoffen te bepalen en af te
spreken de test op heterofiele antistoffen niet meer te gebruiken.
Vervolgbepalingen bij een laag Hb
Door het vermelden van klinische gegevens kan de klinisch chemicus een advies
geven over een vervolgonderzoek, wanneer het Hb erg laag is. In dit verband kan
men ook specifieke afspraken maken over vervolgonderzoek naar hemoglobinopathieën.
Uitvoeren extra bepalingen
Laboratoria bewaren doorgaans het bloedmonster één week na het aanvragen van
de bepaling. Wanneer het noodzakelijk is, kunnen in tweede instantie extra (klinisch
chemische of serologische) bepalingen worden aangevraagd. Een ‘spijtmonster’ is
een serummonster, dat voor bepaalde doeleinden langer kan worden bewaard. Men
kan afspraken maken over de gebruikelijke termijn van bewaren (zodat achteraf extra
bepalingen kunnen worden uitgevoerd), na welke termijn huisartsen specifiek een
bepaling uit een ‘spijtmonster’ kunnen aanvragen en op welke indicaties dit kan
gebeuren. Microbiologische laboratoria bewaren sera doorgaans vele jaren.
Het doorgeven van sterk afwijkende laboratoriumuitslagen
De werkgroep adviseert regionaal werkafspraken te maken over het doorgeven van
sterk afwijkende uitslagen. Door onvoldoende onderbouwing met onderzoek uit de
eerste lijn, is het niet mogelijk hiervoor een landelijke lijst beschikbaar te stellen.
46
Kwaliteitsbewaking van testapparatuur buiten de laboratoriumsetting.
Er zijn steeds meer apparaten en tests beschikbaar om buiten het laboratorium
onderzoek uit te voeren naast of in de buurt van het bed van de patiënt (Point of care
testing (POCT): ook wel bedside testing) bijvoorbeeld op bloedglucose, CRP, INR, Ddimeer, SOA (hiv). Het verdient aanbeveling de eigen apparaten geregeld te laten
testen. Dit kan onder andere met behulp van het laboratorium waarmee de huisarts
samenwerkt. Hiervoor zijn (aanvullende) afspraken en procedures te maken.
Diagnostisch toetsoverleg (DTO)
Naast het gebruik van een probleemgeoriënteerd aanvraagformulier kan de invoering
van de aanbevelingen in de LESA worden bevorderd door het gebruik van
feedbackcijfers in een Diagnostisch Toetsoverleg (DTO).
DTO is een met het Farmacotherapeutisch Toets Overleg vergelijkbaar overleg
tussen het regionale (ziekenhuis)laboratorium en de huisartsen in de regio. De
laatste jaren wordt in steeds meer regio’s DTO georganiseerd voor huisartsengroepen (met name door de huisartsenlaboratoria voor huisartsengroepen
(HAGRO’s). In een dergelijk overleg kunnen zowel nascholing en feedback, als het
maken van regionale afspraken aan de orde komen (5). De feedback van het
regionale laboratorium is geschikt voor het vergelijken van het aanvraaggedrag van
huisartsen met dat van collega’s en om de discussie over verschillen in het
aanvraaggedrag op gang te brengen. Een en ander kan een reden zijn tot het
aanpassen van het aanvraaggedrag.
De Federatie voor Medisch Coördinerende Centra (FMCC) kan een coördinerende
rol spelen bij het aanbieden van regionale ondersteuning.
Referenties
1. Geldrop WJ van, Lucassen PLBJ, Smithuis LOMJ. Een probleemgeoriënteerd
aanvraagformulier voor laboratoriumonderzoek. Effecten op het aanvraaggedrag
van huisartsen. Huisarts Wet 1992;35:192-196.
2. Smithuis LOMJ, Geldrop WJ van, Lucassen PLBJ. Beperking van het
laboratoriumonderzoek door een probleemgeoriënteerd aanvraagformulier. Een
partiële implementatie van NHG-Standaarden. Huisarts Wet 1994;37:464-466.
3. Verstappen WH, Riet G ter, Dubois WI, Winkens R, Grol RP, Weijden T van der.
Variation in test ordering behaviour of GPs: professional or context-related
factors? Fam Pract 2004;21:387-395.
47
4. Anonymus. Wetenschappelijke verantwoording van het landelijk model van een
probleemgeoriënteerd aanvraagformulier voor laboratoriumonderzoek door
huisartsen. Ned Tijdschr Klin Chem 2000;25:1-71.
5. Verstappen WH, Weijden T van der, Sijbrandij J, Smeele I, Hermsen J, Grimshaw
J, Grol RPTM. Diagnostisch toetsoverleg (DTO) vermindert overbodig gebruik
aanvullende diagnostiek door huisartsen. Huisarts Wet 2004;47:127-132.
Werkgroep LESA Rationeel Aanvragen van Laboratoriumdiagnostiek (juli
2013):
De heer dr. E. ten Boekel, klinisch chemicus, namens de NVKC (voorzitter)
Mevrouw J.A.M. van Balen, huisarts, namens NHG
De heer dr. J.L.P. van Duijnhoven, klinisch chemicus, namens de NVKC
De heer dr. B.D. Frijling, huisarts
De heer dr. J.J.H. Hens, klinisch chemicus, namens de NVKC
De heer J.R. van der Laan, huisarts
De heer dr. P.L.B.J. Lucassen, huisarts
Mevrouw dr. K. Mohrmann, klinisch chemicus, namens de SAN
De heer W.P.F. Rutten, klinisch chemicus, namens de SAN
De heer dr. W.H.J.M. Verstappen, huisarts
Mevrouw A.C. de Vries-Moeselaar, wetenschappelijk medewerker NHG
De heer dr. J.F.L. Weel, arts-microbioloog, namens de NVMM
48
Probleemgeoriënteerd aanvraagformulier
❑M
Arts:
❑V
Naam + voorl. ...............................................
Geb. dat.
Kopie rapport
...............................................
Adres
...............................................
Pc + plaats
...............................................
Particulier ❑
[logo NHG/NVKC/SAN/NVMM; adm gegevens
voor
Laboratoriumonderzoek door huisartsen
toevoegen]
Acuut coronair syndroom
Troponine
CK-MB
Acute keelpijn
Verm. mononucleosis: EBV-antilich. (klachten >7 dg)
Opsporing immuunstoornis: Leukocyten, diff.
Algemeen bloedonderzoek
Hb BSE
Glucose (nn) TSH (indien afwijkend vrij T4)
Op indicatie:
eGFR* (vooral ouderen)
ALAT (vermoeden leveraandoening)
Anemie
Chronische ziekte (ACD) ja
nee
Hb, MCV (vervolgdiagnostiek afh. van uitslag):
Micro- en normocytair: Ferritine
Macrocytair:
LDH, reticuloc, tromboc, leukoc, vit.B12, foliumz.
Op indicatie: vervolgonderzoek op Hb-pathie
Controle: Hb
Angina pectoris
Bij vermoeden van anemie of hyperthyreoïdie
Hb
TSH (indien afwijkend vrij T4)
Artritis
Diagnostiek reumatoïde artritis
Reumafactor anti-CCP (beide beperkte waarde)
Diagnostiek jicht: urinezuur
Bij aanvang onderhoudsbehandeling en ter controle:
eGFR* ,, urinezuur
kalium (bij aanvang diuretica)
LDL-chol. (chol.verlager, enkele wkn-3mnd na start)
Driemaand.controle: geen of orale med./1dd insuline:
Glucose (voorkeur nuchter, evt. 2 uur postprandiaal)
2-4 dd insuline: 4 pnts dagcurve, HbA1c (1x/3-6mnd)
Jaarlijkse controle
HbA1c, eGFR*
tot. chol., HDL-chol., LDL-chol., triglyc. (nuchter)
Albumine of albumine/creatinine-ratio (urine)
kalium (diuretica, RAS-remmer)
Diarree
Acuut, bij ernstig ziekzijn: Feceskweek (Salmonella,
Shigella, Campylobact., clostridium difficile)
Indien > 10 dagen: parasitologisch onderzoek
Vermeld klinische symptomen als ziekteduur, koorts
(met/zonder pieken), bloedbijmenging en verblijf
buitenland (waar/wanneer/terug sinds?), recent
antibioticagebruik (welke) en verblijf in instelling.
Monster: zie instructie lab.
Diepe veneuze trombose
D-dimeer
Geneesmiddelentherapie
Lithium, TSH, eGFR* (min. 2x/jr; 12 uur na inname)
Digoxine, K (vermoed. van intoxicatie; voor gift)
Hartfalen
Zfds + nr. ❑
...............................................
keur + inst.
...............................................
Inhalatieallergeenscreeningstest:
allergeenspecifiek IgE
Indien positief uitsplitsen naar onderst. allergenen
huisstofmijt kattenepitheel hondenepitheel
graspollen boompollen berkenpollen
kruidpollen schimmels
Prostaat- en mictieklachten
nitriettest, indien negatief dipslide (uitsluiten
urineweginfectie)
eGFR* PSA (beperkte waarde)
Psychogeriatrie
BSE, Hb, glucose, eGFR*, TSH
Op indicatie:
K, Na (diureticagebruik)
J-GT (vermoeden leveraandoening)
vit. B1, B6, B12, foliumz. (verm. deficiënte voeding)
Schildklierfunctiestoornissen
Diagnostiek: TSH, indien afwijkend vrij T4
Thyreoïditis: BSE, leukoc., vrij-T4
Ziekte v. Graves: anti-TSH-receptor-antistoffen
Controle therapie hypo-/hyperthyreoïdie (combither.):
TSH, vrij-T4
Diagnostiek: (NT-pro-)BNP
Opsporing onderliggende aand en co-morb.:
Hb/Ht, TSH, gluc., CRP, leukoc.,diff.,ALAT, J-GT,
lipidenprofiel
Bij start/controle therapie:
Na, K, eGFR* (bij start, -> 2x/jr)
eGFR*(2 wk na start RAS remmer)
K (2 wk na start diureticu of/ dosering spironol.)
Soa
Chlamydia (cervix-/urethra-uitstrijk: met klachten)
Chlamydia (urine: 1e straals-urine, zonder
klachten)
Gonorroe (uitstrijk cervix en urethra bij )
Gonorroe (1e-straals urine bij )
HIV
Hepatitis B
Lues (controle)
Lues (diagnostiek)
Trichomonas (fluor) Herpes (uitstrijk)
Hemochromatose
Subfertiliteit
Diagnostiek: Transferrinesaturatie, ferritine
Vervolgdiagnostiek bij verhoogde waarden:
ALAT, BSE(CRP), Hb, glucose.
Sperma-onderzoek: zie instr. lab.
CAT: chlamydia IgG-antistoffen
Atriumfibrilleren
Opsporing onderlig. aand.: Hb, TSH, glucose (nn)
Vermoeden hartfalen: (NT-pro)BNP
Controle digoxinegebruik (bij aanvang en jaarlijks):
kalium, eGFR*
Verhoogde bloedingsneiging
Diagnostiek: APTT, PT, trombocyten
Contr. therapie orale anticoagulantia: PT-INR
Cardiovasculair risicomanagement
Risico-inventarisatie :
Preventieconsult: Tot.chol/ HDL-chol.-ratio, , gluc.
CVRM: Tot.chol/ HDL-chol.-ratio, , gluc., eGFR*
Bij aanvang/aanpassing medicamenteuze behandeling
Chol. verlager: LDL-cholesterol ( nuchter, na 3 mnd)
ACE-remmer/ARB, diuretica:
eGFR*, K (herh> 2wkn)
Controle behandeling (jaarlijks)
Hypertensie:
eGFR*, albumine of albumine/creatinine-ratio (urine)
Chol. verlager: LDL-chol. ( nuchter )
ACE-remmer/ARB, diuretica, nierfunctie verlaagd:
eGFR*, K
Risico-inventarisatie DM-2 (1x per 3 jaar)
glucose (bij voorkeur nuchter)
Vermoeden familiaire hyperlipidemie (risicoscore > 6):
Totaal chol., HDL-chol., LDL-chol., triglyceriden
glucose, TSH, ALAT, gammaGT
Coeliakie tTGA
Delier
Opsporing onderliggende aandoening:
BSE/CRP, Hb, gluc., eGFR*, TSH nitriet (urine)
Op indicatie:
Na, K (na braken, diarree, bij diureticagebruik)
J-GT (bij vermoeden leveraandoening)
Ca (bij bedlegerigheid, vermoeden metastasen)
Diabetes mellitus type 2
Diagnostiek en opsporing (1x/3jr)
Glucose ( voorkeur nuchter,evt. 2 uur postprandiaal)
Risico-inventarisatie (nuchter)
HbA1c, tot. chol., HDL-chol., LDL-chol., triglyc.,
eGFR*
Bij aanvang med. behandeling risicofactoren HVZ
TIA
Leveraandoeningen
Diagnostiek leveraandoening: ALAT
Diagnostiek virushepatitis: Hep.A (IgM-anti HAV)
Hepatitis B (HBsAg)
Hepatitis C (anti-HCV)
Glucose (nn)
Chol./HDL-chol.-ratio
Bezinking (bij amaurosis fugax)
Urineweginfecties
Diagnostiek
nitriet (indien negatief: sediment of dipslide)
dipslide/kweek met resistentiebepaling
(gecompl.UWI; persist.kl.bij ongecompl.UWI)
Controle
nitriet dipslide/kweek (zwangeren, kinderen)
Maagklachten
Diagnostiek H. pylori-infectie
Ureumademtest
Serologie H. Pylori
Fecestest
Controle behandeling
Ureumademtest (4 wk na behandeling)
Fecestest ( 4 wk na behandeling)
Serologie (6 mnd na behandeling)
Zwangerschap
Microbiologisch onderzoek
Kweek, banaal
Materiaalsoort…………………………………..
Herkomst mat.:………………………………….
Ziekteverschijnselen:……………………………
Kweek specifiek op……..……………………
Materiaalsoort:………………....……………….
Afnameplaats:…..……………………………….
Ziekteverschijnsel:………………………………
Neonatale icterus
Bilirubine (totaal)
Nieraandoeningen
Diagnostiek nieraandoening:
eGFR*
erythrocyten in urine (teststrook, sediment)
Vervolgdiagnostiek eGFR en albuminurie:
creatinine, eGFR, lipidenspectrum, glucose
Vervolgdiagnostiek metabole complicaties
Hb, kalium, calcium, fosfaat, serumalbumine, PTH
Prenatale screening:
ABO-, RhD bloedgroep, irr. antistoffen
HBsAg, lues, HIV
Hb
(vervolgdiagnostiek afh. van uitslag):
pariteit: ……… à terme datum: ..........
Op indicatie:
Rubella (indien niet gevaccineerd)
Bij bestaande of anamnestische schildklierfunctiest.:
TSH, vrij-T4
TSH-R-antistoftiter (bij hyperthyreoïdie)
Aanvullende informatie…………………………
Overige onderzoeken:
……………………………………………
……………………………………………
……………………………………………
Relevante klinische gegevens en opmerkingen:
…………………….……………………………
………………………….………………………
…………………………………………………………
…………………………………………………………
…………………………………………………………
*eGFR: creatinineklaring
Informatie over bloedonderzoek: www.kiesbeter.nl
Overgevoeligheid
49
The power
of partnership
From the first step of listening to delivering on your
long term goals, we are always there for you.
We don’t just talk about the power of partnership,
we take the time to understand your challenges
and develop solutions which revolve around
your priorities and support your individual
objectives. By working in partnership with your
team we can ensure your laboratory achieves
success in a changing environment.
www.beckmancoulter.com
50
FEEDBACK AANVRAAGGEDRAG EN DIAGNOSTISCH
TOETSOVERLEG
J Trietsch, huisarts - onderzoeker
Implementatie van nieuwe inzichten uit medisch onderzoek gaat vaak moeizaam en
met veel vertraging. De enorme hoeveelheid medische publicaties die jaarlijks
verschijnt is voor artsen en andere gezondheidszorgwerkers niet te overzien.
Onderzoekers, verzekeraars en beleidsmakers in de gezondheidszorg proberen de
overdracht van de nieuwste kennis vanuit de wetenschap naar het veld te faciliteren
(1,2). Het Nederlands Huisartsengenootschap (NHG) startte daarom enkele
decennia geleden al met het ontwikkelen en onderhouden van de standaarden voor
huisartsen, hiAerin is samengevat welk beleid het beste gevoerd kan worden bij
specifieke aandoeningen. Ondanks dat het beleid volgens de NHG standaarden voor
70% gevolgd wordt blijft er sprake van een forse inter-dokter variatie.
30 Jaar geleden zijn huisartsen in Nederland begonnen met bijeenkomsten waarin
een groep van artsen samen met een lokale apotheker afspraken maken over
farmacotherapeutisch beleid in de groep op basis van onder andere de NHG
standaarden. Deze overlegstructuur, het FTO, is inmiddels gemeengoed onder de
huisartsen met een deelname van ruim 95% van de praktijkhoudende huisartsen. De
effecten van deze bijeenkomsten op de kwaliteit van voorschrijven is wisselend en
lijkt hand in hand te gaan met de kwaliteit van de bijeenkomst (3).
Onderzoek van de Universiteit Maastricht laat zien dat een soortgelijk overleg gericht
op diagnostisch aanvraaggedrag van huisartsen, het diagnostisch toetsoverleg
(DTO), een positief effect kan hebben op de kwaliteit van het aanvraaggedrag (4,5).
In 2006 is begonnen met het vervolg van deze studie op te zetten met als doel om te
onderzoeken of brede implementatie van het DTO in de bestaande FTO structuur
haalbaar zou zijn en hoe het effect na overdracht aan het veld zou zijn (6). In deze
cluster-RCT zijn 21 FTO groepen geïncludeerd bestaand uit 206 huisartsen, 39
apothekers en 12 deelnemende laboratoria/ ziekenhuizen in het zuiden van
Nederland. Huisartsen kozen per groep na randomisatie 3 uit 5 klinische
onderwerpen. Zij kregen vergelijkende feedback over diagnostiek en farmacotherapie
met betrekking tot het te bespreken onderwerp in gepaarde bijeenkomsten. Volgens
een gestructureerde agenda werden achtereenvolgens de cijfers en onderlinge
51
verschillen besproken, veranderdoelen vastgesteld en de weerstanden om te
veranderen. Uiteindelijk resulteerde dit in een implementatieplan van de gestelde
doelen voor de eigen praktijk. De eerste bijeenkomst ontvingen de artsen feedback
over hun diagnostische aanvraaggedrag en bespraken dit met als doel te komen tot
afspraken voor de toekomst. Deze bijeenkomst werd begeleid door een deskundige,
vaak een klinisch chemicus. De tweede bijeenkomst ontvingen de artsen feedback
over hun voorschrijfgedrag over hetzelfde onderwerp, begeleidt door de lokale
apotheker. Hierna volgden nog 2 cycli van ieder 2 bijeenkomsten over de volgende 2
onderwerpen. Door de structuur van de bijeenkomsten neemt de kwaliteit en
efficiëntie van de bijeenkomst toe. Doordat de deelnemers afspraken maken over
toekomstig beleid en hieraan een implementatieplan koppelen is de verwachting dat
de inter-dokter variatie zal afnemen binnen de groepen.
De resultaten van de procesevaluatie en de eerste resultaten van het effect op het
diagnostisch aanvraag gedrag zullen besproken worden in de lezing.
Referenties
1. Bero LA, Grilli R, Grimshaw JM, Harvey E, Oxman AD, Thomson MA. Closing the
gap between research and practice: an overview of systematic reviews of
interventions to promote the implementation of research findings. The Cochrane
Effective Practice and Organization of Care Review Group. BMJ 1998;317:465468.
2. Berwick DM. Disseminating innovations in health care. JAMA 2003;289:19691975.
3. Arnold SR, Straus SE. Interventions to improve antibiotic prescribing practices in
ambulatory care. Cochrane Database Syst Rev 2005;4:CD003539.
4. Verstappen WHJM, Weijden T van der, Dubois WI, Smeele I, Hermsen J, Tan
FES, Grol RPTM. Improving test ordering in primary care: the added value of a
small-group quality improvement strategy compared with classic feedback only.
Ann Fam Med 2004;2:569-575.
5. Verstappen WHJM, Weijden T van der, Sijbrandij J, Smeele I, Hermsen J,
Grimshaw J, Grol RPTM. Effect of a practice-based strategy on test ordering
performance of primary care physicians: a randomized trial. JAMA
2003;289:2407-2412.
6. Trietsch J, Weijden T, Verstappen W van der, Janknegt R, Muijrers P, Winkens
R, Steenkiste B van, Grol R, Metsemakers J. A cluster randomized controlled trial
aimed at implementation of local quality improvement collaboratives to improve
prescribing and test ordering performance of general practitioners: Study
Protocol. Implement Sci. 2009;4:6.
52
Implementation Science
BioMed Central
Open Access
Study protocol
A cluster randomized controlled trial aimed at implementation of
local quality improvement collaboratives to improve prescribing
and test ordering performance of general practitioners: Study
Protocol
Jasper Trietsch*1, Trudy van der Weijden1, Wim Verstappen2, Rob Janknegt3,4,
Paul Muijrers3, Ron Winkens1,5, Ben van Steenkiste1, Richard Grol1,6 and
Job Metsemakers1
Address: 1Maastricht University, Dept. of General Practice, School for Public Health and Primary Care (CAPRHI), Maastricht, The Netherlands,
2GP out-of-hours centre, Den Bosch/Eindhoven, the Netherlands, 3OWM Centrale Zorgverzekeraars group, Zorgverzekeraar UA, Tilburg, The
Netherlands, 4Maasland Hospital, Sittard, The Netherlands, 5Diagnostic Centre and department of Integrated Care, Maastricht University Medical
Centre, Maastricht, The Netherlands and 6Radboud University Nijmegen Medical Centre, Centre for Quality of Care Research, Nijmegen, The
Netherlands
Email: Jasper Trietsch* - [email protected]; Trudy van der Weijden - [email protected];
Wim Verstappen - [email protected]; Rob Janknegt - [email protected]; Paul Muijrers - [email protected];
Ron Winkens - [email protected]; Ben van Steenkiste - [email protected]; Richard Grol - [email protected];
Job Metsemakers - [email protected]
* Corresponding author
Published: 17 February 2009
Implementation Science 2009, 4:6
doi:10.1186/1748-5908-4-6
Received: 1 October 2008
Accepted: 17 February 2009
This article is available from: http://www.implementationscience.com/content/4/1/6
© 2009 Trietsch et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
Background: The use of guidelines in general practice is not optimal. Although evidence-based methods to improve
guideline adherence are available, variation in physician adherence to general practice guidelines remains relatively high.
The objective for this study is to transfer a quality improvement strategy based on audit, feedback, educational materials,
and peer group discussion moderated by local opinion leaders to the field. The research questions are: is the multifaceted
strategy implemented on a large scale as planned?; what is the effect on general practitioners' (GPs) test ordering and
prescribing behaviour?; and what are the costs of implementing the strategy?
Methods: In order to evaluate the effects, costs and feasibility of this new strategy we plan a multi-centre cluster
randomized controlled trial (RCT) with a balanced incomplete block design. Local GP groups in the south of the
Netherlands already taking part in pharmacotherapeutic audit meeting groups, will be recruited by regional health
officers. Approximately 50 groups of GPs will be randomly allocated to two arms. These GPs will be offered two different
balanced sets of clinical topics. Each GP within a group will receive comparative feedback on test ordering and prescribing
performance. The feedback will be discussed in the group and working agreements will be created after discussion of the
guidelines and barriers to change. The data for the feedback will be collected from existing and newly formed databases,
both at baseline and after one year.
Discussion: We are not aware of published studies on successes and failures of attempts to transfer to the stakeholders
in the field a multifaceted strategy aimed at GPs' test ordering and prescribing behaviour. This pragmatic study will focus
on compatibility with existing infrastructure, while permitting a certain degree of adaptation to local needs and routines.
Trial registration: Nederlands Trial Register ISRCTN40008171
Page 1 of 14
(page number not for citation purposes)
53
Background
With the ever-growing volume of evidence from medical
research, it has become impossible for physicians to
remain fully up to date. Reviews and guidelines therefore
summarize large quantities of information, making it
more easily available to field workers. In the Netherlands,
general practitioners (GPs) now have access to more than
80 evidence-based medical guidelines developed by the
Dutch College of General Practitioners (NHG). Although
general adherence to these guidelines is approximately
70%, the inter-physician variation is large, and adherence
to certain aspects of these guidelines proves to be difficult
[1-3]. Although there may be sensible reasons to deviate
from guidelines, such as multi-morbidity in a patient, a
physician's level of uncertainty tolerance and patients'
preferences, there seems to be room for improvement. The
inter-physician variation can be regarded as underdiagnosing or undertreating one group of people and at the
same time overdiagnosing and overtreating another
group, both leading to inappropriate care [4]. There is
considerable inter-physician variation in general practice
with regard to test ordering and prescribing [5,6].
Many studies have tried to find evidence for effective
implementation strategies to improve quality of care. A
multifaceted clustered RCT by Verstappen et al. aimed at
optimizing GPs' test ordering behaviour by means of local
quality improvement collaboratives (LQICs), found a
decrease of 8 to 12% in test volumes over a period of six
months [7]. This strategy was tested using six topics for
continuing medical education (CME). Other studies have
tested several implementation strategies to improve test
ordering and prescribing behaviour. Passive dissemination of guidelines or recommendations does not seem to
influence test ordering behaviour. Audit and feedback
have often been used and showed mostly a modest effect
in terms of influencing test ordering or prescribing. The
effect of audit and feedback on adherence to desired practice ranged from -10% to +68% (median +16%) [8-12]. In
other studies, the introduction of a problem-based test
ordering form proved to be a promising tool to improve
test ordering [7,13-18]. Similar effects on volumes of tests
ordered as those in the Verstappen study have been found
for more or less similar multifaceted implementation
strategies [19,20]. Small group peer review using direct
individual feedback seemed to reduce inappropriate prescribing [12,21,22]. Lagerlov found a 6 to 13% improvement in adherence to guidelines for the prescription of
anti-asthmatic drugs and antibiotics for urinary tract
infections in an RCT using reflection on guidelines and
prescription feedback in small groups [23].
The Cochrane Effective Practice and Organization of Care
group (EPOC) systematically reviews studies on implementation strategies to improve quality of care. Their
http://www.implementationscience.com/content/4/1/6
work has generated the general insight that multifaceted
strategies are usually more effective than single interventions [12,24], although this was not entirely confirmed by
an NHS HTA review by Grimshaw et al. [16]. The prevailing insight is that the effect of an intervention is larger
when tailored strategies are used and when barriers to and
facilitators of change are addressed.
Grol has identified in his model of effective implementation six stages in quality of care improvement[25]. In the
first stage, new research findings, new guidelines, experienced weaknesses, or best practices create an opportunity
for quality improvement. In the second stage, after the initial implementation process has been planned, targets for
improvement or change are set. Prior to the actual implementation, the performance, target group, and setting are
analysed. In the fourth stage, the strategies that are to be
used are identified and tested. The implementation is
developed, tested, and executed. Finally, the implementation is evaluated and adapted, if necessary [25]. The
present study will deal with the actual sustainable transfer
of a successful implementation strategy to the field. We
are not aware of published studies testing this process, or
whether effects are sustainable when transferred to the
field. Nor are we aware of published studies on the implementation of a large-scale strategy aimed at influencing
both the test ordering and prescribing behaviour of GPs
simultaneously, using peer review and social influencing
in primary care collaboratives.
In the Netherlands, existing networks of pharmacotherapeutic audit meetings (PTAM) can be used to disseminate
and implement guidelines on test ordering and prescribing. The goal of setting up these meetings by primary care
providers was to improve the quality of their prescribing
behaviour [26]. The local groups usually consist of six to
ten GPs with affiliated community pharmacists [27]. During the meetings, they discuss the choice of drugs in the
context of a specific illness or disease. In recent decades,
this form of CME has gained widespread acceptance
amongst GPs and policymakers in the Netherlands. However, these sessions tend to offer little or no room for discussions on test ordering. Because no other system of
regular meetings exists, the possible underuse, overuse,
and misuse of diagnostic services is not discussed by primary care providers on a regular basis.
The Dutch Institute for the Proper Use of Medicine (DGV)
supports and initiates local or regional implementation of
quality improvement projects on the use of drugs and supports local PTAM groups by supplying them with information and educational materials [28]. Performance levels of
PTAM groups are assessed once a year and rated on the
basis of four levels, level one being the poorest level of
performance and level four the highest. We will use this
Page 2 of 14
54
division into levels as a parameter for pre-randomization
stratification.
2. To determine the critical conditions for effective nationwide implementation.
Participation in PTAM groups by GPs is facilitated by
national and regional support organizations for primary
care, as well as by the government and through incentives
by insurance companies. Attendance at PTAM meetings is
rewarded by accreditation. Currently, approximately 50%
of the group meetings reach the desired level of performance described by policymakers [27]. To reach this level,
groups must at least use feedback on prescribing, create
working agreements, discuss barriers to change, and evaluate working agreements. Most groups are stable and
remain together for 10 years or more, with members
mostly being replaced gradually [27]. Because of the
nature and stability of these groups, they provide an excellent and safe environment for participants to discuss their
own behaviour and barriers to change. We expect this
existing system of PTAM groups will ensure sustainability
of the implementation itself. Therefore, we plan to use
these groups in a large pragmatic trial on the implementation of guidelines, using the strategy previously tested by
Verstappen et al. [7]. However, we will expand the strategy, using social interaction and external influencing as
key approaches for establishing behavioural change, to
both test ordering and drug prescribing. In our view, the
groups will no longer function merely as a PTAM group,
but rather begin acting as LQICs. This trial is expected to
show whether the effects found in less pragmatic trials can
be confirmed. Aiming at both test ordering and drug prescribing, our combination strategy could lead to an even
larger effect because of synergy. We will also evaluate the
costs of implementing the strategy on a large scale.
3. To improve the level of group performance in the participating groups.
4. To reduce undesirable physician variation in test ordering and prescribing; and to reduce underuse or overuse of
specific tests and drugs.
5. To examine the costs of large-scale implementation of
this strategy, and thus to be able to predict future costs for
expansion and maintenance of the strategy.
Research questions
Process
1. Was the strategy implemented as planned?
2. What were the barriers to and facilitators of the implementation of the strategy?
3. Has the level of group performance been improved in
the participating groups?
Effect
1. Do the volumes of tests ordered and drugs prescribed
change in the preferred direction, as described in the
working agreements of the LQICs, compared to baseline?
2. What is the effect of this strategy on GPs' test ordering
and prescribing behaviour in terms of interphysician variation and total volumes of tests and prescriptions with
respect to specific clinical topics, compared to that among
GPs exposed to the same strategy but for other topics?
Objectives and research questions
Hypotheses
We expect that the transfer of the strategy of LQICs to
stakeholders in the field will be feasible. We hereby hope
to create a solid basis for continuation after the end of the
study.
3. Is any gain in the level of group performance predictive
of the effect achieved?
We also expect that large-scale implementation, giving
attention to both test ordering and prescribing behaviour,
will lead to similar changes in performance as those found
on test ordering in the trial by Verstappen et al. [7].
Methods
Successful implementation will be positively related to
the level of group performance of the groups included, in
terms of level of attendance, number of meetings, drawing
up working agreements, discussing barriers to change, and
evaluating working agreements.
Objectives
1. To implement the LQIC strategy in the south of the
Netherlands, stimulating the relevant parties in the field
to take the lead.
Cost
What are the costs of implementing the strategy?
Design and ethics
This multi-centre study will use a balanced incomplete
block design, consisting of two arms (Figure 1). LQICs
will be allocated at random to one of these two arms. All
LQICs allocated to arm A will receive the intervention
with respect to the clinical topics associated with arm A.
All LQICs allocated to arm B will receive the same intervention, but with respect to the topics associated with arm
B (table 1). Each arm will have five different CME topics
to choose from. Each LQIC will choose three different
topics for their discussions, and serve as a control for the
other arm. The GPs will not be aware of the topics they are
serving as controls for, to avoid the Hawthorne effect [29].
Page 3 of 14
55
The Maastricht Medical Research Ethics Committee has
approved this study. All participating GPs will be asked to
sign a written informed consent form.
Population
LQIC groups will be recruited by regional medical coordinators, which are regional health officers or managers
often employed by regional hospitals or primary care laboratories. We have identified 24 organizations offering
diagnostic facilities in the south of the Netherlands. All
organizations will be visited by the researcher and asked
to cooperate. Each medical coordinator then will be asked
to recruit two to four LQIC groups. They will only be
included when all group members consent to participate.
The area from which groups can be recruited will be
restricted to the three southern provinces of the Netherlands (Limburg, Noord-Brabant, and Zeeland) because
these are covered by the insurance companies who provide data for the pharmaceutical database at Maastricht
University (UM). A representative with special expertise in
and knowledge of diagnostic testing, recruited by the
medical coordinator, will attend each LQIC meeting. This
representative will receive copies of the feedback forms of
all GPs in a LQIC, to enable him or her to prepare the sessions. The representative will act as a moderator during
the sessions devoted to diagnostics, after having been
trained to do so (see under 'training'). The medical coordinator will finally also liaise between their diagnostic
centre and the research team. Other stakeholders in our
strategy include community pharmacists, UM, the DGV,
insurance companies, PTAM groups, and individual GPs.
Community pharmacists play a major role in PTAMs in
the Netherlands, providing expertise and sometimes feedback on prescriptions to the participating GPs. Our intervention will leave the role of the pharmacists more or less
unchanged. They provide easily accessible knowledge for
GPs, thus breaking down barriers which might be inherent in distance support such as academic detailing. Like
the medical coordinator, a pharmacist will function as a
moderator in the LQIC. All community pharmacists will
receive training prior to the first session, as described
above. The pharmacists will receive copies of the feedback
forms of all participating GPs in a group, to enable them
to prepare the sessions.
The initiator of this trial is the Department of General
Practice of Maastricht University. The design and maintenance of the database on diagnostics and the data gathering process are coordinated by the first author. The
Maastricht University Centre for Information and Data
Management (MEMIC) will host the diagnostics database,
as they already do for the prescriptions database.
Randomization
LQIC groups will be randomized as such (cluster randomisation). The intervention is aimed at these groups.
Pre-randomization stratification will be performed on
group size and level of group performance using a pre-ran-
Table 1: Modules and distribution over the research arms.
Modules
Arm A
Arm B
Topic
Examples of tests
Examples of drugs
Hypercholesterolaemia
LDL
Statines
Anaemia
haemoglobin
ferro medication
Rheumatic complaints
Waaler-Rose
NSAIDs
Urinary tract infections
Urinary cultures
Antibiotics
Prostate complaints
PSA
D-blockers
Type 2 Diabetes Mellitus
HbA1c
Metformin
Dyspepsia
gastroscopy
proton-pump inhibitors
Chlamydia infections
chlamydia cultures
Antibiotics
Thyroid problems
TSH
Levothyroxine
Perimenopauzal complaints
FSH
Estradiol
For a complete list of all tests and drugs for the modules [See additional file 2]
Page 4 of 14
56
50 groups
stratification
R
Baseline
Intervention
Follow-up
Arm A
Arm B
Baseline
measurements
on all topics
Baseline
measurements
on all topics
Intervention on
topics from
arm A, no
intervention on
topics from
arm B
Intervention on
topics from
arm B, no
intervention on
topics from
arm A
Follow-up on
all topics
Follow-up on
all topics
Figure 1 of randomization and intervention
Flowchart
Flowchart of randomization and intervention.
Page 5 of 14
57
domization questionnaire [See additional file 1] prior to
the intervention. The levels of group performance are as
determined by DGV [28]. This level is a known confounder for an effective intervention on medical education among groups of GPs [30,31]. After stratification, all
groups within a stratum will be randomly allocated to
either arm A or arm B (Figure 1).
Sample size
A sample size calculation is not really possible beforehand, because it is not yet known what working agreements will be created and with respect to what tests or
drugs. The specific targets, incorporated in working agreements will probably be based on extreme overuse or
underuse of certain tests or drugs by some or all group
members. It is possible, for instance, that the group will
decide to eliminate a particular obsolete test or drug or
create a working agreement to decrease or increase the
mean volume of tests ordered or drugs prescribed by 20%,
from 35% to 55%.
The sample size calculation used in this trial is as follows:
to detect an improvement of 20% in a certain target
between groups, assuming an ICC of 0.10 [5], an alpha of
0.05 and a beta of 0.1 and a mean group size of seven, 44
LQICs would be needed. Anticipating a dropout of 10%,
we would need to recruit 50 groups. A population this
large would account for approximately 900,000 registered
patients.
Intervention
Several theories have been postulated on how change in
healthcare can be accomplished, and how effective change
strategies can work in implementation of innovation. In
cognitive theories, professional behaviour is considered
to result from rational processes and experiences from earlier caseloads. In social interaction theories, change of
professional behaviour is thought to be strongly mediated
by peers in a group, the strength of inter-individual ties
within groups, the existence of opinion leaders, and how
much the desired behaviour is consistent with, and fits in,
everyday practice. In total quality management theories,
the use of systematically gathered data is considered to be
crucial to facilitate effective professional development.
These data can then be used in plan-do-study-act cycles
(PDSA cycles) to provide insight into displayed behaviour
and help identify areas where improvement is possible.
This leads to the description of targets. These theories may
overlap or may be complementary. In implementation
science, the use of these theories as a framework is considered obligatory [32]. This intervention therefore will be
multifaceted and consist of audit, comparative graphical
feedback, and small group work with peer review of each
other's performance, discussion of barriers to change,
reaching agreement on future policy, and testing the
agreement. After randomization to arm A or B, each group
can choose from the corresponding set of five clinical topics allocated to that arm, to decide which three topics they
want to discuss. Two balanced sets of topics, one for each
arm, have been defined by the researchers. Each set consists of three major topics, from which the group has to
choose two, and two minor topics, one of which has to be
chosen. Thus, each LQIC will be asked to complete the
entire strategy for three clinical topics of their choice during the intervention period. They are free to schedule extra
meetings on topics not included in this trial, but these
meetings will not be included in the final analysis. Feedback on the topic under discussion will be sent to the
medical coordinator (diagnostic feedback) or local community pharmacists (prescription feedback) two weeks
prior to the test ordering or the prescribing session of the
LQIC, together with the relevant educational materials
(see under 'clinical topics'). The first session, which will
last approximately 90 minutes, will address the diagnostic
test ordering behaviour of the individual GPs and will
have the structure described under 'session structure'
(Table 2). During this session, the GPs will discuss their
diagnostic test ordering patterns and relate them to the
guidelines provided. Individual and group working agreements will be created after barriers to change have been
discussed. The second session will have the same structure, but the subject for discussion will be physicians' prescribing performance. This session will end by creating
group and individual working agreements about preferred
medication. Barriers to change from an individual perspective will again have to be discussed. After this first
topic has been completed, the cycle will be repeated, for a
new topic, as shown in Table 3. At the start of this new
cycle, the group will reflect on the previous agreements,
and revise them if necessary. The working agreements will
then be prepared for further dissemination in the practices. Each session will be chaired by a member of the
LQIC itself. When test ordering is discussed, a local representative from the diagnostic centre will be present, while
a local community pharmacist will be present when pharmacotherapy is discussed. They will act as moderators, not
as chairpersons.
We will test the model and the logistics needed prior to
the large-scale implementation. We plan to do this in a
small pilot study involving five groups of GPs. This pilot
study will run for four months, during which period the
participating GP groups will schedule two meetings. Each
session will be structured according to the method provided by the researchers. The first session will address test
ordering, while the second session will address prescribing. For reasons of efficiency, a set of only three topics will
be used for the pilot study. The topics, which have been
proposed by the project team members, are anaemia, dys-
Page 6 of 14
58
Table 2: Session structure
90 minutes
5 min
Explaining the method/reflection on previous topic
5 min
Critical look at participants' own feedback
5 min
Pairwise/group discussion on inter-individual differences
25 min
Plenary discussion, relating feedback to guidelines
10 min
Pairwise discussion on barriers to change
25 min
Plenary discussion on barriers to change, aimed at problem solving
15 min
Drawing up individual and group working agreements
pepsia, and asthma in combination with chronic obstructive pulmonary disease (COPD).
backs, given the indication. Each module will consist of a
maximum of six easily searchable pages.
Clinical topics
The set of clinical topics the GP groups can choose from
in the main study has been proposed by the authors. After
eligible topics were selected and divided over the two trial
arms, both arms were balanced in terms of the weight of
the topics. The weight depends on the prevalence of the
underlying disease and whether the emphasis within the
topic is on either the volume of tests ordered or the drugs
prescribed. The two sets of topics are also balanced in
terms of subjects, emphasising diagnostic or prescribing
features (Table 1). Each topic includes a number of tests
[See Additional file 2] and drugs [See Additional file 3]
predefined by the project group. For the purpose of feedback and education, these include both well-accepted and
commonly not accepted (or even obsolete) tests and
drugs. Educational materials on each topic will be based
on the relevant national primary care guidelines from the
Dutch College of General Practitioners, guidelines from
the Dutch Institute for Healthcare Improvement (CBO),
and international guidelines if applicable. Guidelines will
be read and 'condensed' into short versions called modules. These modules have been drafted by one of the
authors (JT) and then commented on by an expert on the
topic. Indicative prices for each test and drug will be provided, as well as a short description of its values and draw-
Extraction of feedback data
Data on test ordering behaviour will be extracted by the
regional coordinators from the various databases available at the participating hospital laboratories or primary
care diagnostic centres. Each centre will receive a data fact
sheet prescribing the required data format. This format is
based on rational criteria for laboratory test registration to
facilitate the integration of the individual databases into
one main database. All datasets on diagnostics will be
combined into one newly formed database, to be maintained by UM (Figure 2). Data on prescribing behaviour
will be extracted from the databases of health insurers and
collected into one database, as has already been done at
our institute. This database consists of the reimbursements for prescriptions written by GPs for approximately
5.5 million persons in the south of the Netherlands. Feedback will then be derived from the two main databases
and processed into graphical comparative feedback
reports. Data will be presented as the volume of tests
ordered (e.g., haemoglobin) or defined daily dosages
(DDDs) prescribed per 1000 patients per six months. Participating GPs will receive their data as clustered column
charts, each cluster presenting the data for the individual
GP, the practice in which he or she works, the small group
Table 3: Example of a schedule for the intervention
Topic
GPs
Medical coordinator
Community pharmacist
1. Anaemia
1. Meeting on tests
Moderator
Prepares second session
2. Meeting on drugs
Prepares third session
moderator
3. session on test and drugs (anaemia)
1. session on tests (Chlamydia)
moderator
Present as expert
2. Chlamydia infections
Page 7 of 14
59
Figure
Data
and2 Knowledge flowchart
Data and Knowledge flowchart.
and the wider region. An example of such a graphical feedback report is shown in Figure 3.
LQIC meeting structure
Each meeting will be structured according to a uniform
schedule. After participants have studied the feedback in
pairs or as a group, they will discuss it. Subsequently, the
guidelines as described in the educational materials will
be discussed in relation to the feedback. A plan will then
be formulated to improve the test ordering or prescribing
behaviour. The next step will involve addressing and discussing all the barriers to change at individual and group
levels. Finally, working agreements will be created regarding test ordering and prescribing behaviour for the tests
and drugs discussed. A standardized group meeting structure card will be provided to each LQIC, showing the
structure as recommended by the researchers. However,
groups will be free to adapt the structure to their own preferences or needs.
Training
The participating medical coordinators and local community pharmacists will be trained prior to the first LQIC session, in a two- to three-hour standardized training session
covering three main subjects. The first subject will involve
an explanation of the structure of the trial, the objectives,
the development of the outlines, the source of the feedback data, and the process of data gathering. The second
subject will be the preferred structure for the meetings, the
tools that are to be used, how to read the feedback reports
and relate the feedback to the guidelines. The final subject
of the training session will be how to act as a moderator
instead of a chair during a meeting. Training sessions will
partially be constructed like a LQIC meeting, with the
trainees acting as GPs and the trainer as the moderator.
Variables
Outcome measures
Process evaluation
1. The performance level of the small group collaborative.
Page 8 of 14
60
module A, July-Dec 2007
80
per 1000 patients
70
60
GP
practice n=2
group n=10
region n=30
50
40
30
20
10
0
te
st
A
te
st
B
te
st
C
Figure 3of a graphical comparative feedback sheet
Example
Example of a graphical comparative feedback sheet.
2. Process data such as attendance at meetings, actually
creating working agreements, following the LQIC strategy,
the number of groups that complete participation, and
the number of regions actually participating.
Effect evaluation
1. The volumes of particular tests ordered and particular
drugs prescribed for which the group has agreed that
change, either decrease or increase, would be necessary.
2. The total volumes of tests ordered and drugs prescribed
by the participating GPs for the clinical topics chosen.
3. The inter-physician variation in test ordering and prescribing behaviour for the clinical topics chosen.
Cost evaluation
The costs of implementing the LQIC strategy.
Explanatory variables
We will monitor data that are known to moderate quality
assurance strategies. Therefore the following data will be
gathered prior to the intervention: group size, age and
gender of GPs, type of practice, number of patients registered with the practice, number of patients a GP is
accountable for, number of working hours a week per GP,
number of working hours a week for the group practice as
a whole, distance to the hospital/diagnostic centre,
responsibility for training GP trainees, total number of
GPs collaborating in the practice, whether a GP admits
sales representatives from pharmaceutical firms and if so
how often, involvement in developing national guidelines, and GPs field(s) of special expertise.
All medical coordinators will be asked if problem-based
test ordering forms are used in their region and to send us
a copy of such a form.
Measurements
Prior to randomization, the chair of the group will be
asked to fill out a short pre-randomization form, with
which we will be able to determine the number of GPs in
the group and be able to assess the level of group performance [See Additional file 1]. Data on test ordering and prescribing behavior will be extracted from the existing
databases at baseline (t = 0) and t = 6 months, t = 12
months and t = 18 months. The dataset obtained at t = 0
and the final set will be used for a before-after analysis. A
new questionnaire will be sent to the chair, assessing the
level of group performance after the intervention. After
each meeting, the chair will be asked to fill out a form
with questions about the structure of the session, whether
Page 9 of 14
61
working agreements were created, whether barriers to
change were discussed and if so what the nature of these
barriers was, what educational materials were used, and
the group members' experiences with the strategy.
The process of implementing the strategy in the south of
the Netherlands will be monitored. Participants will be
questioned about their experiences with the strategy. Participating GPs will be asked to report their experiences
with the strategy, and to provide us with the necessary
details on the sessions they have attended. After each session, the targets set by each group will be recorded.
Analysis
Analysis will be based on the intention-to-treat principle.
Data on GPs lost to follow-up will be extracted from the
various databases if possible.
We will analyse covariance using test and drug volumes
during the intervention period as the dependent variables,
and the baseline data and the explanatory variables as
independent variables. The analysis will be repeated using
proportions stemming from prescription performance
indicators, if available. The unit of allocation to the trial is
the LQIC. In larger practices with more than one GP, not
all volumes of tests ordered and drugs prescribed will be
traceable to an individual GP. In these cases, the unit of
analysis will be the practice as a whole. Because of this
unit of analysis error, the data will be analysed using multilevel modelling. Data on drugs and tests will be clustered
to individual GPs at level one, the practice at level two, the
LQIC at level three and the region at level four.
The nature of this study makes it difficult to blind the participants, except for the tests or drugs serving as controls
in the other arm. The data analyst will be blinded for the
allocation result. Costs of the intervention will be calculated. A cost-effectiveness analysis will be based on these
figures. We will use cost minimization analysis from a
societal perspective, assuming that the strategy will reduce
redundant testing and prescribing. If there are signs of
improvement of care (higher scores on the performance
indicators), the impact on health may be estimated by
modelling the future gains and benefits. Data include
costs of coordinating the strategy by the regional contact
group, of preparing feedback reports and of chairing the
GP groups. The costs of the entire test ordering strategy by
Verstappen were 554.70 per GP per six months (three
meetings). The major part of the cost of this strategy consisted of opportunity costs, viz the costs of the GPs' time
spent attending the session. Because GPs were already
attending these meetings and were financially compensated, it seems fair to ignore the opportunity costs. This
results in costs of the test ordering strategy of 92.70. The
gains obtained by improving test ordering behaviour were
301.00 per GP per six months. Introducing the test ordering strategy would save 208.30 (92.70 to 301.00) per GP
per six months [33]. Because prescribing costs are higher,
the cost reductions gained by reducing superfluous prescribing should also be higher.
Time schedule
The intervention period will start in September 2007 and
run through the spring of 2009. Process evaluation will
start when all groups are included. During the intervention, new datasets will be obtained every six months in
order to keep the databases up-to-date for future use in
new sessions.
Discussion
To our knowledge, few studies have been published on
the transfer of effective implementation strategies to the
field. Our strategy has proved to be effective in an earlier
trial on test ordering by GPs in the Netherlands. However,
because this strategy was disseminated and controlled by
academics, it remains unclear how large its effect will be
when transferred to the field. We set up a pragmatic design
in order to test this final step in implementation research,
giving the diagnostic centres a leading role and leaving
GPs much room to adapt and to internalise the strategy.
The project team will act as facilitators to these centres, the
pharmacists involved, and the LQICs. The strategy is targeted first on test ordering and second on prescribing,
which is the natural order followed by GPs when consulted by a patient.
Our strategy is based upon several theories on effective
behaviour change and on effective implementation. These
theories can be identified at several levels of organisation
in our trial. At the level of diagnostic centres and the
LQICs, we expect the innovators and early adopters to join
the trial, which refers to Roger's innovation-diffusion theory [34]. Within groups we expect to see change according
to theories such as Ajzen's theory of planned behaviour
and the PDSA-cycles [25,35]. During a meeting, we expect
to see the preparation for change based on performance
data and actual actions towards change. When new data
will be provided to the groups, we expect reflection on the
goals previously set. The theory of planned behaviour
states that individuals are willing to show change in
behaviour dependent on the perceived control over the
behaviour itself, the attitude of the individual to the
desired behaviour, and the perceived social norms. By
providing graphical comparative feedback, we target at
these perceived social norms. Comparative feedback sets
the norm for a group, and through the phenomenon that
one does not like to be an outlier we expect regression to
the mean with regard to the inter-physician variation. The
moderator who is also an expert on the subject under discussion is expected to act as opinion leader. Furthermore,
Page 10 of 14
62
even a GP from within the group itself can act as a local
opinion leader and thus influence the rest of the group.
The existing PTAM group structure in the Netherland is
widespread and functions reasonably well. However the
need to improve the functioning of these groups is clearly
present. Our strategy is known to improve test ordering
behaviour of GPs, but is not used widely. Transferring
PTAM groups into LQICs gives us the opportunity to add
a discussion on test ordering behaviour to existing discussions on prescribing by GPs in PTAMs. The constitution of
LQICs therefore is not 'old wine in new bottles' but a completely new approach within existing structures.
Several methodological challenges were encountered
when we designed this trial. First, individual GPs are
known to choose topics for CME in which they already
show good performance [36]. This might result in a 'ceiling effect', meaning that little or no improvement in test
ordering or prescribing behaviour would be possible.
However, because the LQIC will have to reach consensus
on the clinical topics they choose, the risk of such a ceiling
effect is probably not very great.
Second, using an implementation strategy on ten different
clinical topics from which GPs can choose introduces
challenges to the sample size calculation. We chose to
leave the LQICs some freedom of choice with regard to
the topics. All clinical topics are well-described in the
national guidelines for each topic. We will use a set of 204
tests and drugs to generate feedback [See Additional files
2] [See additional file 3]. Because we do not know what
agreements local groups will come to, and do not know
beforehand what the desired direction for change is, sample size calculation is very difficult. Because we intend to
improve care by using the national guidelines, we do not
expect to decrease quality of care by this study. However,
it is impossible to predict if change will be towards better
care.
Third, the databases we use are complex, as are the origins
of the data. Most local databases on diagnostics used in
this trial are intended primarily for billing purposes. This
might create problems when extracting data, reading it
into a central database and translating it into feedback. In
the past, no significant problems were encountered when
extracting data from laboratories (personal communication by Verstappen). Data on tests not performed within a
laboratory (e.g., gastroscopy and X-rays), however, are
often stored in separate databases and might not be linked
to a GP but to a patient. In these cases, tracing the GP who
ordered the test is possible but will require an extra effort
from the diagnostic centres. It is possible that recruiting
groups, supplying a moderator for the sessions and implementing this time-consuming data extraction process
might prove to be too much of an effort for the centres.
Most contact persons of the centres, however, have indicated that they were most willing to cooperate and were
aware of the opportunities offered by this trial.
Fourth, the database on prescriptions consists of data
from the large insurance companies in the south of the
Netherlands. Using these records as a basis for feedback
might create several problems. Although most inhabitants
of the southern provinces are insured by one of these companies, prescription data for patients insured with other
companies will not be included in our database on prescriptions. This problem might be solved in the future by
adding more insurance companies to the database.
Another potential problem may be that recording errors
are likely to be present in the databases. Desk staff at local
pharmacies often links a prescription to one of the GPs in
a practice, and often almost all prescriptions for a practice
are thus linked to one physician, even when several physicians collaborate in the practice. This creates an inaccuracy in the database, but only for GPs sharing an office. To
solve this problem, we will also aggregate to an extra level
in these cases, viz the subgroup of GPs sharing an office,
thus creating a fourth column on the graphical feedback
sheet. The last problem we expect to encounter using a
large database on prescription is that we do not know the
indication for which medication was prescribed; these
indications are not known to pharmacists and thus are
not stored in any database. This makes it impossible to
trace a prescription back to a specific disease. By building
a similar database on tests ordered by GPs, we will
encounter this problem as well. We do not however expect
this to be a problem because we will use graphical comparative feedback. All data from all participating GPs are
expected to be equally be affected by this problem and
thus the feedback will be comparable.
Fifth, the tests of the diabetes and hypercholesterolemia
topics partly overlap. We accepted this, however, because
in diabetes, the glucose and HbA1c items are the primary
indicators, whereas cholesterol, LDL, HDL, and the ratio
are the primary indicators in the hypercholesterolemia
topic.
Sixth, we have to be aware of the Hawthorne effect. As discussed above, we chose to use a balanced incomplete
block design to overcome this problem. The complexity of
the strategy, however, would make it more attractive to
use a different design and start the trial in phases. This
would mean that different regions would enrol in the
strategy successively, so we could learn from the early
regions what the weaknesses of our design were and what
we would have to alter. This would create an opportunity
to ameliorate the strategy with each new phase. To this
end, a dynamic wait-listed design could have been more
Page 11 of 14
63
appropriate and beneficial [37]. Conversely, we would
then have had to wait after completing enrolment and
intervening in one region for new data to be added to the
database. The delay would be six months after each
region. This left us with no choice but to start with the
entire population in the same period. In this situation, we
considered the balanced incomplete block design to be
most useful.
Finally, GPs and moderators cannot be sufficiently
blinded in our present design. However, because GPs do
not know what clinical topics are available in the arm they
are not allocated to, we do achieve some level of blinding.
Notwithstanding these methodological challenges, there
are also opportunities in the Dutch healthcare system that
make it attractive to start this trial now. First, the strategy
we intend to use fits in well with the new Dutch healthcare system. After the recent reform, healthcare has turned
into a competitive business, in which financial profits and
market shares may influence decision-making. Our study
might create profiling opportunities for centres, which
might bind GPs more tightly to them, and thus might be
a way for the centres to improve their chances in this market. Finally, diagnostic centres are under increasing pressure from various parties in the healthcare system to
provide feedback to GPs. GPs want feedback to monitor
and claim results when treating chronically ill patients
(e.g., diabetics), while insurance companies want laboratories to provide feedback in order to influence test ordering behaviour, and primary care organizations need GPs'
performance data for various reasons, such as certification.
A preliminary investigation identified 24 eligible diagnostic centres in hospitals, all of which provide diagnostic
facilities to GPs. All were contacted and appointments for
personal visits were made. Two centres were not interested
in participating, and were therefore not visited. Two centres expressed an interest but faced major strategic challenges and found no time to participate. The remaining 20
centres all agreed to participate. One of the participating
centres will not be asked to recruit groups, however,
because it is not linked to a region like the other centres,
which means that knowledge of local PTAM group structures is lacking. This centre will, however, participate in
the large database on diagnostics.
In the south of the Netherlands, health insurance is
offered predominantly by two companies, which insure
the majority of the inhabitants of these provinces. These
insurance companies regularly send updated reports on
prescription data to UM. These files are and will be combined into one research database on prescriptions, maintained by MEMIC. Because the recent health care reform
in the Netherlands, insurance companies have been given
a large role in guarding and improving the quality and
continuity of care. They promote the existence of PTAM
groups in order to improve the quality of care, giving
financial incentives to GPs for attending such group meetings. In some cases, extra incentives are given if working
agreements are created and adhered to. However, the
insurers are unable to evaluate the quality of the group
work. The strategy evaluated in the proposed study should
provide them with a tool to ensure high quality group
meetings.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
JT drafted the manuscript, participated in its design and is
the researcher on the trial. TvdW conceived of the study,
participated in the design and coordination and helped to
draft the manuscript. WV, RJ, PM, JM, BvS and RG all conceived of the study, participated in its design and read and
corrected earlier versions of the manuscript. RW read and
corrected earlier versions of the manuscript. All authors
have read and approved the final manuscript. The Additional Files below (4, 5 and 6) refer to information regarding The Funding approval ZonMw, The Approval ethical
committee and The CONSORT Cluster RCT Checklist.
Additional material
Additional file 1
This file displays the pre-randomization questionnaire as it was sent
to the chair of each LQIC.
Click here for file
[http://www.biomedcentral.com/content/supplementary/17485908-4-6-S1.pdf]
Additional file 2
The impact of local quality improvement collaboratives additional file
2. This file includes all the diagnostic tests used in this trial, the diversion
over the modules and how each item is labelled on the feedback form.
Click here for file
Additional file 3
The impact of local quality improvement collaboratives additional file
3. This file includes all the farmaceuticals used in this trial, the diversion
over the modules and how each item is labelled on the feedback form.
Click here for file
Additional file 4
Funding approval ZonMw. Scanned letter of ZonMw in which the funding of this trial is confirmed.
Click here for file
Page 12 of 14
64
Additional file 5
Approval ethical committee. Scanned letter of the Maastricht ethical
committee, stating that it is not required to fully review this trial by the
committee because Dutch law on research with humans is not applicable.
Click here for file
Additional file 6
CONSORT Cluster RCT Checklist. checklist with reference to page
numbers in this trial concerning the CONSORT statement for clusterRCTs.
Click here for file
[http://www.biomedcentral.com/content/supplementary/17485908-4-6-S6.doc]
13.
14.
15.
16.
17.
18.
Acknowledgements
Funds for this trial were obtained from a unrestricted grant from OWM
Centrale Zorgverzekeraars group, Zorgverzekeraar U.A, Tilburg, the
Netherlands and ZonMw, the Netherlands organisation for Health
Research and Development.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Braspenning JCC, Schellevis FG, Grol R: Kwaliteit huisartsenzorg
belicht [Quality in primary care revised]. Nivel 2004.
Burgers JS, Grol RPTM, Zaat JOM, Spies TH, Bij AK van der, Mokkink
HGA: Characteristics of effective clinical guidelines for general practice. Br J Gen Pract 2003, 53:15-19.
Muijrers PE, Grol RP, Sijbrandij J, Janknegt R, Knottnerus JA: Differences in prescribing between GPs. Impact of the cooperation with pharmacists and impact of visits from
pharmaceutical industry representatives. Fam Pract 2005,
22:624-630.
Harteloh PPM: Kwaliteit van zorg: van zorginhoudelijke benadering naar bedrijfskundige aanpak [Quality of care: from a
care standpoint towards a business management standpoint]. In Maarssen 4th edition. Elsevier/De Tijdstroom; 2001.
Verstappen WH, ter Riet G, Dubois WI, Winkens R, Grol RP, Weijden T van der: Variation in test ordering behaviour of GPs:
professional or context-related factors? Fam Pract 2004,
21:387-395.
Martens JD, van-der-Weijden T, Severens JL, de-Clercq PA, de-Bruijn
DP, Kester AD, Winkens RA: The effect of computer reminders
on GPs' prescribing behaviour: A cluster-randomised trial.
International Journal Medical Informatics 2007, 76:S403-416.
Verstappen WHJM, Weijden T van der, Sijbrandij J, Smeele I,
Hermsen J, Grimshaw J, Grol RPTM: Effect of a Practice-Based
Strategy on Test Ordering Performance of Primary Care
Physicians: A Randomized Trial. JAMA 2003, 289:2407-2412.
Baker R, Falconer Smith J, Lambert PC: Randomized controlled
trial of the effectiveness of feedback in improving test ordering in general practice. Scan J Prim Health Care 2003, 21:219-223.
Sondergaard J, Andersen M, Stovring H, Kragstrup J: Mailed prescriber feedback in addition to a clinical guideline has no
impact: a randomised, controlled trial. Scandinavian Journal of
Primary Health Care 2003, 21:47-51.
Martens JD, Winkens RA, Weijden T van der, de Bruyn D, Severens
JL: Does a joint development and dissemination of multidisciplinary guidelines improve prescribing behaviour: a pre/
post study with concurrent control group and a randomised
trial. BMC Health Serv Res 2006, 6:145.
Winkens RA, Pop P, Bugter-Maessen AM, Grol RP, Kester AD, Beusmans GH, Knottnerus JA: Randomised controlled trial of routine individual feedback to improve rationality and reduce
numbers of test requests. Lancet 1995, 345:498-502.
Jamtvedt G, Young JM, Kristoffersen DT, O'Brien MA, Oxman AD:
Audit and feedback: effects on professional practice and
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
health care outcomes.
Cochrane Database Syst Rev
2006:CD000259.
Thomas RE, Croal BL, Ramsay C, Eccles M, Grimshaw J: Effect of
enhanced feedback and brief educational reminder messages on laboratory test requesting in primary care: a cluster
randomised trial. The Lancet 2006, 367:1990-1996.
Verstappen WHJM, Weijden T van der, Dubois WI, Smeele I,
Hermsen J, Tan FES, Grol RPTM: Improving Test Ordering in
Primary Care: The Added Value of a Small-Group Quality
Improvement Strategy Compared With Classic Feedback
Only. Ann Fam Med 2004, 2:569-575.
Solomon DH, Hashimoto H, Daltroy L, Liang MH: Techniques to
Improve Physicians' Use of Diagnostic Tests: A New Conceptual Framework. JAMA 1998, 280:2020-2027.
Grimshaw JM, Thomas RE, MacLennan G, Fraser C, Ramsay CR, Vale
L, Whitty P, Eccles MP, Matowe L, Shirran L, et al.: Effectiveness
and efficiency of guideline dissemination and implementation strategies. Health Technol Assess 2004, 8:iii-iv.
Grol RPTM, grimshaw JM: From best evidence to best practice:
effective implementation of change in patients' care. The Lancet 2003, 362:1225-1230.
Axt-Adam P, Wouden JC van der, Does E van der: Influencing
behavior of physicians ordering laboratory tests: a literature
study. Med Care 1993, 31:784-794.
Eccles M, Steen N, Grimshaw J, Thomas L, McNamee P, Soutter J,
Wilsdon J, Matowe L, Needham G, Gilbert F, Bond S: Effect of audit
and feedback, and reminder messages on primary-care radiology referrals: a randomised trial.
The Lancet 2001,
357:1406-1409.
Bunting PS, Van Walraven C: Effect of a controlled feedback
intervention on laboratory test ordering by community physicians. Clinical Chemistry 2004, 50:321-326.
Grimshaw JMMP, Shirran LMAM, Thomas RB, Mowatt GMAMBA,
Fraser CMA, Bero LP, Grilli RMD, Harvey EB, Oxman AMDa, O'Brien
MAM: Changing Provider Behavior: An Overview of Systematic Reviews of Interventions. Medical Care 2001, 39:II-45.
Arnold SR, Straus SE: Interventions to improve antibiotic prescribing practices in ambulatory care. Cochrane Database Syst
Rev 2005:CD003539.
Lagerlov P, Loeb M, Andrew M, Hjortdahl P: Improving doctors'
prescribing behaviour through reflection on guidelines and
prescription feedback: a randomised controlled study. Qual
Saf Health Care 2000, 9:159-165.
Freemantle N, Harvey EL, Wolf F, Grimshaw JM, Grilli R, Bero LA:
Printed educational materials: effects on professional practice and health care outcomes. Cochrane Database Syst Rev
2000:CD000172.
Grol R, Wensing M, Eccles M: Improving patient care. The
implementation of change in clinical practice. Elsevier Limited
2005.
Paes AH, Cornips PMJM: Ontstaan en verloop van een farmacotherapie-overleg [the emergence and development of pharmacotherapeutic meetings]. Pharm Weekbl 1987, 122:225-230.
Eimers M, de Groot J: FTO peiling 2005, kwaliteit van farmacotherapieoverleg in Nederland in beeld [PTAM audit 2005,
quality of pharmacotherapeutical audit meetings in the
Netherlands made visible]. DGV Nederlands instituut voor verantwoord medicijngebruik [DGV the Dutch Institute for the Proper Use of Medicine] 2006.
DGV Nederlands instituut voor verantwoord medicijngebruik [DGV the Dutch Institute for the Proper Use of Medicine] [http://www.medicijngebruik.nl]
Verstappen WHJM, Weijden T van der, ter Riet G, Grimshaw J,
Winkens R, Grol RPTM: Block design allowed for control of the
Hawthorne effect in a randomized controlled trial of test
ordering. Journal of Clinical Epidemiology 2004, 57:1119-1123.
Eijk MECv, Paes AHP, Porsius A, Avorn J, Boer Ad: Pre-randomization decisions and group stratification in a randomized controlled trial to improve prescribing. Pharm World Sci 2004,
V26:227-231.
Eimers M, Aalst A van der, Pelzer B, Teichert M, de Wit H: Leidt een
goed FTO tot beter voorschrijven? [Does an effective PTAM
leed to more appropriate prescibing?] (Dutch). Huisarts en
Wetenschap 2008, 51:340-345.
Page 13 of 14
65
32.
33.
34.
35.
36.
37.
Grol RPTM, Bosch MC, Hulscher MEJL, Eccles MP, Wensing M: Planning and Studying Improvement in Patient Care: The Use of
Theoretical Perspectives. The Milbank Quarterly 2007, 85:93-138.
Verstappen WHJM, van Merode F, Grimshaw J, Dubois WI, Grol
RPTM, Weijden T van Der: Comparing cost effects of two quality strategies to improve test ordering in primary care: a
randomized trial. Int J Qual Health Care 2004, 16:391-398.
Rogers EM: Diffusion of innovations 4th edition. New York The Free
Press; 1995.
Ajzen I: The theory of planned behavior. Organizational Behavior
and Human Decision Processes 1991, 50:179-211.
Hobma S: Directed self-learning as approach to continuing
professional development. Maastricht University, General practice; 2005.
Brown CH, Wyman PA, Guo J, Pena J: Dynamic wait-listed
designs for randomized trials: new designs for prevention of
youth suicide. Clin Trials 2006, 3:259-271.
Publish with Bio Med Central and every
scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical researc h in our lifetime."
Sir Paul Nurse, Cancer Research UK
Your research papers will be:
available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright
BioMedcentral
Submit your manuscript here:
http://www.biomedcentral.com/info/publishing_adv.asp
Page 14 of 14
66
REFLECTEREND TESTEN
WP Oosterhuis, arts klinische chemie
AGJM Roos, huisarts
Reflecterend testen is een procedure waarbij de laboratoriumspecialist, op eigen
initiatief, aanvullende testen toevoegt aan een oorspronkelijke laboratoriumaanvraag,
na beoordeling (reflectie) van de resultaten. Hierbij wordt gekeken naar bepaalde
patronen van (afwijkende) laboratoriumuitslagen, in combinatie met eventueel
beschikbare
informatie
uit
het
ziekenhuisinformatiesysteem.
Patiënten
met
voldoende pre-test waarschijnlijkheid op een aandoening komen in aanmerking voor
gerichte, aanvullende diagnostiek. Nadat het onderzoek compleet is, voegt de
laboratoriumspecialist in de meeste gevallen een interpretatief commentaar toe aan
het uitslagrapport. Zo kan de aanvrager de uitslagen gemakkelijker interpreteren. Het
reflecterend testen wordt voornamelijk toegepast bij eerstelijnsdiagnostiek.
Deze vorm van consultverlening (1) vraagt een proactieve houding van het
laboratorium. Het verschilt van ‘reflex testen’, waarbij een van tevoren vastgesteld
testprotocol automatisch wordt doorlopen. Reflecterend testen biedt de mogelijkheid
om richtlijnen te effectueren en met het beschikbare monster kunnen in de meeste
gevallen de aanvullende testen worden uitgevoerd. De voordelen voor de patiënt zijn
een snellere afronding van de diagnostiek en het besparen van een tweede
bloedafname. In de laboratoria van het Verenigd Koninkrijk wordt reflecterend testen
gezien als een integraal onderdeel van de dienstverlening (2). Het laboratorium van
het Atrium Medisch Centrum Parkstad te Heerlen is in juni 2006 gestart met
reflecterend testen bij aanvragen van huisartsen. Uit enquêtes is gebleken dat
huisartsen in onze regio deze service zeer op prijs stellen (3). Buitenlands onderzoek
bevestigt deze bevinding (4). Bovendien beoordelen huisartsen die bekend zijn met
reflecterend testen casuïstiek vaker correct, in vergelijking tot huisartsen waarbij
deze consultverlening niet gegeven wordt (5). Voor laboratoriumspecialisten
Klinische Chemie (i.o.) die de procedure van reflecterend testen in de eerste lijn
willen introduceren c.q. uitvoeren, kan de website www.reflectivetesting.com een
nuttig en bruikbaar hulpmiddel zijn.
67
Aangezien er nog onvoldoende wetenschappelijk bewijs – met betrekking tot de
effectiviteit van reflecterend testen voor het zorgproces van de patiënt – voorhanden
was, hebben we een grote gerandomiseerde klinische trial uitgevoerd. Gedurende
een aantal maanden werden 600 uitslagrapporten verzameld, waarbij reflecterend
testen werd toegepast. Na randomisatie ontvingen huisartsen van 300 patiënten de
resultaten van de oorspronkelijke, door hen aangevraagde, testen (controlegroep).
Huisartsen van de overige 300 patiënten kregen behalve de testuitslagen, ook de
door de laboratoriumspecialist toegevoegde testen en interpretatief commentaar
(interventiegroep). Na een follow-up periode van zes maanden werd informatie
verzameld over laboratoriumonderzoek, behandelingen en/of verwijzing naar een
specialist, ander diagnostisch onderzoek en patiëntspecifieke informatie uit het
huisartsinformatiesysteem (zoals medische voorgeschiedenis en medicijngebruik).
Een expert panel – bestaande uit een laboratoriumspecialist klinische chemie, een
huisarts en een internist – beoordeelde vervolgens de effectiviteit van reflecterend
testen op drie aspecten: 1) Was het reflecterend testen zinvol/nuttig voor de patiënt?,
2) Was er een intentie tot verbetering van het zorgproces?, 3) Was er een
daadwerkelijke verbetering van het zorgproces? Bij discrepanties tussen de
beoordelingen van de panelleden werden de betreffende casussen besproken om tot
consensus te komen. In totaal werd van 350 patiënten informed consent ontvangen.
In een subgroep van 80 patiënten leidden de aanvullende testen niet tot een
interpretatief commentaar. Het vermoeden van een bepaalde combinatie van
afwijkende laboratoriumuitslagen c.q. ziektebeeld werd hierbij niet bevestigd door de
uiteindelijke uitslag. De gegevens van de overige 270 patiënten werden verder
geanalyseerd (interventiegroep: n=148, controlegroep: n=122). Reflecterend testen
werd als zinvol/nuttig beoordeeld in 84% van de gevallen. In de interventiegroep
werd een intentie tot een adequate behandeling c.q. actie door de huisarts
waargenomen in 80% van de gevallen (vs. 56% in de controlegroep; F2 18.38,
p<0.001) (Figuur 1A). Ten slotte is de daadwerkelijke verbetering van het zorgproces
van de patiënt beoordeeld door het panel. In 70% van de gevallen in de
interventiegroep werd een duidelijke verbetering waargenomen (vs. 47% in de
controlegroep; F2 16.27, p<0.001) (Figuur 1B). De percentages liggen hier enigszins
lager dan wanneer we kijken naar de intentie tot verbetering van het zorgproces. Dit
heeft te maken met het feit dat de intentie van de huisarts (om een bepaalde actie te
68
ondernemen, bijvoorbeeld het inzetten van aanvullend onderzoek) goed kan zijn,
maar dat dit uiteindelijk – om welke reden dan ook – niet uitgevoerd is.
Reflecterend testen heeft nooit klachten of bezwaren opgeleverd van individuele
huisartsen of patiënten. Er zijn nu formele afspraken gemaakt met de regionale
huisartsenorganisatie waarin de toepassing van reflecterend testen is geregeld. De
kosteneffectiviteit is een belangrijk aspect dat nader onderzoek verdient. Het
uitvoeren van extra testen bij een bestaande aanvraag is goedkoper dan een tweede
bloedafname. De extra kosten van het laboratorium moeten worden vergeleken met
de winst voor de patiënt als een diagnose eerder wordt gesteld. Ook moet men
hiermee de economische winst verdisconteren van het sneller starten van een
gerichte behandeling, of het stoppen met een onnodige behandeling.
Figuur 1A
Figuur 1B
Referenties
1. Oosterhuis WP, Raijmakers MTM, Leers MPG, Keuren JFW, Verboeket-van de
Venne WPHG, Munnix ICA, Kleinveld HA. Consultverlening: van klinisch
chemicus naar laboratoriumspecialist. Ned Tijdschr Klin Chem Labgeneesk 2009;
34: 214-218.
2. Simpson WG, Twomey PJ. Reflective testing. J Clin Pathol 2004; 57: 239-240.
69
3. Oosterhuis WP, Keuren JFW, Verboeket-van de Venne WPHG, Soomers FLM,
Stoffers HEJH, Kleinveld HA. Eigen inbreng van het laboratorium – huisartsen
positief over ‘reflecterend testen’. Ned Tijdschr Geneeskd. 2009; 153: A486.
4. Barlow IM. Are biochemistry interpretative comments helpful? Results of a
general practitioner and nurse practitioner survey. Ann Clin Biochem 2008; 45:
88–90.
5. Verboeket-van de Venne WPHG, Oosterhuis WP, Waard H de, Sant P van ‘t,
Kleinveld HA. Beïnvloedt ‘reflecterend testen’ het beoordelen van casuïstiek door
huisartsen? Ned Tijdschr Klin Chem Labgeneesk 2011; 36: 272-274.
70
Clin Chem Lab Med 2012;50(7):1249–1252 © 2012 by Walter de Gruyter • Berlin • Boston. DOI 10.1515/cclm-2011-0611
Opinion Paper
Reflective testing: adding value to laboratory testing
Wilhelmine P.H.G. Verboeket-van de Venne1,*,
Kristin M. Aakre2,3, Joseph Watine2,4 and
Wytze P. Oosterhuis1,2
1
Department of Clinical Chemistry, Atrium Medical Centre,
Heerlen, The Netherlands
2
European Federation of Clinical Chemistry and Laboratory
Medicine (EFLM) Working Group on Guidelines
3
Laboratory of Clinical Biochemistry, Haukeland University
Hospital, Bergen, Norway
4
Laboratoire de Biologie Polyvalente, Hôpital de la
Chartreuse, Villefrance-de Rouergue, France
Abstract
Reflective testing is a procedure in which the laboratory specialist adds additional tests and/or comments to an
original request, after inspection (reflection) of the results.
It can be considered as an extension of the authorization
process where laboratory tests are inspected before reporting to the physician. The laboratory specialist will inevitably find inconclusive results, and additional testing can
contribute to make the appropriate diagnosis. Several studies
have been published on the effects of reflective testing. Some
studies focus on the opinion of the general practitioners or
other clinicians, whereas other studies were intended to determine the patient’s perspective. Overall, reflective testing was
judged as a useful way to improve the process of diagnosing
(and treating) patients. There is to date scarce high quality
scientific evidence of the effectiveness of this procedure in
terms of patient management. A randomized clinical trial
investigating this aspect is however ongoing. Cost effectiveness of reflective testing still needs to be determined in the
future. In conclusion, reflective testing can be seen as a new
dimension in the service of the clinical chemistry laboratory to primary health care. Additional research is needed to
deliver the scientific proof of the effectiveness of reflective
testing for patient management.
Keywords: adding on tests; primary health care; reflective
testing.
*Corresponding author: Dr. Wilhelmine P.H.G. Verboeket-van de
Venne, Department of Clinical Chemistry, Atrium Medical Centre,
PO Box 4446, 6401 CX Heerlen, The Netherlands
E-mail: [email protected]
Received September 5, 2011; accepted March 30, 2012;
previously published online April 28, 2012
Background
The core business of the clinical laboratory is to provide
results of tests requested by physicians and other health care
workers. The task of the laboratory can be defined in broader
terms – to help solve diagnostic problems. In the postanalytical phase, laboratory professionals can add value over
the purely analytical service. Their knowledge could be used
in the interpretation of laboratory test results (1). It is no
exception that in a laboratory examination of a patient, abnormal results may be found that could indicate some unexpected
pathology. Recognition and interpretation of abnormal results
by the laboratory specialist may be helpful for physicians and
patients. Examples of disorders typically recognizable by
distinct laboratory findings are hemochromatosis, m-proteins,
hyperparathyroidism, vitamin B12 deficiency, thalassemia,
hepatitis or Gilbert’s syndrome. The laboratory specialist
can take other available (medical) information into account
(e.g., age, gender, previous laboratory test results, and clinical information) when interpreting abnormal test results and
determine whether additional tests are indicated. In most
cases, these tests may be performed with the patient’s material already being available in the laboratory. Comments can
be added to the report to serve the requesting physician. This
process has been called reflective testing (2).
Since the practice of reflective testing is not a common procedure, we recently launched a website (www.reflectivetesting.
com) in order to inform other laboratory specialists in detail
about this procedure (3). With the support of the European
Federation of Clinical Chemistry and Laboratory Medicine
(EFLM) Working Group on Guidelines, this website has been
edited and translated into English.
Reflective and reflex testing
The term reflective testing was chosen because it is discretionary and based on the clinical judgement (reflection) of
a laboratory specialist regarding interpretation of laboratory
results. Reflective testing is different from reflex testing (also
called protocol testing), in which a predetermined test protocol is automatically completed. Examples are the addition of
free thyroxin (T4) when thyroid stimulating hormone (TSH)
is abnormal, or free prostate specific antigen (PSA) in case of
an increased level of total PSA.
However, in cases with multiple abnormal test results, it is
difficult to incorporate additional testing into an automated
protocol. Considering the addition of appropriate tests is not a
simple process, and requires professional, medical experience
Brought to you by | Atrium MC Medische Bibliotheek
Authenticated | 212.178.208.50
Download Date | 8/17/12 3:39 PM
71
1250
Verboeket-van de Venne et al.: Reflective testing
combined with the knowledge of patient characteristics.
Although a laboratory specialist might be dependent on the
limited information on the request form of a general practitioner, the ongoing development of the electronic patient
record allows better assessment of the clinical status of the
patient. An incorporated filter in the laboratory information
system [based on range- or delta checking (4)] can facilitate
the selection of reports that are suitable for assessment.
An important point of attention is the fact that both reflective and reflex testing can be executed in contemporary clinical chemistry laboratories. Reflective testing was introduced
to 155 general practitioners in the area of our hospital in June
2006 concerning only a small selection of patients. From our
own database it has been shown that in 10%–15% (64/512;
average over a 20-day period) of the daily reports abnormal
test results are observed that need evaluation by the laboratory
specialist (5). Additional tests and/or comments are added
in 2%–3% of the daily reports (270 over a 20-day period),
mainly from primary health care requests Reflex testing is
also daily routine in our laboratory, e.g., diagnosing patients
with suspected anemia. Using a single blood sample, reflex
diagnostics is being used to elucidate the cause of the anemia.
A patient-specific, interpretive comment is added to complete
the report.
Studies on the effect of reflective testing
Paterson and Paterson (2) were the first to study reflective
testing quantitatively. They investigated the effect of adding
on either 25-hydroxyvitamin D or total iron binding capacity
in combination with the percentage of transferrin saturation,
in order to confirm vitamin D deficiency or hemochromatosis, respectively. The number of add-on tests needed to obtain
a diagnosis (NND: number needed to diagnose) turned out to
be 4.3 for vitamin D deficiency and 18.8 for genetic hemochromatosis. They highlighted the value of reflective testing
to the requesting clinician in three respects: to help exclude
a diagnosis, to expedite a diagnosis that is fairly obvious,
and to obtain a diagnosis when the original set of results is
equivocal. In another study, the impact and effectiveness of
introducing reflective and reflex testing of magnesium in
severe hypokalemia was investigated (6). Measurement of
magnesium in patients with a potassium concentration ≤ 2.5
mmol/L was consecutively studied during three periods of
6 months (baseline, reflective testing, reflex testing). Diagnosis
of hypomagnesemia significantly increased from 7.7%
(4/52) to 43.1% (31/72) and 69.3% (52/75) with reflective
and reflex testing, respectively. It was concluded that in this
biochemical scenario reflex testing was more effective than
reflective testing. The clinical utility of reflex and reflective
testing was also investigated by Srivastava and co-workers
(7). Five scenarios were prospectively studied for one year:
vitamin D deficiency, hypomagnesemia, hypothyroidism,
hyperthyroidism and hemochromatosis. The main message was that reflex and reflective testing are complementary strategies. Reflex testing is recommended in scenarios
where high efficiency (low NND) can readily be achieved.
The contribution of reflective testing is comparatively greater
when more complex factors need to be considered (e.g., to
diagnose hemochromatosis).
Opinion of professionals and patients
Darby and Kelly (8) conducted a survey to elicit the service
users’ opinion of reflective testing. Ten clinical scenarios,
each involving the possible addition of a specific test, were
circulated to both hospital doctors and general practitioners. The four response options were to add further tests,
phone the clinician and discuss the case, add a comment to
the original results (without consultation) or do nothing. It
was concluded that reflective testing is generally welcomed
by the doctors, provided that the nature and implications of
the specific test are taken into consideration. The results of
this study were confirmed in a Dutch study, with comparable clinical scenarios (5, 9). A difference with the former
study was that the doctors participating in the survey in The
Netherlands were not used to reflective testing as a routine
service. Nevertheless, the results were remarkably similar to
the British study: reflective testing was judged to be useful
by the responding general practitioners in 99% (148/150) of
the cases (5, 10). In 53% (80/150) of the cases reflective testing was reported to have had an effect on the policy of the
general practitioners, in terms of further diagnostics, (change
of) medication or referral to a specialist. The high response
rate of the general practitioners (87%; 77/89) strengthens the
validity of these data and it may be concluded that reflective
testing is considered to be useful. Another study investigated
the influence of reflective testing on the assessment of clinical
case reports by general practitioners (11). A list of 13 cases
was prepared and sent to 56 local general practitioners (which
are used to the procedure of reflective testing including interpretative comments) and 31 general practitioners linked to the
hospital in Den Bosch (which are not familiar with reflective testing). The general practitioners were asked about their
working hypothesis and subsequent action(s) they would take
(e.g., additional laboratory diagnostics, referral to specialist,
medication, other follow-up). The lists were judged by their
agreement with the suspected diagnosis as determined by the
laboratory specialist after adding additional tests. The results
showed a better concordance between the suspected diagnosis
of the laboratory specialist and the actions suggested by the
general practitioners if the general practitioners were familiar with reflective testing (50.8% vs. 38.2%). In conclusion,
reflective testing in primary care had resulted in a learning
effect by general practitioners.
Paterson et al. (12) conducted a study on patients attending
a general practice or a hospital outpatient clinic. They were
asked their views about the practice of add-on testing by the
laboratory specialist. A large majority of patients favored an
approach in which relevant additional tests are performed
without consulting the requesting clinician or patient first.
This is a clear indication that most patients are content to let
professionals add on relevant tests if this is felt to be in their
interest.
72
Interpretative comments
Several studies have been published on the influence of providing patient-specific interpretative comments, without
additional testing. Barlow conducted a survey among general
practitioners and nurse practitioners to analyze whether they
found biochemistry comments on reports helpful. Clinical
comments were added to most endocrine sets of results, glucose tolerance test results and other miscellaneous test results
where interpretation is thought to be of help. There was
overwhelming support for commenting on tests, and most
responders would like to see comments on a greater range of
tests (13). In an additional survey, it was asked whether the
comments actually had had influenced patient management.
In summary, it was concluded that in at least 75% of the cases,
comments either helped or influenced patient management
(14). In a survey studying the value of narrative interpretation
of complex coagulation test panels, physicians indicated that
in approximately 80% of the cases the comments saved them
time and/or improved the diagnostic process (15). Further
documentation supporting the need for interpretative comments was provided by a survey in the US. It was shown that
nearly one in four primary care physicians reported that the
scope of care they were expected to provide was greater than
it should be (16). Among the specialists in the survey, more
than one in three (38%) reported that the complexity or severity of patients’ conditions at the time patients were referred
to them by primary care physicians was greater than it should
be. Additional help from the laboratory specialist (by means
of adding interpretative comments to test results ordered by
general practitioners) could be an important option to tackle
this problem.
It can be argued that interpretative commenting is an
important part of the procedure of reflective testing. In case
of deviating laboratory test results, the laboratory specialist
evaluates the results and decides to add one or more tests.
Most of the time, the report is completed by an interpretative
comment to guide/assist the requesting physician.
Limitations and further research
Several surveys and (observational) studies have been
published on different aspects of reflective testing and most
conclude positively regarding this intervention. However,
these are not considered to present scientific evidence that
reflective testing is related to an improved clinical outcome.
A randomized trial investigating this aspect has recently been
started (17). Preliminary data show that reflective testing
in patients resulted in more adequate actions, compared to
controls with standard care [42% (22/52) vs. 27% (9/33),
respectively] (18).
Another issue is the inter-individual variation between
laboratory professionals, as different individuals can have
different approaches. It has been shown elsewhere, that the
variation between laboratories in the process of authorization
of reports is large (19, 20). The greatest variation between
laboratories occurred in the number of results reviewed in
1251
the clinical validation queue. This varied from 5% to 100%.
The use of post-analytical external quality assessment (e.g.,
asking laboratory specialists to comment on case histories
and distribute feedback reports describing their performances
afterwards) might help to reduce the variance of postanalytical laboratory practice in the future (21, 22).
Establishing an external quality assessment for laboratory
post-analytical activities could be considered when reflective
testing is initiated.
Reflective testing involves the extra costs of additional
tests and personnel time. However, adding tests to an existing order is usually cheaper than a second blood sampling.
Cost savings could also be anticipated due to faster determination of a diagnosis, or by the prevention of making a wrong
diagnosis or performing unnecessary tests. Such cost-benefit
analysis is complicated, and the cost effectiveness of reflective testing has not yet been determined.
Conclusions
Reflective testing can be seen as a new dimension in the service of the clinical chemistry laboratory to primary health
care. Data show that general practitioners generally appreciate this service. They consider reflective testing as useful
and in more than half of the cases this has resulted in subsequent diagnostic testing, (change of) treatment or referral
to a specialist. Additional research is needed to deliver the
scientific proof of the additional value of reflective testing in
primary care for patient management and to determine the
cost-benefit of the procedure. Formal advice on this matter is
also missing.
Conflict of interest statement
Authors’ conflict of interest disclosure: The authors stated that there
are no conflicts of interest regarding the publication of this article.
Research funding: None declared.
Employment or leadership: None declared.
Honorarium: None declared.
References
1. International Organization for Standardization (ISO) 15189:2007.
Medical laboratories – particular requirements for quality and
competence.
2. Paterson JR, Paterson R. Reflective testing: how useful is the
practice of adding on tests by laboratory clinicians? J Clin Pathol
2004;57:273–5.
3. Verboeket-van de Venne WP, Muyrers HH, Pantus JH, Kleinveld
HA, Oosterhuis WP. Website reflective testing in primary care.
Clin Chem Lab Med 2011;49:S663.
4. Simpson WG, Twomey PJ. Reflective testing. J Clin Pathol
2004;57:239–40.
5. Oosterhuis WP, Keuren JF, Verboeket-van de Venne WP, Soomers
FL, Stoffers HE, Kleinveld HA. Eigen inbreng van het laboratorium – huisartsen positief over ‘reflecterend testen’. Ned Tijdschr
Geneeskd 2009;153:A486.
73
1252
6. Jones BJ, Twomey PJ. Comparison of reflective and reflex testing for hypomagnesaemia in severe hypokalaemia. J Clin Pathol
2009;62:816–9.
7. Srivastava R, Bartlett WA, Kennedy IM, Hiney A, Fletcher
C, Murphy MJ. Reflex and reflective testing: efficiency and
effectiveness of adding on laboratory tests. Ann Clin Biochem
2010;47:223–7.
8. Darby D, Kelly AM. Reflective testing – what do our service
users think? Ann Clin Biochem 2006;43:361–8.
9. Oosterhuis WP, Kleinveld HA. ‘Reflecterend’ testen: het
laboratorium ondersteunt de huisarts actief met professioneel vervolgonderzoek. Ned Tijdschr Klin Chem Labgeneesk
2007;32:266–7.
10. Verboeket-van de Venne WP, Oosterhuis WP, Keuren JF,
Kleinveld HA. Reflective testing in the Netherlands: usefulness
to improve the diagnostic and therapeutic process in general
practice. Ann Clin Biochem 2009;46:346–7.
11. Verboeket-van de Venne WP, Oosterhuis WP, Waard H de, Sant P
van ‘t, Kleinveld HA. Reflective testing has a favorable effect on
assessing case reports by general practitioners. Clin Chem Lab
Med 2011;49:S663.
12. Paterson SG, Robson JE, McMahon MJ, Baxter G, Murphy MJ,
Paterson JR. Reflective testing: what do patients think? Ann Clin
Biochem 2006;43:369–71.
13. Barlow IM. Are biochemistry interpretative comments helpful?
Results of a general practitioner and nurse practitioner survey.
Ann Clin Biochem 2008;45:88–90.
14. Barlow IM. Do interpretative comments influence patient
management and do our users approve of the laboratory
‘adding on’ requests? A follow-up General Practitioner and Nurse
Practitioner survey. Ann Clin Biochem 2009;46:85–6.
15. Laposata ME, Laposata M, Cott EM van, Buchner DS, Kashalo
MS, Dighe AS. Physician survey of a laboratory medicine
interpretive service and evaluation of the influence of interpretations on laboratory test ordering. Arch Pathol Lab Med
2004;128:1424–7.
16. St. Peter RF, Reed MC, Kemper P, Blumenthal D. Changes in the
scope of care provided by primary care physicians. N Engl J Med
1999;341:1980–5.
17. Verboeket-van de Venne WP, Oosterhuis WP, Keuren JF,
Kleinveld HA. Effectiveness of reflective testing in primary care
– a randomised clinical trial. Oral communication at the First
European Joint Congress of EFCC and UEMS – Laboratory
Medicine in Health Care.
18. Oosterhuis WP. Reflective testing – is there evidence that it is
worthwhile and for which clinical problems? Clin Chem Lab
Med 2011;49:S110.
19. Prinsloo PJ, Gray TA. A survey of laboratory practice in the
clinical authorization and reporting of results. Ann Clin Biochem
2003;40:149–55.
20. Le Roux CW, Bloom SR. Clinical authorization: what is best for
the patient? Ann Clin Biochem 2003;40:113–4.
21. Challand GS, Vasikaran SD. The assessment of interpretation
in clinical biochemistry: a personal view. Ann Clin Biochem
2007;44:101–5.
22. Li P, Challand GS. Experience with assessing the quality of
comments on clinical biochemistry reports. Ann Clin Biochem
1999;36:759–65.
74
INTERPRETEREN EN BECOMMENTARIËREN VAN
ANEMIE UITSLAGEN
MPG Leers, laboratoriumspecialist klinische chemie
Er is sprake van een anemie wanneer de Hb-concentratie van het bloed lager is dan
de
ondergrens
van
de
referentiewaarde.
Anemie
is
per
definitie
een
laboratoriumdiagnose en geen aandoening. Het kan een symptoom van een
onderliggende ziekte zijn en is een veel voorkomende bevinding in de
huisartspraktijk. Voor een goede behandeling van de anemie is het vaststellen van
de oorzaak uiteraard ontzettend belangrijk. Hierin speelt het laboratorium een
belangrijke rol doordat deze aan de hand van een protocol automatisch aanvullende
testen kunnen uitvoeren ter opsporing van de oorzaak. Het aanbieden van zo’n
protocol bevordert een snellere diagnostiek en minder belasting voor de patiënt.
Dientengevolge zullen patiënten bij wie anemie is geconstateerd adequater
behandeld kunnen worden als de oorzaak van de anemie bekend is. Dit protocol kan
een afgeleide zijn van de NHG-standaard Anemie (2003, wordt op dit moment
herzien)(1) of een zelf ontwikkeld schema. Het NHG-protocol blijkt in de praktijk niet
eenvoudig te implementeren in de laboratoriumpraktijk. Bovendien was een groot
aantal patiënten volgens de NHG-standaard niet te classificeren. Uit een door ons
enkele jaren geleden gehouden enquête onder ziekenhuislaboratoria in Nederland
blijkt dat 64% van de laboratoria de mogelijkheid biedt tot het aanvragen van een
anemieprotocol. Een analyse van de protocollen liet een grote diversiteit aan
diagnostische flowschema’s zien, variërend van een paar bepalingen tot een
uitgebreide analyse van alle mogelijke oorzaken van anemie. Bij 27% van deze
laboratoria met een anemieprotocol voor de eerste lijn werden de resultaten voorzien
van een interpretatief commentaar (2).
In 2007 werd in het Atrium MC Parkstad het anemieprotocol geïmplementeerd zoals
beschreven door Oosterhuis et al. (3). Dit inhoudelijk en logistiek alternatief
stroomschema laat toe dat reflexdiagnostiek met betrekking tot anemie eenvoudiger
implementeerbaar en uitvoerbaar is voor het laboratorium. Bovendien is het mogelijk
om met het voorgestelde stroomschema meervoudige diagnosen te stellen bij een
patiënt, dit in tegenstelling tot het algoritme van de NHG-standaard Anemie. Het
beoogde effect van reflexdiagnostiek bij anemie is een snellere diagnostiek en
75
minder belasting voor de patiënt, aangezien met één bloedafname de mogelijke
oorzaak
van
de
anemie
duidelijk
kan
worden.
Daarnaast
worden
de
laboratoriumresultaten geïnterpreteerd door een laboratoriumspecialist Klinische
Chemie en van een commentaar voorzien. Sinds de implementatie wordt er continu
gewerkt aan het standaardiseren en optimaliseren van het protocol, om zodoende
het aantal gevallen waarin geen oorzaak van het ontstaan van de anemie gevonden
wordt te reduceren tot een minimum (4,5): bij het NHG protocol is dit percentage
52%, bij het alternatief stroomschema zoals destijds voorgesteld door Oosterhuis et
al. 29%, en in de huidige situatie na optimalisatie is dit percentage 14% (3,6).
De komst van nieuwe(re) testen zoals bijvoorbeeld de hemoglobineconcentratie van
de reticulocyten (Ret-He (6) of Chr), RDW, soluble transferrin receptor (sTfR)(6),
methylmalonzuur, holocobalamine, maar ook afgeleiden hiervan zoals de Q-index
(MCV/erytrocytenconcentratie)(7), de sTfR/log ferritine (6) of de transferrine/log
ferritine (8) hebben geleid tot een scala aan parameters die gebruikt kunnen worden
bij de interpretatie van de laboratoriumuitslagen behorend bij de anemiediagnostiek.
Daar enerzijds anemie het gevolg kan zijn van een grote verscheidenheid aan
aandoeningen en anderzijds de huisarts te weinig ervaring heeft om met name al
deze laboratoriumspecifieke parameters goed te kunnen beoordelen, is hier een
belangrijke rol weg gelegd voor de laboratoriumspecialist Klinische Chemie. Door de
laboratoriumuitslagen te beoordelen, te interpreteren, eventueel aanvullende testen
toe te voegen en ten slotte het rapport te voorzien van een interpretatief
commentaar, kan hij/zij de huisarts ondersteunen in het vaststellen van de oorzaak
van de anemie. Daarnaast kan de laboratoriumspecialist overleggen met de huisarts
voor aanvullende gespecialiseerde niet-alledaagse diagnostiek (hemoglobinopathie,
aangeboren
afwijkingen
(enzymafwijkingen,
erytrocytenmembraanafwijkingen,
ijzerinbouwstoornis etc.)).
Referenties
1. Wijk MAM van, Mel M, Muller PA, Silverentand WGJ, Pijnenborg L, Kolnaar
BGM. NHG-Standaard Anemie (M76). Huisarts Wet 2003; 46: 21-29. Rectificatie
algoritme Huisarts Wet 2003; 46: 147.
2. Verboeket-van de Venne WPHG, Oosterhuis WP, Kleinveld HA, Leers MPG.
Anemieprotocollen voor de eerste lijn in Nederland. Ned Tijdschr Klin Chem
Labgeneesk 2010; 35: 111.
3. Oosterhuis WP, Horst M van der, Dongen K van, Ulenkate HJLM, Volmer M,
Wulkan RW. Prospectieve vergelijking van het stroomschema voor
laboratoriumonderzoek van anemie uit de NHG-standaard ‘Anemie’ met een
76
4.
5.
6.
7.
8.
eigen, inhoudelijk en logistiek alternatief stroomschema. Ned Tijdschr Geneeskd
2007; 151: 2326-2332.
Verboeket-van de Venne WPHG, Keuren JFW, Oosterhuis WP, Leers MPG.
Diagnostische waarde van een beknopt anemieprotocol gebaseerd op tien
laboratoriumparameters. Ned Tijdschr Klin Chem Labgeneesk 2011; 36: 275276.
Verboeket-van de Venne WPHG, Oosterhuis WP, Keuren JFW, Ulenkate HJLM,
Leers MPG. Richtlijn NVKC Reflexdiagnostiek bij anemie.
https://www.nvkc.nl/kwaliteitsborging/documents/richtlijn_anemie_def.pdf.
Leers MPG, Keuren JFW, Oosterhuis WP. The value of the Thomas-plot in the
diagnostic work-up of anemic patients referred by general practitioners. Int J Lab
Hematol 2010; 32: 572-581.
Verboeket-van de Venne WPHG, Oosterhuis WP, Leers MPG, Kleinveld HA.
Hemoglobinopathiediagnostiek: de toegevoegde waarde van ‘reflecterend testen’
door laboratoriumspecialisten. Ned Tijdschr Klin Chem Labgeneesk 2010; 35:
211-213.
Castel R, Tax MGHM, Droogendijk J, Leers MPG, Beukers R, Levin M-D,
Sonneveld P, Berendes P. The transferrin/log (ferritin) ratio: a new tool for the
diagnosis of iron deficiency anemia. Clin Chem Lab Med 2012; 50: 1343-1349.
77
Clin Chem Lab Med 2012;50(8):1343–1349 © 2012 by Walter de Gruyter • Berlin • Boston. DOI 10.1515/cclm-2011-0594
The transferrin/log(ferritin) ratio: a new tool for the diagnosis
of iron deficiency anemia
Rob Castel1,*, Martine G.H.M. Tax1,
Jolanda Droogendijk2, Math P.G. Leers3,
Ruud Beukers4, Mark-David Levin5,
Pieter Sonneveld6 and Paul B. Berendes1
1
Department of Clinical Chemistry and Hematology, Albert
Schweitzer Hospital, Dordrecht, The Netherlands
2
Department of Internal Medicine, Albert Schweitzer
Hospital, Dordrecht, The Netherlands
3
Department of Clinical Chemistry and Hematology, Atrium
Medical Centre Parkstad, Heerlen, The Netherlands
4
Department of Gastroenterology, Albert Schweitzer
Hospital, Dordrecht, The Netherlands
5
Department of Hematology, Albert Schweitzer Hospital,
Dordrecht, The Netherlands
6
Department of Hematology, Erasmus Medical Centre,
Rotterdam, The Netherlands
Abstract
Background: Serum ferritin is the best single laboratory test
to diagnose iron deficiency anemia (IDA). Ferritin levels < 20
μg/L are highly specific for IDA, and ferritin levels > 100
μg/L usually exclude IDA. However, ferritin concentrations
between 20 and 100 μg/L are often inconclusive. The objective of this study was to improve the diagnosis of IDA when
ferritin levels are inconclusive.
Methods: We evaluated the predictive performance of
classic (ferritin, mean corpuscular volume, transferrin and
serum iron) and modern [reticulocyte hemoglobin content,
serum transferrin receptor and soluble transferrin receptor
(sTfR)/log(ferr)] iron status parameters to diagnose IDA in
2084 anemic, non-hospitalized patients. The results were
validated in an independent cohort of 274 anemic patients.
Results: In our study population, 29% (595 patients) of the
patients had a ferritin level between 20 and 100 μg/L, hampering diagnosis of IDA. None of the classic or modern
parameters was capable of completely separating the IDA
population from the non-IDA population. However, using a
new parameter, the transferrin/log(ferritin) ratio, the IDA and
non-IDA populations can be completely separated. At a cut-off
value of 1.70, the transferrin/log(ferritin) ratio indicates IDA
in 29% of the patients with inconclusive ferritin levels.
*Corresponding author: Rob Castel, Department of Clinical
Chemistry, Room F-3127, Albert Schweitzer Hospital, Albert
Schweitzerplaats 25, 3318 AT, Dordrecht, The Netherlands
Phone: +31-786523548, Fax: +31-786523156,
Received August 30, 2011; accepted January 15, 2012;
previously published online February 11, 2012
Conclusions: The transferrin/log(ferritin) ratio is a practical
new tool that improves diagnosis of iron deficiency when ferritin levels are inconclusive.
Keywords: anemia analysis; iron deficiency anemia; iron
status parameters.
Introduction
Iron deficiency is a leading cause of anemia (1, 2). Iron deficiency anemia (IDA) results from inadequate iron intake
and/or absorption, or from iron loss caused by bleeding. In
men and postmenopausal women, IDA is commonly due to
blood loss from the gastrointestinal tract, which is caused
by gastrointestinal tract malignancy in 10%–15% of IDA
patients (3, 4). As IDA can be a sign of malignant disease,
it is important to accurately assess a patient’s iron status in
case of anemia. However, despite a wide range of available
laboratory tests, diagnosing IDA often proves to be challenging (1, 5, 6).
Prussian blue staining on aspirated bone marrow particles
is still widely considered as the gold standard for diagnosing
iron deficiency. However, besides its shown limitations (7–9),
it is not quite feasible to perform bone marrow examination
for each anemic patient. Therefore, bone marrow aspirates
have largely been replaced by blood tests to assess a patient’s
iron status. The serum ferritin concentration currently is the
most useful indicator of iron deficiency (10–12). A low serum
ferritin concentration ( < 20 μg/L) confirms iron deficiency,
however, at such low ferritin levels sensitivity is poor
(59%–73%). Being a positive acute phase reactant, ferritin
levels as high as 100 μg/L can occur in iron deficient patients.
As specificity for iron deficiency falls sharply with increasing
ferritin levels, ferritin levels between 20 and 100 μg/L are
often inconclusive (13, 14). Serum ferritin levels > 100 μg/L
usually exclude iron deficiency (15).
Since the introduction of ferritin as a clinical test in the
1970s, several different markers have been proposed to
improve diagnosis of iron deficiency at ferritin levels between
20 and 100 μg/L [e.g., soluble transferrin receptor, soluble
transferrin receptor/log(ferritin) ratio, reticulocyte hemoglobin content and hepcidin]. While most modern indicators of
iron status readily identify uncomplicated iron deficiency,
none is sufficient by itself for the diagnosis of iron deficiency in more complex clinical settings. Modern assays have
not been standardized, and/or are not yet suited for routine
laboratory analysis. Therefore, their clinical significance
and usefulness as an aid in the diagnosis of IDA are limited
(1, 16–22).
Bro
ought to you by | Attrium MC Medische Bibliotheek
78
1344
Castel et al.: A new tool for the diagnosis of iron deficiency anemia
In the present study, we evaluate the efficiency of both
classical and modern laboratory parameters in the diagnosis of IDA in a large population of 2084 consecutive,
non-hospitalized anemic patients presented by the general
practitioner. During our analysis we identified a new practical laboratory parameter for diagnosis of IDA: the transferrin/
log(ferritin) ratio [Tf/log(ferr) ratio]. Our data show that the
Tf/log(ferr) ratio can be used as a practical routine laboratory
test that improves diagnosis of IDA when ferritin levels are
inconclusive.
Materials and methods
Patient data from the Project of Anemia analysis from
the General practitioner to the Albert Schweitzer
hospital (PAGAS)
Over a period of 4 years (2007–2010), we collected data from
2084 consecutive anemic patients (men: aged 18 years and
older; women: aged 50 years and older, to exclude IDA caused
by menstrual blood loss) presented by the general practitioner as
part of the research project PAGAS (Project of Anemia Analysis
from the General Practitioner to the Albert Schweitzer Hospital)
(23). Hemoglobin concentration, reticulocyte hemoglobin content (RetHe) and mean corpuscular volume (MCV) were measured using a Sysmex XE-2100 automated hematology analyzer
(Sysmex Corporation, Kobe, Japan). Ferritin was measured by a
sandwich chemiluminescent immuno-assay on an Immulite 2500
analyzer (Siemens Healthcare Diagnostics, Deerfield, IL, USA),
with a limit of quantitation of 2.0 μg/L. Serum concentrations of
transferrin, soluble transferrin receptor (sTfR), C-reactive protein
(CRP) and iron were measured on an Olympus AU2700 or AU640
analyzer (Olympus Life and Material Science Europe GmbH,
Hamburg, Germany). The number of patients that we included in
the analysis of the RetHe (254 patients), sTfR (74 patients) and
sTfR/log(ferritin) ratio (74 patients) varied depending on when we
had started measuring that parameter in our laboratory.
Data analysis
Data analysis was performed in Microsoft Excel and GraphPad Prism
software. Normality was tested using the d’Agostino and Pearson
omnibus normality test. Statistical significance was determined by
the non-parametric Mann-Whitney test at p-values lower than 0.01.
Results
Frequency distribution of ferritin concentrations
Of the 2084 anemic patients included in our study, 445
patients (21%) had a ferritin level of < 20 μg/L (iron-deficient), 595 patients (29%) had ferritin levels in the range of
20–100 μg/L (inconclusive), and 1044 patients (50%) had a
ferritin level > 100 μg/L (non-iron-deficient).
Figure 1 shows the frequency distribution of ferritin concentrations. There are two frequency peaks; a minor peak at
the ferritin interval of 5–12 μg/L that corresponds to the group
of iron-deficient patients, and a major peak at a ferritin level
of 148 μg/L that corresponds to the group of non-iron-deficient patients. Clearly, there is considerable overlap between
the iron-deficient and non-iron-deficient patient groups at
ferritin levels of 20–100 μg/L. This was to be expected, as
20–100 μg/L is the range at which ferritin concentrations are
often inconclusive.
Identifying iron deficiency with classical iron status
parameters
As Figure 1 suggests the presence of two bell-shaped populations, we set out to find a parameter that would be able to
effectively distinguish these two patient groups. We first evaluated the most relevant classical and modern iron status parameters. To determine how well the classical iron parameters
500
Patient data from the validation cohort
400
Number of patients
350
300
250
200
150
100
50
8
52
14
-9
43
41
18
01
-4
18
3-
78
3
01
3
0
78
34
0-
34
48
814
4
-1
64
8
-6
-2
Definitions of anemia and ferritin cut-off values
28
12
5
2-
5-
12
0
<2
To test the transferrin/log(ferritin) ratio in an unrelated population, we used data of 274 consecutive anemic patients (aged
18 years and older) presented by the general practitioner to the
Atrium Medical Center (Heerlen, the Netherlands) (24). Ferritin
was measured on an Advia Centaur Immunochemistry Analyzer
(Siemens), with a limit of quantitation of 1.5 μg/L. Transferrin
was measured in heparin plasma on a Roche Modular system
(P-module; Roche diagnostics, Hoffmann-La Roche, Inc., Basel,
Switzerland). Hemoglobin concentrations were measured using
a Sysmex XE-2100 automated hematology analyzer (Sysmex
Corporation, Kobe, Japan).
450
Serum ferritin concentration, μg/L
The definition of anemia used throughout this study is a hemoglobin concentration < 130 g/L for male patients and < 120 g/L
for female patients (25). Patients with a ferritin level < 20 μg/L
were judged as iron-deficient; patients with ferritin > 100 μg/L as
non-iron-deficient. Patients with ferritin between 20 and 100 μg/L
we considered as patients in whom iron-deficiency could not be
excluded.
Figure 1 Ferritin frequency distribution.
Frequency distribution of ferritin concentrations in 2084 anemic
patients presented by the general practitioner. Note the frequency
peaks at a ferritin concentration of 5–12 μg/L and at a ferritin concentration of 148 μg/L, which indicate there are two overlapping
populations.
Bro
79
A
4.63
130
107
49.3
56.1
3.77
3.31
92
81
61
2
Serum iron, μmol/L
Diagnosing iron deficiency anemia with modern
indicators of iron status
Transferrin, g/L
30.9
102.7
2.2
1.5
14.5
6.4
4.4
6.4
1.2
3.3
0.9
RetHe, fmol
4.6
1.4
sTfR, mg/L
These data show that all three traditional laboratory tests
distinguish poorly between the patient groups with a ferritin
level < 20 μg/L and > 100 μg/L, as there is large overlap in the
measured data ranges of iron-deficient and non-iron deficientpatients.
1.86
0.75
2.6
2.0
2.04
4.8
2
56
MCV, fL
B
9.2
1345
4.1
1.2
0.4
sTfR/log(ferritin)
Figure 2 Diagnosing iron deficiency anemia with classical and
modern iron status parameters.
(A) The classical iron status parameters MCV, serum iron and transferrin. (B) The modern iron status parameters RetHe, sTfR and
sTfR/log(ferritin). Gray bars correspond to the group of patients with
a ferritin level > 100 μg/L (non-iron-deficient); white bars correspond
to the group of patients with a ferritin level < 20 μg/L (iron-deficient).
The lowest and highest measured values are shown below and above
the bars, respectively. The median values are shown above the thick
black lines.
MCV (mean corpuscular volume), serum iron and transferrin
would separate the iron-deficient and non-iron-deficient patient
groups, we divided our population into two groups: (1) patients
with a ferritin level <20 μg/L (iron-deficient), and (2) patients
with a ferritin level >100 μg/L (non-iron-deficient). Figure 2A
shows the median and extreme values for each parameter measured in both patient groups. As expected, the median MCV of
the iron-deficient patient group is lower (81 fL) compared to
the median MCV of the non-iron-deficient patient group (92
fL) (p<0.01). However, there is considerable overlap between
both groups: 98% of all measured MCVs fall into the overlapping interval of 61–107 fL.
Similarly, the median serum iron concentration is lower
in the iron-deficient patient group (4.8 μmol/L) than in the
non-iron-deficient patient group (9.2 μmol/L) (p < 0.01).
However, the overlap is such that in both the iron-deficient
and non-iron-deficient patient groups serum iron concentrations were measured in the range of 2–49.3 μmol/L, which
includes 99.9% of the data.
The median transferrin concentration of 3.31 g/L in the
iron-deficient patient group is expectedly higher compared
with the median transferrin concentration in the non-irondeficient patient group (2.04 g/L) (p < 0.01). However, 73%
of the iron-deficient and non-iron-deficient patients have a
transferrin level in the range of 1.86–3.77 g/L, the overlapping interval between the two groups.
Next, we tested the ability of three modern iron parameters
to discriminate patients with ferritin < 20 μg/L (iron-deficient) from patients with ferritin > 100 μg/L (non-iron-deficient): the reticulocyte hemoglobin content (RetHe), the
soluble transferrin receptor concentration (sTfR) and the
sTfR/log(ferritin) ratio (Figure 2B). The RetHe is expected
to be lower in patients with iron-deficiency than in non-irondeficient patients, whereas both the sTfR and sTfR/log(ferritin)
ratio are higher in patients with iron-deficiency than in
non-iron-deficient patients.
The median RetHe was 2.0 fmol in the non-iron-deficient
patient group and 1.5 fmol in the iron-deficient patient group
(p < 0.01), with overlap between 1.2 and 2.2 fmol, which
includes 90% of the data. The sTfR median was 6.4 mg/L in
the iron-deficient group and 3.3 mg/L in the non-iron-deficient
group (p < 0.01), with 27% of the data falling into the overlapping interval of 4.6–14.5 mg/L. Finally, the median
sTfR/log(ferritin) ratio in the in the iron-deficient group
was 6.4, compared with 1.2 in the non-iron-deficient group
(p < 0.01), indicating an overlap between 4.1 and 4.4, which
comprised only 4% of the data.
To summarize, the modern parameters sTfR and
sTfR/log(ferritin) ratio perform better than the classical iron
parameters, as judged by the much smaller overlap between
the patient groups with a ferritin level < 20 μg/L and
> 100 μg/L. However, from the data presented in Figure
2B it is apparent that, similar to the classical parameters,
none of the modern parameters is able to completely separate patients with ferritin < 20 μg/L (iron-deficient) from
patients with ferritin > 100 μg/L (non-iron-deficient) in the
population that we analyzed.
A new parameter for the diagnosis of iron deficiency
anemia: the transferrin/log(ferritin) ratio
Neither the classical nor the modern iron parameters that we
tested were able to fully distinguish patients with ferritin < 20
μg/L (iron-deficient) from patients with ferritin > 100 μg/L
(non-iron-deficient). As we aimed at defining a practical
parameter that could be routinely measured in most laboratories, we combined the classical iron status parameters
with ferritin. We found that the transferrin (g/L)/log[ferritin
(μg/L)] ratio [Tf/log(ferr)] is able to completely separate
patients with ferritin < 20 μg/L (iron-deficient) from patients
with ferritin > 100 μg/L (non-iron-deficient) (Figure 3A).
The highest Tf/log(ferr) ratio among the non-iron-deficient
group is 1.70 (median: 0.84), while the lowest Tf/log(ferr)
ratio among the iron-deficient patients is 1.74 (median: 3.49)
(p < 0.01).
Bro
80
1346
A
B
500
450
14.62
Number of patients
400
3.49
300
250
200
150
100
1.70
50
1.74
0.23
8
3
52
-9
43
41
18
01
-4
18
3-
78
14
01
3
0
78
34
0-
8-
34
14
4
48
-1
64
8
-6
28
5
12
-2
12
Tf/log(ferr)
ID+NID
2-
<2
0
5-
0.84
350
C
100
90
80
Sensitivity, %
70
60
50
Transferrin/log(ferritin) ratio
40
Transferrin
MCV
30
SeFe
No discrimination
20
10
0
0
10
20
30
40
50
60
70
80
90
100
100%-Specificity, %
D
100
90
80
Sensitivity, %
70
60
50
sTfR/log(ferritin) ratio
40
sTfR
30
RetHe
No discrimination
20
10
0
0
10
20
30
40
50
60
100%-Specificity, %
70
80
90
100
Figure 3 Diagnosing iron deficiency anemia with a new indicator of iron status: the transferrin/log(ferritin) ratio.
(A) Gray bar corresponds to the group of patients with a ferritin level > 100 μg/L (non-iron-deficient); white bar corresponds to the group of
patients with a ferritin level < 20 μg/L (iron-deficient). The lowest and highest measured values are shown below and above the bars, respectively. ID, iron-deficient; NID, non-iron-deficient. The median values are shown above the thick black lines. (B) Separating populations with
overlapping ferritin levels based on a cut-off value of 1.70 for the transferrin/log(ferritin) ratio. The dotted line is the same plot that was shown
in Figure 1. Using a cut-off value of 1.70 for the transferrin/log(ferritin) ratio, two unskewed, bell-shaped populations can be distinguished
(the continuous curves). Population with transferrin/log(ferritin) ratio > 1.70:skewness:–0.04 and kurtosis:–0.55. Population with transferrin/
log(ferritin) ratio ≤ 1.70:skewness 0.30 and kurtosis:–0.10. (C) Receiver operating characteristic (ROC)-curve of the transferrin/log(ferritin)
ratio [area under the curve (AUC) = 1.00], transferrin (AUC = 0.98), MCV (AUC = 0.85) and serum iron (AUC = 0.73). (D) Receiver operating
characteristic (ROC)-curve of the sTfR/log(ferritin) ratio (AUC = 0.99), sTfR (AUC = 0.94) and RetHe (AUC = 0.87).
Bro
81
1347
74 ± 14 (50 –100)
102 ± 16.1 (48.3 –118)
82 ± 8.8 (56 –104)
16.1 ± 2.0 (12.4 –22.9)
1.2 ± 0.7 (0.3 – 8.9)
15 ± 13 ( < 2 –90)
3.27 ± 0.49 (1.86 – 4.63)
82 ± 12 (47–116)
6.5 ± 4.8 ( < 2 –56.1)
9 ± 7 (2– 81)
339 ± 180 ( < 111 to > 738)
22 ± 13 (3 to > 54)
300 ± 122 ( < 5 –1099)
8.0 ± 11.8 (2.5 –221.2)
9 ± 13 ( < 5 –166)
78 ± 12 (50 –100)
117 ± 12.3 (45.1–129)
92 ± 7.3 (61–124)
14.8 ± 2.0 (11.6 –27.1)
1.3 ± 1.1 (0.2 –21.3)
343 ± 419 (21–3826)
2.08 ± 0.42 (0.75 –3.49)
52 ± 10 (22 – 88)
10.4 ± 6.3 ( < 2 – 42.8)
20 ± 13 (3 –93)
350 ± 193 ( < 111 to > 738)
19 ± 12 ( < 2 to > 54)
275 ± 132 ( < 5 –1357)
9.2 ± 9.9 (1.4 –201)
41 ± 58 ( < 5 –371)
109 ± 9.24 (51.6 –118)
91 ± 7.2 (61–130)
14.8 ± 2.2 (12.0 –27.6)
1.3 ± 1.1 (0.2 –20.9)
258 ± 433 (21–7225)
2.14 ± 0.4 ( < 0.75 –3.77)
54 ± 10 (21–95)
9.8 ± 6.0 ( < 2 – 49.3)
19 ± 12 (3 –92)
376 ± 200 ( < 111 to > 738)
23 ± 14 ( < 2 to > 54)
306 ± 131 ( < 5 –1323)
8.1 ± 5.7 (0.8 –118)
33 ± 53 ( < 5 –341)
69 ± 15 (22 –96)
109 ± 18.7 (51.6 –129)
83 ± 8.9 (56 –107)
16.0 ± 2.3 (12.5 –26.2)
1.2 ± 0.8 (0.4 – 6.8)
17 ± 13 ( < 2 –70)
3.19 ± 0.42 (2.10 – 4.62)
80 ± 11 (53 –116)
7.3 ± 5.5 ( < 2– 44.5)
10 ± 7 (2 –58)
305 ± 167 ( < 111 to > 738)
21 ± 11 (5 to > 54)
276 ± 116 ( < 5 – 693)
7.2 ± 2.5 (1.7–19.3)
9 ± 12 ( < 5 –123)
73 ± 14 (18 –98)
Age
Hemoglobin, g/L
Mean corposcular volume, fL
Red cell distribution width, %
Reticulocytes, %
Ferritin, μg/L
Transferrin, g/L
Iron binding capacity, μmol/L
Serum iron, μmol/L
Transferrin saturation, %
Vitamin B12, pmol/L
Folic acid, nmol/L
Platelets, 10E9/L
Leukocytes, 10E9/L
C-reactive protein, mg/L
Age and laboratory parameters of the whole population divided into male/female and a transferrin/log(ferritin) ratio ≤ 1.70/ > 1.70. Data are shown as: mean ± standard deviation (range).
121–161
80 –100
11–16
< 2.5
20 –150
2.0 –3.6
45 – 80
10 –25
2–50
130 –700
>5
150–400
4.3 –10.0
< 10
137–177
80 –100
11–16
< 2.5
25 –250
2.0 –3.6
45 – 80
14 –28
20 – 60
130 –700
>5
150–400
4.3 –10.0
< 10
Female
Male
Female
Female
Male
Sex
Male
> 1.70
≤ 1.70
> 1.70
≤ 1.70
Transferrin/log(ferritin) ratio
Table 1 Population characteristics.
Intervals of reference
We calculated the Tf/log(ferr) ratio for all of the 2084
patients in our study, the 595 patients with a ferritin level
between 20 and 100 μg/L included. For highest sensitivity, we set the cut-off value for IDA at a Tf/log(ferr) ratio
of > 1.70, thereby creating two bell-shaped populations in
the ferritin concentration frequency distribution (Figure 3B).
One population includes all patients with a Tf/log(ferr) ratio
smaller or equal to 1.70 (skewness: 0.30; kurtosis: –0.10), and
the other population includes all patients with a Tf/log(ferr)
ratio > 1.70 (skewness: –0.04; kurtosis: –0.55). Population
characteristics are shown in Table 1. Receiver operating characteristic (ROC) curves are show in Figures 3C and 3D; transferrin/log(ferritin) ratio [area under the curve (AUC) = 1.00],
transferrin (AUC = 0.98), MCV (AUC = 0.85), serum iron
(AUC = 0.73), sTfR/log(ferritin) ratio (AUC = 0.99), sTfR
(AUC = 0.94) and RetHe (AUC = 0.87).
Of the 595 patients with a ferritin concentration between
20 and 100 μg/L (inconclusive), 174 (29%) had a Tf/log(ferr)
ratio of > 1.70. Thus, at a cut-off value of 1.70, the Tf/log(ferr)
ratio indicates that 29% of the patients with an equivocal ferritin level are iron-deficient. At a cut-off level of 1.74, for
highest specificity, IDA would be indicated in 162 patients
(27%). We found the same cut-off value of 1.70 for the
Tf/log(ferr) ratio when using different cut-off values to define
anemia (men: Hb < 137 g/L; women: Hb < 121 g/L) and iron
deficiency (ferritin < 15 μg/L) (data not shown).
Validation of the Tf/log(ferr) ratio in an unrelated
population of anemic patients
We validated the Tf/log(ferr) ratio in an unrelated population
of 274 anemic patients (24) (see “Materials and methods”
for details). In this population, 97 patients (35%) had a ferritin level < 20 μg/L (iron-deficient), 87 patients (32%) had
a ferritin level between 20 and 100 μg/L (inconclusive), and
87 patients (32%) had a ferritin level > 100 μg/L (non-irondeficient). The highest Tf/log(ferr) ratio in the non-iron-deficient patient group was 1.61, and the lowest Tf/log(ferr) ratio
among iron-deficient patients was 1.89, which are similar to
the ratios that we found in the original population: 1.70 and
1.74, respectively (Figure 3A). ROC-curve analysis showed
an AUC of 1.00 (Figure not shown; compare Figures 3C and
3D).
At the cut-off value of 1.70, we confirmed the two
bell-shaped populations in the ferritin frequency distribution
curve of this unrelated cohort of anemic patients (Figure 4).
One population includes all patients with a Tf/log(ferr) ratio
smaller or equal to 1.70 (skewness: 0.17; kurtosis: –0.41), and
the other population includes all patients with a Tf/log(ferr)
ratio > 1.70 (skewness: –0.29; kurtosis: –0.75).
Of the 87 patients with inconclusive ferritin levels (20–100
μg/L), 36 patients (41%) had a Tf/log(ferr) ratio of > 1.70,
suggesting IDA.
The Tf/log(ferr) ratio and acute phase response
Of the 2084 patients included in our study, we obtained CRP
(C-reactive protein) measurements for 2031 patients: 338
Bro
82
1348
Number of patients
60
40
20
14
-4
18
01
18
378
3
01
3
78
0
034
14
8-
34
48
4
-6
-1
64
-2
8
28
12
12
5-
5
2-
<2
0
Figure 4 Analysis of the transferrin/log(ferritin) ratio in an unrelated anemic patient population.
The dotted black curve is the frequency distribution of ferritin concentrations in 274 anemic patients. The continuous curves are the
two populations can be distinguished based on a cut-off value of
1.70 for the transferrin/log(ferritin) ratio. Note that both curves are
unskewed, and bell-shaped. Population with transferrin/log(ferritin)
ratio > 1.70:skewness:0.17 and kurtosis:–0.41. Population with transferrin/log(ferritin) ratio ≤ 1.70:skewness –0.29 and kurtosis:–0.75.
Compare to Figures 1 and 3B.
(17%) had a CRP level of > 50 mg/L, 778 (38%) patients
had a CRP > 10 mg/L, and 1223 (60%) patients had a CRP
level < 10 mg/L, which in our regional reference population is
considered as no elevation.
As transferrin is a negative acute phase protein and ferritin a positive acute phase protein, the Tf/log(ferr) ratio is
reduced during an acute phase response. Therefore, anemia of chronic disease (ACD) (26) can be distinguished
from IDA by the Tf/log(ferr) ratio. However, acute phase
response may potentially obscure IDA when both conditions occur simultaneously. Nevertheless, in 84 of the
778 patients (11%) with a CRP level of > 10 mg/L, the
Tf/log(ferr) ratio was > 1.70, indicating IDA. These 84
patients with elevated CRP levels included 10 patients (12%)
in whom the CRP concentration was over 50 mg/L. This
shows that the Tf/log(ferr) ratio can identify IDA in accompanying ACD.
Discussion
Accurate diagnosis of IDA is of great importance, as IDA
may be caused by life-threatening disease, such as a gastrointestinal tract malignancy (1–4). Whereas anemia is
straightforwardly determined by measuring the hemoglobin
concentration in blood, there are many laboratory parameters
available to assess a patient’s iron status, all with different
test-specific characteristics and analytical challenges. The
experienced medical specialist would be able to accurately
interpret the combined iron status parameters in specific
clinical cases, however, determining iron deficiency can
often be a challenge for the non-expert clinician. Therefore,
there is a need for a simple yet reliable blood test that can
identify IDA. In clinical practice, a serum ferritin level < 20
μg/L confirms iron deficiency, while iron deficiency is usually ruled out at ferritin levels > 100 μg/L. Ferritin levels
that fall into the gray area between 20 and 100 μg/L are
often inconclusive, as specificity for iron deficiency falls
sharply with increasing ferritin levels (1, 4, 12–14).
In this study we analyzed a large, non-hospitalized population consisting of 2084 anemic patients. Based on the
serum ferritin concentration, we found that 445 patients
(21%) were iron-deficient (ferritin < 20 μg/L), and 1043
patients (50%) were non-iron-deficient (ferritin > 100
μg/L). As many as 29% of all the patients (596 patients)
had a serum ferritin level in the inconclusive range from 20
to 100 μg/L. The frequency distribution curve of the ferritin
concentrations (Figure 1) suggests the presence of two bellshaped populations that largely overlap: the iron-deficient
and non-iron-deficient patient groups. Being able to separate
these overlapping patient populations allows a more accurate and earlier diagnosis of IDA when ferritin levels are
inconclusive.
Our data show that there is considerable overlap in the
measurements of MCV, serum iron, and transferrin between
the patients with ferritin < 20 μg/L (iron-deficient) and
> 100 μg/L (non-iron-deficient) (Figure 2A). The modern
iron status parameters RetHe (reticulocyte hemoglobin
content), sTfR (soluble transferrin receptor) and the sTfR/
log(ferr) ratio performed better than the classical iron parameters, as there is considerably less overlap between the
patients with ferritin < 20 μg/L and > 100 μg/L (Figure 2B).
However, even based on the sTfR/log(ferr) ratio, which has
been suggested as the new gold standard (27), the patients with
ferritin < 20 μg/L and > 100 μg/L could not be completely
separated.
To completely discriminate the iron-deficient from noniron-deficient patient group we combined the transferrin
concentration with the logarithm of the ferritin concentration as the transferrin/log(ferritin) ratio [Tf/log(ferr)]. Using
classical iron status parameters has the advantage that the
tests are well standardized, can be routinely measured,
and are familiar to clinicians. As shown in Figure 3A, the
Tf/log(ferr) ratio fully separates patients with ferritin < 20
μg/L from those with ferritin > 100 μg/L. At a Tf/log(ferr)
ratio cut-off value of 1.70 ( ≤ 1.70: non-iron-deficient; > 1.70:
IDA), two unskewed, bell-shaped ferritin frequency distribution curves are recognized (Figure 3B). We validated the
Tf/log(ferr) ratio in an unrelated population of 274 anemic
patients (Figure 4).
Our data indicate that at a cut-off value of 1.70, the
Tf/log(ferr) indicates that a large group (29%) of the patients
with an inconclusive ferritin concentration are in fact irondeficient. Using different cut-off values to define anemia (men:
Hb < 137 g/L; women: Hb < 121 g/L) and iron deficiency (ferritin < 15 μg/L) did not change the Tf/log(ferr) ratio cut-off
value of 1.70 (data not shown). A potential drawback of using
a parameter based on acute phase proteins is that it might not
identify IDA during an acute phase response. However, in 84
Bro
83
of the 778 patients (11%) with an elevated CRP level ( > 10
mg/L) IDA would still be diagnosed based on the Tf/log(ferr)
ratio.
In conclusion, our results show that the Tf/log(ferr) ratio
has the potential to become a widely used, practical tool
for the diagnosis of IDA that may benefit the many anemic
patients worldwide who are at risk of having IDA.
Acknowledgments
The authors would like to thank Marcel R. de Zoete, PhD, Yale
University School of Medicine, New Haven, CT, for critically reading the manuscript.
Conflict of interest statement
Authors’ conflict of interest disclosure: The authors stated that there
are no conflicts of interest regarding the publication of this article.
Authorship: RC and PBB analyzed the data and wrote the paper.
RC, MGHMT, JD, RB, PS, MDL and PBB are involved in the
Project of Anemia analysis from the General practitioner to the
Albert Schweitzer hospital (PAGAS). MPGL provided the dataset
for validation.
References
1. Clark SF. Iron deficiency anemia: diagnosis and management.
Curr Opin Gastroenterol 2009;25:122–8.
2. Clark SF. Iron deficiency anemia. Nutr Clin Pract 2008;23:
128–41.
3. Rockey DC. Occult and obscure gastrointestinal bleeding:
causes and clinical management. Nat Rev Gastroenterol Hepatol
2010;7:265–79.
4. Killip S, Bennett JM, Chambers MD. Iron deficiency anemia. Am
Fam Physician 2007;75:671–8.
5. Van Vranken M. Evaluation of microcytosis. Am Fam Physician
2010;82:1117–22.
6. Logan EC, Yates JM, Stewart RM, Fielding K, Kendrick D.
Investigation and management of iron deficiency anaemia in
general practice: a cluster randomised controlled trial of a simple
management prompt. Postgrad Med J 2002;78:533–7.
7. Ganti AK, Moazzam N, Laroia S, Tendulkar K, Potti A, Mehdi SA.
Predictive value of absent bone marrow iron stores in the clinical
diagnosis of iron deficiency anemia. In Vivo 2003;17:389–92.
8. Barron BA, Hoyer JD, Tefferi A. A bone marrow report of absent
stainable iron is not diagnostic of iron deficiency. Ann Hematol
2001;80:166–9.
1349
9. Thomason RW, Almiski MS. Evidence that stainable bone marrow iron following parenteral iron therapy does not correlate
with serum iron studies and may not represent readily available
storage iron. Am J Clin Pathol 2009;131:580–5.
10. Cook JD. Diagnosis and management of iron-deficiency anaemia. Best Pract Res Clin Haematol 2005;18:319–32.
11. Cook JD, Flowers CH, Skikne BS. The quantitative assessment
of body iron. Blood 2003;101:3359–64.
12. Schrier SL. Causes and diagnosis of anemia due to iron deficiency. 2009. UptoDate (www.uptodate.com), last accessed
November 2011, last updated December 15, 2011.
13. Guyatt GH, Oxman AD, Ali M, Willan A, McIlroy W, Patterson
C. Laboratory diagnosis of iron-deficiency anemia: an overview.
J Gen Intern Med 1992;7:145–3.
14. Nelson R, Chawla M, Connolly P, LaPorte J. Ferritin as an index
of bone marrow iron stores. South Med J 1978;71:1482–4.
15. Munoz M, Garcia-Erce JA, Remacha AF. Disorders of iron
metabolism. Part II: iron deficiency and iron overload. J Clin
Pathol 2011;64:287–96.
16. Wish JB. Assessing iron status: beyond serum ferritin and transferrin saturation. Clin J Am Soc Nephrol 2006;1(Suppl 1):S4–8.
17. Thomas C, Thomas L. Biochemical markers and hematologic
indices in the diagnosis of functional iron deficiency. Clin Chem
2002;48:1066–76.
18. Punnonen K, Irjala K, Rajamaki A. Serum transferrin receptor
and its ratio to serum ferritin in the diagnosis of iron deficiency.
Blood 1997;89:1052–7.
19. Mast AE, Blinder MA, Lu Q, Flax S, Dietzen DJ. Clinical utility
of the reticulocyte hemoglobin content in the diagnosis of iron
deficiency. Blood 2002;99:1489–91.
20. Malyszko J. Hepcidin assays: ironing out some details. Clin J Am
Soc Nephrol 2009;4:1015–6.
21. Kemna EH, Tjalsma H, Willems HL, Swinkels DW. Hepcidin:
from discovery to differential diagnosis. Haematologica
2008;93:90–7.
22. Thorpe SJ, Heath A, Sharp G, Cook J, Ellis R, Worwood M.
A WHO reference reagent for the Serum Transferrin Receptor
(sTfR): international collaborative study to evaluate a recombinant soluble transferrin receptor preparation. Clin Chem Lab
Med 2010;48:815–20.
23. PAGAS (Project of Anemia analysis from the General practitioner
to the Albert Schweitzer hospital). Last accessed on December,
2012. Last updated December 2011.
24. Leers MP, Keuren JF, Oosterhuis WP. The value of the Thomasplot in the diagnostic work up of anemic patients referred by general practitioners. Int J Lab Hematol 2010;32:572–81.
25. Beutler E, Waalen J. The definition of anemia: what is the lower
limit of normal of the blood hemoglobin concentration? Blood
2006;107:1747–50.
26. Weiss G, Goodnough LT. Anemia of chronic disease. N Engl J
Med 2005;352:1011–23.
27. Swinkels DW, Janssen MC, Bergmans J, Marx JJ. Hereditary
hemochromatosis: genetic complexity and new diagnostic
approaches. Clin Chem 2006;52:950–68.
ro
84
STOLLINGSUITSLAGEN
NCV Péquériaux, arts klinische chemie
Bij een kind van 11 maanden wordt door de huisarts een bloedbeeld aangevraagd.
Bij medische gegevens staat er vermeld: ITP. Het trombocytenaantal is 340 x 109/L.
De laboratoriumspecialist belt de huisarts over de bloedingsneiging. Zijn er
petechieën? De waarnemer vertelt dat er hematomen op de ledematen en op de
borstkast aangetroffen zijn, verder is de anamnese, ook familiair, negatief. De
labspecialist adviseert om de volgende dag het kind opnieuw te prikken voor een
oriënterend onderzoek naar bloedingsneiging (APTT, PT, INR, fibrinogeen). Echter
het kind wordt de volgende dag met spoed door de kinderarts gezien i.v.m.
bijgekomen
slijmvliesbloedingen.
Wederom
overlegt
de
kinderarts
met
de
labspecialist over welke stollingsonderzoeken ingezet moeten worden (ook gezien
benodigde hoeveelheid materiaal). Meteen worden APTT, PT, INR en fibrinogeen
bepaald. De APTT is 87 seconden, de overige onderzoeken zijn normaal.
Aanvullende testen, mengproef, factor VIII, factor IX en von Willebrand factor worden
ingezet. Factor VIII is 1 %, overige onderzoeken zijn normaal. De uitslagen worden
teruggekoppeld met de kinderarts en gerapporteerd met het commentaar: “past bij
hemofilie A”. Het kind wordt verwezen naar een hemofiliecentrum.
Het meedenken door het laboratorium bij het aanvragen en het interpreteren van
laboratoriumaanvragen voor hemostase- en tromboseonderzoek is van essentieel
belang en heeft een grote meerwaarde voor de kliniek. Kennis van hemostase en
trombose is bij de huisartsen en bij de diverse specialismen binnen het ziekenhuis
vaak beperkt, met uitzondering van de afdeling interne geneeskunde (in het bijzonder
de internisten-hematologen).
Gericht aanvragen met behulp van voorgedefinieerde pakketten, oriënterend en
aanvullend onderzoek (Tabel 1), afhankelijk van de klinische vraag, is aan te raden.
De pakketten kunnen op het aanvraagformulier van het laboratorium of digitaal in het
ZIS aangevraagd worden. Is er bij de patiënt sprake van bloedingsneiging of van
trombose, is er sprake van DIS, gaat het over het monitoren van de
85
antistollingstherapie of wil de clinicus een biopsie verrichten in een vitaal orgaan
waarbij lokale hemostase onmogelijk is? Bij een gericht(e) aanvraag/pakket kan een
gericht antwoord/commentaar gegeven worden. Interpretatie van de stollingsuitslagen, zoals beschrijving van de stollings-afwijking, het vaststellen van
interfererende medicijnen (bijvoorbeeld de NOAC’s), advies voor vervolgonderzoek
of aanbeveling voor therapie wordt in het commentaar verwerkt. De commentaren
zijn in het LIS gecodeerd wat meer consistentie oplevert in de rapportage.
Gericht aanvragen, interpretatie en becommentariëren van complexe laboratoriumonderzoeken leveren eenduidige conclusies en vervolgadviezen voor de clinici.
Uiteindelijk is deze consultfunctie door de labspecialist patiëntvriendelijk en
kosteneffectief (o.a. snellere resultaten, zinvolle testen en minder bloedafnames).
Tabel 1
Aanvraagpakketten voor bloedingsneiging/hemostase
(Oriënterend)
onderzoeken
Gecodeerde
commentaren
Bloedingsneiging
Stollingstatus
bij leverlijden
DIS
Biopsie
lever
Biopsie
Nier
ERCP
Anamnese
PT/INR
APTT
Fibrinogeen
Bloedingstijd/PFA
Trombo’s
(vWF)
Geen
laboratoriumaanwijzingen
voor verhoogde
bloedingsneiging
PT
INR
ATIII
Trombo’s
APTT
PT
Fibrinogeen
ATIII
D-dimeer
APTT
PT
INR
Trombo’s
APTT
PT
INR
Trombo’s
Bloedingstijd/PFA
INR
Trombo’s
Geen
aanwijzingen
voor DIS
Geen
contraindicatie
voor
leverbiopsie
Geen contraindicatie voor
nierbiopsie
Vóór
ERCP
overweeg
suppletie
Geen aanwijzingen voor
M. von Willebrand
Normaal oriënterend
stollingsonderzoek
86
Aanvraagpakketten bij tromboseneiging
Tromboseneiging
DVT/LE < 50 jaar
of recidiverend
Tromboseneiging
CVA < 50 jaar
Tromboseneiging
Habituele abortus
Onderzoeken
INR/normotest
(uitsluiten
cumarines)
ATIII
Prot C Ag
Prot C activiteit
Prot S
LAC
aCL
Beta2 GP1
APC (+/- factor V
Leiden mutatie)
Factor II mutatie
(Factor VIII)
aCL
LAC
Beta2GP1
Homocysteine
aCL
LAC
Beta2GP1
Homocysteine
APC (+/- factor V
Leiden)
Gecodeerde
commentaren
Geen aanwijzingen
voor trombofilie of
aanwezigheid van
antifosfolipiden
antistoffen
Antifosfolipiden
antistoffen niet
aantoonbaar
Antifosfolipiden
antistoffen niet
aantoonbaar.
(Indien positief)
Onderzoek
herhalen na 12
weken
(Indien positief)
Onderzoek
herhalen na 12
weken
Tromboseneiging
Antifosfolipiden as
(herhaling na 12
weken)
aCL
LAC
Beta2GP1
Geen aanwijzingen
voor aanwezigheid
van APS
Geen aanwijzingen
voor Factor V
Leiden mutatie
87
Ned Tijdschr Klin Chem 1998; 23: 55-57
Overzichten
Laboratoriumdiagnostiek van hemostase en trombose
J.W.N. AKKERMAN1, E.J. HARTHOORN-LASTHUIZEN2 en J.J.M.L. HOFFMANN3
De Vereniging Hematologisch Laboratoriumonderzoek “VHL”, een onderdeel van de Nederlandse Vereniging voor Klinische Chemie en de Nederlandse
Vereniging voor Hematologie, heeft de auteurs gevraagd te adviseren over het beleid ten aanzien van
laboratoriumaanvragen voor hemostase- en trombose-onderzoek. Redenen hiervoor zijn (i) de grote
verschillen hierin tussen de ziekenhuizen onderling
en de diverse specialismen binnen een ziekenhuis, (ii)
de toenemende druk om vanwege budgettaire redenen het aantal laboratoriumverrichtingen te beperken
en (iii) de voortschrijdende computerisering waardoor “de patiënt achter het monsternummer” door het
laboratorium niet meer wordt herkend en meedenken
door het laboratorium steeds moeilijker wordt. Uiteraard is het advies een compromis tussen wat theoretisch wenselijk is en wat dagelijks praktisch en economisch is. Hierbij is het accent gelegd op een
beperkt en ook in de periferie hanteerbaar pakket. Het
advies is uitgebreid besproken binnen de VHL en het
bestuur van de Nederlandse Vereniging voor Trombose en Hemostase NVTH. Zowel VHL als NVTH
ondersteunen het advies. De opstellers zijn erkentelijk voor de vele adviezen van collega’s binnen en
buiten de VHL en stellen nadere op- of aanmerkingen
zeer op prijs.
Onderzoek naar een bloedingsneiging
Volledige anamnese incl. familie-anamnese en geneesmiddelengebruik
Oriënterend onderzoek
- bloedingstijd
Namens de Vereniging Hematologisch Laboratoriumonderzoek
Academisch Ziekenhuis Utrecht, Afdeling Hematologie1,
Heidelberglaan 100, 3584 CX Utrecht; Groot Ziekengasthuis, Laboratorium Hematologie2, Nieuwstraat 34,
5211 NL ’s Hertogenbosch; Catharina Ziekenhuis,
Algemeen Klinisch Laboratorium3, Michelangelolaan 2,
5623 EJ Eindhoven
Correspondentie: Prof. Dr. J.W.N. Akkerman, Afdeling Hematologie, Academisch Ziekenhuis, Heidelberglaan 100, 3584
CX Utrecht.
Ingekomen:19.08.97
-
telling trombocyten
geactiveerde partiële thromboplastine tijd (APTT)
protrombinetijd (PT)
citraat-plasma invriezen voor vervolg onderzoek
(-70 oC)
Toelichting: Aard en frequentie van de bloedingen bij
de patiënt geven al een eerste indicatie in welk deel
van het hemostase mechanisme een afwijking gevonden kan worden. Zeer belangrijk zijn een eventuele
erfelijke component en de manier waarop de verhoogde bloedingsneiging wordt overgeërfd.
Interpretatie van een verlengde bloedingstijd is niet
mogelijk zonder gegevens over het trombocytenaantal van de patiënt. Indien de bloedingstijd sterker
is verlengd dan men op grond van het trombocytenaantal mag verwachten, is er sprake van een gecombineerde trombocytopenie en trombocytopathie.
De combinatie van APTT en PT geeft een grove indicatie van de werking van de meeste stollingsfactoren
en het optreden van verworven remmers tegen stollingsfactoren. De testen zijn nauwelijks gevoelig voor
afwijkingen in aard en hoeveelheid fibrinogeen,
lichte vormen van hemofilie A of B of van de ziekte
van von Willebrand of stoornissen in de fibrinolyse.
Aanvullend onderzoek
Wanneer bovenstaande testen geen afwijkingen aan
het licht brengen, is aanvullend onderzoek overbodig
tenzij de anamnese dit toch noodzakelijk maakt
(zeker bij aangeboren of erfelijke defecten). Nader
onderzoek naar trombocytopathieën is in eerste instantie gebaseerd op aggregatie-onderzoek, waarbij
door keuze van de juiste stimulatoren een cyclooxygenase deficiëntie (gestoorde arachidonzuur
aggregatie), “storage pool deficiency” (gestoorde
secundaire aggregatie bij o.a. ADP) en Glanzmann
trombasthenie (wel shape change, geen aggregatie)
kunnen worden onderscheiden. Een normaal trombocyten-aantal met normale aggregaties en toch een
verlengde bloedingstijd maken onderzoek naar de
ziekte van von Willebrand noodzakelijk. Nader
onderzoek naar stollingsafwijkingen dient zich te
richten op factordeficiënties (afzonderlijke factorbepalingen) en remmers (titraties). Nader onderzoek
naar een verhoogde fibrinolyse vindt plaats met een
test op D-dimeren en eventueel fibrinogeen-degradatieproducten (1, 2).
Ned Tijdschr Klin Chem 1998, vol. 23, no. 2
88
Onderzoek voorafgaande aan een blinde biopsie in
vitale organen en ERCP
Volledige anamnese incl. familie anamnese en geneesmiddelengebruik
- telling trombocyten
- APTT
- PT
Toelichting: Beoogd wordt slechts ernstige afwijkingen in het hemostasemechanisme aan te tonen. De
bloedingstijd is geen bruikbare maat voor schatting
van het bloedingsrisico (3, 4). APTT en PT dienen
opnieuw als screening van het stollingsmechanisme.
Stollingvertragende medicatie dient uiteraard te worden gestaakt.
Pre-operatieve screening
Toelichting: Bij een blanco anamnese is er geen indicatie voor hemostaseonderzoek.
Diffuse intravasale stolling
- fibrinogeen
- D-dimeren
- fibrinemonomeren (optioneel als D-dimeren negatief zijn)
Toelichting: Het oriënterend onderzoek is in eerste
instantie gericht op het aantonen van verhoogde fibrinolyse en de mate van verbruik in het hemostase systeem, met name van trombocyten en fibrinogeen.
Vervolgonderzoek is gericht op de fase waarin de intravasale stolling zich bevindt.
Vervolgonderzoek
- Factor V
- Antitrombine III
Toelichting: In de beginfase van de intravasale stolling (activatie fase) is de Factor V activiteit dikwijls
abnormaal hoog. In deze fase zijn nog weinig fibrineafbraakproducten in circulatie. In de verbruiksfase
(consumptie coagulopathie) dalen de stollingsfactoren beneden de normale waarden. Factor V en met
name antitrombine III zijn dan verlaagd en de Ddimeer test is sterk positief.
Therapie-controle
Controle coumarine therapie
- PT, uitgedrukt als INR, of
- Trombotest, uitgedrukt als INR
Toelichting: De uitslagen worden uitgedrukt als International Normalized Ratio waardoor de testen onafhankelijk worden van de gebruikte reagentia (5, 6).
Nadelen van de PT zijn de gevoeligheid voor Factor
V en fibrinogeen/fibrine-afbraakproducten. Interferentie door heparine dient te worden uitgesloten indien het gebruikte reagens gevoelig is voor heparine.
De Trombotest blijft een nauwkeurige methode om
de mate van ontstolling door coumarine-derivaten
vast te leggen, mits een stabiel niveau van ontstolling
is verkregen. De test is ongevoelig voor schommelingen in Factor V of de aanwezigheid van fibrinogeen/fibrine-afbraakproducten. Heparine (≤ 2 U/ml)
stoort de test niet in een mate, die de klinische interpretatie beïnvloedt.
Controle heparine therapie
- APTT
- bepaling heparine-activiteit
Toelichting: Naarmate de reagentia voor de APTT beter gestandaardiseerd worden, leidt heparinecontrole
m.b.v. de APTT minder tot onverwachte variaties.
Uiteraard is de test gevoelig voor andere factoren dan
de heparinespiegel, zoals een verlaging van stollingsfactoren of het verschijnen van fibrinogeen/fibrineafbraakproducten zoals die vooral bij diffuse intravasale stolling worden aangetroffen. Ook bij de
overgang van heparine op coumarinetherapie dient
met deze interferentie rekening te worden gehouden.
Bij subcutane heparinisatie is controle door het laboratorium niet nodig. Voor controle van “low molecular weight heparins” is laboratoriumonderzoek eveneens niet noodzakelijk.
Controle van acetylsalicylzuur en gerelateerde medicamenten
Geen routinematige controle geïndiceerd.
Controle op trombolytica
- kortdurende behandeling met relatief hoge doses:
geen laboratoriumcontrole geïndiceerd
- langdurige behandeling met lagere doses: fibrinogeen stolbepaling
Toelichting: Een hoge dosis streptokinase of urokinase
leidt tot volledige uitputting van fibrinogeen en andere
stollingsfactoren en laboratoriumcontrole is weinig
zinvol. Bij de meeste protocollen liggen dosis en duur
van toediening van thrombolytica vast en worden niet
aangepast aan stollings- of fibrinolyseparameters.
Onderzoek naar een tromboseneiging
Geïndiceerd laboratoriumonderzoek:
- PT en APTT
- fibrinogeen stolbepaling
- antitrombine III activiteit
- proteine C en S (eventueel als ratio t.o.v. Factor II
antigeen)
- resistentie tegen geactiveerd proteine C/S (APC
resistentie, indien positief of niet mogelijk door
coumarinetherapie: Factor V-PCR)
- lupus anticoagulans
Toelichting: Een tromboseneiging kan het gevolg zijn
van een verhoogd aantal trombocyten. PT en APTT
89
worden geadviseerd als controle op anticoagulantiatherapie en afwijkingen in het stollingsmechanisme
die de APC-resistentietest beïnvloeden. Bij de verschillende activiteitsbepalingen kan bij afwijkende
waarden immunologische detectie volgen.
Literatuur
1. Nieuwenhuis HK, Sixma JJ, Harker LA and Zimmerman
TS (eds). Measurements of platelet function. Churchill
Livingstone New York 1983; 26-45.
2. George JN, Shattil SJ. The clinical importance of acquired
abnormalities of platelet function. N Engl J Med 1991;
324: 27-39.
3. Channing Rodgers RP, Levin J. A critical reappraisal of
the bleeding time. Sem Thromb Haemostas 1990; 16: 1-20.
4. Lind SE . The bleeding time does not predict surgical
bleeding. Blood 1991; 77: 2547-2552.
5. Editorial. Oral anticoagulant control. The Lancet 1987;
August 29: 488-489.
6. Hirsch J. Oral anticoagulant drugs. New Engl J Med 1991;
324: 1865-1875.
90
Article in press - uncorrected proof
Clin Chem Lab Med 2010;48(3):309–321 2010 by Walter de Gruyter • Berlin • New York. DOI 10.1515/CCLM.2010.061
Review
Laboratory reporting of hemostasis assays: the final
post-analytical opportunity to reduce errors of clinical
diagnosis in hemostasis?
Emmanuel J. Favaloro1,* and Giuseppe Lippi2
1
Department of Hematology, ICPMR, Westmead Hospital,
Westmead, NSW, Australia
2
Clinical Chemistry Laboratory, Department of Pathology
and Laboratory Medicine, University Hospital of Parma,
Parma, Italy
Abstract
The advent of modern instrumentation, with associated
improvements in test performance and reliability, together
with appropriate internal quality control (IQC) and external
quality assurance (EQA) measures, has led to substantial
reductions in analytical errors within hemostasis laboratories.
Unfortunately, the reporting of incorrect or inappropriate test
results still occurs, perhaps even as frequently as in the past.
Many of these cases arise due to a variety of events largely
outside the control of the laboratories performing the tests.
These events are primarily preanalytical, related to sample
collection and processing, but can also include post-analytical events related to the reporting and interpretation of test
results. The current report provides an overview of these
events, as well as guidance for prevention or minimization.
In particular, we propose several strategies for the post-analytical reporting of hemostasis assays, and how this may provide the final opportunity to prevent serious clinical errors
in diagnosis. This report should be of interest to both the
laboratory scientists working in hemostasis and clinicians
that request and attempt to interpret the test results. Laboratory scientists are ultimately responsible for these test
results, and there is a duty to provide both accurate and precise results to enable clinicians to manage patients appropriately and to avoid the need to recollect and retest. Also,
clinicians will not be in a position to best diagnose and manage their patient unless they gain an appreciation of these
issues.
Clin Chem Lab Med 2010;48:309–21.
*Corresponding author: Dr. Emmanuel J. Favaloro, Department
of Haematology, Institute of Clinical Pathology and Medical
Research (ICPMR), Westmead Hospital, SWAHS, Westmead,
NSW, 2145, Australia
Phone: q612 9845 6618, Fax: q612 9689 2331,
Received September 6, 2009; accepted October 15, 2009;
previously published online December 17, 2009
Keywords: diagnostic errors; extra-analytical variables;
hemostasis; hemostasis; post-analytical variables; pre-analytical variables; reporting guidelines.
Introduction
The advent of modern instrumentation, usually interfaced to
the laboratory information system and capable of providing
improvements in test performance and reliability, together
with appropriate internal quality control (IQC) and external
quality assurance (EQA) measures, has led to a considerable
reduction in analytical errors in laboratory diagnostics,
including hemostasis assays. Unfortunately, the reporting of
incorrect or inappropriate test results still occurs, perhaps
even as frequently as in the past. Many of these events can
be related to inappropriate collection or processing of samples, as covered by the common descriptor ‘preanalytical
variables’ (1). In these situations, the resultant test results
might ‘accurately’ reflect the status of the sample being tested, but this sample might not ‘accurately’ reflect the clinical
status of the patient being investigated.
These events can lead to several unwanted clinical consequences, as well as waste valuable resources and place
laboratories at risk (2). To some extent, the seriousness of
the consequences relates to the test being performed and the
experience of laboratory scientists and clinicians in recognizing these issues (1–3).
Within hemostasis, consequences can be serious for errors
related to both routine (‘screening’) coagulation tests and
diagnostic assays. In the case of the former, the result might
adversely influence clinical decisions concerning whether to
undertake further (e.g., ‘specific diagnostic’) testing, unnecessarily delay diagnoses and appropriate triage, and may also
affect anticoagulant therapy decisions. For example, (i) a
‘false normal’ screening test result might prevent further testing for factor assays and incorrectly discount a hemorrhagic
disorder we.g., von Willebrand disease (VWD), hemophiliax,
thus placing the patient at an unjustified risk of bleeding
during operative procedures (i.e., surgery, biopsy, dental
extractions); (ii) a ‘false abnormal’ screening test result
might lead to inappropriate additional costly and time consuming investigations, accompanied by unnecessary anxiety
for the patient; and (iii) a falsely low or high coagulation
test time in a patient being monitored for anticoagulant therapy may lead to subsequent incorrect dosing of anticoagulant
2010/490
Bro
Download Date | 8/16/13 11:50 AM
91
310 Favaloro et al.: Post-analytical reporting in hemostasis
therapy, with risk of thrombosis or bleeding depending on
the direction of change.
Errors related to specialized hemostasis assays may be
even more serious, as these tests are often considered ‘diagnostic’, and cause adverse consequences for both patients
and the healthcare system. Thus, a patient might be diagnosed with a particular disorder, when in fact no abnormality
is present (i.e., ‘false positive’ test result is obtained), or else
a patient with a true disorder might be missed (i.e., ‘false
negative’ test result is obtained). For example, (i) a false
negative antiphospholipid antibody (aPL) or lupus anticoagulant (LA) test result in a patient with the anti-phospholipid antibody syndrome (APS) may lead to lack of
appropriate treatment with anticoagulant therapy to prevent
future thrombosis; (ii) a false diagnosis of VWD may lead
to inappropriate treatment with factor concentrates, or to a
life-long diagnosis of VWD affecting quality of life; (iii) a
patient tested for thrombophilia while on anticoagulant therapy yields a low level of protein C or protein S, leading to
a false diagnosis of deficiency, and consequences thereof.
Preanalytical issues in coagulation
and hemostasis – a summary
This subject has been extensively covered by us in the past
wsee reference (1) for a reviewx, thus, we will provide only
a summary of the main issues here. The modern hemostasis
laboratory performs a large number of distinct tests, often
using a variety of methodologies, and leading to considerable
problems when samples are provided in a non-ideal or
unsuitable manner. Although there are guidelines available
for the proper collection and processing of samples for coagulation and hemostasis testing, and how to manage unsuitable specimens and deciding when to reject unsuitable
specimens (4–6), it is not always clear when a sample
referred to the laboratory is really ‘poor’ or unsuitable. Problems arising from pre-test sample collection, processing,
transportation and storage can be placed broadly within the
category of ‘preanalytical’ factors. Whereas analytical errors
can be avoided by using appropriate test methodologies and
by incorporation of appropriate internal control measures,
preanalytical issues present a more difficult scenario as they
are often outside the control of the laboratory performing the
tests. To date, no reliable quality indicators have been broadly implemented to monitor performance of this essential part
of the total testing process. However, some quality indicators
have been proposed by the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) Working
Group on laboratory errors and patient safety (7).
As noted, preanalytical issues are often difficult for laboratories to detect, and the laboratory professional may simply
not be aware that the sample being tested is inappropriate.
The laboratory would report the test result with the best of
intentions as reflecting an accurate ‘patient-related’ result,
but this may not be the case. The clinician would be even
less aware of the issue of preanalytical variables, and would
base their clinical response on the test result that they
received. For this reason, guidelines for specimen collection
and handling must not only be available to collection staff
and laboratories, but they should be strictly followed and
deviations avoided unless their impact, or lack thereof, on
coagulation and hemostasis testing is known.
Part of the problem in hemostasis relates to the need for
particular blood collection requirements that do not greatly
tolerate variances from the ideal. While most hemostasis tests
require citrate anti-coagulated plasma, some assays require
whole blood and others special processing of plasma (e.g.,
to entirely eliminate cellular components prior to freezing)
(1). Indeed, testing of some assays using non-ideal samples
can generate test findings entirely at odds with the true
patient presentation. In general, serum or EDTA plasma is
unsuitable for hemostasis assays. However, despite most
assays giving inappropriate test results with such samples,
other assays can be performed without apparent consequence
(1). Thus, it is not entirely clear what type of sample has
been presented to the laboratory, especially when this is
received in a secondary aliquot tube (1, 8), and the resultant
test results might be ‘peculiar’ in some cases, but seemingly
reasonable (albeit not necessarily correct) in other cases (1).
Strategies for identifying incorrect samples and dealing
with the many types of preanalytical variables (e.g., poor
collection events, inappropriate processing, transport or storage of samples prior to testing) has been covered previously
(1), and has, therefore, been summarized in Table 1.
Specific problems in laboratory testing
within hemostasis
In addition to the general preanalytical issues described in
the previous section, it is important to recognize that there
exist specific problems in hemostasis laboratory testing, and
these are briefly outlined below.
Routine assays in the coagulation laboratory
These are subject to various preanalytical problems, as well
as timed sampling issues (1). The prothrombin time (PT)
assay is most commonly used for the monitoring of vitamin
K antagonist therapy (VKAT) walso referred to as oral anticoagulant therapy (OAT)x, generally by means of the International Normalized Ratio (INR) (9, 10). The activated
partial thromboplastin time (APTT) assay is most commonly
used to identify deficiencies in the contact factor pathway,
for monitoring of heparin therapy, and for investigation of
LA (11, 12). The thrombin time (TT) assay is primarily used
as a marker of fibrinogen, the major plasma clotting protein,
which can also be quantified by means of a von Clauss procedure. The D-dimer assay measures fibrin breakdown products, and is thus a marker of fibrinolysis, important within
the context of evaluating the possibility of thrombotic events,
such as deep vein thrombosis (DVT) and pulmonary embolism (PE) (13). Each of these assays suffers from individual
preanalytical issues, as well as issues related to standard
employed methodologies. For example, the PT, APTT, TT
ro
92
Bro
Acute phase proteins released
Hospital inpatient being investigated
for thrombophilia post thrombosis
and post initiation of therapy
Hospital outpatient being
investigated for thrombophilia
whilst on therapy
Hospital inpatient being
investigated for thrombophilia
post thrombosis
Patient with history of bleeding
nearing full-term for assessment
of delivery bleeding risk
Into EDTA, serum or lithium
heparin tubes, or contamination
from inappropriate order of
blood draw, or contamination
from patient lines
Sampling of first tube from a
winged collection set
Collection of a large set of
samples, with delay in
collection of coagulation
sample to mixing at the end of
the collection process
Collection of stressed patient
(e.g., young child)
Collection of patient
while on heparin
anticoagulant therapy
while on vitamin K
antagonist therapy
just post a
thrombotic event
Collection of
patient during
pregnancy
Inappropriate
sample collection
Under-filling of blood
collection tubes
Inappropriate sample
mixing
Insufficient or inappropriate
centrifugation leaving cellular
Delayed labeling of blood sample
Incorrect patient sample
processing
Example(s) of event
Preanalytical
event
Release of cellular components post
sample freezing that can cause
(Partial) sample clotting
Plasma sample dilution
EDTA and heparin inhibit
coagulation pathways, and serum
lacks essential components of
hemostasis. However, the outcome
of testing of these samples depends
on the test performed. In most cases,
test results will be completely
erroneous, whilst in other cases there
may be no apparent effect on testing
Reduction of some hemostasis
components and elevation of others
Loss (‘consumption’) of some
hemostasis components and
elevation of others
Vitamin K antagonists lead to
functional defects and deficiencies of
hemostasis components
Heparin inhibits coagulation
pathways
Elevation in some hemostasis
components
Wrong patient tested for assay
General effect(s) of event
Table 1 A summary of the more common pre-analytical problems associated with coagulation and hemostasis testing.
Falsely shortened clotting times
Falsely shortened or prolonged
clotting times
Falsely prolonged clotting times
Falsely prolonged clotting
times leading to further
unnecessary investigation
Short APTT due to elevated FVIII
Short APTT due to elevated FVIII
Prolongs clotting times
Prolongs clotting times investigated
False high fibrinogen
Wrong test result
Example(s) of adverse
effect for routine coagulation
tests
False low level of
antithrombin (hemolysis)
False low factor levels
False low level of
hemostasis components
leading to false diagnosis
of deficiencies
leading to diagnosis of
hemophilia, or false
identification of a factor
inhibitor
Elevated VWF can mask
VWD; reduction in PS;
‘false’ (‘acquired’) APCR
False low AT level, false
negative for aPL
False low PC or PS, false
APCR, false LA
False low AT level, false LA
False high FVIII; false high
VWF can mask VWD
False test result
effect for diagnostic
hemostasis assays
Favaloro et al.: Post-analytical reporting in hemostasis 311
93
APCR, activated protein C resistance; aPL, anti-phospholipid antibody; APTT, activated partial thromboplastin time; AT, antithrombin; LA, lupus anticoagulant; PC, protein C; PS, protein S;
VWF, von Willebrand factor; VWD, von Willebrand disease.
Refrigeration of plasma for
extended time prior to testing,
or use of frost free freezers for
plasma storage
Delays in sample
transport or
inappropriate sample
storage
Loss of coagulation components
Falsely prolonged clotting times
False diagnosis of
VWD, or identification
of incorrect subtype
Falsely shortened clotting times
Activation of coagulation pathways
Refrigeration of whole blood
transport storage
Loss of some hemostasis components
Filtered plasma sample
(previously recommended for
LA testing)
processing
hemolysis (red cell lysates) or
activate coagulation pathways
components within the plasma
fraction
Prolonged clotting times
False diagnosis of VWD,
or identification of
incorrect subtype
or negative LA test
result (platelet lysates)
General effect(s) of event
Example(s) of event
Preanalytical
event
(Table 1 continued)
effect for routine coagulation
tests
effect for diagnostic
hemostasis assays
and von Clauss fibrinogens are clotting assays, and these
may employ optical test systems that can be affected by sample presentation (e.g., hemolysis, lipemia, etc) (14–16).
Additionally, hemolysis and lipemia influence the test results
per se, because they affect the coagulation pathways
(i.e., platelet lysates may activate and promote coagulation).
D-Dimer assays are typically assessed by other assay systems
(i.e., not clot based), but likewise may be influenced by other
sample presentation. The timing of collections will also
influence test results. For example, PT/INR and APTT test
results will be influenced by whether patients are on anticoagulant therapy, and if so, also the stage and extent of
anticoagulant therapy, as well as the type of therapy. For
example, VKAT will influence both APTT and PT/INR, but
primarily the PT/INR. Heparin therapy, while also potentially
influencing both APTT and PT/INR, will primarily affect the
APTT. For the D-dimer assay, there will be an effect based
on the kind and extent of a thrombosis, but these will also
be potentially raised due to other events (e.g., post surgical).
Also, there are many other issues to be aware of, including
specimen collection, transport and processing issues (1,
17–32).
Platelet counts also form a part of the routine screening
panel for investigating disturbances in hemostasis, and several preanalytical variables are known to affect the reliability
of results. These include biological variables, such as physical exercise (33) or circadian rhythm (34), the appropriate
collection (e.g., venous stasis during venipuncture, the bore
needle size, the type of blood collection device) (35–37),
and handling of the specimens (e.g., centrifugation time,
sample stability before analysis) (38, 39). An additional preanalytical problem is EDTA dependent pseudo-thrombocytopenia (40, 41). In vitro EDTA-induced platelet aggregation
is a rare event, affecting 0.09%–0.21% of all hematological
samples, caused by EDTA-dependent exposure of antigenic
determinants of platelet membrane glycoproteins gpIIb–IIIa
and the subsequent reaction of common antibodies (especially cold agglutinins) with these receptors. These events
induce platelet agglutination in vitro and result in spurious
thrombocytopenia and leukocytosis (42). In routine hematological practice, sample collection at 378C and use of alternative anticoagulants, such as buffered sodium citrate, which
requires correction of results by the dilution factor, or a mixture containing trisodium citrate (17 mmol/L), pyridoxal 59phosphate (11.3 mmol/L) and Tris (24.76 mmol/L) (CPT),
are valid alternatives for avoiding EDTA-induced platelet
clumping, and is suitable for automated complete blood
count on most instruments (43). Less frequently, pseudothrombocytopenia might be observed secondary to viral
infections, therapy with glycoprotein IIb/IIIa inhibitors (e.g.,
abciximab), or with olanzapine, mexiletine, or valproic acid
(44).
Deficiencies in protein C, protein S and antithrombin
Protein C, protein S and antithrombin comprise the three
main natural anticoagulants. Deficiencies in these natural
anticoagulants predispose individuals to a high risk of throm-
Bro
94
bosis (45, 46). The primary preanalytical issue related to collection and testing of these natural anticoagulants relates to
inappropriate collection tubes or collection times, such as
while the patient is on anticoagulant therapy or post event
‘consumption’. Factors that influence routine coagulation
assays can also influence clot-based assays for protein C and
protein S, and in some cases chromogenic assays as used for
protein C and antithrombin (1). Antithrombin, in particular,
is influenced by hemolysis. There are also specific assay and
assay sensitivity issues to be aware of (see below).
True well-defined familial protein C and protein S deficiency is very rare, comprising -5% of the thrombophilic
population and -0.5% of normal population. In addition,
true well-defined familial antithrombin deficiency is more
rare, affecting -2% of the thrombophilic population or
-0.2% of the normal population (47). Standard normal
range estimation effects means that ;2% of test samples will
show ‘low’ protein C, protein S or antithrombin, but not
necessarily the same test samples for each test case (45, 48,
49). To this can be added false low levels detected in patients
on heparin and/or warfarin therapy, representing up to 33%
of test cases (45, 48–51). Clearly, the risk of false positive
identification also worsens with poorer patient selection (i.e.,
non-thrombophilic population), and this situation appears to
be worsening with time (45, 49, 52). The presence of LA or
activated protein C resistance (APCR), seen in about 2%–5%
of the general Caucasian population may also interfere with
some clot based protein C and protein S assays (53).
It can be hypothesized that the magnitude of false identification of protein C, protein S, and antithrombin deficiency
exceeds, by several orders of magnitude, the true rate of such
deficiencies (e.g., ;10= false positive to true positive identification in pathology practice is not impossible) (45, 48,
49, 54).
Activated protein C resistance (APCR)
The main preanalytical issue related to collection and testing
of this parameter relates to inappropriate collection tubes, or
collection times, such as pregnant patients, those on anticoagulant therapy or those taking oral contraceptives, or interference from LA. Also, there are specific assay and assay
sensitivity issues (see below).
In stark contrast to protein C, protein S and antithrombin,
APCR is fairly ‘common’ within the Caucasian population,
just like its major ‘cause’, factor V Leiden (FVL) (47).
Indeed, although APCR can be detected in ;25% of the
thrombophilic population, it can also be detected in ;5% of
the normal Caucasian population (47, 52). Like protein C,
protein S and antithrombin, internal audits suggest there is
wide use of thrombophilia investigations in patients without
probable need (45, 48, 49, 52). Moreover, the percent ‘hit’
rate, if we were testing a true thrombophilic population,
would be ;25% of test cases, but internal audits indicate
values closer to 10% (45, 48, 49, 52). When testing finally
reaches a 5% ‘hit rate’, we can be assured that are testing
the general population for thrombophilia. There are several
dangers here. First, such indiscriminate testing costs the
healthcare service millions of dollars. In addition, such indiscriminant testing elevates the risk of false positive cases of
clinically presumptive thrombophilia. As most individuals
with APCR (or heterozygous FVL) will not have thrombosis,
this begs the question of how clinicians are treating patients,
with low to no risk for thrombosis, that are identified with
APCR/FVL, particularly if all we are identifying are background cases of APCR/FVL?
The most common tests used for the assessment of APCR
are based on either the APTT or the Russell viper venom
time (RVVT). The latter will generally detect FVL without
the need for predilution in factor V deficient plasma, whereas
the former will consistently detect FVL only if the assay
incorporates this predilution step (52–55). However, the
RVVT assay, and the APTT assay incorporating predilution
step will generally fail to identify ‘acquired’ APCR, such as
that arising from the presence of increased FVIII or low protein S, as for example, in pregnancy.
Assessment of factor assays and factor inhibitor
assays
Although factor assays comprise a fairly routine part of the
workload for the modern hemostasis laboratory, sample presentation is particularly problematic for these assays. Testing
of EDTA plasma, serum, plasma from a patient on VKAT,
or heparin contaminated plasma, will give rise to different
patterns of test results, which may or may not be apparent
to the laboratory professional depending on the factor assays
actually performed, the methodology used, and the extent of
heparin or EDTA or VKAT ‘contamination’ or effect (1, 56,
57). Such sample presentation may also yield false positive
inhibitor test results. In brief, EDTA plasma primarily affects
FV and FVIII, and will yield low levels on laboratory testing,
as well as abnormal PT and APTT results (1, 56, 57). EDTA
plasma can also yield false positive identification of an inhibitor to both FV and FVIII (56, 57). Heparin contaminated
plasma, either derived from the patient or from incorrect collection, will typically affect the APTT and APTT-based factor assays including FVIII, FIX, FXI and FXII (1, 56, 57).
Depending on the magnitude of heparin contamination, false
identification of a factor inhibitor is also feasible (56, 57).
The presence of LA can also affect APTT and APTT-based
factor assays, namely FVIII, FIX, FXI and FXII, as well as
give rise to false identification of a factor inhibitor (1, 56,
57). VKAT affects both routine screening tests (PT and
APTT), as well as FII, FVII, FIX and FX (1, 56, 57).
Depending on the timing and the extent of therapy, any or
all of these may yield a low result following testing. Serum
tends to be deficient in fibrinogen, FII, FV, FVIII, FIX and
FX (1, 56, 57). Finally, FVIII is an acute phase reactant and
shows increases during times of stress, as well as during
pregnancy.
Assessment of VWD
Several examples related to sample presentation have already
been provided regarding testing for VWD (Table 1). In brief,
Bro
95
serum can lead to loss of primarily high molecular weight
(HMW) von Willebrand factor (VWF), giving a test result
pattern suggestive of type 2 VWD using samples from type
1 VWD or normal individual (58, 59). Filtered plasma, previously recommended for investigation of LA, can also yield
the same kind of test result, as can the use of plasma derived
from previously refrigerated whole blood (59–65). Additional issues include the ABO blood group where VWF can be
up to 30% lower in O-groups; stress, exercise, pregnancy and
age, all of which increase VWF. Also, the menstrual cycle,
with lowest levels seen during the menstrual phase, estrogen
replacement therapy and the circadian rhythm can possibly
decrease VWF (1, 66, 67).
Different test procedures can also be affected, or give rise
to additional concerns. VWF ristocetin cofactor (VWF:RCo)
assays measure a different functional property of VWF than
either VWF antigen (VWF:Ag) or collagen binding (VWF:
CB) assays (67). Thus, test results will not correlate in all
test cases. In general, there are fewer differences in measurable values in normal individuals and patients with type 1
VWD. Thus, values generated by these tests will tend to be
fairly concordant in these cases. In contrast, type 2 VWD
patients show dysfunction of VWF characterized by a wide
range of possible defects. Therefore, VWF values tend to be
discordant when measured by different assays (67). Different
results can be obtained even when dealing with the same
assay. Sometimes these differences can be ascribed to methodological differences. For example, VWF:Ag measured by
latex immunoassay (LIA) will often be different compared
to values obtained by enzyme linked immunosorbent assay
(ELISA) (68). In addition, LIA based assays can be affected
by the presence of rheumatoid factor, and give rise to false
normal values in patients with VWD (67).
An example of the scope of the problem relating to misdiagnosis of VWD because of limited or inappropriate test
selection is worth mentioning. Within Australia, data from
the RCPA external quality assurance program (QAP) indicates that of approximately 55 participant laboratories, the
breakdown of test panels for investigation of VWD varies
widely. About one-third of laboratories perform FVIII:C,
VWF:Ag and VWF:RCo, about one-quarter FVIII:C,
VWF:Ag, VWF:RCo and VWF:CB, and the rest perform
approximately 10 other different test combinations (59–68).
Reports from this QAP have consistently shown that some
test panels (notably incorporating a VWF:CB) will result in
substantially fewer diagnostic errors than one restricted to
VWF:RCo as the only functional VWF assay (68–71). These
problems are not restricted to any geographic area or to general pathology laboratories. A number of recent genetic/phenotypic studies have been reported that have identified error
rates of around 20% for presumed ‘expert’ VWD diagnostic
laboratories in terms of misidentifying type 2 VWD as type
1 VWD (71, 72).
Assessment of primary hemostasis using
the PFA-100
The PFA-100 is a platelet function-screening tool, and is particularly sensitive to VWD (73). A variety of issues can
influence test results (1). Notably, fresh citrate anticoagulated
whole blood is the only appropriate sample, and citrate concentrations can influence test results. Perhaps more importantly is the fact that as a global test of primary hemostasis,
the PFA-100 is sensitive to a wide variety of factors, including hematocrit, platelet count, and anti-platelet medication,
such as aspirin, in addition to its sensitivity to VWD and
platelet dysfunction (1, 73, 74). In addition, factors that
affect the concentrations and function of VWF (apart from
VWD), will also influence the PFA-100, including for example ABO blood group (75). While high sensitivity to these
factors increases the value of the PFA-100 as a primary
hemostasis screening tool, it negatively impacts on its diagnostic ability since an abnormal PFA-100 does not specifically diagnose any particular disorder. Also, the PFA-100 is
not sensitive to mild defects of platelet function and mild
forms of VWD (73). Thus, a normal finding will not always
exclude these defects.
Assessment of platelet function by platelet
aggregometry
There are a large number of preanalytical and analytical
issues related to platelet function testing by platelet aggregometry. The first is the way that blood is collected and how
it is processed and tested. In brief, platelets are very sensitive
to collection and processing artifacts, in particular when testing is performed by light transmission aggregometry which
requires differential centrifugation steps (76). The generation
of false test artifacts may be amplified by adjustments to the
platelet count (76, 77). Another major problem with this test
process is the lack of standardization and the difficulty in
achieving appropriate quality control measures (78, 79). In
regards to lack of standardization, the high variability in agonist usage, both number and concentration is also worth mentioning (78–80).
Assessment of antiphospholipid antibodies
Previously, we gave the example where different VWF test
results might be obtained using different methods. However,
different test results can also be obtained using the supposed
‘same methodology’. A good example is the detection of
anticardiolipin antibody (aCL) by ELISA, where different
laboratories testing the same sample may obtain widely different test results, despite all test methods purporting to test
aCL (81, 82). While the test results and methodologies comprise analytical issues, the choice of which particular method
to utilize might be considered a preanalytical variable. In
addition, depending on the test methods and test panels used
by particular laboratories, different test results might be
reported. This might influence the clinical perception regarding the presence of disease (i.e., post-analytical), and in this
case whether the patient has APS.
Different laboratories and even experts within the field
utilize different tests or methods and test panels for the identification of APS (81). Moreover, there are wide variations
in the detection of solid phase aPL by different commercial
assays (81, 82). Thus, different perceptions will arise among
Bro
96
practitioners regarding the sensitivities and specificities of
different tests and panels for APS. Also, different perceptions
of positive or negative aPL for any given patient will arise
among clinicians, depending on the method, as well as the
test panels, used to identify APS.
Also, different tests have different sensitivities and specificities for ‘liquid-phase’ aPL (i.e., LA testing). The International Society of Thrombosis and Hemostasis (ISTH)
criteria for identification of LA have recently been revised
(12). There are several requirements for testing, and for the
identification or exclusion of LA, including recommendations regarding the type and number of assays to perform.
However, in current practice different laboratories may perform various assays, clinicians may follow a heterogeneous
approach for ordering LA tests we.g., LA screen, a Kaolin
clotting time (KCT), or a dilute Russell Viper Venom Time
(dRVVT)x, and laboratories are often only able to perform
the tests that the doctor actually orders. Thus, for example,
if a doctor only requests a ‘dRVVT’ test, and if the LA is
negative by this test, then the ISTH criteria are not being
followed, and a false negative LA is feasible.
Post-analytical issues
As previously discussed, a laboratory test result often leads
to some clinical action, and this action might have serious
adverse consequences if not appropriate to the clinical situation. While laboratories should not generally direct clinical
action, they should provide as much guidance as possible to
enable clinicians to make the best most informed choices for
patient management (83). In this respect, how the laboratory
reports their test results can also have significant adverse
consequences if clinicians base treatment on the test report.
If the report fails to appropriately identify the significance
or non-significance of test results, or the possibility of false
values arising from preanalytical events and analytical peculiarities, adverse consequences can occur.
The previous section highlighted several issues for consideration. Another example is where a laboratory reports a
test result of a weakly positive IgM aCL without advising
that this has low clinical significance. If the clinician places
too much importance upon the test result, a diagnosis of APS
may be made and anticoagulant therapy may be started without further testing or verification. Another example is reporting values of APCR or protein S without advising that these
tests might be affected by pregnancy or use of oral contraception. Also, misdiagnosis of thrombophilic conditions,
such as due to deficiency of natural anticoagulants is also
possible when testing patients immediately after a thrombotic
episode, due to consumption.
The situation with APS is potentially serious, thus within
the geographic locality of Australasia, a working group was
formed that has since published several guidelines for the
appropriate reporting of aCL and aB2GPI assays (84–87).
Although other hemostasis tests could benefit from specific
guidance, guidelines for such applications are lacking. For
example, for VWD, it is important that laboratories advise
Table 2 Important issues for laboratories and clinicians to consider within the context of extra-analytical issues in hemostasis testing, as
well as some recommendations.
Issue
Consideration/recommendation
Test selection
Select/request the best tests/test processes/test panels for the condition being investigated
Population to be tested and clinical
condition/medication at time of testing
Select the appropriate population/methodology to determine the normal reference range
Only request the test(s) when clinically appropriate and in the right patient at the right time
Sample collection
Proper patient and sample identification
Atraumatic phlebotomy with minimal tourniquet use
Draw 3.2% blue stopper tube first or only after a non-additive tube
Fill tube adequately (no -90% fill)
Adequately and thoroughly mix with tube anticoagulant
Sample transport
Transport promptly at room temperature
Sample processing
Centrifuge within 1 h of phlebotomy to obtain platelet poor plasma (most tests)
Double centrifuge plasma for some tests, namely LA and APCR
Aliquot (in a non-activating secondary tube) immediately following centrifugation for
those tests to be performed later on
Special requirements for some tests, such as platelet function and PFA-100
Sample storage
Test plasma within appropriate time frame; store as required, samples to be tested
subsequently
Sample testing
Select the best test/methodology/test panel for the analyte/parameter being tested
Perform test in timely manner and according to best practice
Result interpretation
Laboratory: provide clinician with appropriate guidance/test interpretation
Clinician: recognize test limitations/extra-analytical issues that may influence test results and
follow local expert laboratory advice
Adapted from reference (1).
Bro
97
Elevated – clot based assay
Elevated – LIA based assay
Any result
Prolongation in screen test/mixing study,
but negative for lupus anticoagulant by
confirmation test
Positive result
Negative
Low/equivocal/positive
IgG negative/IgM positive
Protein S
Activated protein C resistance
Lupus anticoagulant test
Lupus anticoagulant test
Anticardiolipin antibody
Low level
Protein C and/or protein S
Low level
Antithrombin
Low level – some methods (e.g., some clot
based assays)
Positive result
D-dimer
Test result
Test
Bro
IgM anticardiolipin antibodies are less specific than IgG anticardiolipin antibodies for the antiphospholipid
antibody syndrome. Transient IgM aCL may be found in a range of other inflammatory, infectious and
malignant disorders, and rheumatoid factors may also produce false positive results. Repeat testing
(after 12 weeks) is recommended, as is lupus anticoagulant testing
The risk of clinical symptoms in the antiphospholipid antibody syndrome appears to rise with increasing
levels of IgG anti-cardiolipin antibodies. Repeat testing (after 12 weeks) is recommended, as is lupus anticoagulant testing. Transient low level/positive results generally are of questionable clinical significance
Some patients with antiphospholipid antibody syndrome have undetectable anticardiolipin antibodies.
Lupus anticoagulant testing may be indicated
Suggest repeat testing in 12 weeks for confirmation
Lupus anticoagulant (Lupus Inhibitor) not detected. However, screening test suggested potential presence of
other inhibitor type. If patient on anticoagulant therapy (vitamin K antagonist or heparin), please repeat
testing when therapy has ceased. Otherwise, might indicate another inhibitor (e.g., FV or FVIII);
please discuss with laboratory as further testing may be required
Individuals with lupus anticoagulant, factor inhibitors, factor deficiencies, or on anticoagulant therapy may
not provide reliable assay results
wMight need comment regarding possibility of interference from rheumatoid factorx
wMight need comment regarding possibility of interference from LAx
wMight need additional comment regarding possibility of interference by APCR depending on assay/
reagents usedx
Reduced protein C wand/or Sx level detected. Congenital deficiencies of protein C wand/or Sx are very rare.
Low levels of protein C wand/or Sx can occur immediately after a thrombotic episode, with anti-coagulant
or vitamin K antagonist therapy (e.g., warfarin), vitamin K deficiency or liver disease, on hormone
replacement/oral contraceptive therapy/during pregnancy/with nephrotic syndrome wprotein Sx or from a
consumptive coagulopathy, hemodilution, or a blood collection artefact. Please exclude these events and
repeat the test six weeks after cessation of any anti-coagulant therapy
Reduced antithrombin level detected. Congenital deficiencies of antithrombin are very rare. Low levels of
antithrombin may occur immediately after a thrombotic episode, during heparin therapy, in liver disease, or
from a consumptive coagulopathy, hemodilution, in nephrotic syndrome, following L-asparaginase therapy
or a blood collection artefact (including hemolysis). Please exclude these events and repeat the test one
week after cessation of any anti-coagulant therapy
High levels occur immediately after a thrombotic episode. Exclude concomitant non-thrombotic conditions
(pregnancy, cancer, infections, surgery, trauma)
Sample comment
Table 3 Sample interpretative comments for inclusion on test reports to guide appropriate clinical action.
98
Bro
Prolonged closure time (CT) result with
both C/Epi and C/ADP
Normal closure time (CT) result with both
C/Epi and C/ADP
Elevated
Additional add-on comment to any of
above comments
PFA-100
PFA-100
All tests performed by
latex immuno assay
Any test
Please contact the laboratory for further advice
wMight need comment regarding possibility of interference from rheumatoid factorx
Normal closure time (CT) result with both C/Epi and C/ADP. This result will not always discount a
primary hemostasis disorder. If patient being investigated for mucocutaneous bleeding, please discuss with
laboratory, as further testing may be required
Prolonged closure time (CT) result with both C/Epi and C/ADP. Results consistent with any of the
following: very low platelet count, very low hematocrit, recent anti-platelet medication (e.g., aspirin),
moderate to severe platelet dysfunction, and/or moderate to severe von Willebrand disorder. Suggest
medication review and full blood count. Other studies may be indicated; please discuss with laboratory
if required. wNote: comment can be modified as appropriate depending on other test results; e.g., if platelet
count and hematocrit are availablex
Prolonged closure time (CT) result with C/Epi, normal with C/ADP. Results consistent with any of the
following: low platelet count, low hematocrit, recent anti-platelet medication (e.g., aspirin), mild platelet
dysfunction, and/or mild von Willebrand disorder. Suggest medication review and full blood count. Other
studies may be indicated; please discuss with laboratory if required. wNote: comment can be modified as
appropriate depending on other test results; e.g., if platelet count and hematocrit are availablex
Results suggestive of wHemophilia A, Hemophilia A carrier, acquired deficiency, or type 2N von
Willebrand’s disease (depending on test pattern obtained)x. Further studies may be indicated.
walso consider the possibility of serum testedx
Results suggestive of type 2 we.g., 2A or 2B or 2M or pseudo/platelet-type (depending on results)x
von Willebrand disease. Further studies may be indicated; please contact laboratory for advice, or else send
repeat sample for retesting and confirmation. Please note: the following can all provide a false
type 2 VWD test pattern: testing of filtered plasma or serum sample, or testing of plasma after the
refrigeration or storage of whole blood sample at low temperature
wConsider possibility of LA, or EDTA or heparin contaminationx
wDepending on factor and pattern of test results, consider the possibility of comments and further
discriminatory testing related to possible interference with heparin (e.g., low FVIII, IX, XI and XII), or
VKAT (low FII, FVII, FIX, FX) or testing of EDTA plasma (low FV, FVIII) or serum (low FII, FV, FVIII,
FIX, FX)x
Sample comment
Notes: some of these examples are recognized as being too long for practical use; the intention is that they be adapted for specific use in laboratories as per their individual suitability. For
example, text comments related to pregnancy may be irrelevant in a pediatric hospital. Comments in wsquare bracketsx describe additional considerations for the laboratory. Adapted and updated
from reference (1).
Prolonged closure time (CT) result with
C/Epi, normal with C/ADP
Pattern suggestive of type 2 VWD (i.e.,
functional discordance between VWF:Ag
and VWF:CB and/or VWF:RCo, or loss of
high molecular weight VWF multimers)
von Willebrand factor
PFA-100
Positive result
Factor inhibitor
Pattern suggestive of functional
discordance between VWF:Ag and
FVIII:C
Low levels
Factor assays
von Willebrand factor
Test result
Test
(Table 3 continued)
99
on the significance or non-significance of certain test results
to ensure that clinicians do not over diagnose type 1 VWD
or under diagnose type 2 VWD (67, 69). The former might
occur when test values fall below the normal reference range,
while the latter might occur because the laboratory is using
a limited test panel that may miss some forms of VWD. For
thrombophilia tests, it is important to highlight that low test
results for protein C, protein S and antithrombin does not
necessarily mean that the patient has a congenital deficiency,
since the risk of a false positive due to a low value exceeds
by several orders of magnitude the likelihood of a true positive (48, 49, 54). This situation is made worse by poor
patient test selection and inappropriate sampling time points,
such as for example in a patient on anticoagulant therapy.
Although testing for D-dimer is currently regarded as the
mainstay in the diagnostic approach to venous thromboembolism, increased values are frequently observed in a variety
of physiological conditions, such as pregnancy. Also,
increased values are seen in a variety of pathological conditions, such as cancer, infections, trauma, surgery, kidney
and liver failure, etc. Thus, increased values might be commonplace in hospitalized patients (13).
Therefore, the inclusion of interpretative comments on laboratory reports provides added value and greatly enhances
visibility and competency of laboratory activities (1, 45, 88,
89). Such interpretative comments might be particularly
advisable when reporting results of hemostasis testing, given
the large number of potential issues that impact test results,
as highlighted here and elsewhere (1, 45). The laboratory
and clinician also need to consider the appropriateness of the
normal reference range listed on the test report. Another final
but important issue refers to the standardization/harmonization of reported test results. There are several examples
within hemostasis diagnostics. D-dimer testing is a good
example, where multiple units of measure are available we.g.,
ng/mL, g/L or fibrinogen equivalent units (FEU)x (13). This
is not trivial, the conversion factor from FEU/mL to ng/mL
is 0.5, and thus can represent the difference between a negative and positive finding. Another example is VWF and
factor assay testing, where current popular usage is % of
normal, and where alternatives include U/mL, IU/mL, U/dL
and IU/dL.
Conclusions
Preanalytical issues, analytical digressions and post-analytical (mis)interpretation in hemostasis testing can lead to
significant diagnostic error and adverse clinical events, as
well as a largely preventable waste of valuable healthcare
resources. This report has detailed the situation with respect
to a large number of common tests employed in hemostasis.
However, we would be foolish to assume that these issues
are limited to those described in this report. Therefore, Table
2 lists some important issues for laboratories and clinicians
to consider within the context of extra-analytical issues in
hemostasis testing, as well as providing some recommendations for improvements in the testing and interpretation pro-
cess. Some sample interpretative laboratory comments are
also provided in Table 3, as are additional comments to guide
laboratories provide clinically useful information, and not
just ‘meaningless numbers’. In the end, the most useful
advise we can offer is this: be aware of the many issues;
laboratories should select and physicians should request the
best tests and test panels available; perform testing only
when justified and arrange this testing for the correct point
in time for the condition under investigation; collect as much
clinical information as possible; follow the recommendations
of local laboratory experts and specialists; repeat the tests
when these are not in keeping with clinical expectations or
when an abnormal finding is reported; do not release patient
test results when IQC is unsatisfactory, and establish a mutually beneficial clinical-laboratory interface, where both parties can actively collaborate to achieve the best possible
patient outcome.
Conflict of interest statement
Authors’ conflict of interest disclosure: The authors stated that
there are no conflicts of interest regarding the publication of this
article.
References
1. Favaloro EJ, Lippi G, Adcock DM. Preanalytical and postanalytical variables: the leading causes of diagnostic error in haemostasis? Semin Thromb Hemost 2008;34:612–34.
2. Lippi G, Guidi GC. Risk management in the preanalytical phase
of laboratory testing. Clin Chem Lab Med 2007;45:720–7.
3. Lippi G, Guidi GC, Mattiuzzi C, Plebani M. Pre-analytical variability: the dark side of the moon in laboratory testing. Clin
Chem Lab Med 2006;44:358–65.
4. Lippi G, Banfi G, Buttarello M, Ceriotti F, Daves M, Dolci A,
Caputo M, Giavarina D, Montagnana M, Miconi V, Milanesi B,
Mosca A, Morandini M, Salvagno GL, for the Italian Intersociety
SIBioCSIMeL-CISMEL Study Group on Extraanalytical Variability. Recommendations for detection and management of unsuitable samples in clinical laboratories. Clin Chem
Lab Med 2007;45:728–36.
5. Lippi G, Fostini R, Guidi GC. Quality improvement in laboratory
medicine: extra-analytical issues. Clin Lab Med 2008;28:
285–94.
6. Adcock DM, Hoefner DM, Kottke-Marchant K, Marlar RA, Szamosi DI, Warunek DJ. Collection, transport, and processing of
blood specimens for testing plasma-based coagulation assays and
molecular hemostasis assays: approved guideline, 5th ed. Clinical Laboratory Standards Institute. Wayne, PA: CLSI document
H21-A5, 2008.
7. Sciacovelli L, Plebani M. The IFCC Working Group on laboratory errors and patient safety. Clin Chim Acta 2009;404:79–
85.
8. Lippi G, Salvagno GL, Adcock DM, Gelati M, Guidi GC,
Favaloro EJ. Right or wrong sample received for coagulation
testing? Tentative algorithms for detection of an incorrect type
Bro
100
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
of sample. Int J Lab Hematol 2009 Feb 7. wEpub ahead of
printx.
Favaloro EJ, Hamdam S, McDonald J, McVicker W, Ule V.
Time to think outside the box? Prothrombin time (PT), international normalised ratio (INR), international sensitivity index
(ISI), mean normal prothrombin time (MNPT) and measurement of uncertainty (MU): a novel approach to standardisation.
Pathology 2008;40:277–87.
Favaloro EJ, Adcock DM. Standardization of prothrombin time
measurements used to monitor anticoagulant therapy: How
good is your laboratory’s INR? Semin Thromb Hemost 2008;
34:593–603.
Lippi G, Favaloro EJ. Activated partial thromboplastin time:
new tricks for an old dogma. Semin Thromb Hemost 2008;
34:604–11.
Pengo V, Tripodi A, Reber G, Rand JH, Ortel TL, Galli M, et
al. Official communication of the scientific and standardization
committee on lupus anticoagulant/phospholipid-dependent antibodies: update of the guidelines for lupus anticoagulant detection. J Thromb Hemost 2009; 7:1737–40.
Lippi G, Franchini M, Targher G, Favaloro EJ. Help me, Doctor! My D-dimer is raised. Ann Med 2008;40:594–605.
Lippi G, Blanckaert N, Bonini P, Green S, Kitchen S, Palicka
V, et al. Haemolysis: an overview of the leading cause of
unsuitable specimens in clinical laboratories. Clin Chem Lab
Med 2008;46:764–72.
Lippi G, Montagnana M, Salvagno GL, Guidi GC. Interference
of blood cell lysis on routine coagulation testing. Arch Pathol
Lab Med 2006;130:181–4.
O’Brien JR. Relation to blood-coagulation to lipaemia. Lancet
1955;269:690–3.
Lippi G, Salvagno GL, Montagnana M, Guidi GC. Influence
of two different buffered sodium citrate concentrations on
coagulation testing. Blood Coag Fibrinolysis 2005;16:381–3.
Adcock DM, Kressin DC, Marlar RA. Effect of 3.2% vs. 3.8%
sodium citrate concentration on routine coagulation testing. Am
J Clin Pathol 1997;107:105–10.
Lippi G, Franchini M, Montagnana M, Salvagno GL, Poli G,
Guidi GC. Quality and reliability of routine coagulation testing:
can we trust that sample? Blood Coag Fibrinolysis 2006;17:
513–9.
Kiechle FL, Adcock DM, Calam RR, Davis C, Schwartz JG.
So you’re going to collect a blood specimen. An introduction
to phlebotomy. College of American Pathologists. 12th ed.
Northfield, Ill, USA, 2007.
Ernst DJ, Balance LO, Calam RR, McCall R, Smith SS, Szamosi DI, et al. Procedures for the collection of diagnostic blood
specimens by venipuncture. Approved standard, 6th ed. CLSI
document H3-A6. Wayne, PA: Clinical and Laboratory Standards Institute, 2007.
Adcock DM, Kressin DC, Marlar RA. Are discard tubes necessary in coagulation studies? Lab Med 1997;28:530–3.
Lippi G, Guidi GC. Effect of specimen collection on routine
coagulation assays and D-dimer measurement. Clin Chem
2004;50:2150–2.
Adcock DM, Kressin DC, Marlar RA. Minimum specimen volume requirements for routine coagulation testing. Dependence
on citrate concentration. Am J Clin Pathol 1998;109:595–9.
Adcock DM. Sample integrity and preanalytical variables. In:
Preston E, Olsen J, Kitchen S, editors. Quality in laboratory
hemostasis and thrombosis, 1st ed. Oxford: Wiley-Blackwell,
UK, 2009.
26. Lippi G, Salvagno GL, Montagnana M, Guidi GC. Influence
of primary sample mixing on routine coagulation testing. Blood
Coag Fibrinolysis 2007;18:709–11.
27. Lippi G, Salvagno GL, Montagnana M, Poli G, Guidi GC.
Influence of the needle bore size on platelet count and routine
coagulation testing. Blood Coag Fibrinolysis 2006,17:557–61.
28. Lippi G, Salvagno GL, Guidi GC. No influence of butterfly
device on routine coagulation assays and D-dimer measurement. J Thromb Haemost 2005;3:389–91.
29. Lippi G, Salvagno GL, Montagnana M, Guidi GC. Short-term
venous stasis influences routine coagulation testing. Blood
Coag Fibrinolysis 2005;16:453–8.
30. Adcock DA, Kressin DC, Marlar RA. The effect of time and
temperature variables on routine coagulation tests. Blood Coagul Fibrinolysis 1998;9:463–70.
Influence of centrifuge temperature on routine coagulation testing. Clin Chem 2006;52:537–8.
32. Lippi G, Salvagno GL, Montagnana M, Manzato F, Guidi GC.
Influence of the centrifuge time of primary plasma tubes on
routine coagulation testing. Blood Coag Fibrinolysis 2007;18:
525–8.
33. Lippi G, Maffulli N. Biological influence of physical exercise
on hemostasis. Semin Thromb Hemost 2009;35:269–76.
34. Montagnana M, Salvagno GL, Lippi G. Circadian variation
within hemostasis: an underrecognized link between biology
and disease? Semin Thromb Hemost 2009;35:23–33.
35. Lippi G, Salvagno GL, Montagnana M, Franchini M, Guidi
GC. Venous stasis and routine hematologic testing. Clin Lab
Haematol 2006;28:332–7.
Influence of the needle bore size on platelet count and routine
coagulation testing. Blood Coagul Fibrinolysis 2006;17:557–
61.
37. Lippi G, Salvagno GL, Brocco G, Guidi GC. Preanalytical variability in laboratory testing: influence of the blood drawing
technique. Clin Chem Lab Med 2005;43:319–25.
38. Lippi G, Salvagno GL, Montagnana M, Manzato F, Guidi GC.
Influence of the centrifuge time of primary plasma tubes on
routine coagulation testing. Blood Coagul Fibrinolysis 2007;
18:525–8.
39. Lippi G, Salvagno GL, Solero GP, Franchini M, Guidi GC.
Stability of blood cell counts, hematologic parameters and
reticulocytes indexes on the Advia A120 hematologic analyzer.
J Lab Clin Med 2005;146:333–40.
40. Lippi U, Schinella M, Nicoli M, Modena N, Lippi G. EDTAinduced platelet aggregation can be avoided by a new anticoagulant also suitable for automated complete blood count.
Haematologica 1990;75:38–41.
41. Lippi G, Guidi G, Nicoli M. Platelet count in EDTA-dependent
pseudothrombocytopenia. Eur J Haematol 1996;56:112–3.
42. Casonato A, Bertomoro A, Pontara E, Dannhauser D, Lazzaro
AR, Girolami A. EDTA dependent pseudothrombocytopenia
caused by antibodies against the cytoadhesive receptor of platelet gpIIB-IIIA. J Clin Pathol 1994;47:625–30.
43. Lippi U, Schinella M, Nicoli M, Modena N, Lippi G. EDTAinduced platelet aggregation can be avoided by a new anticoagulant also suitable for automated complete blood count.
Haematologica 1990;75:38–41.
44. Berkman N, Michaeli Y, Or R, Eldor A. EDTA-dependent
pseudothrombocytopenia: a clinical study of 18 patients and a
review of the literature. Am J Hematol 1991;36:195–201.
Bro
101
45. Favaloro EJ, McDonald D, Lippi G. Laboratory investigation
of thrombophilia: the good, the bad and the ugly. Semin
Thromb Hemost 2009;35. In press.
46. Lippi G, Franchini M. Pathogenesis of venous thromboembolism: when the cup runneth over. Semin Thromb Hemost
2008;34:747–61.
47. Coppola A, Tufano A, Cerbone AM, Di Minno G. Inherited
thrombophilia: implications for prevention and treatment of
venous thromboembolism. Semin Thromb Hemost 2009. In
press.
48. Favaloro EJ, Soltani S, McDonald J, Grezchnik E, Easton L.
Laboratory identification of familial thrombophilia: do the pitfalls exceed the benefits? A reassessment of ABO-blood group,
gender, age, and other laboratory parameters on the potential
influence on a diagnosis of protein C, protein S, and antithrombin deficiency and the potential high-risk of a false positive
diagnosis. Lab Hematol 2005;11:174–84.
49. Favaloro EJ. Diagnostic issues in thrombophilia: a laboratory
scientist’s view. Semin Thromb Hemost 2005;31:11–6.
50. Florell SR, Rodgers GM. Utilization of testing for activated
protein C resistance in a reference laboratory. Am J Clin Pathol
1996;106:248–52.
51. Somma J, Sussman II, Rand JH. An evaluation of thrombophilia screening in an urban tertiary care medical center: a ‘‘real
world’’ experience. Am J Clin Pathol 2006;126:120–7.
52. Favaloro EJ, Orsag I, Bukuya M, McDonald D. A nine-year
retrospective assessment of laboratory testing for activated protein C resistance: evolution of a novel approach to thrombophilia investigations. Pathology 2002;34:348–55.
53. Favaloro EJ, Bonar R, Sioufi J, Wheeler M, Low J, Aboud M,
Duncan E, Smith J, Exner T, Lloyd J, Marsden K (on behalf
of the RCPA QAP in Haematology). Multi-laboratory testing
of thrombophilia: current and past practice in Australasia as
assessed through the Royal College of Pathologists of Australasia Quality Assurance Program in Haematology. Semin
Thromb Hemost 2005;31:49–58.
54. Favaloro EJ. More on ‘universal’ versus ‘selected’ screening
for thrombophilia: the hidden costs of false-positive diagnosis.
Br J Haematol 2006;134:239–40.
55. Favaloro EJ, Mirochnik O, McDonald D. Functional activated
protein C resistance assays: correlation with factor V DNA
analysis is better with RVVT- than APTT-based assays. Br J
Biomed Sci 1999;56:23–33.
56. Favaloro EJ, Bonar R, Duncan E, Earl G, Low J, Aboud M,
Just S, Sioufi J, Street A, Marsden K (on behalf of the RCPA
QAP in Haematology Haemostasis Committee). Identification
of factor inhibitors by diagnostic haemostasis laboratories: a
large multi-centre evaluation. Thromb Haemostasis 2006;96:
73–8.
57. Favaloro EJ, Bonar R, Duncan E, Earl G, Low J, Aboud M,
Just S, Sioufi J, Street A, Marsden K (on behalf of the RCPA
QAP in Haematology Haemostasis Committee). Mis-identification of factor inhibitors by diagnostic haemostasis laboratories: recognition of pitfalls and elucidation of strategies. A
follow-up to a large multicentre evaluation. Pathology 2007;
39:504–11.
58. Favaloro EJ, Facey D, Grispo L. Laboratory assessment of von
Willebrand factor: use of different assays can influence the
diagnosis of von Willebrand’s disease, depending on differing
sensitivity to sample preparation and differential recognition of
high molecular weight VWF forms. Am J Clin Pathol 1995;
104:264–71.
59. Favaloro EJ. Pre-analytical variables in coagulation testing.
Blood Coag Fibrinolysis 2007;18:86–9.
60. Favaloro EJ, Mehrabani PA, Koutts J. Laboratory assessment
of von Willebrand factor: altered interpretation of laboratory
data, and altered diagnosis of von Willebrand’s disease, as
influenced by the use of different VWF assays and assay conditions. Clin Appl Thromb/Hemost 1997;3:110–8.
61. Favaloro EJ, Mehrabani PA. Laboratory assessment of von Willebrand factor: differential influence of prolonged ambient temperature specimen storage on assay results. Haemophilia 1996;
2:218–23.
62. Favaloro EJ, Mohammed A, Coombs R, Mehrabani PA. Filtered plasma as a potential cause of clinical misdiagnosis: inappropriate testing in a haematology laboratory. Br J Biomed Sci
1995;52:243–8.
63. Favaloro EJ, Nair SC, Forsyth CJ. Collection and transport of
samples for laboratory testing in von Willebrand’s disease
(VWD): time for a reappraisal? Thromb Haemost 2001;86:
1589–90.
64. Favaloro EJ, Soltani S, McDonald J. Potential laboratory misdiagnosis of haemophilia and von Willebrand disorder due to
cold activation of blood samples for testing. Am J Clin Pathol
2004;122:686–2.
65. Böhm M, Täschner S, Kretzschmar E, Gerlach R, Favaloro EJ,
Scharrer I. Cold storage of citrated whole blood induces drastic
time- dependent losses in factor VIII and von Willebrand factor: potential for misdiagnosis of haemophilia and von Willebrand’s disease. Blood Coag Fibrinolysis 2006;17:39–45.
66. Favaloro EJ, Soltani S, McDonald J, Grezchnik E, Easton L,
Favaloro JWC. Reassessment of ABO-blood group, gender and
age on laboratory parameters used to diagnose von Willebrand
disorder (VWD): potential influence on the diagnosis versus
the potential association with risk of thrombosis. Am J Clin
Pathol 2005;124:910–7.
67. Favaloro EJ. Laboratory identification of von Willebrand disease: technical and scientific perspectives. Semin Thromb
Hemost 2006;32:456–71.
68. Favaloro EJ, Bonar R, Meiring M, Street A, Marsden K (on
behalf of the RCPA QAP in Haematology). 2B or not 2B?
Disparate discrimination of functional VWF discordance using
different assay panels or methodologies may lead to success or
failure in the early identification of Type 2B VWD. Thromb
Haemost 2007;98:346–58.
69. Favaloro EJ, Bonar R, Kershaw G, Sioufi J, Thom J, Baker R,
Hertzberg M, Street A, Lloyd J, Marsden K (on behalf of the
RCPA QAR in Haematology). Laboratory diagnosis of von
Willebrand disorder: use of multiple functional assays reduces
diagnostic error rates. Lab Hematol 2005;11:91–7.
70. Favaloro EJ, Bonar R, Kershaw G, Sioufi J, Baker R, Hertzberg
M, Street A, Marsden K (on behalf of the RCPA QAP in Haematology). Reducing errors in identification of von Willebrand
disease: the experience of the Royal College of Pathologists
of Australasia Quality Assurance Program. Semin Thromb
Hemost 2006;32:505–13.
71. Favaloro EJ. An update on the von Willebrand factor collagen
binding (VWF:CB) assay: 21 years of age and beyond adolescence, but not yet a mature adult. Semin Thromb Hemost
2007;33:727–44.
72. Favaloro EJ. Detailed von Willebrand factor multimer analysis
in patients with von Willebrand disease in the European study,
molecular and clinical markers for the diagnosis and management of type 1 von Willebrand disease (MCMDM-1VWD) –
a rebuttal. J Thromb Haemost 2008;6:1999–2001.
73. Favaloro EJ. Clinical utility of the PFA-100. Semin Thromb
Hemost 2008;34:709–33.
74. Crescente M, Di Castelnuovo A, Iacoviello L, Vermylen J, Cer-
Bro
102
75.
76.
77.
78.
79.
80.
81.
letti C, de Gaetano G. Response variability to aspirin as
assessed by the platelet function analyzer (PFA)-100. A systematic review. Thromb Haemost 2008;99:14–26.
Lippi G, Franchini M, Brocco G, Manzato F. Influence of the
ABO blood type on the platelet function analyzer PFA-100.
Thromb Haemost 2001;85:369–70.
Cattaneo M. Light transmission aggregometry and ATP release
for the diagnostic assessment of platelet function. Semin
Thromb Hemost 2009;35:158–67.
Favaloro EJ, Mohammed S. Platelet function testing: auditing
local practice and broader implications. Clin Lab Sci 2009. In
press.
Pai M, Hayward C. Diagnostic assessment of platelet disorders:
what are the challenges to standardization? Semin Thromb
Hemost 2009;35:131–8.
Favaloro EJ. Internal quality control and external quality assurance of platelet function tests. Semin Thromb Hemost 2009;
35:139–49.
Duncan EM, Bonar R, Rodgers SE, Favaloro EJ, Marsden K,
RCPA QAP in Haematology, Haemostasis Committee. Methodology and outcomes of platelet aggregation testing in Australia, New Zealand and the Asia-Pacific region: results of a
survey from the Royal College of Pathologists of Australasia
Haematology Quality Assurance Program. Int J Lab Hematol
2009;31:398–406.
Favaloro EJ, Wong RCW. Laboratory testing and identification
of antiphospholipid antibodies and the antiphospholipid syn-
82.
83.
84.
85.
86.
87.
88.
89.
drome: a potpourri of problems, a compilation of possible solutions. Semin Thromb Haemost 2008;34:389–410.
Reber G, Boehlen F, de Moerloose P. Technical aspects in laboratory testing for antiphospholipid antibodies: is standardization an impossible dream? Semin Thromb Haemost 2008;34:
340–6.
Plebani M. The clinical importance of laboratory reasoning.
Clin Chim Acta 1999;280:35–45.
Wong RCW, Gillis D, Adelstein S, Baumgart K, Favaloro EJ,
Hendle MJ, et al. Consensus guidelines on anti-cardiolipin antibody testing and reporting. Pathology 2004;361:63–8.
Wong RC, Favaloro EJ, Adelstein S, Baumgart K, Bird R,
Brighton TA, et al. Consensus guidelines on anti-beta2 glycoprotein I testing and reporting. Pathology 2008;40:58–63.
Wong R, Adelstein S, Gillis D, Favaloro EJ. Development of
consensus guidelines for anticardiolipin and lupus anticoagulant testing. Semin Thromb Hemost 2005;31:39–48.
Wong RCW, Favaloro EJ. A consensus approach to the formulation of guidelines for laboratory testing and reporting of
antiphospholipid antibody assays. Semin Thromb Hemost
2008;34:361–72.
Plebani M. What information on quality specifications should
be communicated to clinicians, and how? Clin Chim Acta
2004;346:25–35.
Plebani M. Interpretative commenting: a tool for improving the
laboratory-clinical interface. Clin Chim Acta 2009;404:46–
51.
Bro
103
Doing now what patients need next
Roche is een research georiënteerde
onderneming met een gecombineerde
kracht in geneesmiddelen en diagnostica.
Onze personalised healthcare strategie richt
zich op de ontwikkeling van diagnostische
en therapeutische oplossingen, die een
bijdrage leveren aan de kwaliteit van leven
van de patiënt. Een juiste en efficiënte
diagnostiek helpt behandelaars bij het maken
van juiste beslissingen voor de patiënt.
Dat maakt onze inspanningen zo betekenisvol.
Wilt u meer weten over de waarde van
diagnostiek? www.roche.nl
104
FUNCTIETEST UITSLAGEN
JMM Rondeel, arts klinische chemie
De laboratoriumspecialisten in Isala voeren jaarlijks ongeveer 150 poliklinische
functietesten uit (tabel). Deze testen worden aangevraagd in het kader van
diagnostiek en follow-up van endocriene aandoeningen, zoals GH deficiëntie,
acromegalie, gestoorde puberteitsontwikkeling, syndroom van Cushing, bijnierinsufficiëntie, chronisch vermoeidheidssyndroom, hirsutisme en verdenking centrale
hypothyreoïdie. De uitvoering, interpretatie, becommentariëring en rapportage wordt
verzorgd door de labspecialisten zelf. Zij schrijven de recepten uit, voeren de
anamnese, plaatsen het infuus en doen de bloedafnames. Hierin worden zij
ondersteund door twee physician assistants en een doktersassistente. De testen
worden in een apart ingerichte functiekamer nabij het laboratorium uitgevoerd. Alle
testen worden als brief gerapporteerd in het laboratoriumdomein van het elektronisch
patiëntendossier (zie de twee voorbeelden hierna).
Endocriene as
Bijnier
Low dose synacthen
High dose synacthen
CRH
GH
Kind
Volwassene
GTT
Gonade
GnRH
Rest
TRH
Combinatie
Totaal jaar 2012
Indicatie
Aantal
Chronische
vermoeidheid
Hirsutisme
Bijnieras
95
Groeistoornis
Na hypofyse OK
Acromegalie
15
7
6
Puberteitsontwikkeling
15
Centrale hypothyreoïdie
Hypofysefunctie
2
1
146
4
1
105
Testen bij klinisch opgenomen patiënten worden niet uitgevoerd door de labspecialist
maar door de kliniek zelf. De labspecialist becommentarieert en rapporteert deze
testen wel op dezelfde manier als de poliklinische testen. Onder deze testen vallen
bijvoorbeeld de intra-operatieve PTH bepaling, vasten- en dorstproeven en klinisch
uitgevoerde synacthentesten en dexamethason suppressietesten.
Het zelf uitvoeren en coördineren van de poliklinische functietesten biedt grote
voordelen:
-
De testen worden gestandaardiseerd door een beperkt aantal personen
uitgevoerd.
-
De testen worden up to date gehouden wat betreft uitvoering en interpretatie.
-
De tijd tussen aanvraag en volledige rapportage is kort (iets langer dan 1
week).
De tijdsinvestering is daarnaast gering: iedere labspecialist heeft per week gemiddeld
1 test met een werklast van ongeveer 2 uur. De voordelen zijn echter groot:
-
Er is direct patiëntencontact waarbij de labspecialist een grote mate van
autonomie wordt toevertrouwd door de kliniek. Dit geldt voor de uitvoering en
interpretatie van de testen, maar ook voor eigen inbreng en verdere adviezen.
Veel testen worden in een maandelijks multidisciplinair overleg besproken.
-
De expertise in (diagnostiek van) endocriene aandoeningen is toegenomen.
-
De expertise in bloedafname technieken is groot: plaatsen van infusen bij
volwassenen en kinderen en arteriepuncties worden regelmatig uitgevoerd.
-
Zowel aanvragers als patiënten zijn zeer tevreden met de manier waarop deze
testen worden uitgevoerd en becommentarieerd.
Voordeel van het uitvoeren op het KCL zelf is ook dat bijvoorbeeld glucose direct en
snel gemeten kan worden op een bloedgasanalyser en niet met een glucosemeter
zoals op klinische afdelingen gebeurt. Inmiddels zijn meer dan 1000 testen
uitgevoerd waarbij slechts enkele malen complicaties optraden, waarbij de
labspecialist zelf therapeutisch handelt. Hiervoor zijn diverse i.v. medicamenten
voorhanden en een tensiemeter voor volwassenen en kinderen. Bij meer dan 50
uitgevoerde insuline tolerantietesten is 1x een hypoglycemisch coma opgetreden, dat
snel gecoupeerd kon worden met i.v. toediening van een 50% glucose oplossing. Er
106
is 1x een allergische huidreactie opgetreden bij een arginine test tijdens een GH
stimulatietest bij een kind waardoor de test gecoupeerd moest worden en een i.v.
antihistaminicum moest worden toegediend. Bij een GH stimulatietest bij een
volwassene (GHRH+arginine) ontstonden direct na de test ernstige cardiale klachten
waarvoor nadere analyse noodzakelijk was. Milde bijwerkingen komen overigens
veelvuldig voor en zijn vaak karakteristiek voor een specifieke test zoals: slaperigheid
bij clonidine, opvliegers bij CRH toediening en mictiedrang bij de TRH test.
Vasovagale reacties worden een enkele maal gezien. Hoewel tensiedalingen bij
clonidine en hypoglycemieën bij arginine gebruik beschreven zijn, zijn deze nooit
waargenomen in onze praktijk.
Samengevat: het zelf uitvoeren en interpreteren van endocriene functietesten biedt
grote voordelen, waarvan de belangrijkste de toename van endocriene expertise, het
directe patiëntencontact, de bekwaamheid in speciale bloedafnametechnieken (i.v.
infuus, arteriepuncties) en de standaardisering zijn. Hierdoor is het vertrouwen in
zowel de consultatieve als direct patiëntgebonden vaardigheden van de labspecialist
sterk toegenomen.
Referenties
1. Schindhelm RK, Franken AAM, Groeneveld PHP, Rondeel JMM: The low-dose
adrenocorticotropic hormone stimulation test for the diagnosis of adrenal
dysfunction: analysis of 220 consecutive tests performed over a period of 3
years. Ned Tijdschr Klin Chem Labgeneesk 2008; 33: 81.
2. Schindhelm RK, Leur JJCM van de, Rondeel JMM: Salivary cortisol as an
alternative for serum cortisol in the low-dose adrenocorticotropic hormone
stimulation test? J Endocrinol Invest 2010; 33: 92-95.
3. Kokshoorn NE, Smit JWA Nieuwlaat WA, Tiemensma J, Bisschop PH, Groote
Veldman R, Roelfsema F, Franken AAM, Wassenaar MJE, Biermasz NR, Romijn
JA, Pereira AM. Low prevalence of hypopituitarism after traumatic brain injury: a
multicenter study. Eur J Endocrinol 2011; 165: 225-231.
107
Voorbeeld 1: een 7-jarig meisje met borstontwikkeling
Functietest: GnRH
Medicatie: geen
Indicatie: borstontwikkeling bij 7 jarig meisje; pubertas praecox?
Uitgevoerd door: mw S Koekkoek-Meijer, PA i.o.
Tijdstip en dosis van inspuiten (T=0): 9.23u; 0.1 mg GnRH i.v.
Bijzonderheden: geen
Resultaten:
T0
T20
T40
T60
T120
LH
0.1
3.2
2.6
3.2
1.5
FSH
2.5
9.1
10.0
10.2
9.5
Referentiewaarden: *kind puberaal: basaal LH > 1.0 U/L, na GnRH stijging tot > 7.6 u/L (meisje) en LH > FSH
Conclusie: prepuberale oploop van LH, maar sterkere stijging van FSH passend bij premature thelarche; geen aanwijzing voor pubertas praecox.
Dr J.M.M. Rondeel
Voorbeeld 2: een 40-jarige vrouw met het syndroom van Albright, hypogonadisme en een
laag IGF-1
Functietest: GH stimulatietest (GHRH+Arginine)
Medicatie: losec 20 mg 2dd1, domperidon 10 mg 3dd2, insulinepomp sc, calci chew D3 1000/800 1dd1, topicorte emulsie cutaan, oxazepam 1-3- dd 1, nutridrink
Indicatie: 40 jarige vrouw met syndroom van Albright; hypogonadisme; laag IGF-1; GH reserve?
Uitgevoerd door: dr J vd Leur
Tijdstip en dosis van inspuiten (T=0): 9.15u; 0.1 mg GHRH i.v. op T0 en van T0-T30 300 ml 10% Arg-HCl per infuuspomp
Bijzonderheden: krijgt na de test pijn op borst, duizelig, hartkloppingen; RR 124/85; pols 74 RA; glucose op T0 17 mmol/L, op T120 13 mmol/L; doorverwezen naar SEH
Resultaten:
GH
T0
0.9
T30 41.0
T60 29.4
T90 17.9
T120 9.1
Referentiewaarden : *GH (mE/L): stijging tot boven 12 mE/L, maximaal na 60-90 min.
Conclusie: vroege en relatief hoge GH piek die weer daalt: zeker geen GH deficientie. Verwezen naar SEH wegens cardiale klachten.
Dr J.M.M. Rondeel
108
JEI_09_194_Schind*.qxp:.
19-03-2010
12:23
Pagina 92
J. Endocrinol. Invest. 33: 92-95, 2010
DOI: 10.3275/6477
Salivary cortisol as an alternative for serum cortisol in the low-dose
adrenocorticotropic hormone stimulation test?
R.K. Schindhelm, J.J.C.M. van de Leur, and J.M.M. Rondeel
Department of Clinical Chemistry, Isala Clinics, Zwolle, The Netherlands
ABSTRACT. Background: Salivary cortisol is unaffected by
cortisol binding globulin and reflects free serum cortisol
as compared to total serum cortisol. Aim: The aim of the
present study was to compare the salivary cortisol response with the serum cortisol response in a low-dose (1μg) ACTH test in a clinical setting and to determine the optimal cut-off value of salivary cortisol as an alternative to
serum cortisol. Material/subjects and methods: We measured serum and salivary cortisol responses to iv administration of 1-μg ACTH in 51 patients (17 males) referred to
the Department of Clinical Chemistry for ACTH-testing.
Serum cortisol was assessed before, 20, and 30 min after
ACTH-administration, and salivary cortisol was assessed
before and 30 min after ACTH administration. Results:
Mean cortisol at baseline, 20, and 30 min were 0.44 μmol/l
(SD: 0.22), 0.64 μmol/l (SD: 0.24), and 0.70 μmol/l (SD:
0.25), respectively. Median basal salivary cortisol was 8.4
nmol/l [interquartile range (IQR): 3.8-14.2]. Salivary cortisol
at 30 min equaled 35.9 nmol/l (IQR: 21.1-46.2). Basal salivary cortisol was significantly correlated with salivary cortisol at 30 min (r=0.53; p<0.001). Salivary cortisol at 30
min of 23.5 nmol/l had a sensitivity and specificity of 78.1%
and 70.0%, respectively as compared to the serum cortisol
cut-off values of >0.50 μmol/l. Conclusions: The salivary
low-dose ACTH-test yields more dynamic responses than
serum sortisol. However, the sensitivity and specificity of
salivary cortisol are too low to be adequate as an alternative to the serum cortisol measurements. In women on esortisol might
mi
trogen therapy, however, the use of salivary cortisol
be superior to serum cortisol.
(J. Endocrinol. Invest. 33: 92-95, 2010)
©2010, Editrice Kurtis
INTRODUCTION
o serum cortisol (9). In addit
addition, tthe sampling of saliva is
to
easy and may pro
provide a st
stress-free and non-invasive alternative to serum ccortisol in clinical p
pr
practice and in research settin
settings (11). A limited number o
of studies reported on the comparison
arison of salivary cort
corti
cortisol with serum cortest (8, 9, 12). However, these
tisol after a low-dose ACTH tes
udiess were perform
performed in selected populations in a restudies
search setting and
an did not report sensitivity and specia
ficity of the
th measurement of salivary cortisol as compared
to sserum cortisol. Therefore, in the present study we compared the salivary cortisol response to the serum cortisol
response in a low-dose (1 μg) ACTH test in a clinical setting and determined the optimal cut-off value of salivary
cortisol as an alternative to serum cortisol.
The low-dose (1 μg) ACTH stimulation test is used for the
detection of primary or prolonged secondary adrenodrenorenocortical insufficiency, correlates well with the insulin-induced hypoglycemia test (1-3) and is a sensitive test fo
for
nction (1).
). Indeed, a recen
recent
the assessment of adrenal function
meta-analysis showed that
at the low-dose ACTH te
test h
had
the highest sensitivity
compared
to the
sitivity
ity and specificity as compar
co
250-μg ACTH
TH in the diagnosis of ad
adrenal insufficiency
ncyy (4).
changes in total serum
In most
most laboratory assay
assays, change
erum cortisol
i.e. the
the sum of th
the biolog
biologically active
ve cortisol and pro
protein-bound
tein
teinbound
ound cortisol are determined
mined
ined after an ACTH
ACT test.
However,
ver, cortisol
co
binding
ding globulin (CBG) may significantly influence total serum cortisol (5), for instance estrogens may increase
ncrease CBG a
an
and therefore increase total
serum cortisol (6). In sa
saliva, cortisol is present in the unbound form and
d may very well reflect the free cortisol
fraction in serum (7). In line with this conjecture, two studies found higher basal and stimulated serum cortisol in
women who used oral contraceptives as compared to
women who did not use oral contraceptives, whereas salivary cortisol was similar in both groups (8, 9). The use of
oral contraceptives may thus hamper the interpretation of
changes in total serum cortisol (10). Moreover, salivary
cortisol after a low-dose ACTH test has shown a higher
stimulatory response and a lower variability as compared
r
t
i
d
K
e
ic
E
E
S
U
,
L
0
A
1
N
O
0
S
©2
PER
FOR
Key-words: Adrenal insufficiency, cortisol, low-dose, saliva.
Correspondence: R.K. Schindhelm, MD, PhD, MEpi, Isala Clinics, Department of Clinical Chemistry, PO Box 10500, 8000 GM Zwolle, The Netherlands.
Accepted June 25, 2009.
First published online July 28, 2009.
s
i
t
ur
LY
N
O
MATERIALS AND METHODS
Patients
In the present study, 51 consecutive patients (17 males; age:
18-90 yr) were included, who were referred to the outpatient facility of the Department of Clinical Chemistry at the Isala Clinics
in Zwolle, the Netherlands for a low-dose (1-μg) ACTH test by
one of the endocrinologists from the Department of Internal
Medicine. The indications for referral to the laboratory were as
follows: chronic fatigue (no.=37), follow-up after pituitary surgery
(no.=4), hypotension (no.=4), Addison’s disease (no.=2), pituitary cyst (no.=1), pituitary function after septic shock (no.=1),
adrenal function after unilateral adrenal extirpation (no.=1),
adrenal function after chronic cortisone therapy (no.=1). The
low-dose (1-μg) ACTH test was performed by one of the laboratory physicians according to protocol. The nature of the ACTH
test was explained by the laboratory physician; in addition, the
patient received a brochure outlining the procedure of the ACTH
test, and all patients consented to the test. For the present
study, approval from the local Ethics Committee was not re-
109
19-03-2010
12:23
Pagina 93
Salivary cortisol and the low-dose ACTH-test
quired, since the ACTH test was performed in a clinical setting
according to prevailing clinical and diagnostic procedures and
guidelines that are in line with the Helsinki Declaration and regulations for Good Clinical and Laboratory Practice.
Methods
The ACTH tests were carried out at 09:00 h. An indwelling cannula (Y-CAN, Beldico, Marche-en-Famenne, Belgium) was inserted into the cubital vein for blood collection and iv injection
of 1-μg tetracosectide, which was freshly prepared by diluting a
250-μg ampoule of tetracosectide (Novartis Pharma, Nurnberg,
Germany) in normal saline to 1 μg in final a volume of 0.5 ml.
Blood was collected before, and after 20 and 30 min of iv injection and saliva was collected by a saliva collecting tube with
a cotton swab (Salivette, Starstedt, New York, NC) before and 30
min after iv injection. Blood samples were centrifuged after 30
min at 3000 rpm for 10 min, and serum was collected in plastic
tubes and stored at –20 C until analyses. The salivary collecting
tubes were centrifuged at 3000 rpm for 10 min and stored at
–80 C until analyses, as described previously (9, 12). A normal response was defined as a total serum cortisol of >0.50 μmol/l, at
20 or 30 min, according to previously published reports (1, 13).
Laboratory analyses
Serum cortisol was measured by a solid-phase competitive
chemiluminescent enzyme immunoassay (IMMULITE 2000 Cortisol, Siemens Medical Solutions Diagnostics, Los Angeles, CA)
with intra- and inter-assay coefficients of variation (CV) of <10%,
and salivary cortisol was determined by a coated tube radioimmunoassay (Spectria, Cortisol RIA, Orion Diagnostics, Espoo,
spoo,
Finland) with intra- and inter-assay CV of <5.5%.
r=0.72, respectively, both p<0.001). Median basal salivary cortisol was 8.4 nmol/l (IQR: 3.8-14.2 nmol/l), whereas salivary cortisol at 30 min equaled 35.9 nmol/l (IQR:
21.1-46.2 nmol/l). Basal salivary cortisol was significantly
correlated with salivary cortisol at 30 min (r=0.56;
p<0.001). Serum cortisol increased 1.5-fold and salivary
cortisol increased 4-fold. No significant differences in
serum and salivary cortisol were observed between men
and women who did not use estrogens (Table 1). Women who used estrogens (no.=9) had higher basal serum
cortisol as well as salivary cortisol as compared to women who did not use estrogens (Table 1). In contrast,
serum cortisol at 30 min was higher in women who used
estrogens compared to women who did not use oral contraceptives. No difference was found in salivary cortisol
between these 2 groups at 30 min (Table 1). A salivary
cortisol at 30 min >23.5 nmol/l had a sensitivity and
specificity of 78.1% and 70.0%, respectively, as compared
to the serum cortisol cut-off values >0.50 μmol/l which
was used to indicate adequate adrenal function (Fig. 1).
(RO
The AUC of the receiver operating characteristic (ROC)d specificity in
curve was 0.78. The optimal sensitivity and
37) were 82.8% and
an
the “chronic fatigue”-group (no.=37)
nmol/
nmo
62.2%, respectively, with a cut-off value of 23.1 nmol/l
and AUC of 0.77.
atients had a serum cortisol
cortis <0.50
Ten out of the 51 patients
μmo (range:
(ra
μmol/l att 30 min [mean: 0.36 μmol/l
0.05-0.49
mol/l) with a mean basal serum
se
μmol/l)
cortisol of 0.21 μmol/l
μ
ba salivary corti(range: 0.03-0.34 μmol/l)].
The mean basal
nmol/
sol in those 10 pati
patients were 3.6 nmo
nmol/l (range: 0.6-9.1
nmol/ with a salivary cortisol at 30 m
min equaling 17.7
nmol/l)
nmol/ (range: 0.7-39.6
.7-39.6 nmol/l). Whe
When applying the salinmol/l
off value of 23.5 nmol/
vary cut-off
nmol/l, 8 out of 10 patients were
assified
ied with adrena
i
classified
adrenal insufficiency.
One patient was diAd
agnosed with A
Addison’s disease with a serum cortisol at
30 min of <0.05 μmol/l and a salivary cortisol at 30 min of
0 7 nmol/l.
0.7
r
t
i
d
K
e
ic
E
E
S
U
,
L
0
A
1
N
O
0
S
©2
PER
Statistical analyses
Data were presented as mean
ean (SD) or as median [interquar
[inte
[interquartile
range (IQR)] in case
variables, or as
se off non-normally distributed varia
percentages.
were tested by Stuges. Differences between varia
variables w
dent’s
den
t’ss t-tests or by Mann-W
Mann-Whitney U tests, as appropriate.
propriate. The
relation
was
relat
on between variables
va
wa expressed
d as correlation coefficoeffi
cients accord
Spearman. The
specificity, and
cien
cient
according to S
he sensitivity, specifi
specificit
area under
predictive value
nder
d the curve (AUC),
C), an indicator of the p
pr
of a test, were calculated.
cut-off value was designatculated. The best cut-o
ed as the point having the maxim
maximal Youden’s index (Youden’s inmaxima
dex = sensitivity + specific
specificity – 1) (14, 15). All analyses were performed with SPSS version 14.0 (SPSS Inc., Chicago, IL). A 2-sided p-value <0.05 was considered as statistically significant.
FOR
s
i
t
ur
LY
N
O
DISCUSSION
In the present study, we compared serum and salivary
cortisol after a low dose ACTH-test.
Following iv administration of 1-μg ACTH, the stimulatory response of salivary cortisol was higher in comparison
with total serum cortisol. These findings are in agreement
with previous studies that reported a more pronounced
RESULTS
The mean (SD) age of the patients was 49.3 (17.8) yr. Age
was negatively correlated with serum basal cortisol with
borderline significance (r=–0.27; p=0.054), but not to
serum cortisol at 20 min and 30 min (r=–0.19; p=0.19 and
r=–0.13; p=0.36, respectively). Furthermore, age was not
significantly correlated with basal salivary cortisol and
salivary cortisol at 30 min (r=–0.11; p=0.45 and r=0.003;
p=0.98, respectively). Mean basal serum cortisol of the
study population was 0.44 μmol/l (SD: 0.22 μmol/l; range
0.03-1.07 μmol/l). Serum cortisol at 20 and 30 min were
0.64 μmol/l (SD 0.21 μmol/l; range: 0.04-1.07 μmol/l) and
0.70 μmol/l (SD: 0.25 μmol/l; range: 0.05-1.26 μmol/l),
respectively. Basal serum cortisol was significantly correlated with serum cortisol at 20 and 30 min (r=0.79 and
Table 1 - Serum and salivary cortisol levels stratified by sex and
use of oral contraceptives.
Cortisol
Males
No.
17
Females
No estrogens
Estrogens
25
9
Serum cortisol (μmol/l)
Basal
0.40 (0.16)
0.36 (0.17)
0.73 (0.20)a
20 min
0.56 (0.20)
0.60 (0.20)
0.94 (0.17)a
30 min
0.56 (0.17)
0.62 (0.20)
1.00 (0.20)a
7.4 (2.4-11.9)
12.8 (7.0-18.1)b
Salivary cortisol (nmol/l)
Basal
30 min
ap<0.001, bp<0.05
8.3 (4.1-13.5)
31.0 (21.6-48.4) 36.2 (17.0-40.8) 36.6 (21.9-54.9)
compared to women not using oral contraceptives.
110
19-03-2010
12:23
Pagina 94
R.K. Schindhelm, J.J.C.M. van de Leur, and J.M.M. Rondeel
ROC curve
1.0
Sensitivity
0.8
0.6
0.4
0.2
0.0
0.0
0.2
0.4
0.6
0.8
1.0
1-Specificity
Fig. 1 - Receiver operating characteristic (ROC)-curve of salivary
cortisol as compared to serum cortisol of >0.50 μmol/l after lowdose ACTH-testing to indicate adequate adrenal function.
increase in salivary cortisol as compared to total serum
cortisol (8, 9, 12). In addition, the interindividual variability of salivary cortisol seems lower than of serum (9). Both
the higher dynamic response and the lower variability
iability
bility
may therefore enhance the diagnostic performance
erformance of
salivary cortisol in the low-dose ACTH
H test. Previous stud
studopulations
ns in research set
seties performed in selected populations
tings reported no sensitivity
coritivity
ity and specificity of salivary
saliv
co
tisol as compared
red to
o serum cortisol (8, 12
12). Ma
Marcus-Perlman and
d co-workers
o-workers reported a stim
stimulat
stimulated salivaryy cortisol >27.6
27.6 nmol/l in 26 o
out of 28 h
healthy subjects
bjects
jects (8), but
repo
reported
orted
rted no other
othe diagno
diagnostic characteristics
cteristics of the sali
salito serum cortisol. In the
vary cortisol
ortiso as compared
com
t present study,
ttudy, w
we report a lower salivary cortiso
cortisol cut-off value, with a relatively
vely low maximum se
sensitivity and specificity of the salivary
livary cortisol as
a compared to serum cortisol. However, relevant differences were found between
women using oral contraceptives as compared to men
and women who did not use oral contraceptives. Therefore, the salivary cortisol seems only preferable to serum
cortisol in patients with significant changes in CBG that
may hamper the interpretation of serum cortisol after the
low dose ACTH test. We found higher basal and ACTHstimulated serum cortisol in women who used oral contraceptives. However, in contrast to previous studies (6,
8), we observed higher basal salivary cortisol in women
who used estrogens, but no difference in salivary cortisol
after low-dose ACTH testing. This seemingly observed
difference in basal salivary cortisol may be due to the relative low sample size of the women who were on estrogen therapy. Estrogens have been shown to increase
CBG and therefore increase total serum cortisol (6). Indeed, a recent study showed that 3 months of estrogen
therapy significantly increased serum cortisol as compared to placebo. Interestingly, the increase in serum
cortisol was attenuated in women who used estrogens in
combination with progesterone. The exact composition
of the contraceptives used by the patients in our study
was not assessed (16).
Some other limitations have to be taken into account.
In the present study, we included patients with a wide
age distribution (18-90 yr of age). Previous studies have
demonstrated that age may affect serum and salivary
cortisol (17, 18). Salivary cortisol seems to be unaffected
by age (17), whereas serum cortisol may decrease in older age (17). The latter finding is not consistent (18) and
may depend on the study population. In our patient
population only a modest and barely significant negative association was found between age and basal serum
cortisol. Therefore, despite the wide age distribution,
our results are valid for the present study population. Finally, the present set-up of the study did not include a
follow-up regarding the definitive diagnosis of the included patients.
In summary, the salivary cortisol response correlated with
the serum cortisol response in a sample of consecutive
ns
patients of an outpatient sample. However, the sensitivioo low to be
ty and specificity of the salivary cortisol are too
m cortisol meame
adequate as an alternative to the serum
ogen therapy,
herapy, however, th
t
surements. In women on estrogen
the
cortiso
use of salivary cortisol might be superior to serum cortisol.
K
e
ic
NOWLEDGMENTS
EDGMENTS
ACKNOWLEDGMENTS
s
i
t
ur
The
he authors thank Gerda
Ger
Beltman for performing the salivary cortisol
Beltma
measurements.
r
t
i
d
REFERENCES
REF
E
E
S
U
,
L
0
A
1
N
O
0
S
©2
PER
1.
2
2.
3.
FOR
4.
5.
6.
7.
8.
9.
LY
N
O
Tordjman
ordjman K, Jaffe A, Graza
Grazas N, Apter C, Stern N. The role of the
microgram)
adrenocorticotropin test in the evaluation
low
ow dose (1 microgra
microgr
of patients w
with pituitary diseases. J Clin Endocrinol Metab 1995,
1301-5.
80: 13
Rasmuson S, Olsson T, Hagg E. A low dose ACTH test to assess the
function of the hypothalamic-pituitary-adrenal axis. Clin Endocrinol
(Oxf) 1996, 44: 151-6.
Dökmetas HS, Colak R, Kelestimur F, Selçuklu A, Unlühizarci K,
Bayram F. A comparison between the 1-microg adrenocorticotropin (ACTH) test, the short ACTH (250 microg) test, and the
insulin tolerance test in the assessment of hypothalamo-pituitaryadrenal axis immediately after pituitary surgery. J Clin Endocrinol
Metab 2000, 85: 3713-9 [Erratum in: J Clin Endocrinol Metab 2001,
86: 3085].
Kazlauskaite R, Evans AT, Villabona CV, et al; Consortium for
Evaluation of Corticotropin Test in Hypothalamic-Pituitary Adrenal
Insufficiency. Corticotropin tests for hypothalamic-pituitaryadrenal
insufficiency: a metaanalysis. J Clin Endocrinol Metab 2008, 93:
4245-53.
Dhillo WS, Kong WM, Le Roux CW, et al. Cortisol-binding globulin is important in the interpretation of dynamic tests of the hypothalamic--pituitary--adrenal axis. Eur J Endocrinol 2002, 146:
231-5.
Qureshi AC, Bahri A, Breen LA, et al. The influence of the route of
oestrogen administration on serum levels of cortisol-binding globulin and total cortisol. Clin Endocrinol (Oxf) 2007, 66: 632-5.
Kirschbaum C, Hellhammer DH. Salivary cortisol in psychoneuroendocrine research: recent developments and applications.
Psychoneuroendocrinology 1994, 19: 313-33.
Marcus-Perlman Y, Tordjman K, Greenman Y, et al. Low-dose
ACTH (1 microg) salivary test: a potential alternative to the classical blood test. Clin Endocrinol (Oxf) 2006, 64: 215-8.
Simunková K, Hampl R, Hill M, Doucha J, Stárka L, Vondra K.
Salivary cortisol in low dose (1 microg) ACTH test in healthy wom-
111
19-03-2010
12:23
Pagina 95
Salivary cortisol and the low-dose ACTH-test
10.
11.
12.
13.
en: comparison with serum cortisol. Physiol Res 2007, 56: 449-53.
Moisey R, Wright D, Aye M, Murphy E, Peacey SR. Interpretation of
the short Synacthen test in the presence of low cortisol-binding
globulin: two case reports. Ann Clin Biochem 2006, 43: 416-9.
Vining RF, McGinley RA, Maksvytis JJ, Ho KY. Salivary cortisol: a
better measure of adrenal cortical function than serum cortisol.
Ann Clin Biochem 1983, 20: 329-35.
Simunková K, Stárka L, Hill M, Kríz L, Hampl R, Vondra K. Comparison of total and salivary cortisol in a low-dose ACTH (Synacthen)
test: influence of three-month oral contraceptives administration to
healthy women. Physiol Res 2008, 57 (Suppl 1): S193-9.
Dickstein G, Shechner C, Nicholson WE, et al. Adrenocorticotropin
stimulation test: effects of basal cortisol level, time of day, and suggested new sensitive low dose test. J Clin Endocrinol Metab 1991,
72: 773-8.
14.
15.
16.
17.
18.
r
t
i
d
Bewick V, Cheek L, Ball J. Statistics review 13: receiver operating
characteristic curves. Crit Care 2004, 8: 508-12.
Obuchowski NA, Lieber ML, Wians FH Jr. ROC curves in clinical
chemistry: uses, misuses, and possible solutions. Clin Chem 2004,
50: 1118-25.
Edwards KM, Mills PJ. Effect of estrogen versus estrogen and progesterone on cortisol and interleukin-6. Maturitas 2008, 61: 330-3.
Ahn RS, Lee YJ, Choi JY, Kwon HB, Chun SI. Salivary cortisol and
DHEA levels in the Korean population: age-related differences, diurnal rhythm, and correlations with serum levels. Yonsei Med J
2007, 48: 379-88.
Rotman-Pikielny P, Roash V, Chen O, Limor R, Stern N, Gur HG.
Serum cortisol levels in patients admitted to the department of
medicine: Prognostic correlations and effects of age, infection, and
comorbidity. Am J Med Sc 2006, 332: 61-7.
K
e
ic
E
E
S
U
,
L
0
A
1
N
O
0
S
2
PER
s
i
t
ur
LY
N
O
FOR
112
ADVISEREN OVER SPECIALISTISCH ONDERZOEK –
HEMOGLOBINOPATHIEËN
HJ Adriaansen, arts klinische chemie
Consultverlening bij hemoglobinopathieën is in drie fasen van het diagnostiekproces
van belang: het opsporen van dragerschap van een hemoglobinopathie (Hb-pathie),
het stellen van de diagnose Hb-pathie en het geven van advies in geval van een Hbpathie of dragerschap daarvan. In alle drie de fasen heeft laboratoriumspecialist een
belangrijke rol.
Opsporen van dragerschap
Het tijdig opsporen van dragerschap van een Hb-pathie is essentieel om gericht te
kunnen counselen, waardoor ernstige vormen van Hb-pathie kunnen worden
voorkomen of in ieder geval zo vroeg mogelijk kunnen worden gedetecteerd.
Daarnaast kan, bij het bekend zijn van dragerschap van een Hb-pathie, onnodige
ijzersuppletie worden voorkomen. Met het eenmalig meten van het hemoglobine
(Hb), inclusief erytrocytenparameters en een Hb-pathie screening (HPLC of Hbelectroforese) is het mogelijk om vrijwel alle dragers van een Hb-pathie op te sporen.
Een dergelijk protocol wordt in Nederland echter niet aanbevolen. Wel wordt sinds
enkele jaren in de neonatale hielprikscreening middels pcr-analyse onderzoek
gedaan naar de aanwezigheid van de HbS mutatie. Opsporen van dragerschap van
een Hb-pathie gebeurt in de regel met een anemieprotocol en HbA1c-analyse
middels HPLC. Met beide diagnostiekprotocollen worden verschillende typen Hbpathieën opgespoord.
Opsporen van Hb-pathie met het anemieprotocol
Hb-pathieën die aanleiding geven tot microcytaire erytrocyten kunnen in het
algemeen goed met het anemieprotocol worden opgespoord. Een laag MCV, in
combinatie met een hoog aantal erytrocyten en niet afwijkende ijzerparameters past
goed bij een alfa- of beta-thalassemie en/of een HbE-afwijking. In dergelijke gevallen
dient een Hb-pathie screening, zo nodig aangevuld met gerichte DNA-diagnostiek, te
worden uitgevoerd. Uiteraard kan dragerschap van een Hb-pathie samengaan met
een ijzergebreksanemie. Door kritisch te kijken naar het aantal erytrocyten in
113
combinatie met het MCV en - indien beschikbaar - parameters die informatie geven
over de hoeveelheid Hb op single cell niveau kan in veel gevallen de verdenking op
een Hb-pathie, met name een thalassemie, worden uitgesproken. In dergelijke
gevallen dient een Hb-pathie screening te worden uitgevoerd. Indien dragerschap
van een Hb-pathie niet kan worden uitgesloten, dient te worden geadviseerd om na
ijzersuppletie het anemieprotocol opnieuw aan te vragen. De interpretatie van de
resultaten van het anemieprotocol voor het opsporen van dragerschap vereist kennis
en ervaring. Dit kan prima worden gedaan door de laboratoriumspecialist, die
aanvullende diagnostiek kan inzetten, de diagnose kan stellen en gerichte adviezen
kan geven. Dat de interpretatie door de laboratoriumspecialist van belangrijke
additionele waarde is wordt geïllustreerd door onderzoek in ons laboratorium. Het
KCHL van de Gelre ziekenhuizen hanteert sinds 1999 het NHG-protocol anemie.
Bloedbeeld en BSE worden bepaald. In geval van een microcytaire of normocytaire
anemie wordt ferritine toegevoegd en bij een verhoogde BSE serumijzer en
transferrine. Tot en met 2007 werden de resultaten zonder interpretatie aan de
huisartsen gerapporteerd. De huisarts kon vervolgens een Hb-pathie screening
aanvragen. Vanaf 2008 interpreteert een laboratoriumspecialist de resultaten en
voegt zo nodig een Hb-pathie screening toe. In dat jaar werden ruim vier keer meer
heterozygote beta- en alfa-thalassemieën en HbE-dragers opgespoord dan in de
voorgaande jaren. Een groot deel van deze personen was ook in de periode 1999 2007 middels het anemieprotocol onderzocht, maar de Hb-pathie was niet herkend.
Sinds 2008 zijn enkele honderden dragers opgespoord. De laatste twee jaren is het
aantal opgespoorde dragers afgenomen, hetgeen suggereert dat een groot deel van
de dragers inmiddels is opgespoord. Op basis van de demografische gegevens van
het aantal inwoners met een Turkse achtergrond in onze regio lijkt deze aanname
gerechtvaardigd.
Opsporen van Hb-pathie met de HbA1c-analyse
Dragers van structurele mutaties in het beta-gen, zoals HbS, HbC en HbD worden in
het algemeen niet opgespoord met een anemieprotocol. Er is, behoudens in geval
van HbE dragerschap, geen microcytose en het aantal erytrocyten is niet verhoogd.
Opsporen van dergelijke afwijkingen is wel mogelijk met de HbA1c-analyse middels
HPLC. Dit betekent dat detectie van dragerschap alleen bij diabeten plaatsvindt.
Desondanks is dit nuttig voor het opsporen van dragers, direct en via het
114
aanvullende familieonderzoek. Verschillende laboratoria zijn de laatste jaren
overgegaan op immunochemische analyse van het HbA1c. Hb-varianten worden
hiermee niet opgespoord. Indien men de beschikking heeft over een HPLC-analyse
verdient het aanbeveling om de eerste HbA1c-aanvraag op beide platforms uit te
voeren. In geval een Hb-variant wordt gevonden, dient deze te worden getypeerd en
het resultaat met een interpretatie en advies te worden gerapporteerd.
Diagnostiek van Hb-pathieën
Diagnostiek van hemoglobinopathieën vereist kennis van de genetica, de Hb-genen,
de Hb-synthese en de verschillende Hb-pathieën. Met deze kennis is het eenvoudig
om te kunnen bepalen welke analyses dienen te worden ingezet en kan in de grote
meerderheid van de gevallen de diagnose worden gesteld. Het protocol van de VHL
is hierbij bruikbaar. Bij de beoordeling van de analyses dient kritisch te worden
gekeken of er mogelijk sprake is van een combinatie van Hb-afwijkingen. In een
enkel geval zijn de bevindingen niet eenduidig of bijzonder en is analyse door een
expertiselaboratorium noodzakelijk. De laboratoriumspecialist interpreteert de
resultaten, zet aanvullende analyses in en geeft een gericht advies aan de
aanvrager. Het verdient aanbeveling om in het rapport een korte uitleg te geven over
het type Hb-pathie en de ernst van deze aandoening.
Advies bij rapportage van een Hb-pathie
Indien een Hb-pathie of dragerschap hiervan wordt gevonden, is het noodzakelijk om
in het rapport een advies op te nemen over diagnostiek bij verwanten en over nader
onderzoek in geval van een zwangerschapswens. Het genoemde VHL protocol is
hierbij bruikbaar. Standaardteksten kunnen worden gebruikt, maar de gekozen tekst
wordt bij voorkeur specifiek voor de betreffende patiënt geformuleerd: bij kinderen
onderzoek bij ouders, broers en zussen; bij ouderen onderzoek bij kinderen en
eventueel kleinkinderen. De patiënt kan middels een brief worden geïnformeerd. Om
te zorgen dat de relevante informatie ook in de toekomst beschikbaar is, krijgt de
huisarts altijd een kopie van de bevindingen en het advies. Naar aanleiding van de
geboorte van een kind met een homozygote betadelta0-thalassemie, waarvan bij
beide ouders in het verleden dragerschap was aangetoond en gerapporteerd, wordt
in ons ziekenhuis Hb-pathie en dragerschap van een Hb-pathie bij de vitale
115
informatie in het EPD vermeld. Het is duidelijk dat de laboratoriumspecialist ook bij
het formuleren en vastleggen van het advies een essentiële rol heeft.
116
International Journal of Laboratory Hematology
The Official journal of the International Society for Laboratory Hematology
ORIGINAL ARTICLE
INTERNATIONAL JOURNAL OF LABORATO RY HEMATO LOGY
Basic haemoglobinopathy diagnostics in Dutch laboratories;
providing an informative test result
J. O. KAUFMANN*, J. W. SMIT † , W. HUISMAN ‡ , R. N. IDEMA § , E. BAKKER*, P. C. GIORDANO*
*Hemoglobinopathies
Laboratory, LDGA, Clinical
Genetics, Leiden University
Medical Center, Leiden, The
Netherlands
†
Clinical Chemistry, LabNoord,
Groningen, The Netherlands
‡
Clinical Chemistry, Medical
Center Haaglanden,
Leidschendam, The Netherlands
§
Clinical Chemistry, Amphia
Hospital, Oosterhout, The
Netherlands
Correspondence:
J.O. Kaufmann, LUMC, HKG,
Zone S-6-P, PO Box 9600,
2300 RC Leiden,
The Netherlands.
Tel.: +31 71 526 9800;
Fax: +31 71 526 8276;
doi:10.1111/ijlh.12038
Received 3 July 2012; accepted
for publication 9 October 2012
Keywords
Carrier testing, haemoglobinopathy, information, genetic
counselling, diagnostic report,
prevention
S U M M A RY
Introduction: After a first survey in 2001, the Dutch Association of
Hematological Laboratory Research (VHL) advised its members to
adopt a basic protocol for haemoglobinopathy carrier detection and
to provide genetic information with all positive results to allow
health-care professionals to inform carriers about potential genetic
risks. This article reports on the compliance with these recommendations and their consequences.
Methods: Clinical chemists of all 106 Dutch laboratories were
invited to answer a survey on patient population, diagnostic
techniques used, (self-reported) knowledge, use and effect of the
additional information.
Results: The average increase in diagnostic output was over 60%
and the recommended basic protocol was applied by 65% of the
laboratories. Over 84% of the laboratories reported to be aware of
the additional recommendations and 77% to be using them. Most
laboratories with limited diagnostic requests were still sending their
cases to other laboratories and included the genetic information
received from these laboratories in their diagnostic reports. The
effect of information on subsequent ‘family analysis’ was estimated
to be between 26 and 50%.
Conclusions: The present study shows an increase in diagnostic potential for haemoglobinopathy over the last decade, especially in the
larger cities. Low ‘family testing’ rates were mostly found in areas
with lower carrier prevalence or associated with local reluctance to
pass the information to carriers. In spite of a dramatic improvement,
too many carriers are still not informed because of lack of awareness
among health-care providers and more education is needed.
INTRODUCTION
Severe haemoglobinopathies (HbP) like sickle cell
disease (SCD) and thalassaemia major (TM) are the
428
most common recessive conditions in man. Although
carriers of HbP are usually either slightly anaemic or
not affected at all and rarely need therapy for their
condition, children of two healthy carriers have a 25%
© 2012 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2013, 35, 428–435
117
J. O. KAUFMANN ET AL. | BASIC HEMOGLOBINOPATHY DIAGNOSTICS IN DUTCH LABORATORIES
chance of being severely affected. In many endemic
countries, couples at risk are informed about their
situation and offered informed reproductive choice
whereas in non-endemic immigration countries, couples at risk are mostly not detected or not informed in
time for prospective prevention [1]. To offer partner
and family testing, the carrier has to be identified and
informed by the primary-care practitioner or other
health-care provider. Several studies and reviews in
Dutch have been published in local journals to inform
general practitioners (GP’s) and clinical chemists on the
gravity of the problem [2–6]. This has resulted in
improved information and the addition of SCD testing
to the newborn screening program (NBS) in 2007,
focussing on morbidity prevention [6–12]. Although all
newborns with SCD or TM are detected, only affected
newborns are reported and preventively treated. Carriers are considered less important and only carriers of
HbS are reported thus far in The Netherlands [13,14].
Parents of affected children are directly referred for
treatment and receive information by a specialist,
either a paediatrician or a geneticist, whereas parents
of carriers receive the test results and information
only from their GP. The GP is supposed to refer both
parents to the local laboratory and refer for genetic
counselling only when the parents result to be a
couple at risk. National and international studies have
shown the lack of awareness of GP’s on genetics,
which may impede proper information transfer to
these carriers [15–17]. Therefore, risk information
should be an integral part of laboratory diagnostics
when carriers are detected and laboratory reports
Ways
Time
Method
Figure 1. Detecting HbP carriers
through different routes
showing different specialists may
refer for HbP diagnostics leading
to uniform policy not only in
patient care, but also in advice
for family testing, the laboratory
being the central and common
factor.
If positive
429
should inform the unaware GP’s to pass correct information on to the carrier and to take the necessary
steps for partner, parents and family analysis, management and prevention.
As the couple-at-risk detection at the NBS level
comes too late for parents with a newborn, more
effective alternatives must be offered. Early diagnostics in young adults is essential, especially for the
b-thalassaemia carriers who, although detectable, are
overlooked during NBS [18]. Moreover, thalassaemia
carriers who are slightly anaemic, often receive
inappropriate iron therapy. Carriers can easily be
identified upon haematological or ethnic indication if
requested by the GP or midwife. In most cases, the
diagnosis is suspected by measuring the basic haematological parameters and by automatic separation and
estimation of the Hb fractions using high-performance
liquid chromatography (HPLC) or capillary electrophoresis [19–21], facilities that are available in many
laboratories. Finally, carriers are also detected by
chance during diabetes mellitus (DM) monitoring for
which Haemoglobin A1C (HbA1C) is measured on
HPLC [22].
Carrier detection before or in early pregnancy have
been for decades routine facilities in many endemic
countries, and are now becoming available in immigration countries also [23–29]. A flowchart showing
the different routes to detect carriers and to offer
prevention through basic laboratory diagnostics, information and counselling is summarized in Figure 1.
Laboratories play a key role in providing GP’s with a
diagnosis and with sufficient information to confirm or
Upon
Indication
Before
conception
Early
pregnancy
By
screening
During
NBS
Before
conception
By
chance
At DM
control
Early
pregnancy
Basic diagnostic protocol: CBC / HPLC or CE + information
Parents / partner
/ family analysis
Parents / partner / family
analysis
Family /
progeny
And
DNA diagnostics in specialized lab in case of suspected couple at risk
And
Genetic counseling for confirmed couples at risk
and prenatal diagnosis if requested
118
430
exclude genetic risk and laboratories must act in
collaboration with reference centres in case of a presumed couple-at-risk. For this, methods, awareness and
skills at the laboratory level are essential and for this
reason, a first survey of all Dutch laboratories started in
2001[30]. Then, based on the data collected at that
time, the Dutch Association for Laboratory Hematology
Research (VHL) advised its members in 2006 to adopt a
basic diagnostic protocol for carrier detection and to
provide additional information with positive results
(recommendations in www.de-vhl.nl) [30].
These short information texts tailored to fit with
diagnostics in children, young adults and in elderly,
advise GP’s how to proceed with the test results. For
instance, in case of a carrier child or a carrier young
adult, both parents or partners are advised to get
tested as well as closely related family members when
they intend to have (more) children [30]. In case of
elderly carriers, the advice will be to test the next
generations (cascade screening).
Since then, training of clinical chemists and laboratory technicians has been intensified and a better
quality control for HbP diagnostics was introduced.
We present in this study, the development of the
diagnostic potential in the Dutch laboratories and the
effect of information on family testing.
M AT E R I A L S A N D M E T H O D S
Laboratories
Clinical chemists of the 106 registered Dutch
haematological laboratories were invited to answer
an 11-page survey covering several aspects of
haemoglobinopathy diagnostics. Invitations were sent
by both letter and e-mail where possible. The survey
consisted of two sections. The first was focused on
the diagnostic output and the technologies used. In
the second section, we inquired on the use of
additional information provided with the positive
test results and on the effect of information upon
family analysis. The second survey started in 2006
and ended in 2008.
Geographical distribution
Responses were categorized by laboratories from small
or large cities and from geographical regions.
Geographical areas were defined as North, West, East
and South regions of The Netherlands.
Statistical analysis
Some laboratories had merged since the 2001 survey.
Therefore, for longitudinal analysis we have merged
the data assuming that the expertise from the most
advanced location was continued in the new setting.
Results of the second survey were aggregated by city
size and region. The questionnaires were collected in
as MS word document analyzed both in MS Excel
(Microsoft Cooperation, Redmond, WA, USA) and SPSS
16.0-18.0 for Windows (SPSS Inc., Chicago, IL, USA).
R E S U LT S
Response
Of the 106 invited haematological laboratories, 62%
answered the second and more extensive survey, this
is less when compared with the 91% of the first,
much simpler, one-page survey in 2001. We received
both surveys’ questionnaires from 58 laboratories,
55% (58/106) from the first and 88% (58/66) from
the second survey and comparative calculation before
and after VHL recommendations have been made
on this cohort. The other results were calculated from
all questionnaires. The regions are described in
Table 1.
Output
In the comparative cohort, we see a general increase in
the diagnostic output of 64%. Before recommendations, the average number of analysis per year per
laboratory was 133 against 218 in the second survey.
The increase was more dependent on city size than on
region. We observed hardly any increase in requests in
smaller towns (<100 000 inhabitants) where only 30%
of the laboratories reported ‘more than one request per
week’ during the first survey against 31% in the second. Conversely, in larger cities, the annual requests
for diagnostics in the ‘more than one request per week’
category increased from 63 to 90%. Over all, the average number of diagnostic requests per laboratory per
year increased from 185 in the first survey to 457 in
the second. Data are summarized in Figure 2.
119
431
Table 1. Population description of the four Dutch regions
Laboratories 1st survey (N)
Laboratories 2nd survey (N)
Laboratories in larger cities (%) 1st survey data*
Inhabitants (N)
Immigrant from higher risk countries (%)†
Laboratories in larger cities (%) 2nd survey data
North
East
West
South
12
7
25
1 702 020
7.6
42.9
16
11
31.3
3 469 857
12.2
36.4
53
29
60.4
7 258 771
23.1
65.5
25
19
36
3 927 344
11.4
36.8
*Large cities are defined as laboratories in towns with >100 000 inhabitants.
†Immigrants are all first- and second-generation inhabitants born outside The Netherlands, Germany and Belgium.
1st survey
(a)
100
90
80
70
60
% 50
40
30
20
10
0
(b)
Total (N = 81)
North (N = 11)
West (N = 52)
No requests
Once a
month
Once a
week
More than
once a
week
East (N = 16)
South (N = 25)
Frequency requests
100
90
80
70
60
% 50
40
30
20
10
0
2nd survey
Total (N = 66)
North (N = 7)
West (N = 29)
No
Less than Once a
requests once a month
month
Once a
week
More
than
once a
week
East (N = 11)
South (N = 19)
Frequency requests
Figure 2. Requests for thalassaemia and sickle cell diagnostics. The first survey (a) before a report was published
advising addition of information on family testing and (b) second survey after, at the start of neonatal screening
for HbP.
Diagnostic methods
The diagnostic techniques recommended during the
first survey are now implemented by the majority of
laboratories (65%). For instance, in the longitudinal
cohort, HbA2 quantification was performed by 46% of
the laboratories during the first survey and is now used
by practically all (97%) that indicate to perform own
analysis. As shown in Figure 3, almost all laboratories
that indicated to measure HbA2 levels use HPLC (86%).
Another frequently used method is PCR or other DNAbased techniques, with an increase from 5% to 15%
[31–33]. Laboratories receiving one request a week or
less kept using external analysis and expertise.
Patient population and indication
Carrier analysis is performed on children by 91% of
the laboratories, 97% on young adults, whereas
75% reported testing women early in pregnancy in
the different regions (Figure 4a). Testing in the
elderly was slightly higher than the latter with 81%
due to the increase in HbA1C testing. The indications
reported for carrier testing are predominately anaemia, an affected family member and an affected
partner. Ethnicity was reported as indication by
57% of laboratories. Remarkably, 16% of the laboratories (10/63) reported that partners of carriers
have not been tested (Figure 4b). In particular, the
laboratories that still do not add the information to
the positive test results (21%) reported not having
partners of carriers tested. Most striking are the
differences between regions in family- and ethnicitybased diagnostics with significantly lower referrals in
Northern and Eastern regions. Furthermore, medical
indication is most often not reported by the participants from Northern regions with low populations
at risk.
120
432
1st survey
(a)
(b)
100
90
80
70
60
% 50
40
30
20
10
0
Total (N = 68)
%
North (N = 6)
West (N = 37)
East (N = 9)
2nd survey
100
90
80
70
60
50
40
30
20
10
0
Total (N = 60)
North (N = 5)
West (N = 27)
East (N = 6)
South (N = 16)
South (N = 18)
Techniques
Techniques
Figure 3. Diagnostic potential before intervention (a) and after (b) divided in the different regions showing the
changes in diagnostic techniques used in the different regions.
PopulaƟon
(a)
100
90
80
70
60
% 50
40
30
20
10
0
North (N = 7)
West (N = 27)
East (N = 11)
Neonates Children
Young
Early
adults pregnant
women
IndicaƟon
(b) 100
Elderly
South (N = 19)
90
80
70
60
% 50
40
30
20
10
0
North (N = 7)
West (N = 27)
East (N = 11)
South (N = 19)
Anemia
Aīected Aīected Ethnicity
No
family
partner
indicaƟon
Figure 4. Population referred for testing based on population and indication. (a)shows the tested population
according to age category in different regions in The Netherlands and (b) the indication for testing.
Information
During the 2001 survey, 63% of the participants
reported not to be aware of the advised information
texts (Figure 5a). Of the 37% who were aware of the
recommendations, 14% considered eventually using
them and 20% was using or intended to use them.
Only 2% did not want to use these additions [27]. In
the latest study, 23% (14/61) still does not add information texts to the diagnosis whereas 49% does, 15%
uses similar information and 13% trusts on different
methods for transferring information (Figure 5b), for
instance, by the initiative of the specialist. Overall,
when comparing both surveys we found that 85% of
the Clinical Chemists not using the information text
in 2001 are using them now. As shown in Figure 5b,
after recommendation, 77% (47/61) of the questioned
laboratories provided some type of additional information, the advised information text, similar information
texts or information provided by the expert centre or
specialist. The increment was particularly large in the
Eastern region of the country.
Family analysis
The frequency of family analysis after information
could not be answered by many laboratories because
no registration was kept (Figure 6). Of those laboratories who could answer this question, 20% reported
that after providing information, an additional family
member was tested in 0–25% of the cases, 17% of the
laboratories estimated this at 26–50%, whereas 5%
estimated 51–75% and 8% even reported 76–100%
family analysis. Surprisingly, when a carrier is found
121
(a)
100
90
80
70
60
% 50
40
30
20
10
0
InformaƟon 1st survey
(b)
Total (N = 86)
North (N = 10)
West (N = 42)
East (N = 12)
Will not
Intent to add Consider Unaware of
include
advised
inclusion of
the
informaƟon informaƟon informaƟon informaƟon
texts
South (N=22)
100
90
80
70
60
% 50
40
30
20
10
0
433
InformaƟon 2nd survey
Total (N = 61)
North (N = 6)
West (N = 27)
East (N = 10)
Not add
Add advised Add similar Use other
advised
informaƟon informaƟon methods of
conveying informaƟon
informaƟon
South (N=18)
Figure 5. Usage and intention to use the advised information in addition to the test results. (a) first survey intent to
use information texts (b) and the actual use of additional information during the second survey.
Uptake advise family tesƟng
100
90
80
70
60
% 50
40
30
20
10
0
Total (60)
North (N = 6)
West (N = 27)
East (N = 10)
South (N = 17)
Family
No
tesƟng advise
not
given
reported
0–25% 25–50 % 51–75 % 76–100
%
Figure 6. Spin-off of family testing after identification
of a haemoglobinopathy carrier. Estimated
percentage of family testing after the information
and advice were added to the test results of the
propositus.
accidentally, for instance during HbA1C measurement,
the majority of laboratories (92% 57/62) report to
have taken action either by advising the clinician or
to proceed with HbP diagnostics on their own initiative. Finally, 50% of laboratories indicated to test
neonates, which can be explained by the follow-up
procedures after implementation of the national NBS
program.
DISCUSSION
Basic laboratory diagnostics, information and counselling are the three cornerstones for the primary
prevention of HbP. Therefore, one of our major
concerns was to provide the Dutch laboratories with
sufficient knowledge for carrier detection, either by
screening or by regular diagnostics upon indication.
A first survey published in 2006 has shown that
the diagnostic output of Dutch laboratories was limited and that important information for genetic risk
was not reaching the carrier. Therefore, the Dutch
Association for Hematological Laboratory Research
(VHL) recommended the use of a basic protocol for
carrier detection and of genetic information to be
added to all positive results. This second survey was
intended to monitor the developments of diagnostic
output and the use and effect of recommendations in
the last decade.
Due to significant changes both in merging and
laboratory management, longitudinal analysis on all
participants was not possible. We were, however, able
to compare both time points based on geographic and
demographic distribution. Herewith, we were able to
show a significant increase in basic laboratory diagnostics and in collaboration with specialized centres when
compared with 2001. We have also found that more
laboratories are performing a pre-screening before
referring to a specialized centre including cases derived
from DM monitoring. The diagnostic techniques used
have also improved significantly in many laboratories
that now refer to specialized centres for ‘difficult cases’
only. The western region has the highest population
density and number of diagnostic requests which coincides with the highest frequency of immigrants living
in the large cities where populations at risk can reach
levels of over 20% (2008 the Central Bureau of
Statistics ‘Statistics Netherlands’ www.cbs.nl).
122
434
Carriers can be detected on indication (medical or
ethnic), through screening or by chance (i.e. HbA1C
measurement), but the time of detection is essential
for offering prevention. Pre-conception screening as
offered in endemic countries but based on ethnic indication, it would require a difficult political decision
and could bear the risk of stigmatization. Moreover, a
complex separate organization would be needed, especially if multiple independent care providers are
included in a screening and the collaboration of GPs
who are in general reluctant to take this task. Conversely, a far more simple early pregnancy screening
can be implemented as part of the existing Rhesus
and infectious diseases screening offered to all pregnant in most developed countries. The level of
compliance in early pregnancy is high and analysis of
the partner of the diagnosed carriers will allow the
identification of couples at risk in time for prenatal
diagnosis [27]. Moreover, partners and family analysis
at the DNA level in specialized laboratories will define
better information on the type of risk for a better
reproductive choice. To improve this, some expert
centres provide additional information using a letter
addressed directly to the patient that is send to the
physician together with the laboratory results. The
letter reassures the carrier but explains also (to the
patient and the physician) the consequences of
REFERENCES
1. Modell B, Darlison M, Birgens H, Cario H,
Faustino P, Giordano PC, et al. Epidemiology of haemoglobin disorders in Europe: an
overview. Scand J Clin Lab Invest 2007;
67:39–69.
2. Giordano PC, Harteveld CL. [Prevention of
hereditary haemoglobinopathies in The
Netherlands]. Ned Tijdschr Geneeskd 2006;
150:2137–41.
3. Elion-Gerritzen WE, Giordano PC, Haak
HL. [The ‘Anemia in the midwife practice’
standard issued by the Royal Dutch
Organisation of Midwives: a risk of not
recognizing iron deficiency and hemoglobinopathy]. Ned Tijdschr Geneeskd 2002;
146:457–9.
4. Giordano PC, Breuning MH. [From gene to
disease; from hemoglobin genes to thalassemia and sickle cell anemia]. Ned Tijdschr
Geneeskd 2000;144:1910–3.
carrying a recessive trait in case of a couple at risk,
providing information on HbP and links to websites
with additional information. Once standardized letters
matching the specific traits and the patient’s age
group are available, this process is almost effortless.
In conclusion, although it is evident that testing
carriers is essential for prevention, we have, however,
found that these types of analyses are still not
requested in all situations and some laboratories even
reported not seeing carriers at all. We believe that
more efforts are still needed not only by the laboratories, but also by the involved authorities to improve
the availability of HbP testing in young adults and in
pregnant women in particular, by screening at the
national level.
AC K N OW L E D G E M E N T S
We are grateful to M.J. Sander (M.Sc.) for her contribution during the initiation phase of this project
which has been conducted on request of the Dutch
Association for Clinical Hematology Research and the
Dutch Clinical Chemistry Association. The present
study has been conducted according to the ethical
regulations of all institutions involved and no conflicts
of interest were present. Part of the study was supported by the Dutch Zon-Mw project nr. 21000.0105.
5. Daemers DO, Amelink-Verburg MP, Rijnders ME. [The standard “Anemia in first
line obstetric practice” from the Royal
Dutch
Organization
of
Obstetricians
(KNOV): risk for not acknowledging iron
deficiency and hemoglobin abnormalities].
Ned Tijdschr Geneeskd 2002;146:1011–3.
6. Dutch National Health Council. Neonatal
Screening. The Hague: Dutch National
Health Council; 2005.
7. Heijboer H, Van den Tweel XW,
Fijnvandraat K, Peters M. [Recognition of
children with sickle cell disease in The
Netherlands]. Ned Tijdschr Geneeskd 2007;
151:2498–501.
8. Bolhuis PA, Page-Christiaens GC. [The
advisory report ‘Neonatal screening’ from
the Health Council of The Netherlands].
Ned Tijdschr Geneeskd 2005;149:2857–60.
9. de Wert GM. [Neonatal screening: dynamics and ethics]. Ned Tijdschr Geneeskd
2005;149:2841–3.
10. Wilson BJ, Forrest K, van Teijlingen ER,
McKee L, Haites N, Matthews E, et al. Family communication about genetic risk: the
little that is known. Community Genet
2004;7:15–24.
11. Giordano PC. Starting neonatal screening
for haemoglobinopathies in The Netherlands. J Clin Pathol 2009;62:18–21.
12. Lanting CI, Rijpstra A, Breuning-Boers JM,
Verkerk PA. [Evaluation of Neonatal Heel
Prick Screening in Children born in 2007].
TNO report KvL/P&Z 2008.119,2008.
13. Vansenne F, de Borgie CA, Bouva MJ, Legdeur MA, van Zwieten ZR, Petrij F, et al.
[Sickle cell disease in heel injection screening II]. Ned Tijdschr Geneeskd 2009;
153:858–61.
14. Mantikou E, Harteveld CL, Giordano PC.
Newborn screening for hemoglobinopathies
using capillary electrophoresis technology:
testing the Capillarys (R) Neonat Fast Hb
device. Clin Biochem 2010;43:1345–50.
123
15. Oyeku SO, Feldman HA, Ryan K, MuretWagstaff S, Neufeld EJ. Primary care clinicians’ knowledge and confidence about
newborn screening for sickle cell disease:
randomized assessment of educational
strategies. J Natl Med Assoc 2010;102:
676–82.
16. Baars MJ, Henneman L, ten Kate LP. Deficiency of knowledge of genetics and
genetic tests among general practitioners,
gynecologists, and pediatricians: a global
problem. Genet Med 2005;7:605–10.
17. Lubin IM, Caggana M, Constantin C, Gross
SJ, Lyon E, Pagon RA, et al. Ordering
molecular genetic tests and reporting
results: practices in laboratory and clinical
settings. J Mol Diagn 2008;10:459–68.
18. Mantikou E, Arkesteijn SG, Beckhoven van
JM, Kerkhoffs JL, Harteveld CL, Giordano
PC. A brief review on newborn screening
methods for hemoglobinopathies and preliminary results selecting beta thalassemia
carriers at birth by quantitative estimation
of the HbA fraction. Clin Biochem
2009;42:1780–5.
19. Old JM. Screening and genetic diagnosis of
haemoglobin disorders. Blood Rev 2003;17:
43–53.
20. van Delft P, Lenters E, Bakker-Verweij M,
de Korte M, Baylan U, Harteveld CL, et al.
Evaluating five dedicated automatic devices
for haemoglobinopathy diagnostics in
multi-ethnic populations. Int J Lab Hematol 2009;31:484–95.
21. Clarke GM, Higgins TN. Laboratory investigation of hemoglobinopathies and thalassemias: review and update. Clin Chem
2000;46(8 Pt 2):1284–90.
22. Aleyassine H. Low proportions of glycosylated hemoglobin associated with hemoglobin
S and hemoglobin C. Clin Chem 1979;25:
1484–6.
23. Bozkurt G. Results from the north cyprus
thalassemia prevention program. Hemoglobin 2007;31:257–64.
24. Cao A, Rosatelli MC, Leoni GB, Tuveri T,
Scalas MT, Monni G, et al. Antenatal diagnosis of beta-thalassemia in Sardinia. Ann
N Y Acad Sci 1990;612:215–25.
25. Streetly A. A national screening policy for
sickle cell disease and thalassaemia major
for the United Kingdom. Questions are left
after two evidence based reports. BMJ
2000;320:1353–4.
26. Streetly A, Latinovic R, Henthorn J.
Positive screening and carrier results for
the England-wide universal newborn sickle
cell screening programme by ethnicity and
area for 2005-07. J Clin Pathol 2010;63:
626–9.
27. Giordano PC, Plancke A, Van Meir CA,
Janssen CA, Kok PJ, van Rooijen-Nijdam
IH, et al. Carrier diagnostics and prevention
of hemoglobinopathies in early pregnancy
in The Netherlands: a pilot study. Prenat
Diagn 2006;26:719–24.
28. Najmabadi H, Ghamari A, Sahebjam F,
Kariminejad R, Hadavi V, Khatibi T, et al.
29.
30.
31.
32.
33.
435
Fourteen-year experience of prenatal
diagnosis of thalassemia in Iran. Community Genet 2006;9:93–7.
Kaufmann JO, Demirel-Gungor G, Selles A,
Hudig C, Steen G, Ponjee G, et al.
Feasibility of nonselective testing for
hemoglobinopathies in early pregnancy in
The Netherlands. Prenat Diagn 2011;31:
1259–63.
Giordano PC, Smit JW, Herruer M, Huisman W, Pouwels JGJ, Verhoef N, et al.
[Carrier Diagnostics and prevention of
Sickle Cell Disease and Thalassemia Major;
Recommendation of the hemoglobinopathies
workgroup].
Ned.Tijdschr.Klin.Chem.Labgeneesk 2006;31:301–
5.
Phylipsen M, Prior JF, Lim E, Lingam N,
Vogelaar IP, Giordano PC, et al. Thalassemia in Western Australia: 11 novel deletions characterized by Multiplex Ligationdependent Probe Amplification. Blood Cells
Mol Dis 2010;44:146–51.
Munkongdee T, Vattanaviboon P, Thummarati P, Sewamart P, Winichagoon P,
Fucharoen S, et al. Rapid diagnosis of
alpha-thalassemia by melting curve analysis. J Mol Diagn 2010;12:354–8.
Phylipsen M, Chaibunruang A, Vogelaar
IP, Balak JR, Schaap RA, Ariyurek Y,
et al. Fine-tiling array CGH to improve
diagnostics for alpha- and beta-thalassemia
rearrangements. Hum Mutat 2012;33:
272–80.
124
Ned Tijdschr Klin Chem Labgeneesk 2010; 35: 211-213
Hemoglobinopathiediagnostiek: de toegevoegde waarde van ‘reflecterend testen’
door laboratoriumspecialisten
W.P.H.G. VERBOEKET-van de VENNE, W.P. OOSTERHUIS, M.P.G. LEERS en H.A. KLEINVELD
Inleiding
Eén van de vormen van consultverlening door het
klinisch-chemisch laboratorium betreft het toevoegen
van testen en/of commentaar aan een laboratoriumaanvraag (1). Bij deze werkwijze interpreteert de laboratoriumspecialist afwijkende uitslagen en beoordeelt
of aanvullende testen nodig zijn. De doelstelling van
deze werkwijze is om de diagnostiek op zinvolle wijze
te completeren. Bovendien kunnen aanvullende testen
vaak worden uitgevoerd in het al aanwezige bloedmonster, zodat een tweede bloedafname achterwege
kan blijven. In het Verenigd Koninkrijk beschouwt
men deze procedure (ook wel ‘reflective testing’ of
‘reflecterend testen’ genoemd) als integraal onderdeel
van de dienstverlening (2, 3). In juni 2006 is de Afdeling Klinische Chemie en Hematologie van het Atrium
Medisch Centrum Parkstad in Heerlen gestart met het
aanbieden van deze service bij laboratoriumaanvragen
van huisartsen.
Onderzoek heeft aangetoond dat huisartsen in oostelijk
Zuid-Limburg het op prijs stellen dat ons laboratorium
het initiatief neemt om testen en commentaren toe te
voegen (4). Bovendien wordt deze werkwijze vrijwel
altijd als zinvol ervaren. Volgens de betreffende huisartsen wordt het patiëntbeleid in meer dan de helft van
de gevallen op een positieve manier beïnvloed, bijvoorbeeld door een snellere diagnose of behandeling,
eerdere verwijzing naar een specialist, of aanpassing
van medicatie.
Hemoglobinopathieën zijn een groep van erfelijke
aandoeningen waarbij de aanmaak en/of functie van
het hemoglobinemolecuul verstoord is. Door mutaties
in de globinegenen kan de genexpressie verminderd
zijn, zoals bij de α- en β-thalassemieën. Daarnaast
kan ook de structuur van de betreffende genproducten, de globinen, zijn aangetast, waardoor abnormale
hemoglobinen ontstaan, zoals HbS, HbC, HbD en
HbE. Verschillende combinaties van deze relatief vaak
voorkomende erfelijke kenmerken veroorzaken de
ernstige sikkelcelziekte en β-thalassemia major. Hemoglobinopathieën zijn endemisch in landen rond de
Middellandse Zee, Zuidoost-Azië, het Midden-Oosten
en Afrika. Door migratie worden ze echter steeds va-
Afdeling Klinische Chemie en Hematologie, Atrium Medisch Centrum Parkstad, Heerlen
ker ook in ons land gezien. Vroegtijdige en correcte
diagnosestelling, ook van de heterozygote vorm, is
belangrijk. Dit voorkomt de vaak onnodige en soms
schadelijke toediening van ijzerpreparaten. Bovendien
vormt het de basis voor een genetisch advies aan families ten behoeve van primaire preventie.
Onderzoek naar hemoglobinopathie wordt door huisartsen meestal aangevraagd op basis van een belaste
familieanamnese met betrekking tot hemoglobinopathieën. In de huidige studie hebben we onderzocht of
de procedure ‘reflecterend testen’ meerwaarde heeft
bij het opsporen van patiënten met een hemoglobinopathie. De belangrijkste voorwaarden voor aanvullend
onderzoek naar hemoglobinopathie zijn een persisterend microcytair bloedbeeld, al dan niet in combinatie
met een normaal of verhoogd ferritinegehalte, zoals
beschreven in het door ons laboratorium gehanteerde
alternatieve stroomschema voor anemie (5). Bovendien is er vaak, maar niet altijd, sprake van allochtone
herkomst.
Methode
Uit ons laboratoriuminformatiesysteem (LIMS) zijn
eerstelijns aanvragen in de periode 2004-2009 geselecteerd waarbij typering van hemoglobine in bloed is
aangevraagd. Tot en met 2008 is dit uitgevoerd middels HPLC (JASCO-200 serie voor HPLC, HPLC-kolom PolyCAT A Chromsystems GmbH); vanaf 2009
met behulp van capillaire elektroforese (Capillarys 2,
Sebia). Indien de resultaten hier aanleiding voor gaven, is aanvullend DNA-onderzoek ingezet (Sanquin
Diagnostiek Amsterdam, Leids Universitair Medisch
Centrum), meestal naar α-thalassemie. Hierbij werd
het bloed van de patiënt onderzocht op de zeven meest
voorkomende deletietypen voor α-thalassemie. Bij
analyse van de resultaten is vervolgens onderscheid
gemaakt tussen een gerichte aanvraag (via de huisarts)
of een door het laboratorium geïnitieerde aanvraag
(door tussenkomst van de laboratoriumspecialist).
Resultaten
Het totale aantal hemoglobinopathieanalyses neemt
toe sinds de invoering van het ‘reflecterend testen’ in
2006 (figuur 1). Het aantal aanvragen via de huisarts
vertoont een lichte stijging, terwijl het aantal door de
laboratoriumspecialist geïnitieerde aanvragen aanzienlijk toeneemt: dit bedraagt 50% van het totale aantal analyses in 2006 (7/14), 74 % in 2007 (29/39), 69
% in 2008 (27/39) en 75 % in 2009 (43/57).
Ned Tijdschr Klin Chem Labgeneesk 2010, vol. 35, no. 3
125
60
tussenkomst van de laboratoriumspecialist. In totaal
werd 14 keer een Hb-variant gevonden: heterozygoot
HbC, HbE, Hb Lepore of HbS. Bij 7 patiënten werd
een Hb-variant in combinatie met α-thalassemie gevonden. Bij 20% (12/60) van de aanvragen afkomstig
van de huisartsen werden geen afwijkingen vastgesteld, in tegenstelling tot 7% (7/106) van de aanvragen
geïnitieerd door de laboratoriumspecialist. Ten slotte
kon bij 56 patiënten (nog) geen diagnose vastgesteld
worden. Dit had te maken met het feit dat huisartsen
geen nader onderzoek hebben ingezet (of de patiënt
zelf wenst geen nader onderzoek), ondanks het advies
van de laboratoriumspecialist.
43
Aantal analyses
50
40
29
27
30
20
7
10
0
8
9
7
2004
2005
2006
10
2007
12
2008
14
2009
Jaar
Discussie
De procedure ‘reflecterend testen’ leidt tot een aanzienlijke toename van het aantal uitgevoerde hemoglobinopathieanalyses en het aantal gevonden hemoglobinopathieën. Met name β-thalassemie en α-thalassemie,
al dan niet in combinatie met een Hb-variant, worden
veelvuldig opgespoord door tussenkomst van de laboratoriumspecialist. Deze resultaten zijn in overeenstemming met een onderzoek van Adriaansen et al.
(6), waarbij het effect van interpretatie van resultaten
van het anemieprotocol door een laboratoriumspecialist werd bestudeerd. Zij beschreven een verdrievoudiging van het aantal hemoglobinopathieanalyses
ten behoeve van de eerste lijn, door tussenkomst van
de laboratoriumspecialist. Bij deze additionele hemoglobinopathieanalyses werd bovendien veel vaker
een hemoglobinopathie gevonden (73%, 2008) dan
bij de door de huisarts zelf aangevraagde hemoglobinopathie-analyses (32%, 2008; 39%, 2007). In onze
Figuur 1. Het aantal hemoglobinopathieanalyses naar aanleiding van een gerichte aanvraag van de huisarts of een door het
laboratorium geïnitieerde aanvraag (door tussenkomst van de
laboratoriumspecialist); klinische chemie, huisarts.
Behalve het aantal uitgevoerde hemoglobinopathieanalyses is ook onderzocht wat de uitkomst van de
analyses was. Met andere woorden: bij hoeveel van
de uitgevoerde analyses is daadwerkelijk een hemoglobinopathie vastgesteld? In de periode 2004 – 2009
werd bij 52 patiënten een heterozygote β-thalassemie
aangetoond (tabel 1; HA: huisarts, KC: klinische
chemie), het overgrote deel hiervan (n=44; 85 %) op
initiatief van de laboratoriumspecialist. De diagnose
α-thalassemie (homozygoot type 2, heterozygoot type
2 of heterozygoot type 1) werd bij 13 patiënten vastgesteld; wederom in 85% (n=11) van de gevallen na
Tabel 1. Aantal gevonden hemoglobinopathieën in de periode 2004-2009 in het adherentiegebied van het Atrium Medisch Centrum
Parkstad in Heerlen. HA: huisarts; KC: klinische chemie.
β-thalassemie
HA
2004
2005
2006
2007
2
2
1
2
4
16
KC
α-thalassemie
HA
Hb-variant en α-thalassemie
1
Geen afwijkingen
1
10
14
2
2
6
HA
1
2
3
KC
2
1
5
2
2
2
HA
1
KC
Overige
2009
2
KC
Hb-variant
2008
HA
2
1
KC
1
1
HA
1
1
10
2
3
2
5
7
6
1
5
8
KC
Geen diagnose te stellen
HA
(verder onderzoek noodzakelijk)
KC
6
4
14
126
evaluatie over de periode 2004-2009 vinden we eenzelfde trend: het percentage van gevonden hemoglobinopathieën door de laboratoriumspecialist bedroeg
67% (71/106) ten opzichte van 33% bij aanvragen van
huisartsen (20/60). In een studie van Lansbergen et
al. (7) werd over een periode van 12 jaar (1996-2008)
bij 2426 patiënten hemoglobinopathieanalyse aangevraagd. Bij 27% (649/2426) van de patiënten werd
inderdaad een hemoglobinopathie gevonden. Bij dit
onderzoek werden ook de hemoglobinopathieanalyses meegenomen die verricht zijn naar aanleiding van
HbA1c-metingen. In ons onderzoek is deze categorie
buiten beschouwing gelaten, aangezien de primaire
indicatiestelling in dat geval verschilde.
Concluderend kunnen we zeggen dat de procedure
‘reflecterend testen’ van grote toegevoegde waarde kan
zijn in het kader van hemoglobinopathiediagnostiek,
gezien de toename van het aantal gevonden hemoglobinopathieën. Een belangrijk verbeterpunt voor ons laboratorium, dat inmiddels is ingevoerd, betreft het toevoegen van DNA-onderzoek naar α-thalassemie aan
het hemoglobinopathieonderzoek. Hierdoor wordt het
aantal patiënten waarbij (nog) geen diagnose is gesteld
hoogstwaarschijnlijk minder. Met deze aanpassing
hopen we de hemoglobinopathiediagnostiek bij eerstelijns patiënten uit de regio oostelijk Zuid-Limburg
verder te optimaliseren.
Referenties
1. Oosterhuis WP, Raijmakers MTM, Leers MPG, Keuren
JFW, Verboeket-van de Venne WPHG, Munnix ICA,
Kleinveld HA. Consultfunctie: van klinisch chemicus naar
laboratoriumspecialist. Ned Tijdschr Klin Chem Labgeneesk 2009; 34: 214-218.
2. Le Roux CW, Bloom SR. Clinical authorisation: what is
best for the patient? Ann Clin Biochem 2003; 40: 113-114.
3. Simpson WG, Twomey PJ. Reflective testing. J Clin Pathol
2004; 57: 239-240.
4. Oosterhuis WP, Keuren JFW, Verboeket-van de Venne
WPHG, Soomers FLM, Stoffers HEJH, Kleinveld HA.
Eigen inbreng van het laboratorium. Ned Tijdschr Geneeskd. 2009; 153: 2138-2144.
5. Oosterhuis WP, Van der Horst M, Van Dongen K, Ulenkate
HJLM, Volmer M, Wulkan RW. Prospectieve vergelijking
van het stroomschema voor laboratoriumonderzoek van
anemie uit de NHG-standaard ‘Anemie’ met een eigen, inhoudelijk en logistiek alternatief stroomschema. Ned Tijdschr Geneeskd. 2007; 151: 2326-2332.
6. Adriaansen HJ, Remijn JA, Leurdijk HJ, Van Suijlen JDE.
Interpretatie van resultaten van het anemieprotocol door
een laboratoriumspecialist resulteert in een sterke toename
van het aantal nieuw gevonden patiënten met een hemoglobinopathie. Ned Tijdschr Klin Chem Labgeneesk 2009;
34: 90.
7. Lansbergen GWA, Van Rooyen-Nijdam IH, Versteegh FG,
Kok PJMJ, Thomas JMMH, Giordano PC. Evaluatie van
hemoglobinopathiediagnostiek in de regio Midden Holland. Ned Tijdschr Klin Chem Labgeneesk 2009; 34: 91.
127
AUTOMATISERING CONSULTVERLENING:
CONSULTREGISTRATIE IN EEN LIS
M Oostendorp, laboratoriumspecialist klinische chemie i.o.
Consultverlening is een immer belangrijke competentie van de klinisch chemicus,
zoals ook blijkt uit het meerjarenbeleidsplan 2009-2013 “Van meten naar consult”, de
prominente plaats die consultverlening inneemt binnen de opleiding nieuwe stijl en
de recent aangenomen richtlijn “Consultverlening door specialisten laboratoriumgeneeskunde (klinische chemie)”. Naast het geven van een goed advies, is ook de
correcte registratie van een consult in het elektronisch patiëntendossier van belang.
Hierdoor is het zowel binnen het laboratorium als in de kliniek duidelijk wat er is
geadviseerd en worden interpretatiefouten voorkomen.
Binnen het Laboratorium Klinische Chemie en Haematologie van het UMC Utrecht
worden consulten sinds juni 2010 geregistreerd in het laboratoriuminformatiesysteem. Na een pilotfase bestaat sinds januari 2011 de mogelijkheid om consulten
door te sturen naar het EPD van de patiënt. Op basis van praktijkervaringen is de
consulttool verder ontwikkeld. Zo kunnen consulten momenteel zowel via een order
als via een bepaling worden gerapporteerd, is er een autorisatie/supervisiemodule
ontwikkeld voor consulten van de KCio’s en zijn alle consulten van één patiënt bij
elkaar terug te vinden in het laboratoriuminformatiesysteem.
In de lezing zal worden ingegaan op de praktische aspecten van de consultregistratie
via het LIS. Welke problemen kunnen voorkomen en welke verbeteringen zijn er
gaandeweg doorgevoerd? Wat wordt er geregistreerd en wat niet? Hoe is de
traceerbaarheid van de consulten geborgd? Naast het directe belang voor de
patiëntenzorg, levert consultregistratie ook een grote hoeveelheid data op. In de
lezing zal kort worden besproken hoe deze informatie kan worden gebruikt voor
managementdoeleinden.
128
Reinvention
Going beyond
*)-.
the Performance of
the Laboratory
Test
#"&5*)-.6565"65
This is an edited transcript of Dr. Laposata’s keynote address to
laboratory professionals at the ASCP Leadership Exchange, March
20, 2009, in Philadelphia.
I
]'5!)#(!5.)5.&%5.)35)/.5"&*#(!5*"3-##(-5"))-5
."5,#!".5.-.-5(5)5."5,#!".5."#(!5 .,5."5,-/&.5#-5
!(,.85),55&)(!5.#'515.")/!".5."#-51-5."5$)5) 5."5
*.")&)!#-.85/--51".>5.5#-55$)5 ),5&&5) 5/-5#(5&),.),35
'##(85)')(5"-5.)5-.*5/*5(55(50).5 ),5
."5*.#(.51")-5*,.#&5.",)')*&-.#(5.#'5BC5#-5nn5
-)(-5(51")5#-5().5.,.5),,.&385.]-5.%55&))%5.5
!)#(!53)(5."5*, ),'(5) 55&),.),35.-.8
Description of the Problem
5."#(%5.".50,35*"3-##(51")5,#0-55,-/&.5.".5
."5&#(#&5&),.),35"-5!(,.5%()1-51".5.)5)51#."5#.85
"(55-.,.5&0#(!5#(.)5."#-5*,)&'655.")/!".5*"3-##(-5
(5"&*5)(&351#."5."5)0#)/-&35)'*&#.5."#(!-85/.5
'(35*"3-##(-5"05-%5'65 ),52'*&65".5)-5(5
(),'&5&#0,5 /(.#)(5.-.5'(>55Ļ-5,5-)'5-#-5
1",65# 515!05*"3-##(-55&#..&5'),5#( ),'.#)(65#.51)/&5
'%55"/!5&#(#&5#Ŀ,(85
(53)/5#'!#(5."5(.)'#5*.")&)!#-.5-(#(!55
,-.5#)*-35%5.)5."5-/,!)(5(5-3#(!65^
/-.5&&5'5
# 53)/5"055+/-.#)(5)(5.".8_5)/(-5,##/&)/-85(53)/5
#'!#(5!..#(!555-(5(5"0#(!5."5,#)&)!#-.5-365
^]&&5!#05.".5,#!".5%5.)5."5#(.,(#-.65(5# 5.",]-55
+/-.#)(65."3]&&5&&8_5(5&),.),35'##(6515-#'*&35!#05
*"3-##(-55(/',5#(-.5) 5"0#(!5.".5#(.,.#)(5#(51"#"5
15-3-.'.#&&35.%5&&5."5#( ),'.#)(5)/.5."5*.#(.5(5
.&&5."5*"3-##(51".5#.5'(-8
".5# 5151(.5.)5."5*"3-##(-5(5-#5."55
#-5nn5-)(-855.53)/5."5*)*&5#(5."5&#(#&5&5)/&5
-35#.5'#!".55 ,)'5"*,#(5(5'(35*)*&51")51),%5
)(5."5("5#(5)!/&.#)(5)/&5#(.# 351"#"5 .),-5
1"(5 ##(.5'%5."55&)(!85"5*"3-##(-5'35
().5%()165/.53)/5)85
5&&5"05%()1&!5.".5
&#(##(-5(85)'")1515"05
.)5!.5.".5,)!(#.#)(5(5"&*5
&#(##(-5-5*.#(.50).-85)-.5
*.")&)!#-.-5%()15#.]-55*,)&'5#(5
."5&#(#&5&),.),#-85"()&)!#-.-5
,5().5-50#-#&5-5*.")&)!#-.-85Ļ5
*.")&)!#-.-5"05.)5&5."5135(5
)*(5."5)),-5 ),5."()&)!#-.-5.)5
*,)0#5)(-/&..#050#85(35
*.")&)!#-.-5-365^]5&#%5.)55'),5
)(-/&..#05(5&51#."5."5*.#(.-5
(5"&*5*"3-##(-8_5)10,65')-.5
*)*&51")5)5(.)'#5(5&#(#&5
*.")&)!35)(].5,&&35&,(5."5&#(#&5
*.")&)!385"35#-5.".>5./&&365.",5
#-55!))5(5$/-.#ŀ&5,-)(85 53)/5
,55*.")&)!#-.65#.5#-5-#,5.)5!.5*#5
),5)#(!5(.)'#5*.")&)!385
(5."5'#7goof-5."5
!)0,('(.5#5.".5#.51)/&5*35
),5*,) --#)(&5#(.,*,..#)(5) 5-)'5
&#(#&5&),.),35.-.-85)-.5) 5."-5
.-.-5,5'#,)-)*#-5),52'#(.#)(5
) 5!&-5),5!,'5-.#(-85Ļ5!)0,('(.5
,.5)-5(5*/.55^hl_5 .,5
."'85 5."5^hl_5#-5.%(51365.".]-5
1".5."5!)0,('(.5*3-5 ),5)#(!5
."5.-.85/.5# 5."5^hl_5#-565.".]-5
1".5."5!)0,('(.5*3-5."5*,-)(5
),5#(.,*,.#(!5."5.-.85)5.",51-5
')(35#(5."#-5 ),55*.")&)!#-.5#(.,*,.#(!5."5,-/&.85
Ļ#-5,&&35"(!5'35*,.#5-55&#(#&5*.")&)!#-.A5
)/&5./&&35,(5')(35 ),5&#(#&5.#0#.#-5 ,)'5*3),-<5
),5."5ŀ,-.5.#'5#(5'35&# 6535.%#(!5&&5.")-5&)..#(!5.-.5
129
,-/&.-65&))%#(!5.5."'65,#(!5."5'#--#)(5(5#-",!5
().-65(5-3#(!65^5%()151".5."5*.#(.5"-6_551,).5
5*,!,*"5)/.5#.5(55"%5'855#(].5'%5-5
'/"5-5)#(!5(.)'#5*.")&)!365/.5.",51-5(5#()'5
-)/,5(5.".5'55#Ŀ,(855.))%5.".5')(35(5
"#,55,-#(.5)(.)5."5 /&.35(5-/(&3515"5.1)5
*)*&5.)5-#!(5)/.5)!/&.#)(5-./#-5&&5."5.#'85)15
(5)(5*,-)(5)5#.5# 5"5),5-"5.,0&-5),5#-5)(50.#)(>5
2*,.-5,)--5."5&#(#&5,-5,5(65-5#(5(.)'#5
*.")&)!365.)5*,)0#5)(-#-.(.5-,0#85
".5&#(##(-5),,5 ,)'5."5&),.),35(51".5."35
2*.5#-5().5'/"95."351(.5(5/,.5,-/&.85Ļ351(.5."5
-5.)5585Ļ351(.5,)/.#(5.-.-5.)55)'*&.5
^
^
,)--5."5(#.5..-855"055&).5.)5)51#."5,!,5.)5
#!()-.#5.-.#(!5-,0#-85(35*)*&5-35*"3-##(-5),,5
.))5'(35.-.-65/.51".5#-5,&&35"**(#(!5#-5.".5.))5'(35
) 5."5,)/.#(5.-.-5,5#(!5),,85Ļ-5.-.-5,5),,535
*"3-##(-51")5-#'*&35),,5.".550,35385".5!))5
#-5.".>5),52'*&651"(55*.#(.5#-5hk53,-5)&5(5!.-55
&))5&).5#(5"#-5),5",5&!65')-.5&#(##(-5"05()5#51".5
.)5)5 ),55"3*,)!/&#&#.351),%/*85"%5 ),5",#.#&#.35) 5
5.",)').#5,#-%>5".5#-5.".65.),5>5.),55#(>5-5
."5!(.#5.-.5(>5".]-5.#0.5*,).#(55,-#-.(>5
Ļ-5,5,#.#&5+/-.#)(-85(5."5 .5#-5-)5'(35
*"3-##(-5),,#(!5."5.-.-5"05()5&/5/-5.",5#-5$/-.5
.))5'/"5.)5%()1<5)5."35!/--85Ļ35),,55.),55#(-.5
&#(##(-5"05()5#5
")15'/"515%()185.5#-5
.#'5.)5"(!5.".8
1#."#(55,-)(&5./,(,)/(5.#'85Ļ351(.5/-5.)55(#5
.)5."'85/.5.".]-5)/.5#.85",]-5."50#>5",]-5."5
,+/-.5.)5ŀ!/,5)/.51"35."5.-.-5,5(),'&>5&#(##(-5
"05()5#5")15'/"515%()185.5#-5.#'5.)5"(!5.".8
5,5().5)#(!5.".51&&51#."5&#0,35) 5"&."5,5
) 55.),55#(:5."-5,5.1)5.).&&35#Ŀ,(.5.-.-5)5
'(35'#-.%-5,5'5#(5),,#(!5.".50(5# 5."5*.")&)!#-.5
)-5."5.-.5*, .&365."5*"3-##(5(].5!.5.)55#!()-#-85
Ļ/-65.",5#-5&--7."(7)*.#'&5+/&#.35) 5,5(5#(,-5
)-.5.)5'(!5."5*.#(.85Ļ5*.#(.5-.3-5#(5."5")-*#.&5
130
)&/'5h5R5--/5i5R55
/&35hffo5
l8i53-5#(-.5) 5k8n65(5."#-5)-.-5'),5')(3855"05.)5
,%5."53&8
5)(5"55-5) 551)'(51")5"55&)15*,).#(565
(5",5)-..,##(5#(].5%()15.".5*,).#(55!)-5)1(5#(5
0,35*,!((3855.)&5."5*.#(.5.".5# 5-"5#(].5.,'#(.5
."#-5*,!((365-"5)/&5#5 ,)'55&))5&).85)5-"5
.,'#(.5",5*,!((38
(5*.#(.5"55.),55ŀ#(35(5"#-551-5
)(-#-.(.&35*,)&)(!85Ļ5(/,)-/,!)(5#(].5%()151".5
.)5)65-)5"5!05."5*.#(.5 ,-"5 ,)4(5*&-'5(5."(5
*, ),'5."5)*,.#)(85Ļ5ŀ,-.535."5*.#(.51-55
/-5."5-/,!)(5"5!#0(5."5*.#(.5 ,-"5 ,)4(5*&-'5
.".5.))%5,5) 5"#-5 .),55ŀ#(3651#.")/.5%()1#(!5.".5
"5"5."#-5ŀ#(385Ļ5*&-'51-5."(5!#0(5#(5#(-/Ń#(.5
+/(.#.35)(5."5(2.5385Ļ5-)(535"5&5)(5)."5-#-5
) 5"#-5"5(5()15"5(5",&351&%5),5.&%85
Ļ#-5#-51".5"**(-51"(5&#(##(-5)(].5%()151".5
.)5)51#."5."5,-/&.-85")5-/Ŀ,-5# 515)(].5)5."#->5Ļ,5
#-55#!5,."5) 52*,.-5#(5&#(#&5*.")&)!385.]-5().5."5
*.")&)!#-.-51")5-/Ŀ,5# 515)(].5*,)0#50#5)(5&#(#&5
*.")&)!385")5,&&35-/Ŀ,-5#-5."5*.#(.85)')35"-5
.)55."5*.#(.50).855"05.)5,&#45.".51".51],5
*,)--#(!5#-55-'*&5 ,)'55*.#(.85.]-5().5$/-.55-'*&85
(5.".]-5#Ń/&.65-*#&&35#(5."5&,!5'#5(.,-85
Ļ)-5) 53)/51")5"05(5,)/(551"#&5'35"05
(5!(,&#-.-5(5,1(5."5*.#(.-]5&))85'',5.".5
)((.#)(5.".50,3)(5/-5.)5"0>5)153)/],5-#..#(!5#(5
,)(.5) 5(5(&34,5*,)--#(!5gnf5-*#'(-5(5")/,85Ļ5$)5
#-5.).&&35#Ŀ,(.85)'")15#(5."5'#-.5) 5.".6515"05.)5
/(,-.(5.".5.",]-5-.#&&5-)')35)(5."5).",5(51")]-5
,&&35-#%85
".5"**(-5# 5."5*"3-##(5"))--5."51,)(!5.-.5),5
'#-#(.,*,.-5."5,-/&.>5.5.%-5."'5&)(!,5.)5!.5(5(-1,85
Ļ5*"3-##(5#-51)(,#(!51".5.)5)51#."5."5(-1,85(5
*.")&)!#-.-5,5-%5.)5)5.-.-5.".5) .(5"05()5&#(#&5
0&/A*"3-##(-5,5),,#(!5/((--,35-50,35
35#(-.5) 51),%#(!5/*5."5--5.".5."35,&&35(5.)5
0&/.85],5-*(#(!55&).5'),5.#'5)(5.-.-5.".5,5().5
!)#(!5.)5"&*5,"55#!()-#-85Ļ5#'*&#.#)(-5) 5."#-5,5
#''(-65#(5.,'-5) 5)&&,-85
'!#(53)/5,5."5&#(##(65(55hk73,7)&51)'(5
1(.-5.)5-.,.5),&5)(.,*.#0-65/.5-"]-51),,#5.".5# 5
-"5"-55!(.#5,#-%5 .),5 ),5.",)')-#-65#.5)/&5&5.)5
5&).85)5."5*"3-##(5-3-65^5(5.)5),,5."5.-.-5 ),5
"3*,)!/&#&#.385".5,5."3>_5(53)/5!#05."'55
,+/#-#.#)(95/(,-5) 5.-.-5.)5-&.5 ,)'65)(5*!5g65*!5
h65*!5i85^#%5."5,#!".5-#25.-.-8_55Ļ,5#-5()5")*5 ),5&&5
."51&&7#(.(.#)(5&#(##(-85(5#.51#&&55&&5."5'),5
"&&(!#(!5."5(2.5')(."5/-5.",5#-5&13-5'),5(15
#( ),'.#)(85
)')35#(5."5&#(#&5&),.),35%()1-51".5
"3*,)!/&#&#.35.-.-5.)5-&.85)/]5*,)&355&5.)5
#(.# 35 )/,5),5ŀ05.-.-85)5.",53)/5,95)/5%()15#.85Ļ5
&#(##(5"-5()5%()1&!85)15)-5#.5!.5 ,)'53)/5.)5."'>5
,5.",5*).(.#&5-)&/.#)(->5-85.]-5&))%5.5")15.)5
"&*5."'5),,5."5,#!".5.-.85)&/.#)(-5"05(5#(5*&5 ),5
."#-65/.5().5/(# ),'&35**&#5 ),5.5&-.5gk53,-85(5."5
(-1,5#-5,ł25.-.#(!5(51,#..(5.-.5,-/&.5#(.,*,..#)(-85
Reflex Testing
#."5,ł25.-.#(!655*"3-##(5"%-55)25)(55
,+/#-#.#)(5.".5-3-5-)'."#(!5&#%65^3*,)!/&&5
1),%/*65),5*,)&)(!551),%/*85&-5)5&&5."5,#!".5
.-.-5.)5ŀ!/,5#.5)/.8_5Ļ(5."351)/&(].5"05."5#!5
,+/#-#.#)(-85Ļ.]-51".515,.51"(551-5#(5)-.)(95
^"%5."#-5)265(5151#&&5)5."5,-.8_551#&&5.%5)/.5
."5"*,#(5(5-5# 5.".]-5."5.,#0#&5(-1,85 5().651]&&5)5
."#-5'#2#(!5-./365(5# 5#.5!)-5 ,)'5lf5.)5if651]&&5)5."-5
.),-65656565(565(5&))%5 ),55 .),5ŀ#(385
(5# 5."5'#2#(!5-./35)-(].5),,.651]&&5."#(%5)/.55
&/*/-5(.#)!/&(.5),5(55#("##.),85(51"(5.",5#-55
(15&/*/-5(.#)!/&(.5.-.65!/--51")5."351#&&5)'5.)5(5
.,35.)5-&&5#.5.)>5-65."5&),.),35*)*&851],5."5)(-51")5
,5!)#(!5.)5)*.#'#45."#-8
.]-5().5",5.)5-.5/*5,ł25.-.#(!85)/5(5-#!(5."5
&!),#."'-53)/,-&0-85)/5"05.)5!)5 ),5."5#&5)&#35
)''#..5(5-365^,5,53)/,5")#-85)/5(5"))-5
,)'5."#-5#!5.-.5'(/65),51]&&5*/.5#(5&!),#."'-85"#"5
1)/&53)/5&#%>_5))35"-50,5-#5.)5'65^5)(].51(.5
."5&!),#."'-8_50,3)35-#'*&355-#65^&-5*/.5."5,ł25
.-.#(!5#(8_5]05(0,5"5)(5,+/-.5.".51-5./,(5)1(85
)/],5"&*#(!5*"3-##(-5-#&35"))-5."5,#!".5.-.85
Test Result Interpretations
Ļ5).",5(5#-5.)5"&*5*"3-##(-5/(,-.(51".5
.)5)51#."5."5,-/&.-855(5.)5*,)0#5*"3-##(-51#."55
*.#(.7-*#ŀ5(,,.#05 ),5(350&/.#)(51#."5)'*&2#.35
B().5 ),5-65 ),52'*&C65.)5'#'#51".5#-5)(5#(5
(.)'#5*.")&)!35(5,#)&)!385(5(.)'#5*.")&)!365
."35)(].5.%5.".5,-.5#)*-365&))%5.5."5.2.))%65(5
-365^",]-5.".5*,!,*"5)(5/.&5,-.5,#()'>_5(5
131
)&/'5h5R5--/5i5R55
/&35hffo5
."(5#(-,.5.".5.2.5-5."5,*),.85/.5.".]-51".515)5#(5)/,5
/-#(--85)/5"05.)55'),50&/85 53)/5$/-.5-365^(5-(5
#(5."5"')!&)#(55,!#)(6_5.".]-5().50&/78
5"05*,)').5."#-5)(*.5 ),55&)(!5.#'65(5'(35
*)*&5"05-#5.)5'5 .,1,-5.".5."35)5#(.,*,..#)(-85
(55-%5."'5.)5.&&5'5")15."35)5#.85)'5-35."35
*,)0#5."5((5)''(.-85(55&).5) 5*)*&5-35."35)5
#(.,*,..#)(-5&&5365/.5)(&351"(5."35."5*"3-##(-55&&5
."'85)/5(].51#.5/(.#&5.",]-55+/-.#)(65")10,65/-5
5-/,!)(5#-5!)#(!5.)5."#(%5"5),5-"5"-5!).5#.52.&35,#!".85
Ļ5)(*.5) 51#.#(!5 ),55+/-.#)(5.)5)5(5#(.,*,..#)(5
#-5-#&&385)/5(].5)5#.5 ),5."5-#'*&-.5.-.-5.".5,5().5 /,.",5
1),%5/*65&#%50,355(585)/5"05.)5!)5/*5.5&-.5
)(5()."5#(5)'*&2#.385Ļ5 .5#-65."5*"3-##(-5,5!/--#(!85
)-.5*)*&5,5!)#(!5.)55."50#.#'-5) 51".0,5."5
*"3-##(5)-(].5%()185"#"5%()1&!&5*,-)(5#(5."5
&#(#&5&),.),35#-5!)#(!5.)5-.*5 ),1,>5
5#55-/,035#(5gool51"#&551-5.5."5--"/-..-5
(,&5)-*#.&855-(.5)/.5gff5-/,03-5(5,#05jh5
,-*)(--855.))%5(5#(.,*,..#)(5) 55)!/&.#)(5-,#-5
) 5.-.-65-(.5."'5%5.)5."5*"3-##(65(5-#65^ 53)/5
,'',5."#-5-651#&&53)/5(-1,5."-5 )/,5+/-.#)(->_
#,-.5+/-.#)(95"(515!053)/5."5#(.,*,..#)(65#5
3)/5ŀ!/,5)/.51".5."5*.#(.5"5-))(,>5# .37.1)5*,(.5
-#53-85)(5+/-.#)(95#53)/5)5 1,5&),.),35.-.-5
/-515!053)/5(5#(.,*,..#)(>5&')-.5.",7+/,.,-5
-#53-85)165",]-5."5-./((,855-%5."'5# 5."351)/&5
"05'55'#-.%5# 515#(].5!#05."'5."5#(.,*,..#)(85
)/.5.",7+/,.,-5-#53-85)1655,&&35,-*.5'35
)&&!/-85Ļ35,5*"()'(&5*"3-##(-65/.5."#-5-")1-5
.".5#.]-5#'*)--#&5.)5%()150,3."#(!85Ļ5 .5#-65&),.),35
.-.#(!5"-5$/-.5!)..(5.))5#!85#(.37#!".5*,(.5) 5."5
*"3-##(-515-/,035-#515-")/&5%*5*,)0##(!5."5
#(.,*,..#)(-85Ļ#-51-55-/,*,#-#(!&35 0),&5,-*)(-8
5#55-/-+/(.5-/,035#(5hfff65(5*"3-##(-5
,*),.5!#(5.".5."5#(.,*,..#)(-5-.)**5."'5 ,)'5
'%#(!55'#-.%65-"),.(5."5.#'5.)5#!()-#-65(5
,/5&),.),35.-.#(!85)'5&-)5-#5.".5-55,-/&.5) 5."5
#(.,*,..#)(65."35#5().5*, ),'55*,)/,85)'5*.#(.-5
#(].5"05.)5)'5.)5."5")-*#.&85)'5*.#(.-5!).5 1,5
'#.#)(-85)'5*.#(.-5#(].5!.5.,(- /-8
5&-)5----5."5#'*.5)(5"'.)&)!#-.-51"(515
*,)0#5(5#(.,*,..#)(5)(555) 5nn5-)(-85.5ŀ,-.5
-)'5"'.)&)!#-.-51,5/*-.51#."5/-65/-5."351,5().5
!..#(!55)(-/&..#)(651"#"5#-5")15."35'%5."#,5&#0#(!85
.,55 15')(."-65")10,65."35,&#45.".5&&5.")-5
*.#(.-51")5"5&#(!5(5&)..#(!5#-),,-5(5(0,5!).5
5#!()-#-65()15!).5)(5/-515#5."5#(.,*,..#)(65(5
."35!).55, ,,&5."35).",1#-51)/&5().5"085Ļ5*.#(.5
'#!".5"05&52--#0&35 ),5."5(2.5if53,-51#."5()5
#!()-#-855,#!".7-#45)(-/&..#)(85.5'55#Ŀ,(8
)'5'#&5(.,-5-35."35)(].5"05()/!"5--5.)5
#(.,*,.85(5)!/&.#)(65."5ŀ,-.51%515)(&35#5(#(5--5.5
."5--"/-..-5(,&5)-*#.&850,5-0,&53,-5151(.5
.)5if5--55385 5."5-,0#5#-5(0,5#'*&'(.655())35
0,5%()1-5#.-50&/5(5#.5(0,5"-5(5)**),./(#.35.)5!,)18
)*&5,5)'#(!51,5) 5'#&5,,),-5,&.5.)5
&),.),35.-.-85)5")15)-55&),.),35-/*,0#-),5)(.,#/.5
.)5."5-)&/.#)(5) 5."#-5&#(#&5*,)&'>5-55*.")&)!#-.65
5%()155&#..&5#.5)/.5."5*, ),'(5) 5."5.",)'#(5
.#'5#(5)!/&.#)(65/.55)(].5%()15."5."(#&5.#&-855
)/&(].5)5."5.-.85),./(.&36515,5&&5&#(%-5#(55"#(65
(5)/,5)&&.#052*,#(5'%-5/-5*)1, /&85)')35
#(5."5&#(#&5&),.),35%()1-5."5.#&-5.)5*, ),'5."5
.-.52.&35,#!".85(5#.#)(65")10,65."5."()&)!#-.5&-)5
%()1-5.".5."5.",)'#(5.#'5#-5--)#.51#."5&#(!5
(5.".5#.]-5)((.5.)5."5,*.#&-5.#'85Ļ#-5'(-5.".5
."5."()&)!#-.5"-5-)'5&#(#&5%()1&!5) 50&/5.)5."5
*.#(.65-51&&5-5."(#&5%()1&!85)5.",5#-5-)'5/- /&5
)0,&*5#(5)/,5%()1&!85(5.".]-5!,.85(35*.")&)!#-.-5
%()15)(&355&#..&5#.5)/.5."5&#(#&5&),.),385Ļ50,!5
*.")&)!#-.5)-(].5"055!,.5&5) 5%()1&!5)/.51".5
.)5)51#."5.".5.",)'#(5.#'85
)')35"-5.)55."5*.#(.50).85)')35
()185)')351")5%()1-51".53)/5%()15.)35"-5.)5
-365^5%()151".5.)5)5.)5 /,.",50&/.5.".55) 5nn5
-)(-8_55."#(%55-)&/.#)(5"-5.)55#(#0#/&#45 ),5"5
")-*#.&:5."()&)!#-.-65*.")&)!#-.-65(5)/.-#52*,.-5"05
.)5-365^5"05.)5)'5/*51#."55*&(5.)5#(.,*,.5."5.-.5
,-/&.-8_55Ļ#-5#-5."5)**),./(#.35.".515&&5"05(51#.#(!5
),95.")&)!#-.-65&),.),35-/*,0#-),-65'#&5."()&)!#-.-65
(5'#&5."(##(-5&&51),%#(!5-55.'5.)5,"550#.&&35
#'*),.(.5&#(#&5!)&85 515!.5#(0)&05#(5."-5&#(#&5--5
#(5."5135]05-,#65."5*,*.#)(5) 5."5&),.),351#&&5
"(!5(5."50&/5) 5&),.),35*,) --#)(&-5#(5&#(#&5,5
1#&&55("(85
Dr. Laposata is Professor of Pathology and Medicine, Vanderbilt
University School of Medicine; Pathologist-in-Chief and Director of
Clinical Laboratories, Vanderbilt University Hospital, Nashville, TN.
132
AUTOMATISERING CONSULTVERLENING:
RIPPLEDOWN ALS TOOL – IN DE PRAKTIJK
MWM Schellings, laboratoriumspecialist klinische chemie i.o.
Reflecterend testen is een proces waarbij de laboratoriumspecialist de uitslagen van
een laboratoriumaanvraag analyseert en eventueel testen en/of commentaar
toevoegt aan de originele aanvraag. Deze extra service van de laboratoriumspecialist
wordt gewaardeerd en draagt bij aan het verbeteren van het diagnostisch proces (1).
Naast de voordelen van reflecterend testen voor de patiëntenzorg zijn er ook nadelen
te
benoemen
die
grootschalige
invoering
van
reflecterend
testen
kunnen
belemmeren, te weten:
x
De tijdsinvestering van de laboratoriumspecialist
x
Inter-individuele
variatie
tussen
laboratoriumspecialisten
(zowel
voor
toevoegen testen, als voor interpretatief commentaar)
x
De kosten van de extra testen
x
Een moeizame automatisering in het LIS
De laboratoriumspecialisten van het Maxima Medisch Centrum hebben besloten, na
positieve ervaringen vanuit Heerlen en Den Bosch, om het reflecterend testen in te
voeren voor eerstelijnsdiagnostiek. Er werden enkele voorwaarden gesteld: 1) elke
laboratoriumspecialist handelt volgens dezelfde protocollen, 2) de tijdsinspanning
van de laboratoriumspecialist moet zo efficiënt mogelijk zijn en 3) het moet
beschikbaar zijn voor alle aanvragende huisartsen. Aan deze voorwaarden kan
voldaan worden wanneer het reflecterend testen wordt ondersteund door software,
waarbij zowel het aanvragen van een extra test, als het genereren van patiëntspecifiek commentaar automatisch kan worden uitgevoerd.
RippleDown is een ‘expert-system’ waarbij de basis wordt gevormd door een (klein)
aantal regels (rule-based). De meest voorkomende reflextesten en commentaren
kunnen op basis van deze regels gegenereerd worden. Echter, tijdens de
handmatige validatie in de module Validator kan de laboratoriumspecialist een casus
tegenkomen, waarbij het commentaar niet of maar ten dele juist is. Met behulp van
RippleDown
kan
de
laboratoriumspecialist
het
automatisch
gegenereerde
133
commentaar aanpassen en deze wijziging doorsturen naar de module Knowledgebuilder (Figuur 1). In deze module kan de laboratoriumspecialist dan regels bouwen
die van toepassing zijn op de verworpen casus. Op deze manier vindt er een
verfijning en uitbreiding van de regels plaats tijdens het gebruik van RippleDown
(case-based), waardoor er passend commentaar voor elke casus gegenereerd kan
worden. Een overzicht van het gehele validatieproces met behulp van RippleDown is
te zien in Figuur 1.
Figuur 1: Flow-chart reflecterend testen met behulp van RippleDown
RippleDown is geïmplementeerd door de laboratoriumspecialisten van het Maxima
Medisch
Centrum
voor
het
reflecterend
testen
van
eerstelijnsdiagnostiek,
aangevraagd op basis van de probleemgeoriënteerde aanvraag volgens de LESAstandaard. Alle dienstdoende laboratoriumspecialisten valideren de commentaren in
de module Validator en twee laboratoriumspecialisten kunnen de regels aanpassen
134
in de module Knowledge Builder. Validatie van de commentaren gebeurt handmatig
door de laboratoriumspecialist, maar er kan worden ingesteld welk percentage
commentaren nog voor validatie worden getoond, zodat niet alle normale
labuitslagen gezien worden. De invoering van RippleDown heeft ervoor gezorgd dat
het
reflecterend
testen
is
ingevoerd
voor
alle
huisartsen
(>150)
in
het
adherentiegebied van het Maxima Medisch Centrum.
Referenties
1. Oosterhuis WP, Keuren JF, Verboeket-van de Venne WP, Soomers FL, Stoffers
HE, Kleinveld HA. Eigen inbreng van het laboratorium – huisartsen positief over
‘reflecterend testen’. Ned Tijdschr Geneesk 2009;153:A486.
135
The Ideal Laboratory Information System
Jorge L. Sepulveda, MD, PhD; Donald S. Young, MD, PhD
Context.—Laboratory
information systems (LIS) are
critical components of the operation of clinical laboratories. However, the functionalities of LIS have lagged
significantly behind the capacities of current hardware
and software technologies, while the complexity of the
information produced by clinical laboratories has been
increasing over time and will soon undergo rapid
expansion with the use of new, high-throughput and
high-dimensionality laboratory tests. In the broadest
sense, LIS are essential to manage the flow of information between health care providers, patients, and
laboratories and should be designed to optimize not
only laboratory operations but also personalized clinical
care.
Objective.—To list suggestions for designing LIS with the
goal of optimizing the operation of clinical laboratories
while improving clinical care by intelligent management of
laboratory information.
Data Sources.—Literature review, interviews with laboratory users, and personal experience and opinion.
Conclusions.—Laboratory information systems can improve laboratory operations and improve patient care.
Specific suggestions for improving the function of LIS are
listed under the following sections: (1) Information
Security, (2) Test Ordering, (3) Specimen Collection,
Accessioning, and Processing, (4) Analytic Phase, (5)
Result Entry and Validation, (6) Result Reporting, (7)
Notification Management, (8) Data Mining and Crosssectional Reports, (9) Method Validation, (10) Quality
Management, (11) Administrative and Financial Issues, and
(12) Other Operational Issues.
(Arch Pathol Lab Med. 2013;137:1129–1140; doi:
10.5858/arpa.2012-0362-RA)
S
some of these systems have deficiencies that most homeand Web-based software have long overcome, such as
efficient navigation, rapid response, and spelling correction.
Modern clinical laboratories are purveyors of information,
in the form of laboratory results, which may be numbers,
text, graphs, or other images, together with interpretative
data, to assist health care providers in delivering optimal
patient care. The complexity of the information produced by
clinical laboratories has been increasing over time, and with
the advent of large-scale analytic techniques, such as
microarrays and next-generation sequencing, the amount
of data produced will rapidly grow by several orders of
magnitude. Advanced developments in data management
and bioinformatics will need to be incorporated into LIS for
these large data sets to become clinically useful. In addition,
the ability to query large cross-sectional laboratory databases (data mining) is increasingly used to improve the
quality and efficiency of health care delivery. These 2
tendencies mandate an ever-expanding capacity and processing need for LIS and supporting hardware.
Increasingly, the focus of efforts to improve the quality of
laboratory operations is shifting from the analytic phase,
which currently presents few problems, particularly for those
tests performed by highly automated instruments, to
preanalytic and postanalytic aspects of laboratory testing.6
Advanced LIS and associated database and expert systems7
will be critical to the goal of improving the quality of the
extra-analytic aspects of laboratory testing, including the
implementation of paradigm-shifting innovative approaches.
The goal of this article is to list ideas for designing or
improving a state-of-the-art LIS from the perspective of
practicing laboratory professionals, focusing on optimizing
ince the 1970s, laboratory information systems (LIS)
have been critical components of the operation of
clinical laboratories. They were initially developed to collect,
record, present, organize, and archive laboratory results,
often with a focus on generating information for proper
financial management of the laboratory. While information
technology in general is advancing at an increasingly faster
rate, particularly in the hardware domain but also in
software development, LIS have not evolved correspondingly. For example, the current LIS make very limited use of
artificial intelligence approaches such as neural1 or Bayesian2 networks, fuzzy logic,3 genetic algorithms,4 and artificial
immune recognition systems.5 Health care systems in
general can be characterized as conservative and resistant
to change and current health care information systems
(HIS) and LIS are a reflection of this conservative approach.
Despite the potential benefits in cost-efficiency and patient
care improvements possible with well-designed HIS/LIS,
most of these systems lag significantly behind the possibilities afforded by current information technology. In fact,
Accepted for publication August 30, 2012.
Published as an Early Online Release December 5, 2012.
From the Department of Pathology & Cell Biology, Columbia
University, New York, New York (Dr Sepulveda) and the Department
of Pathology & Laboratory Medicine, University of Pennsylvania,
Philadelphia, Pennsylvania (Dr Young).
The authors have no relevant financial interest in the products or
companies described in this article.
Reprints: Jorge L. Sepulveda, MD, PhD, Department of Pathology
& Cell Biology, Columbia University Medical Center, PH1590B, 622
W 168 St, New York, NY 10032 (e-mail: [email protected]).
Arch Pathol Lab Med—Vol 137, August 2013
The Ideal LIS––Sepulveda & Young 1129
136
Modules contributing to the ideal laboratory information system.
the operation of the clinical laboratory and on improving
clinical care by intelligent management of laboratory
information. The focus of this article is to describe desirable
LIS functionalities, while the operational and technical
details on how to achieve this idealized LIS are beyond the
scope of this article. In this work LIS is defined broadly, and
some specific functionalities listed might be provided by
software modules not strictly considered as LIS. These
would include clinical ordering and reporting systems as
components of the HIS, analyzer built-in software, specimen processing and management software (often labeled as
‘‘middleware’’), financial, inventory, and personnel management packages, and others (Figure). In this article we are
interested in describing the desired functionalities independently of which particular software package will be
responsible for their availability. Although the ideas
discussed in this article are mostly applicable to all sections
of the clinical laboratory, some specific issues pertaining to
anatomic pathology, microbiology, molecular and genomic
testing, transfusion medicine, and cell/tissue/blood banking
are beyond the scope of this article.
quality of care should have access to certain information on
all patients. Different levels of security should be available,
and the system should allow users to establish workgroups
with user-definable sets of functions and access to data, as
exemplified in Table 1.
Secure interfaces to the LIS should include advanced login capabilities, for example, by biometric recognition or
radiofrequency identification devices (RFIDs), which minimize keystrokes and log-in time, while providing quick
automated log-out upon leaving the workstation. In certain
secure locations, the system should have the ability to
continuously display live reports of laboratory testing (eg,
pending ‘‘STATs’’ in the laboratory or patient results in the
operating room) without multiple log-in requirements. The
system should have the ability for remote log-in and access
to ordering and reporting systems, for example, via a secure
Web browser, allowing providers and laboratorians to
access the LIS from any location and from mobile and
handheld devices.8,9 The system should allow flexible,
reliable, and informative electronic signatures for authentication of data and documents.
INFORMATION SECURITY
Health care information systems must be secured from
unauthorized internal and external access and preserve the
confidentiality of health records according to applicable law
and regulations without hindering the functionality for
legitimate users. For example, health care providers should
be able to access all the relevant information for their
patients, but not that of other patients unless they are
brought in as consultants. Individuals involved in assessing
TEST ORDERING
Test ordering is the step most amenable to intervention in
order to improve appropriate use of laboratory resources
(laboratory utilization). Test-ordering systems coupled to
intelligent decision-support systems have the potential to
reduce turnaround times and length of stay, and guide
providers toward optimized test utilization,10,11 and can be a
function of either LIS or HIS, sitting at the border between
LIS and HIS. Regardless of which system is used, immediate
1130 Arch Pathol Lab Med—Vol 137, August 2013
The Ideal LIS––Sepulveda & Young
137
Table 1. Categories of Laboratory Information Systems (LIS) Users
Information manager
Health care provider
Technical staff
Manager
Laboratory director
Patient
Full access to all functionalities, including lower-level processes of the system, ability to design scripts and
routines to customize functionalities to local needs.
Order tests, attach comments to orders, define alerts, and view results, with the ability of customizing
reports and interpretative information according to their needs.
Process orders and specimens, perform tests, record results, attach comments to results, and perform quality
and other laboratory management activities.
Produce and review reports and statistics, personnel management activities, inventory, write and review
procedures and other documents.
Ability to design and review all of the activities in his or her area, including access to patient information,
quality management data, document review and management, and cross-sectional reports.
Depending on institutional policy, the system (LIS or HIS) should provide direct access of laboratory test
results, reports, and interpretative comments to patients, eg, through a secure Web-based browser
interface.
Abbreviation: HIS, health care information systems.
feedback must be provided to the user. As in other
situations at the interface between laboratory and clinical
practice, involvement of both clinicians and laboratorians is
essential for the development of the policies and rules
guiding laboratory utilization. The most useful systems are
those that require the health care provider to directly enter
the order in the system, therefore affording the possibility of
interaction between the system and the provider (computerized provider order entry or CPOE systems). An important
consideration for the success of a CPOE system is to
properly design it to maximize usability and match routine
work processes used by providers.12 The following is a list of
desirable functionalities in a test ordering system.
1. The system should receive inputs from the HIS or from
the ordering provider (when the information is
unavailable or inaccurate in the HIS) to include the
following:
a. Ordering provider
i. Name (mandatory).
ii. Specialty.
iii. Address (if in a different location). Interfacing
with credentialing and privileges databases is
desirable to provide the most current provider
information.
iv. Contact media (e-mail address, desk and mobile
telephone numbers, pager number, etc) for
routine notification (mandatory).
v. Contact information for critical result notification
(pager, cell phone, and surrogate contact for
nonbusiness hours), including links to institutional notification cascades/call schedules appropriate to a specific patient (mandatory).
vi. Additional providers/provider team and other
persons legally authorized or desired by care
provider to receive results.
vii. Further notification requests (eg, notify when
results are available, or when results exceed
reference range, critical limits, or custom threshold limit). Ability to establish notification criteria
by institutional, departmental, or other group
policy. Ability to select notification media,
including HIS overriding alert, HIS alert on
patient record, e-mail, short message service
(SMS) text message, automated phone call,
beeper, telefax, and others. For certain critical
areas, for example, operating rooms, new significant results should trigger an audible alert.
Certain notifications (eg, of critical results) must
include a mechanism that ensures fail-safe
notification, returns acknowledgment of receipt
of the information, and allows for escalation of
unacknowledged notifications according to a
preestablished protocol.13
b. Patient information
i. Patient identification (last and first names and
institutional or social security number) or unique
coded audit trail if necessary (eg, for research or
environmental specimens).
ii. Patient demographics, including date of birth/
age, sex (male, female, transgender), race,
ethnicity, and prior names.
iii. Patient location (permanent address and current
location if hospitalized).
iv. Codified diagnoses (preliminary diagnosis by
Diagnostic-Related Groups, International Classification of Diseases (ICD)-9 or ICD-10, where
appropriate) and other relevant clinical information (‘‘reason for study’’).
v. Codified results of nonlaboratory tests.
vi. Height, weight, vital signs.
vii. Medications (with dosing and date/timing of
administration).
viii. Herbal and other supplements.
ix. Diet and meal times (to determine fasting time).
x. Medical procedures applied to the patient,
including surgical interventions and radiologic
procedures.
xi. Gynecologic and obstetric information.
xii. Other pertinent clinical information.
c. Order information
i. Test(s) requested.
ii. Source(s) of specimen requested.
iii. Date/time of order.
iv. Day/time of requested collection (begin, end).
v. Repeat frequency (for standing orders, if institutionally permitted).
vi. Special patient preparation instructions.
vii. Urgency of the test (categories customizable to
institutional needs).
viii. Collection responsibility (patient mail-in, pointof-care, ward or nursing unit, routine phlebotomy rounds, laboratory collection, etc).
ix. Other free-text comments and instructions to
the laboratory.
2. An expert system uses patient information, previous
test results, and clinician input (eg, from a list of
138
Table 2. Items to Be Included in a Test Catalog Entry
Test name and synonyms
Proper specimen with hyperlinks to collection protocols.
Patient preparation requirements, eg, fasting, diets,
medications and herbals to avoid.
Proper timing of collection (time of the day, time relative to
meals, drug administration, etc).
Test charge as determined by hospital administration
(different levels for different patient types as appropriate).
Performing laboratory section.
Links to test performance characteristics.
3.
4.
5.
6.
7.
8.
9.
probable diagnoses) to suggest appropriate tests, test
frequency, and interpretative criteria.
a. Simpler systems may guide the provider to select
from a standardized list of diagnoses and clinical
situations and obtain corresponding guidelines and
clinical pathways with a mechanism for easily
ordering the appropriate tests.12,14
The system has a user-friendly display of the test
catalog (to include testing performed by external
reference laboratories), with available alternative
groupings, for example, alphabetically, by laboratory
discipline, by clinical situation. The menus must be
consistent, complete, regularly updated to maintain
currency, and with standard nomenclature in all HIS
systems interfacing with the LIS. The information for
each entry in the test catalog should display different
user-selected categories and levels of complexity to
include the items shown in Table 2.
The system has the ability to restrict ordering permission by location, diagnosis, provider specialty, etc, for
certain tests.
The system allows definition of tests that require
approval, for example, by a clinical or laboratory
specialist. The approval system should be integrated
with a downstream contact database that automatically
notifies the approver and the ordering provider that
there are pending tests needing approval.
The system has the ability to distinguish research versus
patient-care specimens and enable different billing
procedures (even for different tests on the same
specimen). Research orders should be attached to a
research management system, including the availability
to link to different protocols and research accounts.
An order appropriateness expert system is available
with the functionalities outlined in Table 3.
The ordering system should have the ability to relay
orders to different interfaced systems, for example,
another LIS in another institution or reference laboratory, without manual intervention, so that tests ordered
in one facility can enable specimens to be collected and
accessioned at another location or institution. Ideally,
the catalog of the reference laboratory(ies) should be
available online to the ordering provider, with implementation of institution-specific restrictions and approval processes for ordering, testing, and reporting.
For send-out testing, the system should be able to
generate a shipping manifest with sender, receiver, and
shipping information.15
The ordering system should receive real-time feedback
from the LIS and notify the ordering providers about
the status of the order, including the following steps:
a. Order acknowledged by laboratory.
b. Specimen(s) collected.
Table 3. Desirable Functionalities
of an Order Appropriateness Expert System
The system displays previous relevant test results
(graphically, if required) and pending related orders, with
an opportunity for the provider to cancel the order after
being made aware of such information.
The system has built-in and customizable medical necessity
review and acceptance or rejection criteria, including
criteria for maximum frequency of appropriate ordering
for different situations, eg, by patient location, clinic,
specialist, diagnosis.
The system merges or cancels redundant orders falling
within preestablished criteria (institutionally or nationally
developed). For example, if 2 providers order thyroidstimulating hormone tests in the same week, the orders
are merged and both providers will receive the results. If
a provider orders a hemoglobin A1c test a month after a
result is available in the system, the order is cancelled
and the provider is notified to call the laboratory if an
override is needed.
The system uses available clinical and laboratory inputs to
determine appropriateness. For example, if patient sex is
female and ordered test is prostate-specific antigen, the
order is flagged for cancelling. If patient is receiving
rapamycin and cyclosporine treatment, and only
cyclosporine is ordered, the system asks the provider if
rapamycin measurements are also desired.
The system has the ability to cancel flagged orders with
hard stops to prevent inappropriate work-around of the
rules, while providing a mechanism for the ordering
provider to justify an exception, eg, by asking the
provider to call the laboratory for rule overriding. Certain
order types, eg, associated with research protocols,
should be exempted from appropriateness protocols by
policy.
The expert order appropriateness system should be able to
halt orders that are not associated with a proper
diagnostic code (such as ICD-9 or ICD-10).
Abbreviation: ICD, International Classification of Diseases.
c. Specimen(s) accessioned.
d. Accession(s) activated in laboratory.
e. Analysis completed.
f. Results verified.
g. Results reported; order completed.
10. The system has the ability to split laboratory orders, that
is, one order may comprise multiple tests requiring
multiple specimens and accessions. The system should
have the ability to track the progress and report the
status of each component separately under 1 order.
SPECIMEN COLLECTION, ACCESSIONING,
AND PROCESSING
Appropriate specimen collection and processing is fundamental to the quality of laboratory results, which follow
the well-know principle of ‘‘garbage in, garbage out.’’ An
ideal LIS should have functionalities to optimize specimen
collection and processing, including the following:
1. Specimen collection lists as appropriate to institutional
operation. For example, for each phlebotomy round to a
set of locations, the system produces the appropriate list
of patients to be collected, together with preprinted
accession labels. The list should indicate the most
efficient route to each of the patients, taking into
account the desired testing priority.
2. The system guides the specimen collector with an
online or printed display of proper specimen collection
139
Table 4. Information Entered by the Specimen
Collector That Can Be Useful for Proper Performance
and Interpretation of Certain Laboratory Tests
Specimen number and time of collection for serial
specimens.
Specific site of collection.
Fasting or nonfasting, time of last meal.
Last menstrual period for gynecologic and some
endocrinology tests.
Date/time/dose of last medication (if not available from the
HIS).
Difficulties with specimen collection, eg, prolonged
tourniquet, presence of intravenous lines.
Other relevant clinical information (customizable by test
and prompted by the system).
Abbreviation: HIS, health care information systems.
3.
4.
5.
6.
7.
instructions, in an easy step-by-step format with links
to a full procedure.
The system has the ability to present the collector with a
list of pending laboratory orders and generate unique
bar-coded or RFID labels16 at the bedside upon
scanning the patient identification wristband or other
unique physical patient identifiers.17 The labels generated at the point of collection should include a
minimum of 2 patient identifiers, as well as date and
time of collection, collector identity, urgency of the
order, and as much as possible, the abbreviated names
for tests requested. Use of 2-dimensional bar codes or
RFID labels allows larger amounts of information to be
attached to the specimen. On arrival of the specimen in
the laboratory, the system should be able to recognize
the specimen upon scanning the labels attached to the
specimen container, and initiate testing without further
human intervention, if applicable, for example, in a
robotic specimen-processing automation line.
In addition to automatically recording patient information, location, date and time of collection, and collector
identity, the system should allow the collector to enter
pertinent information in codified or free-text form that
can be useful for proper performance and interpretation
of certain laboratory tests, as exemplified in Table 4.
The system should be able to support bidirectional
interfaces with portable devices for patient identification, specimen accessioning, and point-of-care testing,
including the ability to use wireless connections for data
transmission. Results from point-of-care testing should
be integrated with those from main analyzers while
identifying the source of those results.
A point-of-care test management system should be
available to track instruments, reagents, quality control,
and user identity, training, and competency records.
The system should separately record accessioning of
specimens (ie, matching an order with a physical
specimen), specimen receipt in the laboratory, and
activation of the specimen for analysis. For example, a
phlebotomist scans the patient bar-coded wristband
and chooses an appropriate pending order, the system
records the collection time and accessions the specimen, and the portable device carried by the phlebotomist prints an accession label. The specimen is
collected and the labels are affixed to the specimen
container in the presence of the patient. Upon arrival at
the laboratory reception desk, the specimen accession
8.
9.
10.
11.
labels are scanned to acknowledge receipt by the
laboratory, and then transported to the analytic section
of the laboratory. When placed in an automated robotic
specimen-processing line, the labels are scanned again
and the accession is activated for analysis. Alternatively,
the last 2 steps are merged and the specimens may be
first scanned and activated when placed on a robotic
track. In this manner, laboratory turnaround time can
be differentiated into time from order to collection to
accession to receipt to activation to report. The last
component (activation to report) is the analytic time,
whereas the previous components are preanalytic. It is
important to distinguish the different components of
turnaround time because often only the receipt-toreport processes are under complete control of the
laboratory. Using these time points, ‘‘incomplete lists’’
can be focused on pending orders, on specimens
received in the laboratory, or exclusively on accessions
ready for analysis.
The system should allow deviations of the sequence of
specimen processing described above, according to
institutional policy, for example:
a. Specimens received in the laboratory without an
order or accession, but with appropriate patient
identifiers. Receipt of these specimens in the
laboratory should be acknowledged by the system,
pending arrival of an appropriate corresponding
order. In defined cases, the laboratory staff should
have the ability to enter a paper or verbal order in
the system.
b. Properly identified specimens received in the laboratory with an order, paper or electronic, but without
accession labels. The laboratory acknowledges and
verifies the appropriateness of the order and of the
specimen, and then accessions the specimen and
applies appropriate labels or RFID tags.
c. The system should have the ability to accession and
process nonpatient specimens, for example, animal,
research, or environmental specimens not associated
with a patient, quality control and validation
materials, and most importantly, proficiency-testing
materials. The system should allow selected personnel to assign proficiency-testing materials to one of
many unique virtual patient identities so that the
analyst performing the test is unaware that the
specimens are proficiency-testing materials.
The system should have the ability to de-indentify and
codify specimens for research purposes, and include
database management capabilities for biobanks and
tissue repositories.
The LIS should interface with laboratory automation
management software to ensure that all the preanalytic
requirements stipulated in the ordering process (eg,
centrifugation speed, time, number of aliquots, reflexive
testing) are transmitting to the specimen-processing
system.
The system should be able to track the specimen
location throughout the preanalytic, analytic, and
postanalytic phases, including transportation to various
sections of the laboratory or external sites, and
management of specimen storage.18 The latter includes
functionalities for easy retrieval of the precise specimen
storage location and periodic reports to facilitate batch
disposal of specimens.
140
Table 5. Useful Information Associated
With Reagents and Other Test Components
Name of component
Manufacturer
Catalog number
Lot number
Date/time received in laboratory
Date/time opened and put into service
Initial volume/number of assays
Current volume/number of assays left
Expiration date
Storage requirements
12. The system should be able to generate multiple
specimen aliquot labels that can be scanned to execute
the appropriate testing associated with each aliquot.
This capability should include functionalities for tracking and storing multiple aliquots and slides derived
from a single specimen.
ANALYTIC PHASE
The analytic phase has been the focus of most technologic
developments in clinical laboratory science and is typically
associated with the lowest frequency of errors in the clinical
laboratory. In addition to interfacing with specimen
handling and analytic instrumentation software (often called
middleware) for streamlined processing of analytic requests—including the ability to direct testing to the
appropriate instrument depending on workload, recall
specimens for repeated testing, direct specimen dilutions,
perform reflexive testing, process add-on requests for
additional tests, and record test results and appropriate
comments—the LIS should provide the following functionalities:
1. Track and associate with individual testing records all the
components necessary for testing, particularly for manual assays and those methods associated with laboratory-developed reagents. Information about reagents and
other test components should include the information
shown in Table 5.
2. The appropriate standardized operating procedure for
each test (particularly for manual assays), managed by a
document control system (see below), is easily displayed
or printed upon request by the analyst.
3. The testing instrument is recorded with each patient test.
The analyzer record should include the information in
Table 6 and provide links to online preventive maintenance and service records, with the ability to alert the
user for scheduled maintenance and service. If required
by the laboratory, the instrument manufacturer should
also be automatically notified.
4. The system should produce laboratory-specific workload
lists (‘‘worklists’’) to facilitate batch processing and
resulting of both manual and automated tests and to
track orders that have not been completed. If additional
specimens are received after worklist creation, worklists
should be expandable by scanning the bar code or RFID
tag of the additional specimens.
5. ‘‘Incomplete lists’’ of tests that have been accessioned
but not completed, highlighting those that have exceeded the stated time for the category of the request, should
be displayed on demand, as well as on continuous report
screens, if so desired. Similarly, lists of incomplete or
Table 6. Useful Information Associated
With Each Laboratory Analyzer
Name of instrument
Manufacturer
Serial number
Date placed in service
Expected life
Calibration studies (by test)
Maintenance and repair records
unfulfilled orders should be available on demand or by
schedule with the ability to pinpoint at which point the
failure occurred. Incomplete lists should be able to
include tests sent to reference laboratories. An example
of an incomplete test display with significant clinical
impact is the continuous display in a large screen of
emergency room orders not completed within a predefined time frame, possible with color coding and/or
sorting by age of request, to alert staff to investigate and
process orders or specimens at risk of exceeding
acceptable turnaround time thresholds.
RESULT ENTRY AND VALIDATION
The LIS should not only serve as a repository of laboratory
results generated by the analytic process, but also guide the
analysts into providing high-quality results that are accurate, reproducible, and appropriate to the clinical situation.
Desirable functionalities in result entry and validation
include the following:
1. Ability to record results in various data formats,
including numbers, text with extended characters, and
images, with a flexible data storage approach that avoids
constraining limits on data size.
2. Automated and manual entry and correction of results of
tests performed in interfaced or noninterfaced analyzers
as well as manual tests, with appropriate security levels
applied. Result entry should include options for individual result entry, batch results entry, batch entry by
exception, amended results, appended results, and
intermediate and final results. Results can be entered
either by individual tests, or by panels, with userdefinable panel configuration.
3. The system should allow different levels of result
certification, with the ability to withhold release of
results until approved by a higher-level user, for
example, a supervisor.
4. The system should be able to receive results in a variety
of formats such as tables and graphs from other
laboratories, including external reference laboratories,
through electronic interfaces, for seamless integration of
all laboratory results in the electronic record. An example
of where such combination of data is highly desirable is
for the diagnosis of leukemia, where clinical information
together with hematology, hematopathology, molecular,
and flow data are often needed to make an accurate
diagnosis.
5. The system should be able to use advanced expert
decision support for autovalidation of results.19,20 Autovalidation avoids human intervention in the certification
of laboratory results and is a major driver of efficiency
improvements in laboratory operations.21 The more
sophisticated the system used to perform autovalidation,
141
the lower the probability of reporting an erroneous
result, and the more time is allowed for a human
specialist to examine exceptional results. Inputs used to
arrive at an autovalidation decision include the following:
a. Comparison with results of previous tests in the
patient record (temporal delta checks).
b. Comparison with results of other related tests in the
same or closely related specimens (cross-sectional
check). An example is creatinine versus urea.
c. Checking the specimen for predefined limits of
hemolysis, lipemia, or icterus.
d. Clinical information, including demographics, location (inpatient versus outpatient, type of clinic),
diagnoses, medications, procedures.
e. Results of external and internal quality control.
f. Statistical data on result distribution.19
6. The ability to perform temporal delta checks should
include analysis of temporal data and calculation of rates
of change as well as absolute changes, in reference to
preestablished limits that can vary by patient clinical
information such as demographics, diagnoses, therapies.22,23
7. The expert system should be able to order reflexive
testing based on analysis of results and clinical data,
definable by institutional or laboratory policy and
customizable by the ordering provider, interface with
the specimen-processing and analytic systems, and
append the appropriate codes or comments to the
results.24
RESULT REPORTING
The system should be able to provide a variety of reports
for use in patient care, including standard and userdefinable reports organized by test, test group, date, date
range, ordering provider or provider group, clinic or
specialty, sequential or tabulated cumulative worksheets,
and the following additional capabilities:
1. In addition to the actual value measured, numeric test
results should include display of the following (optional
or mandatory as appropriate):
a. Units of measurement.
b. Reference interval of the appropriate reference
population (user-configurable by a variety of clinical
inputs, including ambulatory versus recumbent, sex,
age, race, body mass, gestational age, menstrual
cycle phase).
c. A measure of individuality25 should be displayed to
guide interpretation of the reference range. For tests
with high individuality, where within-subject variability is much lower than between-subject variability, a comment should be appended that the
individual-based reference changes are more appropriate than population-based reference intervals. For
tests with high individuality and patients with
enough recorded data, the system should be able
to calculate and display an individual-specific
reference range, for example, the central 90% of
previous results, with the ability for the user or the
expert system to exclude from the calculation results
clearly associated with disease.
d. Confidence interval of the results, based on analytic
variability observed at a corresponding level.
Table 7. Useful Flags Associated
With Laboratory Test Results
Results outside the reference interval, with indication of
multiples of upper or lower reference limits.
Results outside of confidence intervals, with indication of
the probability of the change being due to analytic or
biologic variability.25
Results exceeding various levels of medically relevant
thresholds, including multiple tiers of significant and
critical results. The latter should be linked to automatic
notification of providers.
Dynamic change from a previous result (delta) exceeding
user-definable thresholds, eg, exceeding the RCV
interval. The flag could be coded to different levels of
probability of change, eg, ‘‘likely’’ at P . .80, ‘‘more
likely’’ at P . .90, ‘‘very likely’’ at P . .95, and
‘‘virtually certain’’ at P . .99.27
Abbreviation: RCV, reference change value.
2.
3.
4.
5.
6.
7.
e. Alternatively, reference change values, that is, the
interval around the result that is a consequence of
analytic imprecision, within-subject biologic variability, and the number of repeated tests performed.25–27
The user should be allowed to customize the
reference change value interval by selecting a
confidence threshold (eg, 95%) and the appropriate
Z-value for decisions that involve 1-sided (eg,
increase) versus 2-sided (either increase or decrease)
changes.
f. Result-associated flags, listed in Table 7, are available
and thresholds can be predefined by the users.
g. Pertinent comments appended by the analysts.
Report generation that is flexible and configurable by
users, to include both producers (laboratorians) and
recipients (providers, patients) of the test information.
Reports should be made available by a variety of options
to include user-configurable automated secure faxing, emailing, and other electronic text transmission mechanisms.
Sophisticated graphing of laboratory results, ideally with
integration of other appropriate clinical information such
as vital signs, biometrics, medication dose/timing.28–30
Graphing functionalities should match state-of-the art
graphing programs and allow dynamic changes in axis
and scales, histograms, conditional formatting, color
coding, user-definable formulas for calculated results,
and display of multidimensional data. Colored displays
are preferable.
Ability to incorporate in result comments hyperlinks to
pages containing further test information, including
analytic parameters, half-life of toxins, drugs, and certain
other analytes, calculators, clinical guidelines, suggested
follow-up, literature references, and other pertinent data
to help providers interpret the results and use the
information in clinical care.
The system should link pretest and posttest diagnostic
information by displaying positive and negative likelihood ratios for selected diagnoses, based either on HIS
or user input. Upon selection of a particular clinical
condition, the system should display appropriate Bayesian statistics, including sensitivity, specificity, accuracy,
positive and negative predictive values, and receiveroperator curves with links to appropriate references.
Readily accessible display of all possible significant
interferences and causes of abnormal test results,
142
including diseases, herbal supplements, medications.31
This information should be flagged if extracted from data
available in the HIS, and complete lists should be
available for display upon link selection by the user.
8. Intelligent cumulative reports triggered by patientrelated events, such as discharge or outpatient visit, to
facilitate expeditious clinical care. For example, a
decision support to avoid inappropriate discharges due
to unidentified or unaddressed clinically significant
laboratory results has been described.32
9. An expert system should be able to append appropriate
interpretative comments on test results,33–35 taking into
consideration not only the result of the test itself, but also
other pertinent test and clinical information available in
the HIS and a knowledge database updatable with local
information, for example, disease prevalence. Temporal
patterns should be taken into consideration, particularly
for therapeutic drug monitoring and calculation of
clinically useful pharmacokinetic parameters such as
the area under the concentration curve and estimated
elimination half-life.22,23
NOTIFICATION MANAGEMENT
Distribution of results to end users should be defined by a
combination of institutional policy for certain results (such
as ‘‘critical results’’) and user-selected notification mechanisms (eg, printout, fax, e-mail, HIS alert) for routine
reports. A rule-based system should be used to select the
appropriate mechanism and timing for user notification (see
‘‘Test Ordering,’’ above). The notification management
system should have the following capabilities:
1. The LIS should have a sophisticated ‘‘significant result’’
notification system. The system should include multiple
tiers of urgency for significant result notification. For
example, the Massachusetts Coalition for the Prevention
of Medical Errors established 3 levels of notification: red,
orange, and yellow.36 ‘‘Red’’ results are those implying
immediate danger of mortality or morbidity if not rapidly
acted upon. These correspond to the Joint Commission’s
and College of American Pathologists’ (CAP’s) definition
of ‘‘critical test results’’ and mandate direct notification of
a health care provider with the ability to intervene in
patient care, within a maximum time frame set by
institutional policy (usually 15–60 minutes), and require
acknowledgment of the receipt of the information or
‘‘read back’’ process. An example is a potassium level of
2.5 mEq/L (2.5 mmol/L). ‘‘Orange’’ results are highly
significant results that must be acknowledged but are not
immediately threatening to the patient and can wait
several hours (target, 6–8 hours) before notification.
‘‘Orange’’ results include, for example, highly elevated
creatinine, amylase, lipase, and aminotransferase levels.
Notification of the provider should be made by a highpriority process, for example, by a high-priority HIS alert
requiring acknowledgment by the recipient, with a
cascading process of surrogate notification if the
appropriate provider is unavailable. Finally, ‘‘yellow’’
level results may be associated with significant morbidity
or mortality if diagnosis and treatment are not initiated
in a timely manner, but are not immediately threatening.
Yellow results require notification within 3 days and may
include passive methods, such as a HIS alert or chart
note, with mandatory acknowledgment and tracking.
2.
3.
4.
5.
Examples include a high thyroid stimulating hormone
(TSH) or lead level, or a new diagnosis of cancer or
human immunodeficiency virus infection.
Significant result notification should use artificial intelligence and expert decision-support systems for more
relevant identification of true-positive (eg, life-threatening) results, while minimizing false-positive signals (eg,
expected results). The expert system should use the
various inputs previously described for order entry and
autovalidation systems. Even without expert system
intervention, dynamic rules should be used to determine
whether a result is critical. For example, a single
threshold for low hemoglobin level is inappropriate, as
chronic anemia is much better tolerated than acute
anemia. A dynamic threshold to detect a rapid rate of
hemoglobin decline will be more clinically relevant and
will identify patients at risk whose condition may not be
considered critical when using a fixed threshold.22,23,36
In addition to the providers and surrogates defined at the
order entry step, the system uses rule-based notification
of appropriate third parties, such as infection control or
public health departments, depending on the laboratory
result.
Any changes or corrections to laboratory results should
be communicated rapidly to providers, and reports on
interfaced HIS systems should be accurately and
completely updated.
The system should provide the ability for end users to
inquire about laboratory testing, receipt of specimens,
availability of results, with links to appropriate online
information and messaging of laboratory staff if further
information is needed. Search engines should use stateof-the-art technologies allowing for synonyms, misspellings, and advanced Boolean combinations of search
terms.37 Reports of user activity should be available to
laboratory managers for process improvement.
DATA MINING AND CROSS-SECTIONAL REPORTS
The ability to perform queries into the laboratory and
clinical databases is paramount to maximize the efficiency
and quality of the laboratory operation, provide means of
identifying clinical issues affecting a specified population,
perform epidemiologic and public health studies, and case
finding for clinical or research purposes. Advanced data
warehouse and mining capabilities should be available in an
advanced LIS. Examples of useful queries and reports
include the following:
1. Search functions for combinations of laboratory results
and clinical information, such as diagnoses, medications,
and treating specialty, using Boolean logic, producing
reports with user-customizable display of queried and
nonqueried fields such as patient demographics, specimen accession data, and location. Such reports should be
exportable to spreadsheet programs for further analysis.
Ideally, common statistical functions should be available
for aggregate data.
2. Laboratory testing turnaround time reports with the
ability to consolidate or split the various components,
such as order to collection, collection to receipt in
laboratory, receipt to testing, and testing to report, and
the capability to group by accession areas, individual test
or test groups, hours or shifts, employee, patient
location, clinics, providers, etc.
143
3. Surveillance data online reporting to public health
agencies in their required format, using the appropriate
standards.
4. Nosocomial infection tracking and antibiograms reporting the frequency distribution of microbial susceptibility
to antimicrobial agents. Interfacing with pharmacy
records to monitor antimicrobial utilization versus
susceptibility.
5. Laboratory utilization reports by provider, provider
group, specialty, clinics, wards, patient types, diagnoses
and diagnostic groups, ICD-9/10 codes, etc, to include
test type, volumes, and costs per case. The system should
provide appropriate real-time feedback on utilization
data to providers, for example, at the time of patient
discharge.
6. Patient outcome analysis using laboratory data-mining
capabilities and clinical data extracted from the HIS.
Examples of useful parameters to correlate with laboratory testing include mortality, morbidity, hospital length
of stay, and cost of care, grouped by diagnostic groups.
METHOD VALIDATION
Method validation is an important step preceding
implementation of new assays in the clinical laboratory
and is performed periodically in a more summarized mode
to ensure the stability of the assay systems and compliance
with regulatory and accrediting agencies.
1. The LIS should include a module for method validation
with the ability to guide and record the following studies
(including calculation and display of the appropriate
statistics and graphic displays38): assay linearity and
calibration verification; assay intra-run and between-run
precision; comparison with the established methods or
between analyzers; reference range validation; and
interference and recovery studies.
2. The system should alert laboratory staff to the need to
perform validation procedures (eg, biannual linearity
checks and instrument correlations), as appropriate.
2.
3.
4.
5.
6.
7.
8.
9.
10.
QUALITY MANAGEMENT
In the current health care financing environment,
institutions are increasingly focusing on improved quality
and outcomes of patient care to enhance their financial
situation and gain competitive advantages. Quality management for clinical laboratories involves a program to
ensure quality throughout all the aspects of laboratory
operation. More strictly, quality control (QC) refers to
periodic assaying of samples with known reactivity or
analyte concentrations to estimate assay accuracy and
precision. A modern QC program should aim at improving
the accuracy and reliability of laboratory results by
maximizing error detection and minimizing false rejections
of test runs.39 The quality management module should
support accreditation requirements, including CAP, Clinical
Laboratory Improvement Amendments of 1988 (CLIA),40
and International Organization for Standardization (ISO)
15189:2003 standards,41 and include the following functionalities:
1. Quality control protocols and alerting mechanisms
should use thresholds for acceptability, based on the
concept of total acceptable error derived from biological
variation25,42,43 and regulatory requirements. The user
11.
12.
13.
14.
15.
16.
should have the capability of customizing the QC
protocol, based on built-in databases of biological
variability and measured imprecision of the various
tests at relevant clinical decision points, based on the
specific analyzer used to perform the test.
Quality control files for each assay system should record
the following:
a. Information about a particular quality control
material (as described for test components, including lot number, expiration date).
b. Manufacturer or laboratory-assigned control values
for each relevant testing system.
c. Serial quality control test results associated with
each control material and each analyzer.
Each patient test result should be linked to the relevant
quality control result(s) in an easily retrievable record.
The system schedules automated running of quality
control or alternatively alerts appropriate staff to
perform QC tasks.
The system should guide the user in QC rule selection
taking into consideration the total acceptable error and
the analytic performance (precision and bias) of the
testing system.39
Active QC rules and reports should be customizable by
test, test group, analyzer type, laboratory location, and
working shifts.
Sophisticated user-definable display of QC results
should include Levey-Jennings plots and interactive
display of violations of user-selected rules, such as
Westgard rules.
Quality control reports including Levey-Jennings plots
should be easily interpretable so staff can quickly make
critical decisions about test acceptability. Troubleshooting and corrective action guides for QC violations
should be available upon user selection.
The user should be able to customize date intervals and
time scales and aggregate, split, or compare multiple
QC levels, QC lot numbers, test groups, reagent
cartridge, reagent lot number, analyzers, laboratory
sections, and multiple laboratories.
Interfaces to third-party vendors for automated upload
of QC data and real-time download of peer performance data should be available.
The system should provide the ability to document
corrective actions resulting from QC failure in real time.
The system should be able to remove outliers and
erroneous results from QC calculations, based on
appropriate statistical parameters as well as user input.
For certain test runs, as defined by the user, the system
should automatically interrupt analysis or autoverification in case of QC failure and guide staff into
appropriate investigation and decision choices.
The system should have the capability to batch hold QC
results so users do not have to constantly switch screens
to verify QC.
Peer comparison statistics should include range, mean,
median, standard deviations, standard deviation index,
coefficient of variation ratio, Youden charts, and timebased plots and histograms. The system should be able
to import these parameters from external interlaboratory programs.
Alternative means of quality control should be available, including moving averages of normal values, of all
results, or by user-defined criteria. If all results are used
a median value should be presented. Another valuable
144
17.
18.
19.
20.
21.
report for quality monitoring is the display of the
histogram of results by test and various patient
characteristics, with definable flags highlighting deviations from the historical frequency distributions.
The system provides user-definable QC summary
reports for review by supervisory and management
staff with functionality for documentation of review and
corrective actions.
The system should have functionalities to interface with
instrument performance data, temperature-monitoring
systems, water quality parameters, environmental
measurements, and other data pertinent to good
laboratory practice and accreditation requiring periodic
documentation.
The system should manage proficiency-testing (PT)
programs, from inventory control of PT materials to
documenting PT results and investigation of PT failures,
with available online review and certification of PT
results by appropriate management staff. Ideally,
interfacing to external PT program providers should
allow seamless transmission of PT data.
The system manages accreditation requirements online,
including preparation of appropriate documents, for
example, by incorporating checklists and questionnaires
from accrediting agencies in a database allowing for
tracking and documentation of answers to checklist
questions and inspection findings, containing links to
relevant policies, procedures, and other electronic
documents as evidence of compliance. The system
should be capable of capturing and manipulating all
data required for accreditation agencies, such as CAP or
ISO 15189.
The system should have a user-friendly incident, error,
and process improvement tracking mechanism with
sophisticated database, querying, and reporting functionalities.44 The system should allow any user to
initiate reporting of errors and incidents in real time
with an option of anonymity.
ADMINISTRATIVE AND FINANCIAL ISSUES
Management of a modern laboratory requires access to a
variety of data at various levels of consolidation. The LIS
should incorporate advanced administrative and financial
functionalities, including the following:
1. Ability to generate and transmit the necessary forms and
notifications for reimbursement of tests, with the
appropriate test codes (Current Procedural Terminology
[CPT] Systematized Nomenclature of Medicine
[SNOMED], or ICD-9 or ICD-10) selected.
2. Intelligent generation of online and printed regulatory
forms associated with laboratory testing, billing, compliance, and accreditation, such as insurance claims in
Health Insurance Portability and Accountability Act of
1996 (HIPAA) standard transaction formats, Medicare
Waiver/Advance Beneficiary Notice forms, and others,
for use at the point-of-care as well as from administrative locations.
3. Tracking of costs of laboratory operation, including
consumable, labor, amortization, and other fixed costs.
4. Analysis of pricing, profitability, ‘‘make-or-buy’’ decision
tools, and outreach client management capabilities.
5. Workload statistics based on system variables such as
time logged in, number of tests verified, and instrument
raw test counts, as well as user inputs.
6. Ability to produce periodic reports of laboratory productivity and management efficiency, by using the following:
a. Aggregate numbers, such as number of total and
billable tests and interpretations (and who performed
the interpretations), number of full-time employee
equivalents (FTEEs, classified as technical, nontechnical, and scientist/pathologist), hours worked, laboratory costs (broken down by section, variable versus
fixed, etc), number of patients (outpatient visits,
discharges, bed-days, etc), for top-down analyses.
b. Individual cost and productivity analysis per test and
per laboratory workgroup for bottom-up financial
and productivity analyses.
c. User-definable financial and productivity calculations,
for example, cost per billable test, costs per FTEE, paid
hours per FTEE, number of tests, and costs per patient
(different categories), with comparison with benchmarks and the ability to customize granularity (eg,
whole laboratory, laboratory section, accession group,
or individual test).
7. Inventory and materials management including functionalities for automated ordering from selected suppliers and real-time tracking of budget.
8. A document management system fully integrated in the
LIS with readily accessible procedure manuals from
testing information menus, and mechanisms for periodic
notification, review, and certification by appropriate
parties. The document control system should allow
scanning and proper linking of pertinent documents,
such as reagent and QC package inserts, test requisitions, and reports from outside laboratories.
9. Personnel management capabilities to include interfacing with human resource databases, labor-cost accounting, and tracking of credentials, competency,
continuing education training, and performance appraisals. Competency training and credentialing records
should be linked to the ability of the user to complete
defined tests or test groups, using the LISs. Ideally, the
system would provide an interface to online continuing
education and role-based competency-training modules.
OTHER OPERATIONAL ISSUES
Other desirable functionalities of an ideal LIS include the
following:
1. The system should have enough capacity to record large
datasets and interface with legacy systems (in real time
or through import functions) to capture historical
laboratory data, with the goal of storing lifelong results
on each patient. Capabilities for handling large genomic
data sets, while providing meaningful reports and ‘‘justin-time’’ education to clinical providers,45 will be
increasingly necessary in future LIS.
2. The system should capture industry standards for
coding, billing, document generation, and interface
formats, such as CDC, HL7 CDA1/2, XML, ASC X12,
LOINC, SNOMED-CT, ICD-9, or ICD-10, as appropriate for each data type. Mapping dictionaries for
interconversion between different standards should be
available as appropriate.
3. The user interface and navigation should be intuitive
and user friendly.
145
4. The system should minimize the number of keystrokes
required for all activities (use automatic return where
possible).
5. The system should have uniformity for similar tasks
within the software, for example, using ‘‘enter’’ for all
programs rather than clicking ‘‘OK’’ for some and
double clicking for others.
6. All screens and reports should be printed and exported
in appropriate document text, spreadsheet, or graphic
formats.
7. Fully functional text editor in text entry fields with rich
text and common word-processing functionalities.
8. Appropriate backup data capture and retention with
rapid retrieval in the event of system failure.
9. Auditing capabilities should track changes in all data
(results, QC, patient information, etc) made by any
user.
10. Interfaces with single or multiple HIS systems should
be flexible in data formats and fully functional with
appropriate routines available for testing the functionality of the interfaces before use in order to meet enduser expectations.
11. The interface with HIS should allow real-time updating
of the LIS with pertinent information, such as patient
location and current provider, and conversely, laboratory data should be rapidly available in all interfaced
HIS systems. Particular focus should be placed on the
interface between the LIS and the pharmacy software
for drug and test selection, detection and prevention of
drug reactions and drug-laboratory interactions, monitoring of drug levels, etc.31
12. The system should integrate instant messaging, forum,
online meeting, and social-networking capabilities to
enhance communication among laboratorians and with
laboratory users.
13. The LIS should be capable of performing multiple
functions simultaneously with imperceptible impact on
its speed.
14. Where manual activities are involved, these should be
accomplished with a minimal number of keystrokes and
waste of time and energy, with no degradation of LIS
performance regardless of workload.
SUMMARY
In this article we listed a considerable number of features
desirable in future LISs, aimed at improving the quality and
cost-efficiency of patient care by optimizing the operation of
clinical laboratories and most importantly, the interface
between health care providers and the clinical laboratories.
Laboratory information systems are critical for proper
packaging of the information produced by clinical laboratories to be optimally used by clinical providers. We envision
the LIS as replacing humans in most activities that allow the
option of human error. Humans would interact with the LIS
through user-friendly interfaces designed with a lean
approach to optimize efficiency and maximize productivity.
The ideas listed in this work have been variably implemented in currently available LISs, but considerable effort in
incorporating combinations of artificial intelligence, expert
systems, advance database, data mining, and other state-ofthe-art information technologies must be used to arrive at a
comprehensive, fully functional, user-friendly, and clinically
useful LIS.
References
1. Temurtas F. A comparative study on thyroid disease diagnosis using neural
networks. Expert Syst Appl. 2009;36(1):944–949.
2. Le QA, Strylewicz G, Doctor JN. Detecting blood laboratory errors using a
Bayesian network. Med Decis Making. 2011;31(2):325–337.
3. Basciftci F, Incekara H. Design of web-based fuzzy input expert system for
the analysis of serology laboratory tests. J Med Syst. 2011;36(4):2187–2191.
4. Mantzaris D, Anastassopoulos G, Adamopoulos A. Genetic algorithm
pruning of probabilistic neural networks in medical disease estimation. Neural
Netw. 2011;24(8):831–835.
5. Kodaz H, Özsen S, Arslan A, Günes S. Medical application of information
gain based artificial immune recognition system (AIRS): diagnosis of thyroid
disease. Expert Syst Appl. 2009;36(2, pt 2):3086–3092.
6. Plebani M. Towards quality specifications in extra-analytical phases of
laboratory services: what information on quality specifications should be
communicated to clinicians, and how? Accred Qual Assur. 2006;11(6):291–296.
7. Blick KE. Decision-making laboratory computer systems as essential tools
for achievement of total quality. Clin Chem. 1997;43(5):908–912.
8. Park S, Parwani A, Satyanarayanan M, Pantanowitz L. Handheld computing
in pathology. J Pathol Inform. 2012;3:15.
9. Serdar MA, Turan M, Cihan M. Rapid access to information resources in
clinical biochemistry: medical applications of Personal Digital Assistants (PDA).
Clin Exp Med. 2008;8(2):117–122.
10. Westbrook JI, Georgiou A, Lam M. Does computerised provider order
entry reduce test turnaround times: a before-and-after study at four hospitals. Stud
Health Technol Inform. 2009;150:527–531.
11. Baron JM, Dighe AS. Computerized provider order entry in the clinical
laboratory. J Pathol Inform. 2011;2:35.
12. Yasini M, Duclos C, Lamy JB, Venot A. Facilitating access to laboratory
guidelines by modeling their contents and designing a computerized user
interface. Stud Health Technol Inform. 2011;169:487–491.
13. Parl FF, O’Leary MF, Kaiser AB, Paulett JM, Statnikova K, Shultz EK.
Implementation of a closed-loop reporting system for critical values and clinical
communication in compliance with goals of the Joint Commission. Clin Chem.
2010;56(3):417–423.
14. van Wijk MA, van der Lei J, Mosseveld M, Bohnen AM, van Bemmel JH.
Assessment of decision support for blood test ordering in primary care: a
randomized trial. Ann Intern Med. 2001;134(4):274–281.
15. Byrne CM, Rudin RS, Johnston DS, Pan EC. Assessing the value of
laboratory electronic data interchange in the Department of Veterans Affairs.
AMIA Annu Symp Proc. 2010;2010:76–80.
16. Shim H, Uh Y, Lee SH, Yoon YR. A new specimen management system
using RFID technology. J Med Syst. 2011;35(6):1403–1412.
17. Hill PM, Mareiniss D, Murphy P, et al. Significant reduction of laboratory
specimen labeling errors by implementation of an electronic ordering system
paired with a bar-code specimen labeling process. Ann Emerg Med. 2010;56(6):
630–636.
18. Emmerich KA, Quam EF, Bowers KL, Eggert AA. The combination of
specimen tracking with an advanced AutoLog in a laboratory information system.
J Med Syst. 1998;22(3):137–145.
19. Oosterhuis WP, Ulenkate HJ, Goldschmidt HM. Evaluation of LabRespond,
a new automated validation system for clinical laboratory test results. Clin Chem.
2000;46(11):1811–1817.
20. Fuentes-Arderiu X, Castineiras-Lacambra MJ, Panadero-Garcia MT. Evaluation of the VALAB expert system. Eur J Clin Chem Clin Biochem. 1997;35(9):
711–714.
21. Goldschmidt HMJ. A review of autovalidation software in laboratory
medicine. Accred Qual Assur. 2002;7(11):431–440.
22. Post AR, Harrison JH Jr. Temporal data mining. Clin Lab Med. 2008;28(1):
83–100, vii.
23. Post AR, Harrison JH Jr. PROTEMPA: a method for specifying and
identifying temporal sequences in retrospective data for patient selection. J Am
Med Inform Assoc. 2007;14(5):674–683.
24. Antill ML. A Reflexive Testing Expert System Based on Fuzzy Logic.
Cleveland, Ohio: Cleveland State University; 1992.
25. Fraser CG. Biological Variation: From Principles to Practice. Washington,
DC: AACC Press; 2001.
26. Kroll MH. Multiple patient samples of an analyte improve detection of
changes in clinical status. Arch Pathol Lab Med. 2010;134(1):81–89.
27. Fraser CG. Improved monitoring of differences in serial laboratory results.
Clin Chem. 2011;57(12):1635–1637.
28. Regeniter A, Steiger JU, Scholer A, Huber PR, Siede WH. Windows to the
ward: graphically oriented report forms—presentation of complex, interrelated
laboratory data for electrophoresis/immunofixation, cerebrospinal fluid, and
urinary protein profiles. Clin Chem. 2003;49(1):41–50.
29. Hoeke JO, Bonke B, van Strik R, Gelsema ES. Evaluation of techniques for
the presentation of laboratory data: support of pattern recognition. Methods Inf
Med. 2000;39(1):88–92.
30. Bauer DT, Guerlain S, Brown PJ. The design and evaluation of a graphical
display for laboratory data. J Am Med Inform Assoc. 2010;17(4):416–424.
31. Schiff GD, Klass D, Peterson J, Shah G, Bates DW. Linking laboratory and
pharmacy: opportunities for reducing errors and improving care. Arch Intern
Med. 2003;163(8):893–900.
32. Mathew G, Kho A, Dexter P, et al. Concept and development of a
discharge alert filter for abnormal laboratory values coupled with computerized
146
provider order entry: a tool for quality improvement and hospital risk
management. J Patient Saf. 2012;8(2):69–75.
33. Winstanley T, Courvalin P. Expert systems in clinical microbiology. Clin
Microbiol Rev. 2011;24(3):515–556.
34. Smythe GA, Drew CM. REPCAT: desktop expert system for interpreting and
validating laboratory data for pheochromocytoma diagnosis with the database
application Omnis 7. Clin Chem. 1997;43(1):134–140.
35. Van Hoof V, Wormek A, Schleutermann S, Schumacher T, Lothaire O,
Trendelenburg C. Medical expert systems developed in j.MD, a Java based expert
system shell: application in clinical laboratories. Stud Health Technol Inform.
2004;107(pt 1):89–93.
36. Hanna D, Griswold P, Leape LL, Bates DW. Communicating critical test
results: safe practice recommendations. Jt Comm J Qual Patient Saf. 2005;31(2):
68–80.
37. Blechner M, Kish J, Chadaga V, Dighe AS. Analysis of search in an online
clinical laboratory manual. Am J Clin Pathol. 2006;126(2):208–214.
38. Westgard JO. Basic Method Validation. Madison, WI: Westgard QC, Inc;
2008.
39. Westgard JO, Darcy T. The truth about quality: medical usefulness and
analytical reliability of laboratory tests. Clin Chim Acta. 2004;346(1):3–11.
40. Centers for Medicare & Medicaid Services. Clinical Laboratory Improvement Amendments (CLIA). http://www.cms.gov/Regulations-and-Guidance/
Legislation/CLIA/index.html?redirect¼/clia/. Modified May 3, 2012. Accessed
May 7, 2012.
41. Burnett D. ISO 15189:2003–quality management, evaluation and continual improvement. Clin Chem Lab Med. 2006;44(6):733–739.
42. Fraser CG, Petersen PH. Quality goals in external quality assessment are
best based on biology. Scand J Clin Lab Invest Suppl. 1993;212:8–9.
43. Oosterhuis WP. Gross overestimation of total allowable error based on
biological variation. Clin Chem. 2011;57(9):1334–1336.
44. O’Kane M. The reporting, classification and grading of quality failures in
the medical laboratory. Clin Chim Acta. 2009;404(1):28–31.
45. Hoffman MA, Williams MS. Electronic medical records and personalized
medicine. Hum Genet. 2011;130(1):33–39.
147
CONSULTVERLENING: ACHTEROMZIEN EN VOORUITBLIKKEN
JW Janssen, laboratoriumspecialist klinische chemie
Inleiding
Consultverlening is een actief proces waar je als laboratoriumspecialist klinische
chemie een directe rol in vervult. Maar het invullen van die rol is niet altijd even
logisch geweest. Deels uit tijdgebrek, deels uit onvermogen en soms uit een
combinatie van beiden. Tijdgebrek dat eind jaren 90 naar voren kwam toen er een
tekort aan klinisch chemici was. En van het aantal taken dat de klinisch chemicus
moest vervullen viel de arbeidsintensieve consultverlening vaak als eerste af. Maar
ook deels uit onvermogen. In lang niet alle opleidingsinstituten konden de KCio-ers
kennismaken met een structurele wijze van consultverlening. Laat staan dat
consultverlening een vaste plaats in het opleidingscurriculum had verworven.
Achteromzien
Binnen de NVKC heerste rond de millenniumwisseling al langer het gevoel dat het
vak onvoldoende werd ingevuld met louter op het laboratorium passen (als een goed
huisvader). Daar wil ik niemand mee te kort doen, maar het vak was op die wijze niet
toekomst bestendig. Nieuwe wegen moesten worden ingeslagen. Resultaten van
brainstorm sessies binnen de vereniging en het werken aan de herstructurering van
de opleiding analoog aan de opleiding voor medisch specialisten gaven de input voor
het meerjarenbeleidplan 2009-2013 “Van meten naar consult, van chemisch naar
medisch”. In de wandelgangen werd dit beleidsplan “van C naar M” genoemd. De
grondgedachte van het beleidsplan was: Verleg het accent van je taken als klinisch
148
chemicus. Namelijk van activiteiten vooral binnen het laboratorium naar je taak als
gids in diagnostica land voor alle aanvragers van laboratorium diagnostiek.
Financiën versus inhoud
Maar ook het bezuinigingsbeleid van de overheid vroeg onze aandacht: Breng een
halt toe aan de stijgende kosten in de gezondheidszorg. En kijk ook scherp naar een
reductie van de kosten van laboratoriumbepalingen. Daar zijn consultancybureaus
gretig op ingesprongen en wij werden snel overspoeld met allerlei plannen om het
aantal laboratoria landelijk terug te brengen tot een achttal of minder.
Ook
bestuurders
van
ziekenhuizen
werden
geconfronteerd
met
meerdere
bezuinigingsronden. Door sommige bestuurders van ziekenhuizen werd dit vertaald
in het niet honoreren van de uitbreiding van klinisch chemici, in het sterk omlaag
brengen van budgetten voor de laboratoria, door openlijk over outsourcing van
laboratorium bepalingen te spreken en daarmee de laboratoriumspecialist klinische
chemie (en zijn mede medisch ondersteunende collega’s) te dwingen de taak van
politieagent op zich te nemen. Die invulling was niet hetgeen wij voor ogen hadden
met de term “Gepast gebruik van laboratorium onderzoek”. De inhoudelijke invulling
vanuit de NVKC was meer om intensief met onze klinische collega’s te overleggen
over de zin en onzin van het aanvragen van laboratoriumbepalingen bij de
verschillende ziektebeelden. En er was ook noodzaak om dat te doen, omdat al
enige tijd het vak klinische chemie binnen het medisch curriculum nagenoeg
verdwenen was. En medische studenten niet meer van een klinisch chemicus (op
een enkele uitzondering na) onderricht in de klinische chemie kregen. Dat had
belangrijke gevolgen voor de jong opgeleide artsen. Zij gingen het klinisch chemisch
laboratorium zien als een black box waarbij de verpleging de aanvragen deed, soms
149
zelfs het materiaal afnam, en zij als behandelaar, de uitslagen op beeldschermen
gerapporteerd kregen. Volstrekt onkundig van het feit dat er laboratoriumspecialisten
klinische chemie aanwezig zijn voor nadere ondersteuning. Kortom onze focus moest
gericht worden op consultverlening door de laboratoriumspecialist klinische chemie.
Consultverlening
Het programma van vandaag is daar een mooi voorbeeld van. Wij hoeven elkaar als
laboratoriumspecialisten niet meer van het belang van consultverlening te
overtuigen. We zijn er voldoende van doordrongen dat dat noodzakelijk is. Maar
informatie over de wijze, de hoe-vraag, waarop we met elkaar invulling aan
consultverlening geven kunnen we niet voldoende met elkaar delen. Wat dat aangaat
heb ik een hoge pet op van de creativiteit en inventiviteit van alle collega’s. Al die
afzonderlijke pareltjes van consultverlening kunnen tot een schitterend parelsnoer
bijeen gebracht worden.
De KCvD en de labbabbal
Aan de hoe-vraag wil ik ook een kleine bijdrage leveren. Binnen het Sint Franciscus
Gasthuis hebben we sinds 2002 actief invulling gegeven aan de functie klinisch
chemicus van dienst. Een 24/7 functie die een positief effect gehad heeft op onze
serviceverlening naar de aanvragers toe.
Een aantal praktische afspraken hebben ons als vakgroep geholpen om de functie
KCvD verder invulling te geven:
x
Werkplek KCvD direct op het laboratorium, dicht bij de receptie.
x
Eén oproep specifiek voor KCvD.
x
Eén mobiel dienstnummer.
150
Maar ook de nascholing hebben we opnieuw vormgegeven. In een programma met
de collega’s van de andere medisch ondersteunende afdelingen verzorgen we
laagdrempelige praktische scholingsmomenten. Hierin trekken we gezamenlijk op
met de radiologen, ziekenhuisapothekers en medische microbiologen.
De titels waaronder deze bijeenkomsten plaatsvinden: Kweek van de week, Preek
van de apotheek, Betoog van de radioloog en Labbabbal spreken voor zich. Deze
scholingsmomenten vinden wekelijks plaats. ‘s Ochtends direct na de overdracht bij
de interne geneeskunde, kindergeneeskunde, gynaecologie en chirurgie. Met een
lagere frequentie doen we hetzelfde bij de neurologie, urologie, SEH en anaesthesie.
Beide activiteiten, KCvD en de scholingsmomenten, hebben ons geholpen om een
nadere invulling aan consultverlening te geven. Vooral op praktische zaken gericht
en met een hoge frequentie gegeven. Maar het heeft meer opgeleverd. In de
continue gesprekken met onze klinische collega’s bleek al snel dat hun richtlijnen niet
altijd dezelfde informatie bevatten. Richtlijnen voor de gynaecologie ten aanzien van
het gebruik van het urinesediment bevat niet dezelfde informatie als de richtlijn van
de kinderartsen. In beide gevallen was er geen klinisch chemicus betrokken bij het
opstellen van de richtlijn. Daar liggen wel onze kansen. Maak consultverlening niet
afhankelijk van toevallige kontakten maar geef het een structurele invulling.
Vooruitblikken
Maar laat ik vooruit blikken. Op welke wijze kunnen wij in de toekomst onze bijdrage
verder vormgeven?
Daarvoor maak ik graag gebruik van onderstaand schema.
151
De patiënt centraal en meerdere zorgverleners met ieder hun eigen bijdrage om de
patiënt heen en met een open communicatie naar elkaar toe. De intensieve
contacten van de zorgverleners en uitgebreide consultverlening over en weer zullen
de zorg voor de patiënt ten goede komen. Maar de individuele beroepsbeoefenaar
kan dit niet alleen. Hij/zij zal door zijn wetenschappelijke vereniging ondersteund
moeten worden met richtlijnen. Richtlijnen die de wetenschappelijke vereniging op
haar beurt weer opgesteld heeft samen met andere wetenschappelijke verenigingen.
En ook dat samenwerkingsverband moet niet van ad hoc kontakten afhankelijk zijn.
Deelname aan een gestructureerde vorm zoals de Raad Kwaliteit van de OMS (Orde
van Medische Specialisten) zal voor alle betrokken verenigingen een stap voorwaarts
zijn. Onze bijdrage aan de richtlijnen van andere wetenschappelijke verenigingen zal
veel energie vragen maar op de werkvloer (lees: tijdens de consultverlening) van
grote waarde zijn. Een tweede aspect dat wij in de toekomst verder moeten
uitbreiden
betreft
de
onderbouwing
van
de
informatie
die
door
de
laboratoriumspecialist klinische chemie over een analytisch resultaat gegeven kan
worden.
Wat betekent déze uitslag voor déze patiënt. De analytische validatie van nieuwe
testen wordt door ons zorgvuldig uitgevoerd maar de klinische validatie blijft hier bij
152
achter. “Waarom wordt de implementatie van nieuwe laboratoriumtesten niet met
dezelfde strengheid uitgevoerd als de testen voor nieuwe geneesmiddelen?”, zo
vraagt O’Kane zich af in een recent editorial van de Annals of Clinical Biochemistry
(2013;50:293-295). “Is het wel verantwoord naar de patiënt toe dat geavanceerde
analytische testen door iedereen uitgevoerd kunnen worden en al dan niet voorzien
van een interpretatie aan de aanvrager verstrekt worden?”. Hebben wij als
wetenschappelijke vereniging hier niet een taak om de krachten te bundelen? De
vraag stellen is hem bijna beantwoorden. Maar dat kunnen we als wetenschappelijke
vereniging niet alleen. Samen met andere wetenschappelijke verenigingen en
ondersteund door de diagnostica industrie kunnen we hieraan vorm geven. Goed
uitgevoerde klinische evaluaties geven een schat aan informatie die in de
gezamenlijke richtlijnen verwoord kan worden. Daarmee vormt deze informatie een
sterke basis voor de consultverlening in alle individuele kontakten. Daar zijn alle
laboratoriumspecialisten mee gebaat.
De PAOKC van vandaag is een goed voorbeeld hoe we elkaar kunnen informeren
over consultverlening. Daarmee geven we invulling aan een belangrijk aspect van
ons vak. En dat kunnen we niet vaak genoeg doen.
153

syllabus

Transcription

Similar documents

WORKING FOR RARE AND GENETIC DISEASES

2015 – VIEW JUNIOR SELECTIONS

Sparender bestraling van de axilla?

W NEWSFLASH nr°17 HOSPITHERA,

Abstracts Najaarsvergadering 6 november 2009

Cubital tunnel

Klik hier - ThromboGenics

Stress en pijn in vissen

ONederlands Tijdschrift voor - Nederlands Tijdschrift voor Orthopaedie

De afdeling Medische Oncologie, eerder afdeling

Medische tests (beoordeling stand van de wetenschap

Year Report Julius Center 2007