Human Resource Truths

Mike Astion, M.D. Patient Safety
Acknowledgements
Patient safety in the clinical lab
• Bio-Rad
• Numerous colleagues in
the Dept of Lab Medicine,
Univ of Washington
• Debbie Su, MT
Michael Astion, MD, PhD, HTBE
Professor and Director, Reference Laboratory
Services
Dept Laboratory Medicine, University of Washington
• Dr. James Hernandez,
Linda Nesberg, Mayo Clinic
• Peggy Ahlin, Bonnie
Messinger, ARUP
• AACC / ClinLabNews/Dr.
Nancy Sasavage
Overview / Conclusions
• QI is enhanced by ...
• Increasing personal accountability
• aligning $ incentives with quality
• loving the patient
Human Resource Truths
• The foundation of quality in clinical laboratories
is a commitment to measure and the courage to
perform carefully chosen, strong interventions,
especially automation.
• You can do it!!
• Thanks.
Just Culture: 3 errors and the response to them
HR Truth #1:
Effective managers support a “low blame”
workplace, but don’t believe “blame-free”
is achievable. They choose a highly
accountable, “just” culture supported by
metrics.
1. Honest human errors (most errors belong here)
• Response: fix system
2. Errors due to at-risk behavior (e.g. short-cutting)
• Response: fix system; then coach, monitor
3. Errors due to recklessness (e.g., impairment, intentional)
or incompetence (error rates >> peer error rate)
• Response: punish
Just culture: interview with David Marx. Lab Err Pat Safety. 2005; 1(6)2-4.
Mike Astion, M.D. Patient Safety
#2: Self evaluation is mostly useless and
can lead to patient harm. Objective
competency assessment of critical tasks /
knowledge is the path to patient safety.
Objectivity holds us accountable.
Who should be blamed and disciplined?
•
tech with data entry error?
•
miscommunicating tech?
•
forgetful doctor?
•
nurse mislabeler?
•
All of above?
•
None of the above unless intentional, impaired, or error rate
>> peer rate
The inaccuracy of self-assessment: Use objective
competency assessment, not self-assessment
Example: Objective Competency Assessment
100
Percentile
•
Self-assessment
Actual
performance
0
Bottom
2nd
3rd
Top
Kruger J, Dunning D. Unskilled and unaware of it…J Personality and Social
Psychology. 1999;77:1121-1134.
Haun DE, et al. Assessing the competence of specimen-processing personnel.
Laboratory Medicine. 2000;31:633-637.
Reed et al. 2 year study of patient safety competency using
online method (www.medtraining.org).
– 875 staff from 29 labs used online competency exams
– Each person completed 40 questions over 2 years
– 5 topic areas covered
• workplace culture
• classifying errors
• how to reduce errors
• prioritizing QI
• general concepts
Reed et al. A 2-Year Study of Patient Safety Competency Assessment in 29
Clinical Laboratories. Am J Clin Path 2008; 129: 959-962.
Weakest Area: How to Reduce Errors
“Which will most significantly reduce pipetting errors?”
• 69% chose “robotic pipetting after usability testing”
• 29% selected the weaker: “Warning label near pipettes…”
• 2% chose the worst: “Quietly tell techs to be careful.”
$ and commonsense
Mike Astion, M.D. Patient Safety
Furniture,
Fountains
Loving the Patient
Euphemisms are a form of communication
that contributes to disconnection.
• “occurrence”
• “variance”
• “adverse event”
$
of
fe
e
$ $ $$
$$ $ $
C
Art
Misaligned financial ($) incentives take away
from patient safety. Projects are most likely
to succeed when $ incentives line up with
patient safety. Example: Automation.
Misaligned $, Example 1:
Furniture, Flooring, Coffee Shops, Waterfalls
and Artwork are NOT patient safety
$
$
Patient
Safety
Disconnection from patients harms patients.
Disconnection is caused by being one step
removed from patients and by poor
communication in discussing patient harm.
“Estimates vary regarding the frequency of
physician disclosure, but a reasonable estimate is
that about 30% of harmful errors are disclosed to
patients. When caregivers choose to disclose, they
choose their words too carefully.”
©Jennifer Astion
“Welcome to the sentinel event”
From: Current Concepts in the Disclosure of Serious Medical
Errors to Patients: An Interview with Thomas Gallagher, MD. Clin
Lab News, January 2008
www.aacc.org/publications/cln/2008/July/Pages/0708_safety1.aspx
Mike Astion, M.D. Patient Safety
When patients come to lab meetings
“The experience created a palpable sense of
community in the lab. For weeks following these
sessions, we got comments from employees saying
how good the experience had been. Our employees’
focus returned to the patient. In closing, we would like
to share this advice with CLN readers. If you are still
skeptical about bringing patients into the lab, trust us.
Your skepticism will evaporate the moment you meet
your first patient and hear their story.”
Letter to the Editor: Clin Lab News, April 2010
Cathy Groen, MT (ASCP) and Corinne Fantz, PhD
www.aacc.org/publications/cln/2010/april/Pages/safety2.aspx
When lab techs visit patient care units
“Actually seeing a baby undergoing ECMO, while
a dedicated care team labored to provide care,
affected the technologists deeply. It only took
about 20 minutes to create a dramatic shift in
their thinking by connecting them to these tiny
patients and their care providers. This change
has improved teamwork between lab staff and
nursing and created a fertile environment for
carrying out other quality improvements.”
Letter to the Editor: Clin Lab News, October 2009
Kim Skala, MT (ASCP) and Corinne Fantz
www.aacc.org/publications/cln/2009/october/Pages/1009_safety4.aspx
6 ways to overcome disconnection
1. Feedback patient
outcomes to lab staff
4. Participate in apologies
to patients.
2. Patients attend lab
meetings.
5. Work more directly
with nursing. Hear and
share their stories.
3. Have lab staff attend
patient rounds
Implementing strong interventions to
↓ errors requires the measurement of
quality and the monitoring of behavior.
6. Speak in stark, plain
language about patient
safety.
Disconnection from patients and care providers as a latent error in
pathology and laboratory medicine: An interview with Dr.
Stephen Raab. 2009. Clinical Laboratory News. 35(4): 14-15.
Example 2, measuring and monitoring:
Ex 1, monitoring: I would not drink milk directly
from container if I knew I was being monitored.
Call center project from a large, de-identified hospital.
• Before: “Laaaaab!!”
• After: “This is the clinical laboratory, can I help you”
20%
10%
0%
©Istock photo
Before
After
Abandon call rates before and after call center. FTE = constant.
Mike Astion, M.D. Patient Safety
Abandoned Call Rate (Target 3%)
Average Speed of Answering Calls
(Target < 10 sec)
Choose strong interventions to
↓errors even though they are
risky, time consuming, and
difficult to implement.
Weak, easy, low risk interventions in the Lab
• Training (alone)
• Telling people to be
more careful
• Double checks without
accountability
• Warning labels
• Memos
MEMO 4/7/10
To: All Lab Staff
From: Dr. BigEgo, Lab Director
Re: Lab errors
Stop making errors
immediately. Be more careful
when performing tests, and
when entering data. All of you
need to go to keyboard training
to learn how to type. For god
sakes, do a better job. You all
don’t know what you are doing.
www.patientsafety.gov/
Stronger interventions
Intervention
Example(s)
∆ Physical plant
Automation zone
Major software
enhancement
•CPOE, LIS-EMR interface
•Autovalidation
Eliminate steps
•Analyzer consolidation
•Direct tube sampling
Industrial engineers
•IE design of courier systems, staffing
Equipment
standardization
•One glucometer for POC sites
•Front-end automation
Advanced design for
manual work
• ↓ log in errors by load leveling and
isolation, specialization, ↑supervision
/accountability and redundant data entry.
Strong Interventions:
Robots are Good and Bigger
Robots are Even Better
Mike Astion, M.D. Patient Safety
Small Robots: Immunoassay Consolidation
Big Robots: Total Laboratory Automation decreases
many kinds of errors.
LEAN: ↓steps,
↓motion,
↓specimen
requirements, ↓TAT
↓
↓
↓
↓
EIA
Western Blot
Gel diffusion
18th century
methods
loved by
Lab Director
IFA
The lab of future puts many more assays into this
type of automated facility.
Automated Immunoassay
Instrument on Automated Line
Robots ≠ perfect.
Errors in the highly automated environment
are much less common. But, errors in the
highly automated environment can harm
many patients quickly.
(TLA at Ohio State, photo courtesy Dr. Mike Bissell)
↓ Risks associated with strong interventions
by usability testing / site visits
•
For everything from
reference lab interfaces,
call centers, analyzers…
•
Usability testing: not
always possible, but great
if you can do it
•
Site visits better than
talking on phone
Cell phone usability testing
Example: Strong intervention supported by
measurement.
Gaining quantitative control over specimen collection,
transportation and processing.
Project begins around 2004. At that
time, the metric that was collected and
reviewed was attendance.
Mike Astion, M.D. Patient Safety
Quality Dashboard: Current State (2010)
Phlebotomy
•
•
•
•
% AM results available by AM rounds
Productivity per FTE
Error rates (mislabel, wrong tube,
contamination rate)
Relabel rate
Couriers:
•
•
arrival time
Missing data re arrival time
Processing:
•
•
•
•
•
•
tally of each type of work by shift
Log in errors for each type of req
Accuracy of cancel / credit
% of errors that harm patients
# specimens processed per hour per
FTE
# specimens decanted per shift
Call Ctr:
• Tally inbound and outbound
calls
• time to 1st answer
• abandon call rate
• FTE: time logged into system
Human Resources
• Attendance
• Fraction of time a supervisor is
present on a shift
• Counseling status
Cause and Effect Diagram:
Potential contributing factors to serious processing errors
Policy &
procedure
Staffing
Training
Organization/Management /
Regulatory
Finance ☺
• Receipts per outside client
• 3rd party payer report card
Work environment
Teamwork factors
Process/Design IT/
Technology
Hospital patients: % of test results available for morning
rounds (8:30AM). Key to success was single piece flow.
Processing: Logging Work Type by Shift – Hospital B
Weekdays
Data Range: 12/01/09 to 12/31/09
Shift
Midnight
0000-0800
BD Check
IF
Partial IF
144
44
272
19
Day Shift
0800-1600
22
101
218
127
Evening
1600-0000
13
164
199
126
Shift
Manual
272 high risk
requisitions
per weekday
Saturdays
Data Range: 12/01/09 to 12/31/09
BD Check
IF
Partial IF
Manual
Midnight
0000-0800
131
28
254
15
Day Shift
0800-1600
11
1
136
45
Evening
1600-0000
10
126
157
14
73 high risk
requisitions
per Saturday
Mike Astion, M.D. Patient Safety
Specimen processing errors:
Weekdays and Weekends are not the same
Reducing errors when work must be manual
Specimen Processing Log in Errors
Measuring Productivity of Specimen Processing
• Isolate the high risk work
• Standardize the work
• Specialize the work to a small group of highly
trained people who are tightly monitored.
• Remove time constraints from the group
• ↓ batch size, smooth work flow because people
tend to work too fast if they see a huge batch of
work in front of them.
• Redundant data entry by 2 staff members
Courier On Time Arrival at Hospital B:
Monthly Summary of Arrival Time: Route A
(Total 7 Routes Mar 09’ – Jan 10’)
ROUTE Mar-09 Apr-09 May-09 Jun-09
A
95% 100%
95%
95%
B1
na
na
na
77%
B2
59%
32%
81%
41%
C
91%
59%
86%
86%
D
91%
86%
95% 100%
E
95%
82%
95% 100%
80% 100%
Weekend
60% 100%
Weekend
H1
H2
H3
H4
I1
I2
I3
I4
I5
Jul-09 Aug-09 Sep-09 Oct-09 Nov-09 Dec-09 Jan-10
100% 100% 100% 100% 100% 100% 100%
64%
75%
59%
95%
83%
93%
89%
30%
52%
70%
86%
85% 100% 100%
87%
75%
79% 100% 100% 100% 100%
100% 100%
94% 100% 100% 100% 100%
95%
84% 100%
91%
95% 100% 100%
100% 100% 100% 100%
75% 100% 100%
67%
0%
0% 100% 100%
75% 100%
96%
71%
86%
86%
94%
94% 100%
100%
81%
90%
90%
88%
88% 100%
89%
83%
93%
93%
88% 100% 100%
91%
94%
94%
94%
91% 100% 100%
100% 100% 100%
100%
94% 100%
75% 100% 100%
11% 100% 100%
6% 100% 100%
Target : no later than 18:45 pm
Mike Astion, M.D. Patient Safety
Monthly Summary of Arrival Time:Route B
Target: no later than 15:00 pm
How many QI philosophies do I need, and
how many tools within each philosophy?
Philosophies
TQM
Six Sigma
Tools
Fishbone
Diagram
Process
Map
CQI
Checklist
Lean / Toyota
Production
System:
(Please
rename me)
RCA/
FMEA
Example(s)
∆ Physical plant
Automation zone
Major software
enhancement
•CPOE, LIS-EMR interface
•Autovalidation
Eliminate steps
•Analyzer consolidation
•Direct tube sampling
Industrial engineers
•IE design of courier systems, staffing
Equipment
standardization
•One glucometer for POC sites
•Front-end automation
Advanced design for
manual work
• ↓ log in errors by load leveling and
isolation, specialization, ↑supervision
/accountability and redundant data entry.
Backbone is the most important
philosophy and tool. Backbone failure
hinders patient safety.
Time/
Motion
Studies
Running
away
Crying
©Istock photo
Etc…
Conclusions
• The foundation of quality in clinical laboratories
is a commitment to measure and the courage to
perform carefully chosen, strong interventions.
• The chances of succeeding are improved by
enhancing accountability, choosing projects
where $ incentives align with quality, and
attacking disconnection.
• You can do it!! Robots can help.
• Thanks.
Stronger interventions
Intervention