should one or two embryos be transferred in ivf?

Transcription

- A health technology assessment
Danish Health Technology Assessment 2005; 7 (2)
2005
SHOULD ONE OR TWO EMBRYOS BE
TRANSFERRED IN IVF?
Danish Centre for Evaluation and Health Technology Assessment
SHOULD ONE OR TWO EMBRYOS BE TRANSFERRED IN IVF?
- A health technology assessment
Hans Jakob Ingerslev1, Peter Bo Poulsen2, Ulrik Kesmodel3, Astrid Højgaard1, Anja Pinborg4,
Tine Brink Henriksen5, Jens Seeberg6, Lars Ditlev Ottosen1
1. Fertility Clinic, Skejby Sygehus, Aarhus University Hospital
2 MUUSMANN Research & Consulting Co.
3 Institute of Public Health, Department of Epidemiology, University of Aarhus
4 Fertility Clinic, Juliane Marie Centre, Rigshospitalet, Copenhagen University Hospital
5 Department of Paediatrics, Skejby Sygehus, Aarhus University Hospital
6 HRA Consult & Department of Antropology and Etnography, University of Aarhus
Danish Health Technology Assessment 2005; 7 (2)
Should one or two embryos be transferred in IVF? A health technology assessment
C National Board of Health, Danish Centre for Evaluation and Health Technology Assessment
Key words: health technology assessment, HTA, Single embryo transfer, IVF, Twins, SET
Language: English with summary in Danish
URL: http://www.sst.dk
Version: 1.0
Version date: 20051017
ISBN (electronic version): 87-7676-208-4
ISSN (electronic version): 1399-2481
Format: Pdf
Design: 1508 A/S and National Board of Health
Published by: National Board of Health, Denmark, november 2005
This report should be referred as follows:
Ingerslev HJ, Poulsen PB, Kesmodel U, Højgaard A, Pinborg A, Henriksen TB, Seeberg J, Ottosen LD
Danish Health Technology Assessment 2005; 7(2)
Copenhagen: National Board of Health, Danish Centre for Evaluation and Health Technology Assessment, 2005
Series Title: Danish Health Technology Assessment
Series Editorial Board: Finn Børlum Kristensen, Mogens Hørder, Leiv Bakketeig
Series Editorial Manager: Stig Ejdrup Andersen
For further information please contact:
National Board of Health
Islands Brygge 67
DK-2300 Copenhagen
Denmark
Phone: π45 72 22 74 00
E-mail: cemtv/sst.dk
Home page: www.dacehta.dk/www.cemtv.dk
The publication can be down loaded free of charge at: www.dacehta.dk
Foreword
In the last decades the twin birth rate has more than doubled in Denmark and in other European
countries. Fertility treatment is assessed to explain 70% of the increase and in vitro fertilization
(IVF) makes up about half of this.
In Denmark two embryos (double embryo transfer, DET) are transferred in most IVF treatments.
Two embryos increase the chance of pregnancy but also of having twins. It is well-known that
multiple births are associated with increased risk for both mother and child, first of all because of
risk of premature delivery.
The 1997 guideline from the Danish National Board of Health recommends that a maximum of
two fresh embryos should be transferred in the same treatment. The couple should also be involved
in the decision-making process and therefore needs to receive the necessary information. The purpose of the guideline is to minimize the number of transferred embryos and at the same time ensure
that the treatment meets the couples’ desire of having a child.
The Danish Council of Ethics and the Danish Ministry of the Interior and Health have asked for
more studies concerning consequences of different IVF-strategies to be able to improve the basis
for decision-making.
This HTA-project contributes with the information in demand by discussing the consequences of
implementing an obligatory single embryo transfer (SET) policy versus optional two embryo transfer in Denmark. The focus is on consequences for pregnancy rates, the couples preferences and
knowledge and on economical and organisational aspects.
One conclusion is that an enforced SET-policy would be in conflict with the patients’ interests and
wishes. Most couples have a desire for more than one child. Furthermore, the couples wish to
achieve children by twin pregnancy rather than by successive single pregnancies. This is also the
case when the couples are informed of the increased risk of complications in twin pregnancy.
From an economical perspective a DET-policy is more cost-effective per delivery and per child
compared with a SET-policy. However, a DET-policy involves higher cost for the health care sector
and society than SET due to higher costs for antenatal care, neonatal care and production lost.
The report cannot unambiguously recommend that in future only one embryo should be transferred
in IVF. But the report explains the consequences of using one transferring method in preference to
the other. By this it contributes with relevant information to decision-makers at clinical and political
level, as well as to the infertile couples.
The report is published in DACEHTA’s series ‘‘Danish Health Technology Assessment’’ and has
gone through an editorial process and external peer-review before publication in the series.
November 2005
Finn Børlum Kristensen
Director
4
Table of contents
Summary in English
7
Dansk resume
13
1
19
Introduction
2 Multiple birth rates in Europe during the last two decades
21
3 Risks associated with multiple pregnancy and birth
3.1 Pregnancy
3.2 Delivery
3.3 Malformations
3.4 Neonatal complications
3.5 Long-term consequences
3.6 Vanishing twins
22
22
22
23
23
24
24
4 Factors influencing the twin birth rate
4.1 Natural conception
4.2 Assisted conception
4.3 Elective single embryo transfer (eSET)
4.4 Results of randomization between one and two embryos in an unselected population of patients
in their first cycle at the Fertility Clinic, Aarhus University Hospital, Skejby Sygehus
4.4.1 Introduction
4.4.2 Materials and methods
4.4.3 Results
4.4.4 Discussion
4.5 Tools for selection of patients for eSET
4.5.1 Prediction models for pregnancy and twin risk following IVF/ICS
4.5.2 Other tools to identify treatment cycles for eSET, simple inclusion criteria
26
26
26
27
5 What do the infertile couples prefer: Single or double embryo transfer – a single child or twins – and why?
5.1 Introduction
5.2 Results of a qualitative study – Patient perspectives
5.2.1 Methodology
5.2.2 Findings
5.2.3 Discussion and conclusion
5.3 Results of a questionnaire study
5.3.1 Introduction
5.3.2 Material and methods
5.3.3 Results
5.3.4 Discussion
5.4 Patient perspectives- discussion and conclusions
38
38
38
38
39
48
48
48
49
50
56
57
6 Cost-effectiveness of SET versus DET strategies
6.1 Introduction
6.2 Purpose
6.3 Method – type of analysis
6.4 Material
6.4.1 Data collection
6.4.2 Data on resource use
6.5 Results
6.5.1 Cost of an IVF singleton versus IVF twins
6.5.2 Cost-effectiveness of SET- versus DET-policies
6.6 Sensitivity analyses
6.6.1 IUI couples excluded from the analysis
6.6.2 Neonatal Intensive Care
6.6.3 Cost of cerebral palsy
6.6.4 Production lost changed
6.6.5 Direct costs changed
6.7 Discussion
58
58
58
58
59
59
60
60
60
67
71
71
71
72
72
72
73
29
29
29
30
33
34
34
35
5
7 Organisational consequences of SET
7.1 The process
7.1.1 Treatment numbers
7.2 Estimates of economic consequences of SET
7.2.1 Public health care clinics
7.2.2 Private fertility clinics
7.2.3 Savings in expenses to twin pregnancy, delivery and neonatal care following SET
7.2.4 Economic balance in case of SET
7.2.5 Considerations on public funding of IVF in Denmark
7.3 Staffing and education
7.4 Communication and culture
7.5 Discussion and conclusions on organisational perspectives
75
75
75
77
77
78
78
79
79
79
79
80
8 Discussion and conclusion
82
References
88
Appendix 1. Information on the choice between one or two embryos and on twin pregnancies
Appendix 2. Information omkring valget mellem et eller to æg og om tvillingegraviditet
Appendix 3. Sensitivity analyses – cost-effectiveness ratios
93
95
97
Ordforklaring
98
6
Summary in English
In vitro fertilization (IVF) and intrauterine insemination (IUI) are major assisted reproductive
techniques used for infertility treatment. These treatments were introduced in Denmark in the
beginning of the eighties. Presently around 10.000 IVF treatments and a similar number of IUI
treatments are performed yearly in Denmark. In 2004 around 4% of all Danish children resulted
from IVF treatments. Within the public health care system three IVF treatments are offered to all
infertile couples, who do not have children together and where the female is ∞40 years of age.
Accordingly, most Danish counties offer free treatment to one child only, unless the treatment
results in twins or higher multiples or if embryos are frozen. All counties offer free transfer of frozen
embryos derived from treatments in the public health care system – also to a second child.
In Denmark two embryos have been transferred in most cases until now (double embryo transfer,
DET). Thus, in 2004 DET was performed in 68% of all treatments, single embryo transfer (SET)
in 27% (either electively (eSET) or due to the fact that only one embryo was available), and three
embryos were transferred in 5%. Obviously DET increases the likelihood of twins. It is well-known
that twin pregnancies are associated with increased risk of complications in pregnancy, first of all
an increased risk of premature delivery. Premature newborns have increased mortality – depending
upon the gestational age – and increased demands for neonatal care and risk of long term sequelae.
In 2004 26.4% of deliveries following IVF treatment in Denmark were twin deliveries, compared
with approximately 1% following natural conception.
A SET policy may either be obligatory (comprising all treatments) or elective (eSET), e.g. comprising women with a high chance of pregnancy, either based on a voluntary informed decision or
obligatory to the good prognosis group.
The dilemma for the infertility practitioner is to create a balance between the desire to maximize
the opportunity for the infertile couple to have a child and the need to minimize the risk of harm
to the future child. This is also an issue for the decision-makers, since the dilemma has consequences
for resource allocation and the society as a whole.
Purpose of this health technology assessment
The overall purpose of this health technology assessment (HTA) report was to describe the scientific
knowledge and to contribute with information and input to the decision-makers on the question
on an obligatory single embryo transfer policy in Denmark. Accordingly, this HTA evaluates if a
SET policy compared with the present DET policy would reduce the chance of pregnancy, what
the attitudes of the infertile couples are to SET and fewer twin pregnancies, and finally what the
organisational and economical consequences would be.
Materials and methods
The comparison between a SET and a DET policy was based upon the systematic and comprehensive framework of an HTA covering each of the following aspects: the technology, the patient, the
economy, and the organisation.
The principal question whether one or two embryos should be transferred in IVF was elucidated
by an analysis of the literature and by a randomized study comparing results following transfer of
a single fresh embryo (SET) with transfer of two embryos (DET) in an unselected population of
infertile patients. Secondly, this HTA evaluated patient attitudes to this question by the literature,
7
and a qualitative interview study, and by a mailed survey. The organisational consequences of
introducing SET in Denmark were analysed in terms of likely changes in organisational processes
such as workload and flow of patients, and effect on staffing. Finally, the possible economic consequences in the public health care sector of introducing SET in Denmark were evaluated by a health
economic analysis. The data for this analysis were based upon published clinical studies comparing
SET with DET and on the basis of own data on costs associated with pregnancy and delivery of
singletons and twins. Thus 213 couples with a pregnancy following IVF filled in cost diaries during
one year from establishment of the pregnancy (ultrasound in week 8) and until three months after
delivery.
Results
The technology
The increase in the twin birth rate by 2.4 fold in Denmark and similar figures from other European
countries during the last decades seem to be caused partly by increasing age of women at start of
reproduction (1/3) and partly to consequences of fertility treatment (2/3). In Denmark, the age of
an average primipara was 24.1 years in 1973 and 28.6 years in 2003. Twinning rates increase 4fold from 15 to 37 years of age. Fertility treatment seems to explain 2/3 of the recent national
increase in twin rates, with an equal contribution from IVF and other infertility treatment (IUI
and simple hormone stimulation). Half of that is associated with IVF/ICSI treatment. Accordingly,
since IVF contributes to 1/3 of the total increase of twinning rates, transfer of a single embryo at
a time could result in a certain reduction of the risk of twinning rates following IVF/ICSI, but
hardly without an expected reduction of pregnancy rates and fewer children born.
Published randomized controlled studies (RCT) of single embryo transfer (SET) versus double
embryo transfer (DET) to selected groups with a predicted good prognosis for pregnancy have
shown an overall reduction in the pregnancy rates from an average of 48.1 to 31.3% with a
reduction in twin rates from an average of 34.9% to 2% after transfer of fresh embryos. However,
in the largest largest Scandinavian multicentre RCT, successive replacement of a frozen and thawed
embryo in an eSET group increased the cumulative pregnancy rate to a level, which was similar to
that obtained by transfer of two fresh embryos in the DET group.
Surprisingly, the present randomized study of transfer of one or two fresh embryos in an unselected
population did not reveal any difference in pregnancy rates. Observational studies from Finland
and Belgium indicate unchanged pregnancy rates following introduction of single embryo transfer
to selected groups of patients (eSET). Data from Sweden similarly show on a national basis that it
has been possible to implement eSET to a large proportion of all patients without lowering pregnancy rates. Such data may indicate lack of an otherwise expected increase in pregnancy rates.
Despite use of elaborate prediction models it seems difficult theoretically to identify a group of
patients with a predictable high chance of pregnancy and risk of twins for selection to eSET without
significant consequences in terms of overall pregnancy rates on one side and limits effects on total
twin rates on the other.
Maternal and neonatal risks
Multiple pregnancy puts both the mother’s and the child’s health at risk. Complications of pregnancy occur more frequently in twin than in singleton pregnancies. Preeclampsia occurs with a 2.4
fold increased risk (13.4% vs. 6.2%) in IVF/ICSI twin vs. singleton pregnancies. Other complications of pregnancy such as hypertensive disorders, thrombo-embolism, urinary tract infections,
anaemia, and vaginal-uterine haemorrhage are more frequent leading to increased risk of extended
periods of bed rest, and hospitalization. Twins have an increased risk of being born preterm. In a
Danish register study, the crude percentages of children born prior to 37 completed weeks of
gestational age were 43.9% in IVF twins and 7.3% in IVF singletons the odds of birth prior to 32
completed weeks was increased 7-fold with crude frequencies being 8.5% in IVF twins vs. 1.3% in
8
IVF singletons. These risks result in increased use of medication to prevent preterm labour and
surgical procedures such as Caesarian section (52.9%). European data indicate a maternal mortality
of 4.4 per 100.000 in singleton pregnancies and 10.2 per 100.000 for multiple pregnancies.
Some malformations appear at a higher frequency in both spontaneously conceived and in IVF
twins compared to singletons. The risk increase is modest and requires very large studies to be
demonstrated. Further, some of these risks changes may at least partly be secondary to preterm
delivery.
Neonatal morbidity in terms of neonatal intensive care unit (NICU) admission is 3.8 fold increased
in IVF twins (56.4%) compared to IVF singletons (25.4%) and twins on average spent 9 days
more in the NICU than singletons. Neonatal mortality seems similar or even slightly lower in IVF
than in spontaneously conceived twins and Danish results revealed a significantly lower risk of
death during the first year of life. However, perinatal mortality was twice as high in IVF twins as
in IVF singletons (20.7 vs. 11.0 per 1000).
In a large Danish study neurological sequelae and cerebral palsy occurred with the same prevalence
in IVF twins, spontaneously conceived twins, and IVF singletons. However, cerebral palsy seems
increased in IVF singletons compared with singletons from the normal population. This may be
explained by the vanishing twin phenomenon, which is a twin foetus that disappears after documented foetal cardiac activity, occurring in 10.4% of live-born IVF singleton pregnancies. Survivors
of a vanishing co-twin had a significantly increased risk of low birth weight, preterm delivery and
mortality compared with other singletons, but also compared with children following an IVF pregnancy with only one detectable foetus from the first ultrasound.
The patient
A qualitative study
The infertile couples were well informed about the treatment process and they knew about the risk
of preterm delivery in twin pregnancies, but usually not detailed statistics. They had a broad view
on the issue of twin pregnancy including projected quality of life for the family-to-be and intended
reduction of future risks. Risk information that basically questioned the infertility treatment tend
to be ignored or seen as unwelcome, whereas information that confirms (or at least not questions)
what the couple desires easily can be accepted. One of the important factors in decision-making
was the widespread desire for more than one child. Very positive values were attached to having a
sibling. The physically and psychologically stressful character of the IVF treatment as well as social
stressors and the fact that IVF treatment was only offered to obtain the first child for the couple
influenced positively the wish to achieve children by twin pregnancy rather than by successive single
pregnancies. Most couples found the total load of these factors to be considerable and to constitute
a sufficient reason to prefer a twin pregnancy. The present results strongly indicated that an obligatory single embryo policy would be in conflict with patient interests and wishes.
The survey
The present study revealed that the majority of infertile couples prefer twins (59%) to one child at
a time (38%), but a larger majority (79%) planned to have two embryos transferred in the next
treatment. Accordingly, the preference of DET is not only explained by a wish to have a high
success rate and thus avoiding having more treatments, but reflects a deliberate wish to have twins
in the majority of couples. Reasons for wishing twins were desire for siblings, mutual pleasure
between siblings or less specific positive attitude to twins. A single child was preferred (42%) due
to risk of a difficult pregnancy with twins, and risk to the child or mother.
The information given on twins was evaluated as very satisfactory, satisfactory or fair by nearly all
couples, but only a little less than half had received oral counselling. Only 12% stated that their
decision on number of embryos transferred was due to advice by the fertility clinic whereas one
third described the decision as their own choice. The general impression was that more specific and
9
organised information is needed. In bivariate analyses in the present study there was no association
between opting for twins and having received information or feeling well informed. The desirability
of twin gestations described in three different scenarios with increasing risks for mother or child
revealed that the median desire for a twin gestation decreased with increasing risk. As risk aversion
could be demonstrated by wording the statistical result as a chance to one half of the couples and
risk to the other half. Accordingly, infertile couples seem to be affected by information offered on
the twin issue.
Treatment-related physical and psychological stress seemed to be an important factor for wishing
twins, whereas economic motives for twin preferences seemed of less importance. Only 27% would
opt for SET if combined with four reimbursed embryo transfers or even an unlimited number of
treatments or if treatments for a second child were reimbursed.
The Economy
The obvious advantage associated with a SET policy is a lower rate of the more expensive multiple
pregnancies compared with DET, where the twin rate is around 25%.
However, in the present study the SET-policy did not show to be more cost-effective compared
with the DET-policy. The cost per clinical pregnancy using SET was DKK 131,446 compared with
DKK 115,321 with DET, and the cost per delivery was similarly higher using SET (DKK 149,833)
compared with DET (DKK 120,324). However, including frozen embryos the cost-effectiveness
was almost similar for the two alternatives.
The difference between SET and DET increased in terms of cost per child born, being DKK
148,204 in case of SET compared with DKK 93,265 with DET.
The lower cost-effectiveness ratios by DET is due to the fact that DET is more effective compared
with SET, since DET results in a higher clinical pregnancy rate, higher delivery rate and more
children born – partly due to twins. At the same time DET is more expensive per patient treated
due to higher costs associated with delivery and neonatal care. Accordingly, the extra cost per patient
(incremental cost-effectiveness ratio) by using the more efficient DET policy is around DKK 82,000
per additional delivery and around DKK 50,000 per extra child delivered following DET. However,
the extra costs per delivery and per child born with DET do not seem high.
The conclusion from this Danish cost-effectiveness analysis comparing a SET-policy with the traditional DET-policy used today is that the cost per delivery or the cost per child is lower using
DET compared with SET. However, DET involves higher total costs per woman treated, i.e. costs
for antenatal care, delivery, neonatal care and production lost due to the higher frequency of twin
pregnancies and deliveries, and this involves higher costs for the health care sector and society than
SET.
Organisational perspectives
If the average cumulative pregnancy rate for patients undergoing IVF should be maintained following a change from DET to SET more IVF treatments are necessary, i.e. hormone stimulations and
oocyte pick-ups.
More straws are to be frozen in SET than in DET, and more embryo thaws are to be performed.
Finally, more transfers of frozen-thawed embryos are necessary in case of SET.
It is difficult precisely to calculate what the specific consequences of these changes will be in respect
to logistics and need for resources, but estimates were done.
It was estimated that introduction of a SET policy has the following consequences in Denmark,
supposing maintenance of an unchanged cumulative pregnancy rate by increased number of cycles
10
offered to compensate for a lower pregnancy rate in SET and still treatment for the first child
only:
a.
b.
c.
d.
e.
500-1000 more hormone stimulations, oocyte pick-ups, and fresh embryo transfers per year.
Freezing of 900-1800 extra straws containing a single embryo.
900-1800 more transfers of frozen embryos per year.
18 more transfers of frozen embryos per year to obtain a second child.
756 more hormone stimulations and oocyte pick-ups, and fresh embryo transfers per year to
obtain a second child in private fertility clinics following an obligatory SET policy.
An estimate of economic consequences in case SET is introduced in Denmark is that the public
health care system will carry an extra burden of around DKK 14.9-25.9 mio. depending upon what
the conditions will be for introduction of SET. The savings in respect to reduced obstetrical and
neonatal care represent between DKK 12 and 24 mio. Based upon these assumptions, introduction
of SET seems to represent a change in the public expenses and budgets associated with IVF varying
between a saving of DKK 9.1 mio. (best case) and increase in expenses of DKK 13.9 mio. (worst
case). An extension of public funding to treatments for child number two could represent a compensation for the lost chance of having two children in one treatment – and may help to decide on
SET in a situation, where the couple are responsible for the choice between SET or DET.
Introduction of SET will necessitate employment and education of extra staffing, which is realistic
to carry out within a time frame of six months to a year.
Any non-obligatory SET policy will impose new or enhanced tasks of informing the infertile couple
on risks associated with twin pregnancy and delivery against consequences for their chances to
obtain pregnancy to help them decide to have one or two embryos transferred. This represents
undoubtedly a change in culture and priority of the professionals treating the patients.
Respect for patient autonomy is retained by leaving the decision to the couple by defining strict
rules on the content of the information on the risks of twin pregnancy and delivery and the national
recommendations to the patient, and may be preferable to more rigid alternatives such as obligatory
SET or strictly defined eSET. Ultimately it is the couple that carries the consequences of choosing
either one or two embryos.
Conclusion
Overall, the present HTA analysis allows the following answers to the HTA questions asked concerning consequences of SET versus DET in respect to technology, patient, organisation and
economy:
1. To what extent does an unselected SET instead of DET reduce pregnancy rates in IVF?
All previous randomized studies have shown that elective single embryo transfer reduces the pregnancy rate per fresh cycle significantly. However, the present randomized study did not reveal any
difference, possibly due to random variation or an unquantifiable difference in embryo quality.
Observational data from other studies have indicated similarly that it is possible to maintain unchanged pregnancy rates following introduction of single embryo transfer to selected groups of
patients.
2. What is the basis for the couple’s decisions on the twin question, and how would an obligatory single embryo
transfer policy be in keeping with the interests of the infertile couple?
Previous studies and our own data on patient attitudes revealed a strong desire for twins among
couples undergoing fertility treatment. Patients formed this decision on the basis of an evaluation
of the social, psychological and physical discomfort related with IVF treatment combined with the
11
wish to have more than one biological child. Accordingly, an enforced single embryo transfer policy
would be in conflict with patient interests and wishes.
3. What organizational consequences are expected in case of introduction of an obligatory single embryo transfer
policy to replace the present DET policy?
Introduction of SET seem to necessitate employment and education of extra staffing and seem to
represent a change in the public expenses to health care associated with IVF between a saving of
DKK 9.1 mio. (best case) and increase in expenses of DKK 13.9 mio. (worst case). Any eSET
policy will create increased challenges to information and councelling of patients in connection
with the choice between one and two embryos balancing risks associated with twin pregnancy and
delivery against chance to obtain a pregnancy.
4. What are the expected health economic consequences for the society following an obligatory single embryo
transfer to replace the present DET policy as judged from an expected lower pregnancy rate and a reduced
consumption of resources with respect to delivery, neonatal service etc.?
In the present study the SET-policy did not show to be more cost-effective compared with the
DET-policy, which is more effective (higher clinical pregnancy rate, higher rate of delivery and
children), but also more expensive (higher delivery cost and neonatal intensive care costs). However,
the extra costs per delivery and per child born with DET do not seem high.
The question ‘‘Should one or two embryos be transferred in IVF?’’ is not easy to answer. Respect
for patient autonomy should be considered against economic aspects including the effectiveness of
these rather physically and psychologically stressful treatments. Complications and long term sequelae associated with preterm delivery derived from twin pregnancies resulting from transfer of two
embryos should also be taken into account. The ultimate answer to this question is to be given by
the decision-makers. The present health technology assessment report is hopefully a help in this
process.
12
Dansk resume
Ufrivillig barnløshed behandles med såkaldt kunstig befrugtning, som er en samlebetegnelse for
inseminationsbehandlinger (IUI) og ægtransplantationsbehandling (IVF). Disse behandlinger blev
introduceret i Danmark i begyndelsen af 1980’erne. Antallet af IVF behandlinger ligger nu på knap
10.000 årligt, og tallet er stort set det samme for IUI. Det skønnes, at omkring 4% af alle danske
børn i 2004 er resultat af IVF behandling. I Danmark betaler det offentlige tre IVF behandlinger
til ufrivilligt barnløse par, der ikke har fælles børn, og hvor kvinden er under 40 år. De fleste danske
amter tilbyder således kun ét barn på det offentliges regning, med mindre behandlingen resulterer
i tvillinger eller flere børn, eller kvinden har nedfrosne æg fra tidligere ægudtagninger. I alle amter
lægges eventuelle frosne æg gratis op – også til barn nummer to.
Ved IVF behandling i Danmark har man hidtil i de fleste tilfælde oplagt to befrugtede æg (på
engelsk kaldet for double embryo transfer, DET). I 2004 blev der oplagt to befrugtede æg ved 68%
af behandlingerne, ét æg (på engelsk betegnet single embryo transfer, SET) ved 27% (enten på
grund af et reelt valg, eller fordi der kun var ét æg til rådighed), og tre æg ved 5% af behandlingerne.
En toægspolitik øger selvsagt sandsynligheden for at føde tvillinger. Det er velkendt, at tvillingegraviditeter er forbundet med en øget risiko for komplikationer, først og fremmest i form af for tidlig
fødsel. For tidligt fødte børn har nedsat chance for at overleve, øget behov for behandling og pleje
i perioden efter fødslen og øget risiko for langtidsfølger. I 2004 var 26,4% af fødsler efter IVF
behandling tvillingefødsler, hvilket er væsentligt højere sammenlignet med naturligt opståede tvillingegraviditeter (ca. 1%).
En enkeltægspolitik kan enten tilrettelægges som et obligatorisk pålæg, dvs. at der kun må lægges
ét befrugtet æg op i livmoderen ved alle behandlinger eller som en elektiv ordning (eSET), hvor
kvinder med særlig god graviditetschance og tvillingesandsynlighed kun får ét befrugtet æg lagt op
efter anbefaling og information eller som en tvungen ordning.
Da oplægning af to optimalt udviklede befrugtede æg alt andet lige giver større graviditetschance
end ét æg, står lægen som klinisk beslutningstager i øjeblikket overfor et dilemma ved at skulle sikre
en passende balance mellem maksimering af graviditetschance for det ufrivilligt barnløse par og
minimering af risikoen for det kommende barn. Da valget desuden har samfunds- og ressourcemæssige konsekvenser er det derudover relevant, at beslutningstagere på det amtspolitiske og nationale
niveau forholder sig til problemstillingen.
Formålet med denne medicinske teknologivurdering
Det overordnede formål med denne medicinske teknologivurdering (MTV) er at afdække den
videnskabelige baggrundsviden og dermed bidrage med information til beslutningstagerne, når de
skal tage stilling til det overordnede spørgsmål, om der skal være en obligatorisk enkeltægspolitik i
Danmark. Det er derfor i denne MTV undersøgt, om en enkeltægspolitik giver en mindre chance
for graviditet end den sædvanlige toægspolitik, hvad parrenes holdning er til enkeltægspolitikken
og færre tvillingegraviditeter, og endelig hvad de organisatoriske og økonomiske konsekvenser er af
at indføre en enkeltægspolitik i Danmark.
Materiale og metoder
Sammenligningen mellem enkeltægspolitikken og toægspolitikken er baseret på en MTV ramme,
hvor elementerne teknologien, patienten, organisationen og økonomien er belyst. Det overordnede
spørgsmål »om der skal transfereres et eller to æg ved IVF« er søgt besvaret dels ved en analyse af
den internationale litteratur om emnet, dels ved egne undersøgelser i form af en lodtrækningsunder-
13
søgelse, som sammenligner graviditetschancen efter oplægning af henholdsvis et og to befrugtede
æg i en ikke-selekteret gruppe af ufrivilligt barnløse kvinder. Projektet har desuden søgt at vurdere
ufrivilligt barnløse pars holdninger til MTV spørgsmålet gennem en litteraturgennemgang, en kvalitativ interviewbaseret undersøgelse samt en spørgeskemaundersøgelse. De organisatoriske konsekvenser af en eventuel enkeltægspolitik i Danmark er belyst ved en analyse af de organisatoriske
ændringer, som kan forekomme i forhold til arbejdsprocesser og patientflow samt bemanding ved
overgang fra toægspolitik til enkeltægspolitik. Endelig er de økonomiske konsekvenser af en enkeltægspolitik i Danmark belyst ved en sundhedsøkonomisk analyse. Datagrundlaget for denne analyse
baserer sig dels på publicerede kliniske studier, der sammenligner enkeltægspolitikken med toægspolitikken, og dels på egne opgørelser af omkostningerne ved henholdsvis enkeltfødte og tvillinger,
idet 213 par, der var blevet gravide efter IVF-behandling, udfyldte omkostningsdagbøger gennem
et år fra graviditetstidspunktet (dvs. fra skanningen i 8. uge) og indtil tre måneder efter fødslen.
Resultater
Teknologien
Den mere end fordoblede tvillingehyppighed gennem de sidste årtier i Danmark og andre lande i
Europa skyldes, dels at kvinder med aldersbetinget stigende tvillingerisiko føder deres børn senere,
dels at barnløshedsbehandling anvendes mere end tidligere. I Danmark var den gennemsnitlige alder
for en førstegangsfødende 24,1 år i 1973 og 28,6 år i 2003. Tvillingehyppigheden øges 4 gange fra
15-årsalderen til 37 år. Fertilitetsbehandling synes at forklare 2/3 af stigningen i tvillingehyppigheden, med et ligeligt bidrag fra IVF behandlinger og anden fertilitetsbehandling (IUI og simpel
hormonbehandling). Da IVF behandling således bidrager til 1/3 af den samlede stigning i tvillingehyppigheden vil oplægning af et æg ad gangen kunne forventes at bidrage til en vis reduktion af
tvillingehyppigheden, men næppe uden lavere graviditetsrater og færre børn født som konsekvens.
Offentliggjorte randomiserede, kontrollerede studier af enkeltægspolitikken overfor toægspolitikken
i patientgrupper med en god graviditetschance har vist, at graviditetschancen efter oplægning af
friske æg (ikke frosne) falder fra 48,1% (toægspolitikken) til 31,3% (enkeltægspolitikken) med et
fald i tvillingehyppigheden fra 34,9% til 2%. For patienter, der har to gode æg, fandt man i en stor
skandinavisk multicenterundersøgelse, at oplægning af et friskt æg efterfulgt af et frosset og tøet til
en udvalgt gruppe patienter gav næsten samme graviditetschance som to friske æg lagt op i samme
behandling.
Lidt overraskende viste nærværende randomiserede studie ikke forskel i graviditetschancen efter
oplægning af et eller to friske æg i en uselekteret patientgruppe. Observationelle studier fra Finland
og Belgien tyder på uændrede graviditetsrater efter introduktion af enkeltægspolitik til udvalgte
patientgrupper (eSET). Data fra Sverige viser ligeledes, men på nationalt plan, at det har været
muligt at indføre enkeltægspolitikken til en meget stor andel af patienterne, uden at graviditetsraten
er blevet lavere. Sådanne data kunne dog også fortolkes, således at graviditetsraterne uden enkeltægspolitik ville være steget i samme periode.
Trods brug af avancerede prædiktionsmodeller synes det vanskeligt teoretisk at identificere en gruppe
med en forudseelig særlig stor graviditetschance og sandsynlighed for tvillinger med henblik på
elektiv enkeltægspolitik (eSET) uden konsekvenser for den samlede graviditetschance på den ene
side og tilstrækkelig virkning på den totale tvillingerate på den anden.
Risici for mor og barn
Flerfoldsgraviditet er forbundet med øget risiko for mor og barn. Graviditetskomplikationer forekommer hyppigere i tvillingegraviditeter end i graviditeter med et enkelt barn. Svangerskabsforgiftning (præeklampsi) forekommer med end dobbelt så hyppigt (13,4% vs. 6,2%) hos IVF/ICSI
tvillingegravide sammenlignet med enkeltbarnsgravide. Andre komplikationer så som forhøjet blodtryk, blodpropper, urinvejsinfektioner, blodmangel og blødning i graviditeten forekommer hyppige-
14
re og fører til hyppigere sygemelding og hospitalisering. Tvillinger har øget risiko for at blive født
for tidligt. I et dansk registerbaseret studie var hyppigheden af børn født før 37 fulde uger 43,9%
hos IVF tvillinger og 7,3% hos IVF enkeltfødte og risiko for fødsel før fulde 32 uger var øget 7
gange med en hyppighed på 8,5% hos IVF tvillinger overfor 1,3% hos IVF enkeltfødte. Disse risici
fører til øget brug af medicin for at forebygge for tidlig fødsel men også for at forhindre følgerne
af for tidlig fødsel hos barnet samt operative indgreb så som kejsersnit (52,9%). De europæiske tal
for mødredødelighed er 4,4 per 100.000 hos enkeltbarnsgravide og 10,2 per 100.000 for flerfoldsgravide.
Visse misdannelser forekommer hyppigere både hos naturligt opståede og IVF tvillinger sammenlignet med enkeltfødte. Nogle af disse misdannelser kan være sekundære til for tidlig fødsel.
Risikoen for at blive indlagt på en afdeling for for tidligt fødte børn er 3,8 gange øget hos IVF
tvillinger (56,4%) sammenlignet med IVF enkeltfødte (25,4%), og de tilbringer 9 dage flere i en
sådan afdeling end enkeltfødte. Den neonatale dødelighed (antal levendefødte børn der dør indenfor
de første 28 levedøgn per 1000 levende fødte) er den samme eller lidt lavere hos IVF end hos
spontant undfangede tvillinger. Danske resultater har vist en signifikant lavere risiko for død i første
leveår. Den perinatale dødelighed (i dette arbejde defineret som antal dødfødte efter udgangen af
28 fulde svangerskabsuger og dødsfald i første leveuge per 1000 fødte) var dobbelt så høj hos IVF
tvillinger som enkeltfødte (20,7 overfor 11,0 per 1000).
I en stort dansk undersøgelse fandt man, at neurologiske følger og spastisk lammelse forekom med
samme hyppighed hos IVF tvillinger, spontant undfangne tvillinger og IVF enkeltfødte. Den øgede
hyppighed af spastisk lammelse hos IVF enkeltfødte sammenlignet med enkeltfødte fra normalbefolkningen kunne skyldes, at tvillingegraviditeter er hyppigere end tvillingefødsler, da det ene foster
i en tvillingegraviditet ikke sjældent (10,4%) går til grunde. Den overlevende i en sådan oprindelig
tvillingegraviditet har en signifikant øget risiko for lav fødselsvægt, for tidlig fødsel og død sammenlignet med andre enkeltfødte børn – men også sammenlignet med børn af IVF graviditeter, hvori
der fra starten kun har kunnet påvises et enkelt foster.
Patienten
Det kvalitative studie af ufrivilligt barnløse par
Ufrivilligt barnløse par var velinformerede om behandlingsprocessen, og de kendte til risikoen for
for tidlig fødsel ved tvillingegraviditet, men ikke til detaljeret statistik. De havde en bred tilgang til
tvillingeproblematikken, som både inkluderede livskvalitet og fremtidig risiko. Undersøgelsen viste,
at informationer om risiko, som stiller spørgsmålstegn ved behandlingen, er patienterne tilbøjelig
til at ignorere og anser dem for uvelkomne, mens information, der bekræfter, hvad parret ønsker,
accepteres. En af de vigtige faktorer bag beslutningsprocessen er det udbredte ønske om mere end
ét barn. Der er meget positive værdier knyttet til det at have søskende. Den fysiske og psykiske
belastning i forbindelse med behandlingerne, de sociale stressfaktorer og den kendsgerning, at der
kun bliver tilbudt hjælp til ét barn i offentligt regi, påvirker ønsket om at få tvillinger snarere end
to børn ved hinanden efterfølgende graviditeter. De fleste par finder den samlede belastning af disse
faktorer betydelig og en tilstrækkelig grund til at ønske tvillinger. Nærværende resultater peger på,
at en obligatorisk enkeltægspolitik vil være i klar modstrid med patienternes interesse og ønske.
Spørgeskemaundersøgelsen til ufrivilligt barnløse par
Undersøgelsen viste, at de fleste barnløse par foretrækker tvillinger (58%) frem for et barn ad gangen
(38%), men det store flertal (79%) havde planlagt at få to æg lagt op i næste behandling. Hos
flertallet af parrene forklares præference for to æg således ikke blot af et ønske om at have størst
mulig graviditetschance og derved undgå flere behandlinger. Det skyldes også et overvejet ønske
om at få tvillinger hos flertallet af parrene. Begrundelserne for at ønske tvillinger er ønsket om
søskende, gensidig glæde mellem søskende eller en mindre specifik positiv holdning til tvillinger. Et
enkelt barn blev ønsket af 42% begrundet med risiko for en vanskelig graviditet med tvillinger og
risiko for mor og barn.
15
Den givne information blev vurderet som meget tilfredsstillende, tilfredsstillende, eller nogenlunde
tilfredsstillende af næsten alle par, men kun lidt mindre end halvdelen havde modtaget mundtlig
rådgivning. Kun 12% angav, at deres beslutning vedrørende antal æg skyldtes rådgivning på fertilitetsklinikken, og en tredjedel angav, at det var deres eget valg.
Det generelle indtryk er, at der er behov for mere specifik og organiseret information. Der er i
nærværende studie ikke nogen sammenhæng mellem ønske om tvillinger og det at have modtaget
information eller at være godt informeret.
Ønsket om tvillinger beskrevet i tre forskellige scenarier med stigende risiko for mor og barn viste,
at ønsket om tvillinger blev mindre med stigende risiko beskrevet. En uvilje overfor risiko blev
afsløret ved at formulere det statistiske resultat som et chance-scenarium til halvdelen af deltagerne
og som et risiko-scenarium til den anden halvdel. Gruppen, som blev præsenteret for risiko-scenariet, var mere forbeholdne overfor et ønske om tvillinger end gruppen, der blev præsenteret for
chance-scenariet. Infertile par synes således påvirket af den information, de bliver givet om tvillingespørgsmålet.
Behandlingsrelateret fysisk og psykisk belastning synes at udgøre en vigtig faktor for at ønske tvillinger, mens økonomiske motiver for tvillingepræferencer synes at betyde mindre. Kun 27% af respondenterne ville ønske en enkeltægspolitik, selvom de havde fire gratis behandlinger i offentlig regi
eller endog et ubegrænset antal til rådighed, eller hvis behandling til barn nummer to var gratis.
Økonomien
Den umiddelbare fordel ved enkeltægspolitikken er en lavere hyppighed af dyre tvillingegraviditeter
sammenlignet med toægspolitikken, hvor tvillingehyppigheden er så høj som omkring 25%. Men
den sundhedsøkonomiske analyse, der sammenligner omkostningerne ved enkeltægspolitikken med
toægspolitikken, viser dog ikke, at enkeltægspolitikken er mere omkostningseffektiv sammenlignet
med toægspolitikken. Således opgøres omkostningen per klinisk graviditet til 131.446 kr. ved enkeltægspolitikken, mens den kun er 115.321 kr. per klinisk graviditet ved toægspolitikken. De tilsvarende resultater i forhold til fødsler viser ligeledes højere omkostninger per fødsel ved enkeltægspolitikken (149.833 kr.) sammenlignet med toægspolitikken (120.324 kr.). Ved medtagelse af frosne
æg bliver omkostningseffektiviteten ved de to alternativer dog tilnærmelsesvis den samme.
Forskellen mellem enkeltægspolitikken og toægspolitikken øges, når omkostninger per barn født
opgøres, idet omkostninger per barn født ved enkeltægspolitikken er 148.204 kr., mens den kun er
93.265 kr. ved toægspolitikken.
Årsagen til de lavere cost-effectiveness ratioer ved toægspolitikken er, at den er mere effektiv sammenlignet med enkeltægspolitikken, idet toægspolitikken resulterer i højere klinisk graviditetsrate,
højere fødselsrate og flere børn – sidstnævnte på grund de ekstra børn, der fødes som tvillinger ved
toægspolitikken. Samtidig er toægspolitikken dog også dyrere for den gennemsnitlige patient behandlet på grund af højere udgifter til især fødsel og neonatal indlæggelse. Ekstra omkostninger
per patient (inkrementale omkostnings-effektivitetsratio) ved i dag at benytte den mere effektive
toægspolitik er således omkring 82.000 kr. per ekstra fødsel opnået og omkring 50.000 kr. per
ekstra barn, der fødes med toægspolitikken. Dog synes ekstra omkostningerne per fødsel eller per
barn født med den mere effektive toægspolitik ikke at være ekstraordinært høje. Konklusionen af
den sundhedsøkonomiske analyse er, at omkostningen målt både per fødsel og per barn født er
lavere ved toægspolitikken sammenlignet med enkeltægspolitikken. Imidlertid er de samlede udgifter
til toægspolitikken højere, som følge af højere udgifter til graviditet og fødsel, neonatal indlæggelse
og arbejdsfravær (produktionstab) begrundet i tvillingefødslerne. Toægspolitikken belaster derfor
sundhedsbudgettet og samfundet mere end enkeltægspolitikken ville gøre.
16
Organisatoriske perspektiver
Hvis den gennemsnitlige kumulative graviditetsrate for IVF patienter skal opretholdes ved skift fra
en toægspolitik til en enkeltægspolitik vil der være behov for flere IVF behandlinger, dvs. flere
hormonstimulationer og ægudtagninger. Flere strå skal fryses ved enkeltægspolitikken sammenlignet
med toægspolitikken, flere skal tøs, og i sidste ende skal der foretages flere ægoplægninger. Det er
svært præcist at beregne, hvad de specifikke konsekvenser er for logistik og budget, men der er gjort
et forsøg herpå.
Det blev skønnet, at indførelse af en enkeltægspolitik i Danmark vil have følgende konsekvenser,
hvis en uændret kumulativ graviditetschance skal opretholdes ved et øget antal behandlinger for at
kompensere for en lavere graviditetsrate ved enkeltægspolitikken. I skønnet er begrænsning af behandling til det første barn i offentlig regi opretholdt:
a.
b.
c.
d.
e.
500-1000 flere hormonstimulationer, ægudtagninger og friske ægoplægninger per år.
Nedfrysning af 900-1800 ekstra strå, hver med ét befrugtet æg.
900-1800 flere ægoplægninger per år.
18 flere oplægninger af frosne æg per år til barn nummer to.
756 flere hormonstimulationer, ægudtagninger og friske ægoplægninger per år til barn nummer
to på privatklinik ved indførelse af obligatorisk enkeltægspolitik.
Et skøn over, hvad indførelse af enkeltægspolitikken vil medføre i form af ekstra udgifter for den
offentlige sundhedssektors budget, er af størrelsesordenen 14,9-25,9 mio. kr., afhængig af under
hvilke betingelser enkeltægspolitikken indføres. Omvendt vil besparelsen for sundhedsbudgettet på
grund af reducerede udgifter til obstetrisk og neonatal omsorg repræsentere et beløb af størrelsesordenen 12-24 mio. kr. På grundlag af disse antagelser vil indførelse af enkeltægspolitikken repræsentere en ændring i offentlige udgifter og budgetter varierende mellem en besparelse på 9,1 mio. kr. (i
bedste fald) og en øget udgift på 13,9 mio. kr. (i værste fald).
En udvidelse af det offentliges tilbud til behandling til barn nummer to kunne repræsentere en
kompensation for tabet af muligheden for at få to børn på én gang. En sådan ændring kunne
befordre et valg af enkeltægspolitikken, men udgifterne hertil er ikke medtaget i denne beregning.
Som det også følger af beregningerne af de udgiftsmæssige konsekvenser vil en indførelse af en
enkeltægspolitik i Danmark nødvendiggøre ekstra personale og uddannelse heraf, hvilket dog kan
klares inden for et halvt til et helt år.
Afhængigt af om en enkeltægspolitik indføres som en obligatorisk ordning for alle barnløse, der
IVF behandles i offentligt regi, eller det indføres som en frivillig, ikke-obligatorisk ordning vil det
stille yderligere krav til information til de barnløse. Således kan respekten for det barnløse pars
autonomi opretholdes ved i form af en frivillig, ikke-obligatorisk ordning at overlade valget til
parret, idet det dog samtidig vil nødvendiggøre øget information og rådgivning omkring risici
forbundet med tvillingegraviditet og -fødsel sammenholdt chancen for graviditet. En information
og rådgivning der må være veldefineret i forhold til dets omfang og karakter. En øget frekvens af
disse valgsituationer for parret vil utvivlsomt medføre behov for ændring af kultur og prioritering
for personalet på fertilitetsklinikkerne.
Konklusion
MTV-analysen tillader følgende svar på de stillede undersøgelsesspørgsmål i forhold til konsekvenser
af enkeltægspolitikken for teknologi, patient, organisation og økonomi sammenlignet med den
eksisterende toægspolitik:
17
1. I hvilken grad vil den kliniske graviditetsrate blive reduceret ved uselekteret oplægning af ét befrugtet æg i
livmoderen i stedet for to i forbindelse med IVF behandling?
De hidtil offentliggjorte lodtrækningsstudier har vist, at oplægning af ét befrugtet æg til udvalgte
grupper af kvinder reducerer graviditetschancen per oplægning af friske æg. Nærværende lodtrækningsundersøgelse viste imidlertid ikke nogen forskel i graviditetsraten mellem en étægsgruppe
og en toægsgruppe. Årsagen hertil var muligvis tilfældig variation eller en ikke-påviselig forskel på
ægkvalitet i de to grupper. Observationelle data fra andre studier har vist, at det er muligt at
fastholde uændrede graviditetsrater ved indførelse af enkeltægspolitik til udvalgte grupper af kvinder
(eSET).
2. Hvilke beslutningsmæssige vilkår har patienterne for at tage stilling til spørgsmålet om tvillingegraviditeter, og
i hvilket omfang vil en påbudt enkeltægspolitik harmonere/være i strid med patienternes interesser?
Tidligere studier og vores egen undersøgelse viser et udbredt ønske om tvillinger blandt par, der er
i behandling for ufrivillig barnløshed. Parrene tog stilling til spørgsmålet på grundlag af en vurdering
af sociale, psykiske og fysiske ulemper ved IVF-behandling i en situation, hvor der er et udbredt
ønske om at få mere end ét biologisk barn. En påbudt enkeltægspolitik i den nuværende situation
ville derfor være klart i strid med patienternes interesser og ønsker.
3. Hvad vil de organisatoriske konsekvenser (f.eks. konsekvenser for IVF-klinikker) være ved ændring fra en toægspolitik til en enkeltægspolitik?
Indførelse af en enkeltægspolitik i Danmark vurderes at kræve ekstra personale og uddannelse heraf
og skønnes at medføre en ændring i de offentlige udgifter til sundhedsvæsenet af en størrelsesorden,
der kan variere mellem en besparelse på 9,1 mio. kr. (bedste fald) og en øget udgift på 13,9 mio.
kr. (værste fald). Hvis en enkeltægspolitik indføres som en frivillig, ikke-obligatorisk ordning (eSET)
vil det derudover stille yderligere krav til personalet om information og rådgivning af de barnløse
par omkring risici forbundet med tvillingegraviditet og -fødsel sammenholdt med chancen for
graviditet ved oplægning af hhv. ét eller to æg op.
4. Hvad vil de sundhedsøkonomiske konsekvenser for samfundet være af indføre en enkeltægspolitik sammenlignet med en toægspolitik, vurderet ud fra forventet lavere graviditetschance, modsat et forventet reduceret
ressourceforbrug i forbindelse med fødslen, neonatal indlæggelse, m.m.?
I den foreliggende undersøgelse viste enkeltægspolitikken sig ikke at være mere omkostningseffektiv
sammenlignet med toægspolitikken, der er mere effektiv pga. højere fødselsrate og flere børn. Toægspolitikken er imidlertid også dyrere (højere udgifter til fødsel og neonatal indlæggelse). Ekstra
omkostningerne per fødsel eller per barn født med den mere effektive toægspolitik synes dog ikke
at være ekstraordinært høje.
Spørgsmålet »Skal der lægges ét eller to æg op ved IVF?« er således ikke let at besvare. Respekt for
patientens selvbestemmelsesret må vægtes i forhold til økonomiske aspekter, herunder effektiviteten
af disse forholdsvis fysisk og psykologisk belastende behandlinger. Komplikationer og langtidsfølger
forbundet med for tidlig fødsel afledt af tvillingegraviditet som følge af oplægning af to æg må også
inddrages. Det ultimative svar på dette spørgsmål må gives af beslutningstagerne. Nærværende MTV
rapport kan forhåbentlig facilitere denne beslutningsproces.
18
1 Introduction
In Denmark potent fertility treatment was introduced in the early 1980’ties with the in vitro
fertilization technique (IVF) and with intrauterine insemination with husband or donor semen,
both most frequently combined with ovarian stimulation.
Mathematically, the probability of a pregnancy following fertility treatment is correlated to the
number of embryos transferred or follicles induced, but biologically this correlation is not absolute
since embryos of different quality do not have equal implantation potential (Templeton and Morris
1998). Another mathematical fact is that transfer of more than one embryo increases the risk of
multiple pregnancy which is associated with increased risks of adverse outcome of pregnancy, delivery and long term consequences for the children. Thus, the ethical dilemma for the infertility
practitioner is to create a balance between the desire to maximize the opportunity for the infertile
couple to have a child and the need to minimize the risk of harm to the future child and the family
(Shenfield 2003).
Public funding for infertility treatment influences the decisions of both the couple and the doctor
in this dilemma. In the USA, there is no public funding for IVF and insurance coverage is limited.
Guidelines are provided by the American Society of Reproductive Medicine, however no national
legislation concerning assisted reproductive technology (ART) exists. A significant proportion of all
transferrals include more than three embryos and even more than five embryo transfers are practised.
The rate of twin and high order multiple births was 53% in 2002 (Wright et al. 2004).
National figures from Europe are rather heterogeneous. The latest published data from 2000 show
that the proportion of single embryo transfers varied between 8% (Hungary) and 27% (Finland),
double embryo transfers from 16.7% (Greece) to 84% (Sweden), triple embryo transfers from 4%
(Sweden) to 51% (Ireland) and transfer of four or more from 0% (Denmark, Iceland, Sweden, UK)
to 43% in Ukraine (Andersen et al. 2004).
Historically, transfer of four embryos was practised in Demark during the first years, but at an early
stage the number of embryos was reduced to three and subsequently triple embryo transfers were
allowed in poor prognosis patients only. In Denmark the 2004 figures for IVF/ICSI show that of
7795 transfers 27.4%, 67.5% and 5.2% were single, double and triple embryo transfers, respectively.
A total of 2056 clinical pregnancies treatments were established with 72.7% singletons, 26.4% twins,
and 0.9% triplets. In terms of expected delivered children, these figures correspond to a total of
3176 children with 1084 singletons, 1084 twins and 36 triplets (http://www.fertilitetsselskab.dk/).
In Europe during the last couple of years focus has been on the high incidence of twin pregnancies
(25%) associated with transfer of two embryos, resulting in a proportion of twin children from
IVF treatments of 40%.
Finland has been pioneering a development towards single embryo transfer (SET), resulting in a
decrease in the multiple birth rate following IVF from 26% in 1995 with a SET rate of 16% to
14% with a SET rate of 39% with apparently unchanged pregnancy rates (Tiitinen and Gissler
2004).
Nevertheless the recent properly designed Nordic randomised study showed that even in selected
patients, SET yield lower pregnancy rates than double embryo transfer (DET), which may be
compensated, however, by the subsequent transfer of a frozen embryo derived from the same treatment cycle (Thurin et al. 2004).
Physicians have a political and ethical responsibility to inform policy-makers about the most appropriate infertility treatments. Without equal or fair access to public funding for infertility treatment,
19
the pressure is increased on the family and the physician to maximize the chances of a live birth
with the potential to ignore the consequences of multiple pregnancies. Public funding for a specified
number of cycles that would give the patients a chance of a healthy singleton birth should minimize
these pressures. Additionally policy makers should be aware of the consequences of multiple pregnancies and the potential cost-effectiveness of infertility treatment aiming at a singleton pregnancy
(Shenfield 2003). In Denmark three IVF cycles with embryo transfers and derived frozen embryo
transfers (FER) are offered free of charge to infertile couples, who have no previous children in
their relationship if the woman is below 40 years of age.1 Accordingly, in principle only one child
is provided. However, a second child free of charge is provided in case of a twin pregnancy, and
transfer of frozen embryos derived from the IVF cycle resulting in the first child are transferred free
of charge for a second child, but the majority of couples do not have surplus frozen embryos.
The purpose of the present health technology assessment (HTA) report was to elucidate the consequences of obligatory single embryo transfer versus optional two embryo transfer in Denmark
according to the Danish definition of a HTA (Statens Institut for Medicinsk Teknologivurdering
2000) with specific focus on pregnancy rates, the women’s/couples preferences/knowledge and on
economical and organisational aspects. The following HTA questions were formulated:
2. On what basis do infertile couples decide about the twin question, and how would an obligatory
single embryo transfer policy be in keeping with the interests of the infertile couples?
3. What organisational consequences are expected in case of introduction of an obligatory single
embryo transfer policy?
4. What are the expected health economic consequences for the society following an obligatory
single embryo transfer as judged from an expected lower pregnancy rate and a reduced consumption of resources with respect to delivery, neonatal service etc.?
1
Except in the counties of Storstrøm, Roskilde and Vestsjælland were treatment to a second child is allowed.
20
2 Multiple birth rates in Europe
during the last two decades
In most European countries the twin birth rate remained nearly constant lingering around 1:83
until the mid 1980s and then gradually increased with a rapid rise from 1990 to 1996 (Imaizumi
et al. 1998). From the first IVF child was born in 1978 the secular changes in twinning rates
highlight the substantial effect the introduction of ART, performed in a relatively small group of
women has caused on the overall national twin birth rates.
This pattern was most pronounced in Denmark, where the total increase in the twin birth rate
during the past two decades reached 2.4 fold (Official statistics of Denmark). This is partly explained by the liberal access to infertility treatment termed assisted reproductive technology (ART),
which includes both IVF and other types of infertility treatment such as hormonal stimulation with
insemination with donor or husband’s semen (IUI) and other types of hormonal stimulation for
anovulation. Denmark has the highest number of IVF cycles performed per inhabitant in Europe
(Andersen et al. 2004).
Similar changes in twinning rates have been observed in the other Scandinavian countries i.e. a
1.9-fold increase in Sweden between 1973 and 2000 and a 2.2-fold increase in Norway from 1974
to 2002 (Official statistics of Sweden, Official statistics of Norway).
However, IVF is not the sole contributor to the increasing twin birth rates. Increasing child bearing
age and other types of infertility treatments plays a significant role also. A Swedish register study
estimated that one third of the rise was explained by increasing childbearing age, one third by ART
procedures other than IVF and one third by IVF procedures (Bergh et al. 1999). This was further
documented in a Danish study, in which the adjusted population based twinning rate from 198994 increased 2.7-fold and was almost exclusively observed in women aged Ø30 years and was
limited to dizygotic (DZ) twinning (Westergaard et al. 1997). Accordingly, implementation of
obligatory single embryo transfer can be expected to contribute to less than one third reduction of
the increase in twinning rates in the Nordic countries. Intervention against multifollicular hormonal
stimulation associated with IUI and treatment of anovulation as well as stimulation to family
building earlier in life is necessary to obtain a more comprehensive reduction in twinning rates.
Double embryo transfer is now practised in most European countries. The triplet rate has become
rather low for IVF patients but twin rates have remained fairly constant, lingering around 25%
while the overall pregnancy rates have been maintained (ESHRE Campus Course Report 2001,
Andersen et al. 2004). Overall, the proportion of IVF singletons and twins in Europe 2000 was
56.7% and 38.7%, respectively (Andersen et al. 2004).
Finland was the first country voluntarily to implement elective single embryo transfer and the first
to show a decline in the national twin birth rate from 17.1 in 1998 to 14.9 per 1000 births in
2004 (Tiitinen et al. 2003, Official statistics of Finland). In Sweden a revision of the guidelines
from ‘‘The National Board of Health and Welfare’’ (see http://www.sos.se) on ART was passed in
December 2002, declaring that in general only one embryo should be transferred in IVF. However,
in cases with a scientifically and empirically low risk of twin pregnancy two embryos may be
transferred, but only following information about the risks associated with twin pregnancy and the
couple should be offered counselling with a paediatrician. This has resulted in 67% SET among all
fresh single embryo transfers in Sweden in 2004, while maintaining an unchanged delivery rate of
27% per transfer and a decrease in the IVF multiple birth rate to 5.6% (Bergh et al., Läkartidningen
2005, in press).
21
3 Risks associated with multiple pregnancy and birth
3.1 Pregnancy
Multiple pregnancy puts the mother’s health at risk. This is due to an increased risk of hypertensive
disorders, thrombo-embolism, urinary tract infections, anaemia, and vaginal-uterine haemorrhage
(placental abruption, placenta praevia) (Senat et al. 1998). There is also an increased risk of extended periods of bed rest, hospitalization, medication to prevent preterm labour and corticosteroids, surgical procedures such as Caesarian section, and preterm labour. European data indicate
a maternal mortality of 4.4 per 100.000 in singleton pregnancies and 10.2 per 100.000 for multiple
pregnancies (Sebire et al. 1998).
In spontaneously conceived pregnancies the incidence of preeclampsia is higher in twin than in
singleton conceptions (Coonrod et al. 1995, Santema et al. 1995, Campbell et al. 1999). This was
confirmed also in IVF pregnancies, where the risk of preeclampsia was 2.4-fold increased for IVF/
ICSI twin (13.4%) vs. singleton pregnancies (6.2%) (Pinborg et al. 2004a). Further, two other
studies have shown higher morbidity in IVF twin than singleton pregnancies in terms of pregnancy
induced hypertension and intrahepatic cholestasis and higher maternal hospitalisation rates (Klemetti et al. 2002, Koivurova et al. 2002b).
In a Danish study the risk of sick leave and hospitalisation in pregnancy was 6.8 and 3.5-fold
higher in IVF/ICSI twin than singleton pregnancies (Pinborg et al. 2004a). Accordingly, IVF twin
pregnancies seem to carry a higher maternal morbidity than IVF singleton pregnancies not solely
caused by more careful precautions being taken in IVF pregnancies.
3.2
Delivery
Previous studies have shown that the risk of preterm delivery and low birth weight in IVF
pregnancies is higher than in the general population (Bergh et al. 1999, Dhont et al. 1999,
Schieve et al. 2002, Westergaard et al. 1999). The poorer outcome in IVF pregnancies is mainly
explained by the higher multiple birth rates with twin births as far the predominant contributor,
albeit higher order multiple births account for considerably worse outcome. Recently national
cohort studies and two meta-analyses have shown that IVF singletons carry a higher risk of preterm delivery and low birth weight than spontaneously conceived singletons (Bergh et al. 1999,
Helmerhorst et al. 2004, Jackson et al. 2004, Schieve et al. 2002). This higher risk has not been
recovered in IVF vs. spontaneously conceived twins, where obstetric outcome is similar (Helmerhorst et al. 2004, Dhont et al. 1999, Pinborg et al. 2004b). As monochorionic (MC) twins are
associated with higher morbidity this may be explained by a lower rate of monochorionic twins
following IVF (1-2%) compared to spontaneous conceptions (20%) (Derom et al. 2001, Sebire
et al. 1997, Loos et al. 1998).
There is one study published which compares IVF and spontaneous twins of the same zygosity
(Lambalk et al. 2001). It shows that when only dizygotic twins are compared a similar risk is seen
in ART twins compared to spontaneous twins.
As a rough estimate IVF twins are born with an average gestational age three weeks earlier than
IVF singletons and with a mean birth weight about 1000 g lower (Pinborg et al. 2004c). In the
Danish register study the age- and parity adjusted odds ratio of birth prior to 37 completed weeks
(preterm delivery) was 10-fold increased (OR 9.9, 95%CI 8.7-11.3). The crude percentages of
preterm deliveries were 43.9% in IVF twins and 7.3% in IVF singletons and odds ratio of birth
prior to 32 completed weeks was increased 7-fold (OR 7.4, 95%CI 5.6-9.8) crude frequencies
being 8.5% in IVF twins vs. 1.3% in IVF singletons. Similar results were obtained for low and
very low birth weight (Pinborg et al. 2004c).
22
Several studies have shown that caesarean section (CS) rates are considerably higher in IVF twin
than singleton pregnancies with 2-3-fold increased relative risks from about 50% to 20% in singletons with considerable variations between the countries (Dhont et al. 1999, Westergaard et al.
1999, Klemetti et al. 2002, Koivurova et al. 2002b, Pinborg et al. 2004c). Even compared with
twins after spontaneous conception crude CS rates are higher in IVF twins 52.9% vs. 42.7% in
the Danish National twin cohort study (Pinborg et al. 2004b). However, after age- and parity
adjustment, this increased risk disappeared OR 1.1 (1.0-1.2). This is in line with the findings in
the Australian review, where the relative risk of CS in IVF vs. control twins was 1.2 (1.1-1.3)
(Helmerhorst et al. 2004).
3.3
Malformations
A recent Australian review reported that the pooled odds ratio of major birth defects in IVF vs.
spontaneously conceived children was 2.0 (1.5-2.7) and for all defects (majorπminor) 1.4 (1.3-1.5).
The consequence of these increased risks is dependent on the baseline prevalence of birth defects
in the background population, which is 1% for major malformations detected at birth in Denmark
(unpublished data from the National Board of Health). With a baseline prevalence of 1%, number
needed to harm (NNTH) is 100 children. Hence to deliver one surplus child with a major malformation 100 IVF children are to be born. In the seven reviewer-selected studies, OR for singletons
only was 1.4 (1.2-1.5) (Hansen et al. 2005). The lower risk in the analysis restricted to IVF
singletons indicates that twinning may have some influence on the overall increased risk of malformations in IVF infants. In accordance, a Swedish register study on 736 ICSI singletons and 400
twins found that the stratified risk of malformations at birth in ICSI vs. spontaneously conceived
children decreased from an OR of 1.8 (1.2-2.6) to 1.2 (0.8-1.8) after adjustment for twins (Wennerholm et al. 2000). The excess risk in ICSI children could be explained to a large extent by conditions
associated with preterm delivery, i.e. patent ductus arteriosis (PDA) and undescended testes. This
is in agreement with the findings of the Danish national twin birth cohort where an increased total
malformation rate in IVF/ICSI twins vs. singletons could be eliminated by excluding PDA, which
is strongly associated with preterm delivery (Pinborg et al. 2004c).
Taken together some types of malformations are known to appear at a higher rate in spontaneously
conceived twins than in singletons such as neural tube defects, hydrocephaly, PDA and alimentary
tract defects (Källen 1986, Doyle et al. 1991). Some of these excess risks have also been confirmed
in IVF twins, but the risk increase is modest and requires very large studies to be detected. Furthermore, the risk of specific malformations may at least partly be secondary to preterm delivery.
3.4
Neonatal complications
Not surprisingly, neonatal morbidity in terms of neonatal intensive care unit (NICU) admissions is
considerably higher in IVF/ICSI twins than in singletons. The Danish national twin cohort study
revealed that IVF/ICSI twins had a 3.8-fold increased risk of admission to the NICU compared
with IVF singletons (56.4% vs. 25.4%). This risk decreased to 1.8 after adjustment for preterm
delivery (Pinborg et al. 2004c). Furthermore, IVF twins spent on average 9 days more in the NICU
than singletons and the frequency of neonatal admission of more than 7 days duration was 75%
vs. 45% in IVF singletons, the corresponding frequency of neonatal admission of a duration 28
days or more was 28% and 10%, respectively (P∞0.001) (Pinborg et al. 2004c).
Based on the literature neonatal mortality is similar to or perhaps slightly lower in IVF than in
control twins (Bergh et al. 1999, Dhont et al. 1999, Helmerhorst et al. 2004). Our national cohort
study revealed no significant difference in perinatal mortality between IVF and control twins.
However, the rate of live-born, who died within the first year of life, was significantly lower among
IVF twins (10/1000) compared with control twins (15/1000) with the vast majority dying in the
23
neonatal period (Pinborg et al. 2004b). If neonatal mortality in IVF twins is reduced, it is probably
related to the lower frequency of MC twinning. As expected perinatal mortality in IVF twins was
twice as high as in IVF singletons; 20.7 vs. 11.0 per 1000 (Pinborg et al. 2004c).
3.5
Long-term consequences
Only two population-based controlled studies on neurological sequelae in IVF twins have been
published; a Swedish and a Danish both stratified by gender and year of birth and enrolled 2060
and 3393 IVF twins, respectively (Strömberg et al. 2002, Pinborg et al. 2004d). Similar adjusted
risks of cerebral palsy in IVF vs. control twins were provided in both studies, while the only risk
factors of cerebral palsy were male sex and prematurity or low birth weight, whereas maternal age
had no influence (Strömberg et al. 2002, Pinborg et al. 2004d).
The Danish study yielded similar prevalence rates of neurological sequelae and cerebral palsy in
IVF twins, control twins and IVF singletons [8.8; 9.6; 8.2] and [3.2; 4.0; 2.5] per 1000 children,
respectively (Pinborg et al. 2004d). However, in a Danish register study on singletons born between
1995-2001, the rate ratio of cerebral palsy in IVF vs. non-IVF singletons was 1.8 (1.2-2.8) (0.33%
vs. 0.19%) (Lidegaard et al. 2005). Thus, it seems that the same prevalence rate of cerebral palsy
observed in IVF twins and singletons is attributable to a higher rate of cerebral palsy in IVF than
spontaneously conceived singletons. This is in agreement with results from the Swedish study, where
IVF singletons carried an increased risk of 2.8 (1.3-5.8) as compared with singletons from the
general population (Strömberg et al. 2002). In the Swedish study, however, the crude rates of
cerebral impairment seemed higher in IVF twins vs. IVF singletons, thus 2.2% versus 1.4% for
neurological sequelae and 0.73% versus 0.37% for CP, but no statistical comparisons were conducted. A Danish national postal survey on four-year old Danish IVF children showed that special
needs (ergo or physiotherapy, speech therapy or a special remedial teacher) were present in significantly more IVF twins than singletons (9.9% vs. 6.1%) and speech therapy was provided to 6.4%
vs. 3.2% (Pinborg et al. 2003a). After adjustment for low birth weight, a similar risk of having
special needs were seen in IVF twins and singletons, whereas IVF twins were still more likely to
receive speech therapy than singletons (OR 2.0, 95%CI 1.1-5.0) (Pinborg et al. 2003a). In line
with this, maternal rating of their offspring’s speech development was significantly poorer in IVF
twins than singletons even after adjustment for low birth weight. Also Strömberg et al. (2002)
found increased requirement of treatment in childhood disability centres in children born after
IVF/ICSI compared to controls, even among singletons (OR 1.4 (1.0-2.1)). Still, it should be noted
that the rates of the major disabilities mentioned are counted per mille.
The most important clue concerning long-term outcome in IVF twins is that the 3-4 fold increased
risk of cerebral palsy in spontaneously conceived twins vs. singletons is not found in IVF twins vs.
IVF singletons (Pinborg et al. 2004d, Scher et al. 2002). However, this may be due to a higher risk
of cerebral palsy in IVF singletons compared with spontaneously conceived singletons (Lidegaard
et al. 2005, Strömberg et al. 2002).
3.6 Vanishing twins
The disappearance of one of two gestational sacs or embryos after documented foetal heart activity
is known as the vanishing twin phenomenon. The vanishing twin phenomenon occurs not only in
relation to foeti papyracei, but twin material can also be reabsorbed without leaving any trace
(Landy and Keith 1998). As the literature on the poorer outcome in IVF singleton became evident,
the question about vanishing twins in IVF pregnancies arose. In spontaneously conceived twin
pregnancies, late intrauterine death of one twin has considerable influence on the risk of morbidity
and mortality in the surviving co-twin (Pharoah et al. 2000, Scher et al. 2002). A Danish multicentre cohort study on 8542 clinical pregnancies detected by early ultrasonography, reported that
24
10.4% of live-born IVF singletons was a twin gestation in early pregnancy (Pinborg et al. 2004e).
These survivors of a vanished co-twin carried a 2.1 and 2.3-fold increased risk of very low birth
weight (∞1500 g) and very preterm delivery (∞32 weeks), respectively and a three fold increased
mortality rate. Birth weight and gestational age was correlated with the time of onset of spontaneous
reduction, the later onset the worse outcome. In accordance two recent papers revealed that birth
occurred significantly earlier in singleton pregnancies with two gestational sacs than in those with
one (Dickey et al. 2002, Lancaster et al. 2004). Moreover, a large US register study found that the
risk of low birth weight was higher the higher the number of foetal hearts on early ultrasonography
for both singletons and twins (Schieve et al. 2002).
25
4 Factors influencing the twin birth rate
4.1 Natural conception
In natural conception it is important to distinguish between factors of importance to dizygotic
(DZ) or monozygotic (MZ) twinning rates. For DZ twinning rates both race, nutrition and inheritance plays a role with almost twice the risk in women, whose mother or sister had DZ twins (Tong
et al. 1998). Furthermore, maternal age has a considerable effect on the DZ twinning rate, which
increases more than 4-fold from 15 to 37 years, followed by an abrupt decline (Tong et al. 1998).
Independently of maternal age, the twinning rate also increases with increasing parity (Tong et al.
1998). In Denmark, the age of an average primipara was 24.1 years in 1973 and 28.6 years in
2003. The average age at delivery has increased by 0,1 year per year during the last decades (Danish
Birth Register 2003, http://www.sst.dk/publ/tidsskrifter/nyetal/pdf/2004/23_04.pdf ).
The incidence of MZ twinning is largely independent of the above-mentioned factors. However, it
has been shown that a mother who is a MZ twin, has an increased risk of having MZ twins,
indicating some kind of genetic predisposition to MZ twinning (Tong et al. 1998).
4.2 Assisted conception
It has been estimated that one third of the recent rise in the Swedish twin birth rate was explained
by increasing childbearing age, one third by ART procedures other than IVF and one third by IVF
procedures (Bergh et al. 1999).
Overall, the most recent European IVF/ICSI data for 2000 showed a distribution of deliveries with
25.336 (73.6%) singleton, 8.396 (24.4%) twin, 674 (2.0%) triplet, and 13 (0.04%) quadruplet
(Andersen et al. 2004). For comparison, 560 multiple pregnancies were established in Denmark in
2004 by IVF/ICSI out of a total of 2.056 clinical pregnancies (542 (26.4%) twins and 18 (0.9%)
triplets (http://www.fertilitetsselskab.dk/)).
ART procedures other than IVF include intrauterine insemination by husband’s semen or by donor
(IUI) and ovarian stimulation of anovulatory women. There is no international monitoring of
pregnancies following IUI and only recently a voluntary national reporting of IUI treatments in
Denmark has been established. The results of 2004 comprised 9118 IUI cycles resulting in 1.357
clinical pregnancies of which 15.5% were multiple (181 (13.3%) twins, 28 (2.1%) triplets and 1
(0.09%) quadruplets) (http://www.fertilitetsselskab.dk/). There are no national or international data
available on multiple pregnancy rates following other kinds of ovarian stimulation. It is well known,
however, that following induction of ovulation with clomiphene citrate the multiple pregnancy rate
is in the order of 8%, and with gonadotrophins of 10 to 40% (Wolff 2000).
A British study concluded that older age, the presence of tubal infertility, four or more previous
IVF attempts, and long duration of infertility all significantly reduced the odds of a birth and the
odds of multiple births following IVF. Furthermore, a previous live birth increased a woman’s odds
of a birth but not of multiple births (Templeton and Morris 1998). They also showed that the
higher the number of fertilized eggs the higher the likelihood of a live birth and that the number
of fertilized eggs were more important in determining the chance of a live birth than the number
of embryos actually transferred into the uterus. If more than four eggs had been fertilized and were
available for transfer, the woman’s chance of a birth was not diminished by transferring only two
embryos. Transfer of more than two embryos, however, increased the risk of multiple births.
By a multivariate analysis Strandell and co-workers found that the number of good quality embryos
transferred, female age, tuba indication and the number of previous IVF cycles was the only independent risk factors of multiple births (Strandell et al. 2000). They calculated that multiple birth
26
rates could be reduced from 26% to 13% in all IVF births if single embryo was performed in
selected patients with high risk of multiple births (50% of all cycles). Concomitantly, the total birth
rate would decrease from 29% to 25%. The authors claimed that this could be completely restored
by one additional transfer with a single frozen embryo (Strandell et al. 2000).
Though the contribution of MZ twinning to ART twin birth rates is negligible (1-2%), it has been
shown that ovarian stimulation more than doubles the MZ twin birth rate (Tong et al. 1998).
Gonadotrophin treatment, rather than micromanipulation including ICSI, zona drilling and assisted
hatching, has been claimed as a major contributing factor for this (Schachter et al. 2001). In-vitro
conditions also plays a role, since the incidence of MZ twinning is 3-fold increased in blastocyst
transfer compared to cleavage-stage transfer (Milki et al. 2003).
4.3 Elective single embryo transfer (eSET)
Two reviews of pregnancy rates following single embryo transfers in cycles using fresh embryos and
in cycles using frozen-thawed embryos have been published recently (Gerris 2005, Bergh 2005).
Randomized controlled trials
The results of the six randomized controlled trials (RCT) in elective single embryo transfer (eSET)
versus double embryo transfer (DET) are listed in Table 1. Five of the RCTs have relatively small
sample sizes making firm conclusions difficult (Gerris et al. 1999, Martikainen et al. 2001, Lukassen
et al. 2005, Gardner et al. 2004, Van Montfoort et al. 2004).
TABLE 1
Randomized controlled trials comparing pregnancy rates in elective single embryo transfer (eSET) with double embryo transfer (DET)
Author
Year
Publication
Country
No. of patients
eSET
DET
Gerris*,.
1999
Belgium
26
27
Ongoing pregnancy/live birth rate, %
eSET
DET
p-value/
RR (95%CI)
38.5%
74.0%
1.8 (1.1-2.9)
Twin birth rates, %
eSET
DET
Inclusion criteria
10%
30%
NS
4.5%
39.0%
36%
NS
0
60.9%c
76.0%
NS
0/14
37%
(26%)
47.4%
331
39.7%a
43.8%
0.3
0.8%
36.1%
100
(34%)a
36%
NS
2%
33%
∞34 yrs, 1. IVF/ICSI cycle
Ø2 high quality embryo
∞36 yrs, 1.IVF/ICSI cycle
±4 high quality embryos
∞35 years, 1.IVF/ICSI
Ø2 high quality embryo, FSH∞10IU/L
FSHÆ10IU/L
Ø10 follicles±12 mm at day of HCG
∞36 yrs, 1. or 2. IVF/ICSI cycle
∞38 years, 1. IVF/ICSI cycle
Martikainen*,.
2001
Finland
144
70
32.4%
47.1%
Lukassen*
2005
Netherlands
54
53
41%b
Gardner.
2004
US
23
25
Thurin.
2004
Sweden
330
Van Montfoort
2005
Netherlands
100
a
OPR in eSET group with one fresh eSET and if no live birth then a frozen eSET, in brackets OPR with only one fresh embryo transfer in the eSET
group.
b
OPR in eSET group with one fresh eSET and if no live birth then a fresh eSET, in brackets OPR with only one fresh embryo transfer in the eSET
group.
c
Only blastocyst transfers.
* Studies included in the Cochrane review by Pandian et al. 2004. Only preliminary results of the Lukassen study were included in the Pandian review
(Lukassen et al. 2002).
.
Studies included in the review by Gerris et al. 2004 and Bergh 2005.
Taken together, the mean pregnancy rate after SET of fresh embryos in four truly prospective
randomized trials was 31.3% with 2.0% twins and 48.1% after DET with 34.9% twins (Gerris et
al. 1999, Bergh 2005, Martikainen et al. 2001, Gardner et al. 2004, Thurin et al. 2004).
Only three randomized trials full-filled the selection criteria for inclusion in a recent Cochrane review.
These studies were all limited by their small sample size (Pandian et al. 2005). Furthermore, none
included subsequent single frozen embryo transfers (Gerris et al. 1999, Lukassen et al. 2002, Martikainen et al. 2001). The review concluded that live births and pregnancy rates following single embryo
transfer are lower than following double embryo transfer as are the chances of multiple pregnancy
27
including twins. However, the authors stated that it was unlikely that the conclusions were robust
enough to catalyse a change in clinical practice due to insufficient design and size of the studies. The
review did not include the recent large RCT from Scandinavia (Thurin et al. 2004).
The aim of the largest RCT on eSET vs. DET was to show equivalence concerning live birth
between the two strategies; one fresh single embryo plus one frozen-thawed SET versus one fresh
DET (Thurin et al. 2004). In this multi-centre study 661 women ∞36 years, undergoing their first
or second IVF/ICSI cycle with at least two embryos of high quality, were randomized to either
DET (nΩ331) or eSET (nΩ330). Restricted to the fresh eSET cycle the pregnancy rate was significantly lower 33.6% in the eSET vs. 52.61% in the DET group (P∞0.001). This is in accordance
with the recent Cochrane review yielding a clinical pregnancy rate in two vs. one embryo transfer
of OR 2.1 (95%CI 1.2; 3.5), live birth rate OR 1.9 (95%CI 1.1; 3.2) and multiple pregnancy rate
10.0 (95%CI 2.6; 38.2) (Pandian et al. 2004). However, including pregnancies following one
transfer of a single frozen-thawed embryo in the eSET group, similar pregnancy rates were obtained
in the two groups with an ongoing pregnancy rate of 39.7% in the eSET and 43.8% and in the
DET group (PΩ0.31), respectively. There was only one twin birth in the eSET group and a twin
rate of 33.1% in the DET group. The protocol of the study did not consider possible additional
pregnancies derived from FER cycles in the DET group.
To summarize the results of the RCTs, eSET shows satisfactory pregnancy rates in good prognosis
patients, however lower than after DET. Further twin rates are significantly lower after eSET and
the Thurin study as the main contributor emphasized the importance of a well-functioning freezing
programme (Thurin et al. 2004).
Observational studies
In the first report from Finland pregnancy rates in 74 eSET and 742 DET were similar 29.7% vs.
29.4%, if at least two good embryos were available for transfer (Vilska et al. 1999). Another
retrospective study from a single entity in Finland from 1997-2001 on 1871 IVF/ICSI cycles
showed that the number of eSET increased from 11 to 56%, while the multiple delivery rate
dropped markedly from 25 to 5% maintaining a relatively stable overall pregnancy rate, mean
34.0% (Tiitinen et al. 2003). The same experience was reported from a single entity in Belgium,
where the proportion of eSET increased from 1.5% to 17.5% from 1997 to 2002, while the overall
pregnancy rate was fairly constant on 34% and the twinning rate dropped to 14% (De Sutter et
al. 2003). Finally two reports from the same Belgian group showed the same encouraging results
with pregnancy rates of 30-40% after eSET (Gerris et al. 2002, 2004).
To summarise the observational studies indicate that similar pregnancy rates are obtained after eSET
and DET. However, as the good prognosis patients receive eSET while poor prognosis women
receive DET, the two groups are not comparable. Nevertheless, the Swedish data from 2004 showed
unchanged national delivery rates of 27% and a dramatic decrease in multiple birth rates to 5.6%
with a SET rate of 67% (Bergh et al. Läkartidningen 2005, in press).
The overall conclusion seems to be that introducing eSET to good prognosis patients diminishes
the twin birth rate considerably. Observational studies and national data from Sweden indicate that
ongoing pregnancy rates of 30-40% per transfer can be maintained despite a high rate of eSET. A
better selection of embryos for transfer or other concomitant changes may explain such data. Since
the transfer of two good embryos always yields more pregnancies than the transfer of one good
embryo unchanged pregnancy rates in observational studies may also indicate lack of an otherwise
expected increase in pregnancy rates (Bergh 2005). There is no doubt that a well-functioning
freezing programme is of paramount importance, since more good embryos are frozen following
eSET and more focus should be placed upon the cumulative pregnancy rate following fresh eSET
and the derived transfer(s) of single frozen-thawed embryos from each stimulation cycle and subsequent oocyte pick-up rather than estimating success rates upon transfer of multiple fresh embryos
(Tiitinen et al. 2001, Gerris et al. 2003, Tiitinen et al. 2004, Thurin et al. 2004).
28
4.4 Results of randomization between one and two embryos in an
unselected population of patients in their first cycle at the Fertility Clinic,
Aarhus University Hospital, Skejby Sygehus
4.4.1 Introduction
The consequences of an obligatory SET policy in terms of impact on pregnancy rates are not
known.
As discussed previously, the literature on SET describes results of randomized controlled trials of
double embryo transfers versus single embryo transfers in selected, good prognosis populations
(eSET). Observational studies are likewise limited to similar comparisons. In a single recent Dutch
study 300 less selected women (age ∞41 years, first cycle, two fertilised oocytes) were randomised
to SET or DET. The ongoing pregnancy rate was 21% following SET and 38% following DET
(van Montfoort et al. 2005). Accordingly we planned a RCT study comprising all infertile couples
commencing their first IVF treatment.
4.4.2 Materials and methods
Patients
All couples referred to their first IVF treatment during the period 1.4.2002-1.4.2004 were invited
to participate in the study. Reasons for exclusion were few: Previous IVF treatment, male factor
necessitating testicular retrieval of spermatozoa, and prior decision on single embryo transfer due
to patient wish or medical indication (previous preterm delivery, preeclampsia, placental abruption,
cervical incompetence).
Candidates for participation were informed about the study in a leaflet mailed to them together
with the general written information about the treatment. Further, the study was presented at
advisory meetings with groups of new patients and during visits with nurses or doctors at the
fertility clinic. Finally, patients were asked to participate at the beginning of their treatment. Participation implied that on the day of embryo transfer, patients were randomized (block randomization,
6 patients in each block) by the sealed envelope method to either one or two embryo transfer. The
study was not blinded to either patient or staff.
Patients allocated to transfer of one embryo were offered an extra compensatory treatment cycle in
case pregnancy was not achieved within the three reimbursed transfers allowed in Denmark.
All patients gave written informed consent after counselling, and the study was approved by the
Regional Ethics Committee. The primary outcomes were clinical pregnancy five weeks after embryo
transfer, and live birth rate as a result of one treatment cycle. The main outcome measures were
based on the transfer of fresh embryos only. Subsequent analyses included frozen embryos from the
randomized treatment cycle, but the result of the randomization procedure was not used when
transferring thawed embryos.
Treatment protocol
In all cycles down regulation was performed from cycle day 21 with BuserelinA nasal spray 0,1 mg
six times per 24 hours. Following a menstrual bleeding, stimulation was started two weeks later
with a starting dose of 100 or 150 IU PuregonA per day, individually adjusted. First follicular
ultrasound was done on the 10th stimulation day. When 3-4 follicles of Ø18 mm were present
HCG (ProfasiA or PregnylA, 10,000 IU) was given 37 hours before oocyte pick-up. For OPU,
patients were given i.v. Phentanyl 0.25 microgram, and a paracervical block with lidocaine. OPU
was performed ultrasound-guided transvaginally with a double-lumen needle (CookA, 16G) in
oocyte collection media (MedicultA) with flushing of each follicle as necessary. Luteal phase support
was given with progesterone vaginally (ProgestanA, 100 mg¿4) and continued in case of a positive
pregnancy test for an additional three weeks.
29
Measurement of serum HCG was performed two weeks following embryo transfer. A level of Ø20
IU was considered a positive test. Pregnant patients were scanned routinely 5 weeks after embryo
transfer. Clinical pregnancy was defined as a live intrauterine pregnancy at this scan. The implantation rate was defined as number of gestational sacs divided by the number of embryos transferred.
Laboratory procedures
Oocyte handling and ICSI procedure was performed as described previously (Ingerslev et al. 2001).
Embryo quality was scored as 1.0 if the embryo had blastomeres of equal size without fragmentation, 2.0 in case of unequal blastomeres without fragmentation, 2.1 if fragmentation was ∞10%,
2.2 with 10-19% fragmentation, 3.1 with 20-49% fragmentation and 3.2 with ±50% fragmentation.
Statistical analysis
The sample size necessary for the present study was calculated on the basis of an expected average
ongoing pregnancy rate of 25% per DET cycle started, of which 25% (6.25% in total) would be
twin pregnancies. For SET, we assumed that the 6.25% obtaining a twin pregnancy with DET
would also become pregnant with SET, and of the remaining 18.75% obtaining a pregnancy with
DET, half (9.375%) would become pregnant with SET. To detect a difference between a pregnancy
rate of 25% (as expected in the DET group) and a pregnancy rate of 15% or less (approximately
that expected in the SET group) it was estimated that 250 women would be needed in each group,
i.e. a total of 500 women (aΩ0.05, bΩ0.2). It was anticipated that the rate of participation would
be 75% and that this could be reached within a time period of two years.
All data were entered prospectively into a database. SPSS (Statistical Package for the Social Sciences),
version 11.0, and Stata version 8 were used for data analyses. Comparison of frequences was made
by c2-test, means (∫SD) by t-test if data followed a normal distribution, otherwise by MannWhitney U-test. Intention to treat principle was applied.
4.4.3 Results
Patients
Out of 777 patients, 168 (21.6%) were willing to participate in the study and fulfilled the inclusion
criteria for participation. They were randomized to transfer of one embryo (84) and two embryos
(84), respectively. The study was stopped after the planned two year period, since it was not realistic
to expect inclusion of the intended number of patients in the study (500) within a reasonable time
period.
The characteristics of participants and non-participants are shown in Table 2, which shows, that
the study groups seems to be a representative sample of the total group of 777 patients, except that
the participants were significantly older (32.1 vs. 31.2, pΩ0.007)) and more often the cause of
infertility among non-participants was classified as ‘‘other’’ (Table 2).
30
TABLE 2
Comparison between participants and non-participants. Aarhus, Denmark, 20022004
Mean female age, years (SD)
Mean body mass index
Median FSH, day 3-5
Median length of infertility (months)
Cause of infertility
Unexplained
Damaged tubes
Anovulation
Endometriosis
Male factor
Other
Missing
Female age (years)
19-24
25-29
30-34
35-39
40π
Smoking (number of cigarettes/day)
0
1-10
11π
Missing
Alcohol (standard drinks/day)
0
1-5
6π
Missing
Coffee (cups/day)
0
1-5
6-10
±10
Missing
Treatment type
IVF
ICSI
Median number of eggs retrieved
Score of best embryo*
1
2
3
Positive HCG
Yes
No
Clinical pregnancy achieved
Yes
No
Birth
Yes
No
Median (mean) gestational age at birth
Participants
N
168
32.1 (3.8)
163
24.0 (4.5)
128
5.6
162
29.0
Non-participants
N
609
31.2 (4.0)
565
23.7 (4.3)
352
5.6
557
26.0
p-value
42
34
6
18
68
0
0
25.0%
20.2%
3.6%
10.7%
40.5%
0.0%
132
107
34
64
228
43
1
21.7%
17.6%
5.6%
10.5%
37.5%
7.1%
0.013†
4
47
79
37
1
2.4%
28.0%
47.0%
22.0%
0.6%
29
225
238
112
5
4.8%
36.9%
39.1%
18.4%
0.8%
0.099
108
29
27
4
65.9%
17.7%
16.5%
396
81
87
45
70.2%
14.4%
15.4%
0.506
128
30
0
10
81.0%
19.0%
411
100
1
97
80.3%
19.5%
0.2%
0.846
58
88
17
0
5
35.6%
54.0%
10.4%
0.0%
240
274
39
5
51
43.0%
49.1%
7.0%
0.9%
0.129
113
55
168
67.3%
32.7%
5
404
205
609
66.3%
33.7%
5
0.822
91
39
10
65.0%
27.9%
7.1%
260
138
44
58.8%
31.2%
10.0%
0.373
55
113
32.7%
67.3%
196
413
32.2%
67.8%
0.892
47
121
28.0%
72.0%
154
455
25.3%
74.7%
0.481
40
128
37
23.8%
76.2%
39 (37.7)
136
473
135
22.3%
77.7%
39 (38.0)
0.675
0.007
0.431
0.703
0.232
0.711
0.728
* Best embryo score: 1Ωvan den Abbeel morphology score of 2.2 or better and 4 cells present; 2Ωvan
den Abbeel morphology score of 2.2 or better but not with 4 cells present or score worse than 2.2
but with 4 cells present; 3Ωvan den Abbeel morphology score worse that 2.2 and not 4 cells present.
†
If leaving out the group ‘other’, pΩ0.786.
31
Comparing participants randomized to one and two embryos, respectively no significant differences
were found in baseline characteristics, but more of the embryos available were of high quality in
the group randomized to one embryo (73.9% vs. 56.3%, pΩ0.06) (Table 3).
TABLE 3
Comparison between randomization groups. Aarhus, Denmark, 2002-2004
Mean female age, years (SD)
Mean body mass index
Median FSH, day 3-5
Median length of infertility (months)
Unexplained
Damaged tubes
Anovulation
Endometriosis
Male factor
Smoking (number of cigarettes/day)
0
1-10
11π
Missing
Alcohol (standard drinks/day)
0
1-5
Missing
Coffee (cups/day)
0
1-5
6π
Missing
Treatment type
IVF
ICSI
Median number of eggs retrieved
Score of best embryo*
1
2
3
One embryo
N
83
32.5 (3.7)
82
23.6 (4.3)
65
5.4
81
26.0
Two embryos
N
84
31.8 (3.8)
81
24.4 (4.7)
63
5.7
81
30.0
p-value
20
19
4
9
32
23.8%
22.6%
4.8%
10.7%
38.1%
22
15
2
9
36
26.2%
17.9%
2.4%
10.7%
42.9%
0.832
52
18
11
3
64.2%
22.2%
13.6%
56
11
16
1
67.5%
13.3%
19.3%
0.254
67
12
5
84.8%
15.2%
61
18
5
77.2%
22.8%
0.224
31
42
7
4
38.8%
52.5%
8.8%
27
46
10
1
32.5%
55.4%
12.0%
0.627
55
29
84
65.5%
34.5%
5
58
26
84
69.0%
31.0%
5
51
13
5
73.9%
18.8%
7.2%
40
26
5
56.3%
36.6%
7.0%
0.221
0.279
0.131
0.259
0.588
0.06
* Best embryo score: 1Ωvan den Abbeel morphology score (Staessen et al., 1990) of 2.2 or better and 4
cells present; 2Ωvan den Abbeel morphology score of 2.2 or better but not with 4 cells present or
score worse than 2.2 but with 4 cells present; 3Ωvan den Abbeel morphology score worse that 2.2 and
not 4 cells present.
Randomisation
Randomisation resulted in two groups with 84 patients in each. In the SET group, two patients
changed their mind and decided to have two embryos transferred. Deviation from intended randomisation was more frequent (p∞0.001) in the DET group, because only one transferable embryo
was available (17/84 (20.2%)). No suitable embryos were available for transfer in 18% (15/84) and
15% (13/84) in the SET and DET group, respectively.
Positive pregnancy test
Respectively 32% (27/84) in the SET group and 33% (28/84) in the DET group had a positive
S-hCG (pΩ0.869). Of those with a positive S-hCG, 2 in the SET group and 3 in the DET group
only achieved a biochemical pregnancy, while 1 and 2, respectively, miscarried before 8 completed
weeks of gestation.
Pregnancies and births
The subsequent proportion of ongoing clinical pregnancies was comparable in the two groups
(Table 4). The overall proportion of clinical pregnancies among participants in the study was
32
comparable to that of non-participants (Table 2). Restricting the analysis to those participants, who
had the intended number of embryos transferred – according to randomisation – did not change
this outcome (34% (23/67) vs. 37% (20/54), pΩ0.757). Similarly, restricting the analysis to those
cycles, where an embryo score of 1 (2.2 or better) was found, also yielded comparable proportions
of clinical pregnancies in the two randomisation groups (39% (20/51) vs. 38% (15/40), pΩ0.867).
Finally, restricting analyses to those, who had at least one embryo, transferred also yielded comparable results (data not shown).
A twin pregnancy occurred in 10 pregnancies in the DET group, but in none in the SET group.
Accordingly, the implantation rates were 34% (24/71) and 26% (33/125) in the SET and DET
group, respectively (pΩ0.273).
The proportion of women carrying their pregnancy to birth was also comparable in the two groups
(Table 4). Even so, median gestational age was shorter in the DET group (37 vs. 39 weeks, pΩ
0.022). Among the deliveries in the SET group, 74% were spontaneous, 5% induced, and 21% were
caesarean section, while in the DET group the percentages were 48%, 10%, and 43%, respectively.
TABLE 4
Pregnancies and births within randomization groups. Aarhus, Denmark, 2002-2004
N
Clinical pregnancy achieved
Yes
No
Delivery
Yes
No
One embryo
%
95%CI
N
Two embryos
%
95%CI
p-value
24
60
29%
71%
19-39
61-81
23
61
27%
73%
18-38
62-82
0.864
19
65
23%
77%
14-33
67-86
21
63
25%
75%
16-36
64-84
0.717
Subsequent cycle with frozen embryos
A total of 46 women subsequently underwent a cycle with thawed embryos from the randomized
cycle, 30 in the SET group, and 16 in the DET group. Seven of the women had achieved a clinical
pregnancy in the randomized cycle, and 3 of them had carried the pregnancy to a birth.
In the SET group, 13 had two embryos transferred, seven had one embryo transferred, and ten had
no suitable embryos for transfer. In the DET group, seven had two embryos transferred, one had
one embryo transferred, and eight had no suitable embryos for transfer.
Four women in the SET group, and three women in the DET group achieved a positive S-hCG,
and two in each group subsequently achieved a clinical singleton pregnancy. None of the women
had become pregnant in the randomized cycle.
The four pregnancies resulted in one induced delivery in the SET group at 39 weeks of gestation,
and one spontaneous delivery in the DET group at 38 weeks of gestation.
4.4.4 Discussion
The present study is the first to randomize between SET and DET in an unselected population of
infertile patients. Previously, both observational and randomized studies have analysed selected
patient groups with ‘‘good prognosis’’, i.e. those with better pregnancy rates than average. While
the observational studies indicated that similar pregnancy rates are obtained in good prognosis
patients after eSET and DET (Vilska et al. 1999, Tiitinen et al. 2003, De Sutter et al. 2003, Gerris
et al. 2002, 2004), the randomized studies all showed a lower mean pregnancy rate after SET of
fresh embryos (Gerris et al. 1999, Bergh 2005, Martikainen et al. 2001, Gardner et al. 2004,
Thurin et al. 2004).
33
On this background, it is even more surprising that the present study showed no difference in the
proportion of clinical pregnancies between patients who had one or two embryos transferred.
This is indeed difficult to explain since mathematically, the transfer of two embryos should represent
a better chance of pregnancy than only one. The implantation potential of embryos transferred in
the SET group was higher in the SET group than in the DET group, but insignificantly so. Rather
many of the planned DET cycles ended with single embryo transfer because of only one transferable
embryo available. Violating the intention to treat principle by analysis of those cycles only where
the intended number of embryos were transferred – according to randomization – did, however,
not change the conclusion. The rather high number of planned DET cycles with only one embryo
was due to a very soft stimulation regimen during the study period. A significantly higher fraction
of top quality embryos in the SET group could explain the better implantation potential in that
group compared to the DET group. However, restricting the analysis of the clinical pregnancy rate
in the two groups to cycles where only top quality embryos were transferred, still no differences
between the two groups were observed. Accordingly, it is hard to believe that this may have contributed to the better implantation rate in the SET group.
The low participation rate of only 21% might have compromised the interpretation of the present
results. However, participants seemed to be a representative sample of the infertile patients referred
to the clinic during the study period in terms of background characteristics. The fact that only 168
(instead of the planned 500) were recruited into the study resulted in limited power to detect any
significant differences between the two groups. Yet, in order to detect as statistically significant a
two percent difference between the proportions of clinical pregnancies or delivery rates, respectively,
approximately 10,000 participants would have been required in each group, i.e. far more than even
the originally intended sample size. In any case, the randomisation seemed to be successful, except
that more patients in the SET group had a top embryo transferred, and the very small differences
between the two groups with regard to clinical pregnancies and deliveries would not have been
significantly different even with the intended number of participants.
Nevertheless, random variation, or an unquantifiable difference in embryo quality are presently the
best theoretical explanations of the results.
The participation rate was lower than expected despite the fact that patients allocated to SET were
offered an extra, compensatory treatment cycle (with the possibility of DET) in case pregnancy was
not achieved within the three reimbursed transfers allowed in Denmark. However, retrospectively
the low participation is in accordance with the patients’ attitudes expressed in the survey in our
clinic (see later), showing that more than two thirds of the patients prefer DET and three reimbursed cycles to SET with one or more compensatory treatment cycles – or even reimbursement of
treatments for a second child.
4.5 Tools for selection of patients for eSET
4.5.1 Prediction models for pregnancy and twin risk following IVF/ICSI
So far the search for predictors among commonly registered variables in ART databases have not
resulted in significant progress in the identification of strong predictors of pregnancy.
Retrospective statistical analysis of fertility databases has resulted in several models which calculate
predictors of pregnancy and twin pregnancy following IVF/ICSI. Among large numbers of observations and many different variables, surprisingly few embryo and patient specific variables have
proved to be statistically significant predictors of pregnancy. In general, the probability of pregnancy
is statistically associated with the number of embryos available for transfer, embryo quality judged
as developmental stage and fragmentation, and female age (Wheeler et al. 1998, Strandell et al.
34
2000, Hunault et al. 2002, Thurin et al. 2005, Ottosen et al. in prep). Age and quality and number
of embryos available are interrelated parameters.
TABLE 5
Significant risk factors of pregnancy and twin pregnancy
Study
Wheeler et al. 1998
Strandell et al. 2000
Hunault et al. 2002
Thurin et al. 2005
Ottosen et al. (in prep)
Significant risk factors for pregnancy (twin and singleton) (ORΩodds ratio)
Female age and embryo score. Logistic regression coefficients significant at pΩ0.05
Female age, embryo quality, previous attempts (ORƒ1 at pΩ0.05)
Female age, embryo quality (1), number of retrieved oocytes, transfer day
(ORƒ1 at pΩ0.05)
Cycle number, embryo developmental stage, treatment type (IVF or ICSI),
ovarian sensitivity (ORƒ1 at pΩ0.05)
Female age, embryo quality (1), Basal FSH (ORƒ1 at pΩ0.05)
Data material, study type
Retrospective cohort study of 795 cycles
Retrospective cohort study of 642 first cycles
Strictly selected group. 661 cycles included in
eSET study, Thurin et al. 2004
1 Based on developmental stage, and morphology score.
The risk factors identified are relatively weak in all the models (although significant at a 0.05 level)
(Table 5). As most models have embryo quality and female age as statistically significant risk factors
it is not surprising that simple guidelines, based on the same variables, define a good prognosis
group with nearly the same sensitivity as the sophisticated statistical models published so far. At
Skejby Fertility Clinic, a prediction model was generated by logistic regression analysis of 1675
fresh cycles performed from 2000-2003, each cycle with 2 embryos transferred (Ottosen et al. in
prep.). The model can calculate the probability of pregnancy of an individual treatment cycle based
on female age, BMI, basal FSH, and embryo quality. Pregnancy was defined as foetal heart beat
detected by transvaginal ultrasound at gestation week 7. The ability of the model to identify a good
prognosis group was compared with typical simple inclusion criteria for eSET. This comparison
revealed that no important risk factors other than age and embryo quality were present in the
database information, so a prediction model will in all probability only perform slightly better than
typical inclusion criteria when applied on future cycles. Elective SET criteria (age Æ36 and at least
two top quality embryos based on developmental stage and embryo morphology) identified a good
prognosis group of 403 from a total of 2193 fresh two-embryo transfer cycles. The pregnancy rate
in this group was 47.9%. If the logistic regression model was used to identify 403 cycles with the
best prognosis, out of the total 2193 cycles, then the pregnancy rate within this group was only
slightly higher, 50.5%. The prior probability of pregnancy among all the 2193 two-embryo transfer
cycles was 34.6%. Thus, even though the size of the good prognosis cycle group only comprised
about 20% of the total number of two-embryo transfer cycles, just about half of the cycles defined
as good prognosis cycles resulted in a pregnancy, whether the group was defined by the prediction
model or eSET criteria.
Accordingly, sophisticated mathematical models using all known significant risk factors will not
perform significantly better than simple criteria used in previous eSET studies.
4.5.2 Other tools to identify treatment cycles for eSET, simple inclusion criteria
So far elective SET versus DET has primarily been studied in good prognosis treatment cycles.
(Gerris et al. 1999, 2002, Tiitinen et al. 2003, Thurin et al. 2004, Martikainen et al. 2004, Gardner
et al. 2004, Lukassen et al. 2005, Montfoort et al. 2005). Simple inclusion criteria were used.
However Tiitinen et al. (2003) partially used exclusion criteria. The good prognosis treatment cycle
was defined differently in the published observational and RCT studies, and the rationale behind
inclusion criteria is rarely discussed in detail. Neither is the proportion of included cycles relative
to the total number of cycles consistently reported.
35
TABLE 6
Criteria for allocation to eSET in different studies
Study
Gerris et al. 1999
Thurin et al. 2004
Martikainen et al. 2004
Gardner et al. 2004
Age
criteria
Æ33
Æ35
Æ36
None
Lukassen et al. 2005
Æ35
Montfoort et al. 2005
Æ37
Embryo criteria
Other criteria
Two top quality embryos
At least two good quality embryos
At least one top quality
None
First treatment cycle
First or second treatment cycle
First or second treatment cycle
Day 3 FSHÆ10 mIU/ml
E2∞80 pg/ml, Norm. endometrial cavity
±10 follicles±12 mm
No previous failed treatment cycles,
Basal FSH∞10 IU/l
First three cycles
At least one excellent embryo and
one good embryo, judged on fragmentation
Two transferable, at least one good quality
% included of total treatment
cycles meeting criteria
∑20% included
34,4% included
24% included
Not reported
Not reported
Not reported
The number of cycles eligible for eSET only comprised a minor fraction of the total number of
treatment cycles (Table 6). An estimation of the fraction of eligible cycles, based on an analysis of
all fresh treatment cycles performed from 2000-2003 at Skejby Fertility Clinic showed that using
ageÆ36 and two top quality embryos (∞10% fragmentation and 4 blastomeres on day two) for
selection, only 11% (403/3685) met the typical eSET criteria. Among cycles with two transferable
embryos, that is cycles where a choice between SET and DET could be made, (nΩ2193, embryo
morphology score 3.1 or better and at least 2 blastomeres on day 2) 18.4% (403/2193) of the
cycles were eligible for eSET (Ottosen et al. in prep). However, due to soft ovarian stimulation
protocols, the number of oocytes (6-7 per oocyte pick-up) was somewhat lower than in most clinics
the fraction of patients with two transferable embryos may seem low. Nevertheless, commonly used
selection criteria for eSET candidates exclude a major proportion (in this case 81.6%) of fresh
cycles with two transferable embryos, where multiple pregnancies still is a potential risk. The recent
Cochrane study by Pandian et al. (2004) similarly indicated that despite rigorous criteria resulting
in selection of one third or less of the total number of fresh cycles, eSET still results in reduced
pregnancy rates, and leaves a large group where multiple pregnancies still can occur at a significant
rate.
A comparison between eSET in a good prognosis group and an obligatory SET policy on cumulative pregnancy rates including both fresh and all transfers of frozen and thawed embryos has not
been reported.
However, using age Æ36 years and two high quality embryos as selection criteria a retrospective
analysis of the Skejby Database allows an estimate of the overall consequence of eSET in a good
prognosis group (Ottosen, unpublished data). In the RCT study by Thurin et al. (2004) the individual embryo implantation rate in the DET and SET cycles were nearly identical (28.5% vs. 27.6%).
The implantation rates were identical in the two groups. Hence, it may be reasonable to assume a
hypothetical 50% loss of gestations in the good prognosis group allocated to eSET. This would
equal a total loss of 55 out of 759 pregnancies (Table 7). This may be an over estimation and is
based on the assumption that the two top quality embryos had the same implantation potential, as
seen in the study by Thurin et al. (2004), and that implantation of two embryos are mutually
independent events in case of DET (Trimarchi 2001). If we further assume that in case of eSET,
the twin pregnancies in the good prognosis group would have been singletons, the twin rate would
be reduced by 34.7% from 239 to 156 of the total number of pregnancies, among all the treatment
cycles with two transferable embryos. However, the number of pregnancies would have decreased
7.2%, with 55 from 759 to 704, and the total number of children would have decreased from 998
to 860 (if all went to term). Considering the fact that most families want more than one child, the
eSET policy would have left an additional demand for up to 138 babies. These numbers corresponds well with the findings by Strandell et al. (2000) who, based on a prediction model, calculated
that among 2107 cycles, a 56% reduction of the twin rate (from 158 to 70) could be achieved at
the expense of a 14.7% decrease in pregnancy rate (births) (from 611 to 520). The trade off has
to be considered between eSET in a rigorously defined small good prognosis group resulting in a
limited loss of pregnancies and a limited reduction of multiple pregnancies versus eSET in a large
36
broadly defined good prognosis group resulting in a considerable loss of pregnancies followed by a
substantial reduction of multiple pregnancies.
TABLE 7
True results from 2193 fresh two-embryo transfer cycles from the
Skejby database, and hypothetical results if eSET had been applied (see text)
Cycles
No eSET (true results)
eSET (hypothetical outcome)
2193
2193
Total
pregnancies
759 (34.6%)
704 (32.1%)
Twin
pregnancies
239 (31.5%)
156 (22.1%)
Recent studies show that in practice, the loss of pregnancies following application of eSET in some
cases is less, probably due to the fact, that the two embryos eligible for transfer may not have the
same implantation potential although both classified as good quality embryos (Tiitinen 2003).
Better predictors of pregnancy could improve existing prediction models and improve inclusion
criteria. Tools for identification of the best embryos are the likely key to this. Aneuploidy screening
may be such an instrument, but also more knowledge may be helpful about physiological characteristics such as amino acid profiles (Houghton et al. 2002) and oxygen consumption (Shiku et al.
2001, Houghton et al. 2004). Also more detailed patient characteristics could improve the ability
to predict pregnancy among ART patients and improve treatment success rate while reducing frequency of multiple births. Life style factors, which are known to increase time to pregnancy among
naturally conceiving couples (Hassan et al. 2004) may add information, but inclusion of life style
factors such as smoking and body mass index (BMI) added little – if anything – to estimations of
OR of pregnancy in the Skejby database (Krogh 2004 in prep.). Further, biochemical and genetic
markers of endometrial receptiveness and physiological markers such as utero-ovarian vascular impedance may be promising (Hoozemans 2004, Ozturk 2004).
In conclusion, it is difficult on the basis of prediction models to identify a group of patients with
a predictable high chance of pregnancy and risk of twins for selection to eSET without significant
consequences in terms of overall pregnancy rates on one side and limited effects on total twin rates
on the other. However, it is interesting that in Sweden in 2003 58.5% of all fresh transfers were
SET with an apparently unchanged clinical pregnancy rate of 35.3% with an 11.7% multiple rate
(Andersen AN, personal communication). Such data indicate that biology does not always follow
mathematical rules, but it cannot be excluded that in case of an unchanged DET policy the clinical
pregnancy rate may have increased.
37
5 What do the infertile couples prefer:
Single or double embryo transfer – a single child
or twins – and why?
5.1
Introduction
According to the Danish definition of health technology assessment (HTA), not only the technology
itself and its cost-effectiveness should be evaluated, but also patients’ attitudes towards the technology. Patient attitudes to a new or changed technology are one of the key issues in a health
technology assessment and should be considered by the decision-makers.
This chapter describes patient perspectives on the issue of transfer of one or two embryos in IVF
treatment. The first part of the chapter provides results of a qualitative study and the second part
results of a survey among infertile patients in treatment by IVF/ICSI at the Fertility Clinic at
Aarhus University Hospital, Skejby Sygehus. In addition to the independent contribution of the
qualitative study, the results contributed to the development of the design and content of the
questionnaire in the quantitative survey.
Several questions are relevant in connection to patients’ attitudes towards transfer of one or two
embryos in IVF/ICSI. The key issue is of course to elucidate preferences for one or two embryos
and to disclose the character of informed decision such preferences are based upon, i.e. how are
these couples informed, what influences their attitudes, what are the motifs for the decision etc.
Within the framework of an HTA it is important to establish knowledge about the degree of
convergence/divergence between patients’ preferences and medical and economical considerations.
5.2 Results of a qualitative study – Patient perspectives
The present qualitative study was undertaken with the dual objective of understanding:
H
H
the circumstances under which patients make decisions about the issue of twin pregnancy; and
the extent to which an obligatory single embryo policy would be in harmony/conflict with
patient interests.
These objectives were pursued under three headings: a) communication and information; b) risk
perception; and c) decision-making processes, each of which had a separate set of research questions.
A comprehensive unpublished report has been prepared in Danish on the findings of the qualitative
study (Seeberg 2003).
5.2.1 Methodology
Semi-structured interviewing was conducted with 18 couples undergoing IVF/ICSI treatment. In two
cases, only the woman could be interviewed. Interviews were primarily conducted in the couples’
homes, and they lasted approximately one hour each. Additional interviews were conducted with four
physicians and three nurses at the Fertility Clinic. All interviews were audio-recorded and transcribed.
Participant observation was conducted over five weeks during February to May 2002 at the Fertility
Clinic to obtain necessary background information, identify participants, and observe communication patterns.
Data was coded using the Nvivo 2.0 software package. An analytical framework based on narrative
theory was used to understand unfolding patterns of motivation and decision-making in relation
38
to a perceived future with a live-born child as a result of the IVF treatment (Mattingly 1998,
Seeberg 2005).
Due to the extensive interviewing process and interaction with couples both at the clinic and in
their homes, and the systematic approach to qualitative analysis, the study is perceived to have a
high internal validity (Bernard 1994). No reason was found to believe that the couples who took
part in the sub-study deviated in important ways from the patient group as a whole. External
validity would be limited to clinics working under similar conditions, i.e., public hospital clinics
with treatment offered to get the first child only.
5.2.2 Findings
5.2.2.1 Participants
Eighteen couples took part in the sub-study. Most were in employment, a few were students. In
terms of physical distance to the clinic, six couples were living in or around Aarhus, whereas twelve
couples were distributed in Jutland within a radius of 177 km. Age ranged from 26 to 40 years for
women and 26 to 45 years for men. There was considerable variation in the number of eggs
harvested and embryos available for transfer, actually transferred and frozen (Table 8).
TABLE 8
Age and IVF fertilization for participants (NΩ18)
Age, woman (yrs.)
Age, man (yrs.)
Oocytes harvested
Embryos available for transfer
Embryos transferred
Embryos for freezing
Min
26
26
2
0
0
0
Max
40
45
14
11
2
9
Mean
33.1
32.2
6.5
3.3
1.3
1.2
The reason for infertility was poor semen quality in nine cases. Five cases were related to the woman
(endometriosis, fibroids, tubal factor), and in four cases the reason was unknown.
At the time of interviewing, three couples were awaiting embryo transfer in the present treatment
cycle, ten couples awaited the outcome, and five had a negative outcome. Eleven couples were in
their first treatment cycle, three couples were in the second cycle, and four couples had had more
than two treatment cycles.
Out of the 18 couples, ten had been asked to participate in the clinical study, i.e. randomization
between transfer of one or two embryos. Five had agreed and five refused to participate in the
clinical study.
5.2.2.2 Treatment experiences
In order to understand couples’ views on the possibility of twin pregnancy it is necessary to
appreciate their investment of time to achieve a pregnancy. This process can be divided into three
phases: Phase I includes any number of years from the wish arose to have a child, over the
realization that there was an infertility problem, medical investigations and other techniques that
may have been unsuccessfully attempted, before being referred to IVF treatment. Once referred,
they went through a process of medical investigation and consultation at the Fertility Clinic,
attending an advisory meeting, and signing up for treatment at a given menstrual cycle. This last
part of Phase I may take from four to ten months. Phase II is the actual treatment phase, involving new routines of medication including injecting oneself (or having the partner inject), and
harvesting of oocytes and transfer of embryos. Phase III begins with a new period of waiting
before a pregnancy test can be performed to get the result of the treatment. If the test is positive,
the woman is referred to relevant antenatal care like any other pregnant woman. If it is negative,
the couple is returned after a minimum pause of one month to the pre-Phase I selection process
39
to get access to treatment at the onset of a later menstrual cycle, provided they have not reached
the limit of reimbursed treatments.
All couples were interviewed in the last part of Phase II or the early part of Phase III, and the
central topic of discussion – transfer of one or two embryos – was central to Phase II. The treatment
undergone in Phase II was characterized by the participants as one of intense hope, but also one of
stress related to actual or feared adverse effects, experiences of pain and psychological instability.
IVF treatment involves many visits to the clinic, and this could be difficult to align with a fulltime job, particularly if the couple felt a need to keep the IVF treatment secret.
The issues of importance to the couples changed over the three treatment cycles that they could
normally receive to achieve pregnancy. Table 9 shows the themes that ‘fill the most’ in the narratives
of the couples over the three treatment cycles.
TABLE 9
Key themes in treatment narratives by treatment cycle
Treatment cycle 1
Background story
Oocyte pick-up
Hormone stimulation
Attending the clinic
Vaginal ultrasound
Physical pain
Treatment cycle 2
Embryo freezing
Attending the clinic, clinic being busy
Waiting
Oocyte pick-up
Vaginal ultrasound
Not becoming pregnant
Treatment cycle 3 or more
Hormone stimulation
Embryo freezing
Psychological stress
Waiting
Not becoming pregnant
This pattern can be tentatively quantified on the basis of the amount of talking that couples invest
in the themes during interviews.2 Figure 1 shows the three themes that most markedly tended to
disappear from the stories of participants in later treatment cycles compared to the first treatment
cycle. The background story of how difficult it was to arrive at IVF treatment tended to fade, as
did the excitement about vaginal ultrasound and the fear linked to the oocyte pick-up.
2.
The graphs presented in Figure 1 and 2 are calculated on the basis of interview transcripts and express text coded for a specific theme relative
to total text of interviews with couples in a given treatment cycle. While it seems reasonable to assume that themes that participants wish to
discuss at length during interviewing are important to them, it is important to note that the opposite is not true: An understated point expressed
in one sentence can be potentially very significant.
40
FIGURE 1 Downward trends, selected themes over three treatment cycles
At the stage of the first treatment cycle, few couples questioned that they will become pregnant as
a result of the IVF technology. The background story is one of how they arrived at this treatment,
which they are positive will solve their problem, and they can afford to focus on the novelty of the
treatment. The stories at this stage are full of hope and some couples are of course fortunate to
obtain pregnancy in the first attempt. For those who continued to the second and perhaps the third
treatment cycle, the focus during interviews changed dramatically. Figure 2 shows how the theme
of freezing of embryos with its implied hope of a second chance gains importance as a positive
marker.
FIGURE 2 Upward trends, selected themes over three treatment cycles
41
Overall, however, the tone of the interviews changed in a negative direction, with hope fading as
the possibility of not becoming pregnant gained more prominence, followed by frustration at waiting, fatigue and psychological stress (Figure 2). While the interviews at the time of the first treatment cycle contained stories of hope and optimism, the second cycle stories were characterized by
suspense and fear, and the third cycle stories were marked by frustration and despair.
It should be noted that these experiences did not express dissatisfaction with the clinic and the
staff. Almost all couples expressed widespread satisfaction with the clinic, and the staff was generally
characterized as both very professional and very friendly. These positive circumstances were very
well appreciated but they could not take away the disappointment if the outcome of the treatment
was not positive.
5.2.2.3 The twin question
It is clear from the above discussion that IVF treatment is normally arrived at after prolonged
attempts to get a child, and that to follow the treatment in itself poses difficulties for the couples.
In addition, the initial hope that IVF treatment will in fact result in a child is gradually changed
into a more realistic perception that this might not be the outcome. It is in this context that the
general preference for transfer of two embryos and for twin pregnancy should be understood.
Most couples had a strong desire to have children, rather than a child. They imagined a future with
siblings playing together as the happy ending of the fertilization process they found themselves to
be in. This imagined future carried the motivation to undergo the difficulties related to IVF treatment. While the couples were generally very satisfied with the services they received, many saw IVF
treatment as being cumbersome, emotionally hard and practically difficult. Whenever the treatment
was considered difficult, this experience was evaluated against the promise of child-bearing and
parenting in the near future. If they could have two children within one treatment cycle, this was
the much preferred option compared to two cycles. In this context, they would also take age into
consideration as an important factor. Having waited for a long period of time for the first round
of treatment, the exact timing of the next treatment was uncertain, and whether at that age the
woman would still be able to become pregnant or in worst case eligible for treatment. Furthermore,
at the time of interviewing, the couples had only access to treatment to the first child. A second
child would have to be pursued through treatment at a private clinic. While many said this was
not a deciding factor, for some this would be a substantial out-of-pocket expense.
Accordingly, twin pregnancy and birth was generally seen by the couples as the ideal option and
they preferred transference of two embryos not only to increase chances of pregnancy per se, but
to increase chances of a twin pregnancy.
5.2.2.4 Information and communication
The above discussion outlined couples’ experiences leading to and during IVF treatment. Another
important factor influencing their perceptions and decision-making in regard to the question of
transferring one or two embryos was the knowledge base that most people primarily obtained in
the contact with the clinic.
The clinic used three complementary communication strategies:
H
An obligatory advisory meeting for all couples prior to onset of treatment
H
Written materials (handed out and available on the Internet)
H
Consultations with doctors and nurses
The advisory meeting
The advisory meeting is organized at Skejby Sygehus and held at regular intervals according to
patient flow. Information is provided by a doctor, a nurse, a lab technician and a secretary. The
latter three groups primarily give practical information, whereas the doctor also provides background
information on causes of infertility, criteria of inclusion for IVF treatment, hormonal treatment,
42
possible side effects, and chances for pregnancy (including twin pregnancy). The couples are informed that smoking and overweight reduce their chances of pregnancy. Complications related to
twin pregnancies are not emphasized, whereas triplet pregnancy is described in high-risk terminology.
The meeting serves as a summary of the written information, thereby ensuring that all couples have
received basic information in at least one medium prior to onset of treatment. In addition, the
couples, finding themselves in a big auditorium with one hundred other couples in the same
situation, learn that they are not alone.
Written information material
All couples receive a 43-pages booklet called ‘Ægtransplantation’ (IVF/ICSI) that describes the
treatment process in detail (Fertilitetsklinikken 2004). In addition, it covers topics such as hormonal
treatment and possible side effects, embryo freezing, ICSI, etc.
The booklet mentions an increased possibility/risk of twin pregnancy related to transfer of two
embryos and states that approx. 30% of all IVF pregnancies are twins. It says that twin pregnancies
not always pass normally and that there is a risk of preterm delivery. Couples are encouraged to
request transfer of only one embryo if they cannot accept the increased risk of twins. The rare
possibility of triplets is mentioned, as is the option of fetal reduction. Couples who participate in
special projects or procedures receive additional written information (ibid.).
Consultations with doctors and nurses
Prior to initiating hormonal treatment with SuprecurA nasal spray on the 21st day after onset of
menstruation, the couple is called for a comprehensive consultation with a nurse. They are shown
around in the clinic and are then informed in detail about the treatment process, the use of the
nasal spray and the injection pen and other treatment related issues and procedures. The nurse
routinely also discusses vaginal ultrasound, issues of smoking and diet, mood fluctuations, chances
for pregnancy and the risk that treatment could be unsuccessful. After this, there is adequate time
for the couple to ask questions. Frequently, the woman will air her fears related to the oocyte pickup. The nurse will attempt to play down the unpleasantness of the procedure, saying for example
that ‘‘it is much less painful than giving birth to a child’’ or telling a story about a husband who
felt much more discomfort than his wife and had to lie down.
This 21-day-consultation is the main opportunity for the couple to discuss their concerns face-toface with staff. In addition, time is set aside for brief consultations in connection with oocyte pickup, embryo transfer and vaginal ultrasound. Acute questions are generally managed through a
telephone consultation with a doctor or a nurse.
Four treatment-related risks are routinely informed to all couples. These include ovarian hyperstimulation, other adverse effects from hormonal treatment, risk of infection related to the oocyte
pick-up, and risk of multiple pregnancy.
Staff and patients agree that there is a division of labour concerning communication where nurses
are seen as following the couple closely in terms of overall well-being, emotional problems and
practical treatment issues, whereas the doctors are slightly less accessible and mainly seen as providing problem-based information on issues of specific medical knowledge.
The style of communication is generally positive and often jocular, thereby creating a relaxed atmosphere that downplays the couple’s nervousness concerning potentially unpleasant procedures or
unwarranted outcomes. Most staff sees the couples as potentially vulnerable to information that
raises serious doubt about a positive treatment outcome, and it is generally avoided to raise problems
before they arise.
43
Statistical information forms a part of the information, primarily in connection with overall chances
of pregnancy and the possibility/risk of twin pregnancy. The ultimate cumulative chance of pregnancy after application of all relevant infertility treatment modalities at the time of the study was
stated as 75% and for IVF treatment as 60% (three treatment cycles with embryo transfer). While
the figures tended to stick in the minds of patients, the difference of denominator between the two
figures was not always clear to the couples.
Patients’ assessment of information and communication
The study revealed widespread satisfaction among the users with the level of communication at the
clinic. Most felt they had been well informed and that they were part of a shared project together
with the staff. In addition to the relaxed style of communication prevalent in the clinic, this also
reflects that the couples were in fact directly involved in the treatment process, including injection
of medicine, which necessitated a good level of communication. The environment in the clinic was
seen by patients as overall supportive and safe.
As mentioned above, couples were generally in an emotionally vulnerable situation. Some couples
experienced that doctors expressed differing or contradictory opinions about their case. While such
experiences were uncommon and perhaps not completely avoidable, they colored the assessment of
the treatment for the concerned couples.
In general, information from other sources did not play a major role and was seen as secondary to
the information obtained at the clinic. One exception to this general pattern was sharing of personal
experiences with couples (friends and family) who had twins. Such sharing of experiences was seen in
important ways to support the general wish to have twins as an outcome of IVF treatment in the group.
The study did not identify any specific needs for structural changes of the existing communication
practice concerning the treatment process, including information concerning risks.
5.2.2.5 Risk and uncertainty
This section further explores how risks and uncertainties in connection with IVF treatment are
conceived by the couples. The issue of risk of twins is analyzed within a framework of general risk
perception and statistical risk information.
Statistical risk
On the basis of interpretation of the information given by the clinic, patients perceived the chance
of pregnancy in IVF treatment to be 60-70%. This meant that the risk of failure – i.e. that
pregnancy would not be achieved – was 30-40%. However, the couples felt that statistics did not
translate into useful information for their specific case, as it is not possible to be 60% pregnant.
For the couple to undergo the emotional stress and physical and practical discomfort involved in a
potentially prolonged IVF treatment process, they needed to believe that they belonged to the
successful group. Above, Figure 2 indicates that the idea of belonging to the possibly unsuccessful
group of couples who do not achieve pregnancy was established only after the first failed attempt.
This ‘‘zero-child risk’’ constituted the most important risk for the couples and they were willing to
accept a number of other risks if they believed that it would reduce this fundamental uncertainty.
One remarkable finding was that statistical information, in the interpretation of many couples, was
negatively emotionally charged. Information concerning risk about not becoming pregnant, or
about complications in twin pregnancy and delivery, was not seen as useful information. The
motivation to undergo the treatment built fundamentally on a hope that it would result in a happy
family future with children, and statistical risk information threatened this hope and could thereby
reduce motivation. Many couples expressed the need to ‘think positively’ rather than worry about
the statistics. Accordingly, an implicit consensus had developed among couples and clinical staff to
‘take things (i.e., complications of treatment) as they come’.
44
While one should be careful not to conclude that patients fundamentally do not wish to be informed about chances for successful treatment or significantly increased risks during pregnancy, it
is important to note that there is a delicate balance beyond which detailed risk information may
be harmful. It is also important to note that risk information means something else for the recipient
than for the sender. The recipient will need to evaluate and reconcile the information with the
personal project of undergoing IVF treatment. Therefore, risk information that basically questions
the project will tend to be ignored or seen as unwelcome, whereas information that confirms (or at
least not questions) what the couple desires easily can be accepted.
This is illustrated by the contrast between risk perception in twin and triplet pregnancy. Among
the 18 couples, 15 wanted transfer of two embryos, and 14 of these preferred to get twins. Transfer
of two embryos was generally not just a strategy to increase chances of a pregnancy at the risk of
getting twins; it was a deliberate strategy to increase chances of twin pregnancy. The couples did
not focus on possible complications during pregnancy and delivery. Rather, they were looking
forward to the time after delivery, where siblings would enjoy each other and where they would
have avoided another round of IVF treatment, at an older age with reduced chances of success,
with expenses to be paid out-of-pocket in the private health sector. Some couples were, however,
also concerned about potential practical and medical problems related to twins, and a few couples
preferred to get only one child. Table 10 provides an overview of the reasoning involved in this
pattern.
TABLE 10
Patients’ reasoning for and against twin pregnancy
For twin pregnancy
Twins enjoy each other when they grow up
H Positive perception of twins (‘‘double joy’’)
H Many twin pregnancies and deliveries are successful
H Would like to have more children but wants to reduce number of:
– Pregnancies
– IVF-treatments
H Would like to have more children but difficult to have one at the time due to
– Age
– Costs
H
Against twin pregnancy
Pregnancy can be difficult
H Delivery can be preterm
H Twins can be difficult to manage practically and financially after the delivery
H
Table 10 illustrates that merely deciding the issue of twin pregnancy on the basis of statistics
concerning risks of adverse health outcomes for mother or child would be insufficient from the
patients’ perspective. Couples had a broader view on the issue of twin pregnancy that also included
projected quality of life for the family-to-be and intended reduction of future risks.
It would be incorrect to see this near-elimination of risks related to twin pregnancy and delivery as
caused by insufficient information. On the contrary, risk information was received and subsequently
interpreted in such a way that it was contained and did not pose a risk for the overall wish and
motivation of the couples. In contrast, risk information on triplet pregnancy and delivery did not
pose a threat to the motivation to undergo IVF treatment and could therefore be easily accepted,
as shown in Table 11.
45
TABLE 11
Patients’ reasoning for and against triplet pregnancy
For triplet pregnancy
Positive perception of triplets (only one couple) (‘‘triplets, that would
also be nice’’)
H Possibility of having three children (only one couple) (‘‘Taking my age
into consideration, and what you subject your body to, I’d like to make
the most of it’’)
H
Against triplet pregnancy
Extensive strain on the family (interviewees used words like ‘terrible’, ‘too big a mouthful’, ‘twice as hard as twins’, ‘not fortunate’, ‘logistical problems’, ‘only two breasts’,
‘not normal’)
H Great risk of congenital malformations, abortion, still birth (‘too much’, ‘dangerous’,
‘handicapped’, ‘little chance that it goes well’, ‘small and frail’)
H Risk to the woman (‘complications’, ‘not with my health’, ‘much more physically demanding’)
H Foetal reduction (‘they’d have to remove one of them’)
H
While only one couple in fact would have liked to get triplets, all others agreed that the involved
statistical risks were too big for children, mother and subsequent family life. In short, there was no
contradiction between the statistical risk information and the couples’ IVF treatment project. Still,
the issue of fetal reduction was perceived to be extremely difficult to deal with in the context of a
hypothetical triplet pregnancy, and many were not able to take a decision unless circumstances
forced them to.
Experienced risk
In contrast to the statistical risk discussed above, couples identified other risks and uncertainties
that they had to deal with pragmatically during the course of the IVF treatment. These are important because they influence the evaluation of the IVF treatment per se and therefore also the attitudes
towards the choice between a possibility of one IVF treatment resulting in twin pregnancy versus
two successive IVF-treatments.
Information management played a major role for most couples, both at work and vis-à-vis friends
and family.
An example of work-related risk in connection with IVF treatment was provided by a woman who
worked in a company that was being merged with another company at the time of the IVF treatment. In the general climate of possible dismissals as a consequence of the merger she was very
nervous that her colleagues would find out about her treatment, and she had invented a complicated
web of excuses and lies to make it possible to take the nasal spray at the fixed timings during
workdays and to be able to travel to the hospital when required. While this case was extreme, it
was not unique in the sense that other women also very nervous about the attitude at the workplace
concerning the treatment and the need to take leave on particular days with little or no possibility
of flexibility. Others, in contrast, found support from colleagues and did not face problems in this
connection. However, in cases where the woman would feel that her job or future career opportunities may be jeopardized by simply undergoing IVF treatment, twins would clearly be a more
attractive possibility than successive IVF treatments if more than one child is desired.3
Information management vis-à-vis friends and family varied from total openness to complete secrecy. Couples who were open about the IVF treatment could also benefit from social support.
However, if the treatment was not successful they would have to share the disappointment, and for
some this added to the burden of the treatment as they felt they should not only be able to cope
with their own feelings but should also be able to comfort others – particularly their parents. In
contrast, one couple had decided to share the information about their treatment with none. When
the treatment turned out to be unsuccessful, they felt terribly isolated in their despair but found it
was impossible to change their strategy at that stage. Most couples, however, shared the information
with those closest to them and hid it from others, thereby dividing the world into those who knew
and those who did not. Some couples also felt uneasy – or felt that others became uneasy – when
they were together with friends who had children, and they decreased social contact with these
friends. Related to this issue were feelings of inadequacy, particularly in cases where infertility had
3
A Danish Supreme Court decision from 2002 voted that women cannot be dismissed from their workplace because of IVF treatment.
46
been diagnosed in one person. Information management posed a problem for some, but not for
all. In the first group, how, when and with whom to share information could be very real stress
factors related to IVF treatment.
Another threat formulated by some couples concerned the couple’s relationship. If the treatment
would be unsuccessful, they would in most cases be unlikely to get children who were biologically
related to both. This threat was directly related to the ‘zero-child risk’ but IVF treatment was
perceived to be the last chance. If it failed, the couple would need to consider options like adoption
or semen/egg donation – or, as one woman said, ‘‘Who says he would not find a new and younger
wife and get children with her?’’
The above-mentioned social risks and uncertainties influenced the overall experience of the couples
when they assessed how difficult (or easy) it was for them to undergo IVF treatment. Thereby, they
also influenced the weighing of the risks related to twin pregnancy compared to the risks of two
rounds of IVF treatment.
5.2.2.6 Decision-making processes
The involvement of the male partner in the treatment process was negotiable and varied considerably across couples. Some men gave injections and were present at all or almost all consultations at
the clinic, while other men were involved in less obvious ways, or perhaps less involved. For the
female partner, this could not be the case. Obviously, IVF treatment is first and foremost a treatment
of the woman, as it is she who subjects her body to oocyte pick-up, hormone stimulation and risks
of adverse effects. This basic fact influenced the decision-making processes in important ways.
Often, the man would feel that he necessarily was placed on the sideline and could not do much,
while the woman (in principle and to whatever extent) was suffering for them both. In general,
this led to dynamics, where the most important issue for the man was the concern to the woman,
and she would have a kind of veto on all important decisions. Her limits would determine the
treatment course. This could, however, place a certain pressure on the woman to be ‘brave’ and not
to show pain or concerns about possible side effects from hormonal treatment. This was a precarious
situation that could also be influenced by moodiness caused by the hormonal treatment. While the
IVF treatment in different ways affected the psychological dynamics in the couple’s relationship,
most couples seemed to have been well prepared for this and found the necessary support from
friends and families. These issues were frequently also touched, even if jokingly, at the clinic, where
staff made the couple understand that their reactions were common.
The clinical staff played a very important role for all treatment-related decisions, and there was an
outspoken degree of trust from the patients. In connection with transfer of one or two embryos,
this question was aired at an early stage of the counseling and information provided to the couple,
but the final decision was made at the time of embryo transfer. Usually the lab technician asked
about this in the presence of a physician and a nurse, once it was known that two good embryos
were available.
The fact that there is almost always convergence between the advice provided and the desires of
the couple is likely to strengthen the trust in the staff among the patients.
Couples were generally grateful that they were offered IVF treatment to address their infertility
problem. However, they were also concerned about the tendency they saw in mass media and met
among friends to define infertility problems as an individual (not a societal) problem. Some saw
infertility as an epidemic and sought its causes in environmental pollution, which, to them, justified
the availability of IVF treatment in the public health system. Some also felt that the stagnating
population size should cause public concern and that twin pregnancies therefore should not be
discouraged. In general, there was a strong consensus among the couples that the decision whether
to transfer one or two embryos, whenever possible, should be left solely to the couple.
47
5.2.3 Discussion and conclusion
The infertile couples were well-informed about the treatment process and they knew about the risk
of preterm delivery in twin pregnancies. Usually, they did not know detailed statistics about the
increased risk to the fetus or the mother in twin pregnancies compared to singleton pregnancies,
but they had a broad view on the issue of twin pregnancy that also included projected quality of
life for the family-to-be and intended reduction of future risks.
This near-elimination of risks related to twin pregnancy and delivery did not seem to be caused by
insufficient information. On the contrary, risk information on twins was received, but subsequently
interpreted in such a way that it was contained and did not pose a risk for the overall wish and
motivation of the couples. Further, risk information on triplet pregnancy and delivery did not pose
a threat to the motivation to undergo IVF treatment and could be easily accepted. Accordingly,
risk information that basically questions their infertility treatment will tend to be ignored or seen
as unwelcome, whereas information that confirms (or at least not questions) what the couple desires
easily can be accepted.
One of the important factors in decision-making was the wide-spread desire for more than one
child. Very positive values were attached to having a sibling. Two factors influenced positively the
wish to achieve children by twin pregnancy rather than by successive single pregnancies:
H
H
The IVF treatment process was experienced as stressful, and it was increasingly stressful the
longer it lasted. Couples reported stress factors related to a) physical discomfort and side effects
from the treatment; b) psychological discomfort caused by the treatment; and c) social stressors
related to partner, relatives, friends and workplace. Most couples found the total load of these
factors to be considerable and to constitute a sufficient reason to prefer a twin pregnancy.
Free IVF treatment in a public clinic was only offered to obtain the first child for the couple.
The present results strongly indicate that an obligatory single embryo policy would be in conflict
with patient interests and wishes.
From a patient perspective, it would be desirable to have both the option of transfer of two embryos
and the possibility of treatment for child number two, because this would establish a genuine choice
between 1) pursuing two successive singleton pregnancies and 2) pursuing a twin pregnancy, if the
couple wished to have two children.
These results will be discussed together with the results of the survey with the perspective of patient
autonomy in respect to SET.
5.3
Results of a questionnaire study
5.3.1 Introduction
Several studies have shown that a large proportion of infertility patients desire multiple pregnancies
(Pinborg et al. 2003b, Child et al. 2004, Ryan et al. 2004). It has been proposed that a proportion
of the multiple gestations is a result of sparse knowledge of neonatal complications among obstetricians and inadequate information among fertility patients (D’Alton et al. 2004), which cannot
outweigh the desire for an ‘‘instant family’’ (Child et al. 2004). However, no effect of additional
information on the acceptability of SET could be seen in a study by Murray et al. (2004).
It was the purpose of the present study to evaluate infertile couples’ attitudes towards the choice
between transfer of one or two embryos, to twinning and their acceptance of the associated neonatal
risk. We also aimed to elucidate which factors have an impact on the patients’ choices including
the influence of perception of risk. Finally, we wished to evaluate how patients balance the ac-
48
ceptability of a hypothetical policy of SET with the option of having more single embryo transfers
versus the existing policy of three reimbursed double embryo transfers.
5.3.2 Material and methods
All IVF/ICSI couples referred to the Fertility Clinic at Aarhus University Hospital, Skejby Sygehus,
in September 2004 were approached by a mailed questionnaire.
The questionnaire was designed by the group to cover issues of possible importance regarding
acceptability or wish for twins. Results of the qualitative study described in section 5.3.1 were used
in this process. Questionnaires were anonymized but coded to allow tracing of non-responders.
In a pilot study, 12 patients and partners were interviewed after completing the questionnaire. They
were invited to comment on the questions and on any subject which they found relevant. It was
confirmed that the topics were considered important and final alterations were made.
The patients were at all stages of IVF or ICSI treatment. Thus, some were waiting for their first
treatment; others had completed one or more treatments with a positive or negative result concerning clinical pregnancy.
Enclosed with the questionnaire were a stamped addressed envelope and a letter describing the aims
of the study, that anonymity was guaranteed, and that the participation was voluntary. Both patient
and partner were invited to self-complete a questionnaire. The respondents were encouraged to
complete the questionnaire without consulting the partner.
After two weeks, non-responders were contacted again by mail once.
5.3.2.1 The questionnaire
A 56-item questionnaire was developed for the women. The first section in the questionnaire to
women ascertained fertility information, parity, history of infertility, and present family structure.
The second section related to the degree to which the respondents desired twin pregnancies, and
for what reasons. They were also asked to point out their sources of information and to evaluate
their satisfaction concerning information on multiple pregnancies on a scale as follows: very satisfactory, satisfactory, fair, unsatisfactory, and very unsatisfactory. Questions in the third section dealt
with treatment-related stress, physical pain and side effects on a five-point Likert-type scale in the
following way: unacceptably severe, severe, acceptable, mild, none. The importance of a question
was measured on a four-point scale: very important, important, not very important, and unimportant.
In the fourth section, the respondents were asked about their future preferences for either SET or
DET if the number of reimbursed IVF treatments was increased or unlimited.
The fifth section ascertained demographic information; social position was measured in a standardized way including school education and job position. Furthermore, the respondents were asked
about their openness concerning their treatment.
Finally, the respondents were presented with three scenarios describing potential pregnancy complications (preterm delivery before 32 completed weeks of gestation, preeclampsia, neurological sequels, and stillbirth). In the first scenario, the risks for the complications were given as 1%, 2%,
0.2%, and 0.8%, respectively; in the second scenario, the risks were respectively 10%, 5%, 0.7%,
and 5%, and in the last scenario the risks were respectively 40%, 25%, 30%, and 20%. The middle
risk scenario represents the actual risk of a twin pregnancy and the last scenario mimics the risks
of a triplet pregnancy according to Malone et al. (1998), Grobman et al. (2001), and National
Vital Statistics Report (2001). However it was not revealed to the respondents how these scenarios
were constructed. For each scenario, the respondents were asked to indicate their wish for a twin
49
pregnancy under the conditions described in the scenarios by using a 5-point Likert scale from 1
(very undesirable) to 5 (very desirable).
To test whether the mode of risk presentation influenced the willingness to take risks, half of the
questionnaires included the above risk scenarios, while the other half of the questionnaires included
an alternative risk presentation where the reciprocal figures were given and the word ‘‘chance’’ was
used instead of ‘‘risk’’ describing the chance of having a gestation without the mentioned complications. No questionnaires contained both types of scenarios.
Based on the questionnaire for the women, a 41-item questionnaire was developed for their partners,
leaving out questions on cause and length of infertility, previous pregnancies and births, pain and
stress related to hormonal treatment etc. Otherwise the question and answer categories were identical. The presentation of risk vs. chance was identical within couples.
5.3.2.2 Statistics
For bivariate analyses the chi squared test was used for categorical data. Continuous data following
a normal distribution were analysed by t-test and ANOWA, while continuous data not following a
normal distribution were analysed by Mann Whitney U-test and Kruskal-Wallis test. For multivariate analyses of risk factors for dichotomous outcomes (e.g. wish for singleton vs. twin pregnancy),
logistic regression models were used, and ordinal (ordered) logistic regression was used to evaluate
risk factors for outcomes with more than two, ordered categories (e.g. the risk scenarios). Because
the attitudes and wishes of two partners cannot be considered independent observations, robust
variance estimates taking into account this non-independence were obtained. SPSS (Statistical Package for Social Sciences), version 11.0, and Stata version 8 were used for data analyses.
5.3.3 Results
Of the 588 couples invited, 5 women and 8 men returned the questionnaire unanswered, and 169
women and 176 men never returned the questionnaire, leaving a total of 414 women (70.4%) and
404 men (68.7%) for analyses.
5.3.3.1 General questions
Demographic characteristics of the participants are presented in Table 12.
50
TABLE 12
Baseline characteristics of participating men and women. Aarhus, Denmark, 2004
Mean age, years (SD)
Employment status
Employed
Unemployed
Student
Education, years
None
1
⁄2-1
3-4
±4
Nulliparous (no biological children)
Children with current partner
Yes
No
Know a family with a twin birth
Respondent is a twin
Unexplained
Damaged tubes
Poor sperm quality
Anovulation
Endometriosis
Male sterilized
Female sterilized
Median duration of infertility, months (range)
Previous twin birth
Previous preterm delivery ∞37 weeks
Previous stillbirths
Previous child with congenital malformation or disease
Women
(NΩ414)
32.3 (4.2)
Men
(NΩ404)
35.4 (5.9)
p-value
74.0%
16.9%
9.1%
85.9%
10.8%
3.3%
∞0.001
16.8%
25.8%
44.8%
12.8%
85.6%
13.8%
16.1%
52.3%
17.9%
80.4%
0.001
4.1%
95.9%
46.3%
1%
3.8%
96.3%
37.7%
2.5%
0.772
∞0.001
0.223
0.014
0.099
33.1%
17.6%
39.4%
10.6%
11.1%
3.6%
3.4%
42 (0-240)
0.9%
7.5%
0.6%
0.8%
5.3.3.2 Wish for twins
The majority of the patients and their partners preferred having twins (58.7%) to having one child
at a time (37.9%), while only 3.5% claimed to be indifferent. Within couples, most agreed on
their preference (74.6%). Among 60% of the couples who disagreed, the woman preferred twins
while the man preferred a singleton.
In bivariate analyses, the preferences for twins was positively associated with female gender, anovulation, not having endometriosis, not having unexplained infertility, not having biological children
and/or step children, being unemployed, number of working hours, short school education, and
short length of further education. There was no association with age, other causes of infertility,
duration of infertility, having adopted children, having children with current partner, parity, previous spontaneous or induced abortions, previous insemination or IVF treatment, previous embryo
transfer, having received information, feeling well informed, or the extent to which family, friends,
colleagues or others were informed about the treatment.
In a logistic regression model (preference for twins versus a singleton) including the variables that
were significantly associated with preference in the bivariate analyses, only the variables in Table 13
were significantly associated with twin preference. For some variables (e.g. school, employment),
information was missing for some individuals. Inclusion in the regression model of missing values
as a separate category for each variable with missing information did not substantially or significantly change the estimates.
51
TABLE 13
Factors associated with twin preference.
Multivariate logistic regression analysis
Gender
Male
Female
Biological children
No
Yes
Step children
No
Yes
School education
Æ10
±10
Anovulation
No
Yes
Endometriosis
No
Yes
Unexplained infertility
No
Yes
OR
95%CI
1.00
1.65
1.25–2.18
2.01
1.00
1.28–3.16
1.90
1.00
1.16–3.09
1.53
1.00
1.06–2.21
1.00
2.05
1.08–3.90
1.00
0.42
0.23–0.77
1.00
0.63
0.41–0.97
Among those who preferred to have twins, the primary reasons mentioned by each respondent for
wanting twins are shown in Table 14. No substantial or significant differences between male and
female respondents were found (data not shown).
TABLE 14
Primary reason for wanting twins (NΩ476)
I want to be sure that my children have siblings
I have a positive attitude towards having twins
I want as few IVF treatments as possible
Age plays a role
Twins are a joy for each other
I want as few pregnancies as possible
I cannot/will not pay for a 2nd child in a private clinic
I want as few deliveries as possible
Other
Not stated
Primary reason (%)
111 (23.3)
106 (22.5)
92 (19.3)
55 (11.7)
35 (7.4)
9 (1.9)
7 (1.5)
4 (0.8)
26 (5.5)
31 (6.5)
The primary reason mentioned for wanting one child at a time in the group preferring to have one
child at a time are shown in Table 15. There were differences between men and women: Men more
often pointed out twinning as a risk for the mother (26.4% vs. 6.8%, p∞0.001) and for the
marriage (5.7% vs. 0.8% pΩ0.020), while women were more concerned about the fetal risks than
the men (35.3% vs. 15.5%, p∞0.001).
52
TABLE 15
Primary reason for wanting one child at a time (NΩ307)
There is a risk for the fetuses
There is a risk for the mother
Twin pregnancies can be troublesome
I want more than one pregnancy
Twins would make it difficult to make ends meet in our everyday life
Twins would be a problem for our relationship
I have been advised not to carry a twin pregnancy for medical reasons
With my current work, it will be difficult to have twins
I want to deliver more than once
It is too expensive to have twins
Other
Not stated
Primary reason (%)
74 (24.1)
55 (17.9)
34 (11.1)
25 (8.1)
18 (5.9)
11 (3.6)
11 (3.6)
9 (2.9)
4 (1.3)
1 (0.3)
37 (12.1)
28 (9.1)
The above results are well in line with the fact that only 3.8% of the infertile patients would like
to be the parent of only one biological child, while 52.5% preferred to have two biological children,
and 43.7% wanted to be the mother or father of three or more biological children.
5.3.3.3 Specific questions related to the counselling
The information given on twins by the fertility clinic during treatments was evaluated as fair,
satisfactory, or very satisfactory by 95% (312/329) of the respondents. Nevertheless, only 41.6%
(340/818) stated that they had received oral counselling on advantages and disadvantages related
to twin pregnancy.
Among those 340 respondents who confirmed having had oral information on the possible complications of twin pregnancies 239 (70.3%) were informed by a doctor, and 169 (49.7%) by a nurse
at the Fertility Clinic, while 19 (5.6%) had had the information at an advisory meeting at the
clinic. A few patients had the information from another doctor (2.6%), their family (3.2%), or
friends (2.4%).
Among the 803 respondents, 620 (77.2%) found counselling on possible risk of twin gestations
important. Among the 465 respondents who denied having received any information, (72.7%)
found information important, and of the 338 respondents who confirmed having had the information, 83.4% found it important (p∞0.001). Furthermore, need of (more) information on twin
pregnancies was indicated by 32.3% (108/334) stating that they had received oral counselling on
twin pregnancies, whereas this was the case for 70.6% (326/462) stating that they had not received
oral counselling on twin pregnancies (p∞0.001).
To evaluate if the need for information was satisfied in other manners, the patients were asked to
point out their sources of information about twins. Furthermore they were encouraged to indicate
from where they would prefer to get information (Table 16).
53
TABLE 16
Information about twin pregnancies (NΩ818). There were no substantial or significant
differences between women and men (data not shown)
There are twins in our family
Know family with twins
From other IVF/ICSI couple
Conversation with a doctor at the clinic
Conversation with a nurse at the clinic
Advisory meeting at the clinic
Own G.P
Other contact with the Health Care System
Media
Internet
Patient association
Information leaflet
Through my work/education
Other
Not stated/don’t know
How did you obtain your
knowledge about twins?
Total
N
(%)
233
(28.5)
316
(38.7)
113
(13.8)
198
(24.2)
168
(20.6)
201
(24.6)
20
(2.4)
15
(1.8)
327
(40.0)
127
(15.5)
6
(0.7)
115
(14.1)
74
(9.0)
71
(8.7)
24
(2.9)
From where would you prefer to get
information about twin pregnancies?
Total
N
(%)
–
–
–
–
216
(26.4)
572
(70.1)
509
(62.2)
228
(27.9)
64
(7.8)
–
–
34
(4.2)
77
(9.4)
18
(2.2)
190
(23.3)
41
51
(5.0)
(6.3)
Of those 236 who had completed a treatment cycle, 83 (35%) had one embryo transferred, and
153 (65%) had two embryos transferred. The respondents were asked to point out the reason for
the number of embryos replaced (Table 17). Almost all the patients knew they had a choice. A
minority of 10.5% were directly advised to have either one or two embryos by the medical staff at
the Fertility Clinic. The most common answer was that the decision was their own.
TABLE 17
Primary reason for number of replaced embryo at the last embryo
transfer (NΩ812*)
Reason
It was my choice
Only one embryo was eligible for replacement
The Fertility Clinic advised me to have two embryos replaced
The Fertility Clinic advised me to have only one embryo replaced
It was never discussed
I took part in a randomized study
I did not know I had a choice
Not stated
Primary reason (%)
238 (29.1)
106 (13.0)
72 (8.8)
14 (1.7)
19 (2.3)
23 (2.8)
11 (1.3)
323 (39.5)
* 9 women and 3 men gave more than one answer and were excluded from the analysis.
5.3.3.4 Attitudes associated with well known risk
Table 18 and 19 show how the respondents scored the desirability of twin gestations described in
three different scenarios, where the risk figures given were those corresponding to singleton, twin,
and triplet pregnancies as described above. The analyses were restricted to the 705 women and
men, who answered the questions relating to all three scenarios. In Table 18 the figures in the
scenarios were presented as risk of complications, whereas in Table 19 the figures in the scenarios
were presented as chance of avoiding the complications described. The median desire for a twin
pregnancy decreased with increasing risk, from four in scenario 1 to three in scenario 2 and one in
scenario 3. When presented to the actual risk scenario for twin gestations the willingness to take a
risk was positively associated with desire for twins (ORΩ4.54 (3.23-6.38)) and negatively associated
with age of 35π years (ORΩ0.58 (0.36-0.94)). Patients and partners did not differ in their desire
for a twin pregnancy within each scenario (data not shown). When comparing the method of
providing risk information, i.e. risk scenario vs. chance scenario, the group presented to the risk
version was significantly more reluctant to desire twins than the group encountered with the recipro-
54
cal chance figures in the middle scenario (ORΩ4.91 (3.41-7.01)). Gender was not correlated to risk
taking attitude (data not shown). In the twin scenario, 60.9% (217/356) of the respondents presented to the chance version and 30.4% (106/349) of the respondents presented to the risk version
found a twin pregnancy within the scenario described very desirable or desirable.
TABLE 18
Scores for desirability of a twin pregnancy within three different risk
scenarios (NΩ349)
Singleton scenario
Twin scenario
Triplet scenario
Likert score (1Ωvery undesirable and 5Ωvery desirable)
1 (%)
2 (%)
3 (%)
4 (%)
5 (%)
31 (8.9)
20 (5.7)
57 (16.3)
105 (30.1)
136 (39.0)
56 (16.0)
86 (24.6)
101 (28.9)
63 (18.1)
43 (12.3)
255 (73.1)
54 (15.5)
23 (6.6)
6 (1.7)
11 (3.2)
TABLE 19
Scores for desirability of a twin pregnancy within three different chance
scenarios NΩ356
Singleton scenario
Twin scenario
Triplet scenario
Likert score (1Ωvery undesirable and 5Ωvery desirable)
1 (%)
2 (%)
3 (%)
4 (%)
5 (%)
9 (2.5)
9 (2.5)
46 (12.9)
78 (21.9)
214 (60.1)
10 (2.8)
35 (9.8)
94 (26.4)
98 (27.5)
119 (33.4)
141 (39.6)
93 (26.1)
73 (20.5)
28 (7.9)
21 (5.9)
The respondents were asked to evaluate stress associated with the treatment in terms of side-effects
of hormone treatment, pain at oocyte retrieval and general physical and psychological stress. Only
13.6% reported IVF treatment not to be stressful (Table 20). The importance of physical and
psychological stress upon the decision between transfer of one or two embryos increased systematically across five categories of perceived physical and psychological stress (Test for trend, p∞0.001)
(Table 20).
TABLE 20
Estimation of physical and psychological stress associated with IVF
treatments (NΩ737)
Degree of physical and
psychological stress
Unacceptable severe
Severe
Acceptable
Mild
No stress
How important is the of physical and psychological stress
for your decision on number of embryos to be transferred?
Important
Not important
Total
N (%)
N (%)
N (%)
6 (2.8)
1 (0.2)
7 (0.9)
69 (31.9)
90 (17.3)
159 (21.6)
101 (46.8)
241 (46.3)
342 (46.4)
25 (11.6)
104 (20.0)
129 (17.5)
15 (6.9)
85 (16.3)
100 (13.6)
5.3.3.5 Preference for the future
A large majority of 78.5% (625) expressed wish to have two embryos transferred in a future
treatment, while only 6.2% (49) wanted SET.
Among the 229 respondents who preferred one child at a time, 81.2% (186) planned to have DET
in their next treatment. In comparison, 98.6% (413/419) of the respondents who preferred twins
would opt for DET in their next treatment.
5.3.3.6 Number of reimbursed cycles and choice between one or two embryos
Being presented with a statement of equal chances of having a childbirth the respondents were
asked about preferences for either 4 IVF treatments with a SET protocol and 5% chance of twins,
or 3 IVF treatments with a DET protocol and 25% chance of twins. A majority of 73.2% (575/
785) would prefer the latter option. A similar proportion (73.4%) would opt for DET even if the
55
number of government funded IVF treatments for the first child was unlimited. When the patients
were asked to decide between SET and DET if reimbursed IVF treatment were to be a future
option for a second child, as many as 67.7% (531/784) respondents still deemed DET as their
preferred treatment option.
5.3.4 Discussion
The present study revealed that the majority of infertile couples currently in treatment prefer twins
(58.7%) to one child at a time (37.9%), but a larger majority (78.5%) planned to have two embryos
transferred in the next treatment. Accordingly, the preference of DET is not only explained by a
wish to have a high success rate and thus avoiding more treatments, but reflects a deliberate wish
to have twins in the majority of couples. The present proportion of couples preferring twins seems
high. Other studies have shown that the proportion of fertility patients preferring twins were 14%
(Kalra et al. 2003), 20.3% (Ryan et al. 2004), 38.9% (Child et al. 2004) and 32% (Murray et al.
2004), respectively. In the last study 93% denied minding twins. The higher number of Danish
patients wanting twins seems to be a robust observation, as a national survey among IVF/ICSI
mothers of twins and singletons showed that 84.7% of twin mothers and 62.3% of singleton
mothers preferred to have twins (Pinborg et al. 2003b).
Preference of DET is in agreement with present practice in Denmark. The acceptance of an existing
policy is well known among fertility patients (Murray et al. 2004) and was also seen in an earlier
study in our clinic concerning the preference of two different treatment protocols (Højgaard et al.
2001).
The wish to have twins was not associated with age as also Pinborg et al. (2003b) found in a
Danish national survey – but in contrast to the findings of Child et al. (2004). Reasons for wishing
twins were desire for siblings, mutual pleasure between siblings or less specific positive attitude
towards twins. A single child was preferred (42%) due to difficult pregnancy with twins, and risk
to the child or mother.
The information given on twins was evaluated as very satisfactory, satisfactory or fair by nearly all
couples, but only a little less than half had received oral counselling. It is difficult to evaluate to what
extent the counselling about twins was sufficient. Only 12% stated that their previous decision(s) on
number of embryos transferred was based on advice from the fertility clinic, whereas one third
described the decision as their own choice. Nearly all knew they had a choice. In spite of the fact
that more than half of the couples had found information in the media and the internet only a
minority found these sources preferable. Very few had actively sought information by using patient
associations. It seems that an important source of information is case based by knowing families
with twins, but highest priority had personal counselling at the Fertility Clinic. Accordingly, the
general impression is that more specific and organised information is needed. However, the question
is whether information would change the wish to have twins. In the analyses in the present study
there was no association between opting for twins and having received information or feeling well
informed. Other studies have shown that infertility patients seem to be rather unaffected by perceptions of a high risk associated with twins (Kalra et al. 2003). Furthermore, attitudes towards eSET
seemed independent of methods of information provision in a randomized study by Murray et al.
(2004). Perception of risk appears to be strongly dependent upon wording. The desirability of twin
gestations described in three different scenarios with increasing risks for mother or child revealed
that the median desire for a twin gestation decreased with increasing risk. When comparing the
method of providing risk information, i.e. a risk scenario vs. a chance scenario the group presented
to the risk version was significantly more reluctant to desire twins than the group encountered with
the reciprocal chance figures. As risk aversion could be demonstrated by changing the mode of risk
presentation, infertile couples still seem to be affected by information offered on the twin issue.
In Denmark three IVF/ICSI embryo transfers are reimbursed. Further treatments in case no pregnancy is obtained or treatments to achieve a second child must take place in a private clinic. In
56
bivariate analyses, the preferences for twins were positively associated with being unemployed, number of working hours, short school education, and short length of more advanced studies. Nevertheless only a few stated that they could not pay for a second child in a private clinic. Accordingly,
economic motives for twin preferences seemed of less importance. In contrast, Kalra et al. (2003)
suggested that treatment cost could be one of the hidden factors explaining the desire for twins.
The fact that only a minority would opt for SET if combined with four reimbursed embryo
transfers or even an unlimited number of treatments or if treatments for a second child was reimbursed strongly suggest that economic considerations are not at play. In a previous survey patients
seemed to prefer the simplicity and short duration of a soft stimulation regimen in spite of drawbacks such as a high risk of cycle cancellations and accordingly the necessity of more treatment
cycles (Højgaard et al. 2001). Thus treatment related stress may be an important factor together
with a sincere – but hardly knowledge based – wish to have twins.
Acknowledgements
We would like to acknowledge the volunteering pilot couples, the patients participating in the study
and the staff at the Fertility Clinic.
5.4 Patient perspectives – discussion and conclusions
The results of the qualitative study and the survey are concordant by revealing a strong desire to
have at least two children. This translated into a wish to have twins among couples undergoing
fertility treatment. Many associated positive values with twins. Nevertheless both studies also showed
that experienced physical and psychological stress associated with the treatment strongly influenced
the preference for twins. Both motives are in accordance with the fact that in the survey patients
did not focus on cumulative chances of pregnancy following a sequence of treatment cycles since
more reimbursed treatments to compensate for a lower pregnancy rate following SET were not
attractive.
Coping with information was interesting in both studies. Despite an overall high satisfaction with
the information given, very little specific knowledge was acquired by the patients. The qualitative
study even indicated that risk information was neutralised by the couples in order to avoid that
risk figures would pose a threat to the overall wish and motivation of the couples. Such a repressive
mechanism may explain why we did not find any association between opting for twins and having
received information or feeling well informed and that some data indicate little effect of information
(Murray et al. 2004). Still, wording of information did influence choices in three different risk
scenarios. A positive wording with emphasis on chances rather than risks could inspire future
guidelines for written and oral information.
Both studies indicate that an obligatory single embryo policy would be in conflict with patient
interests and wishes. More carefully prepared information on twin pregnancy seems to be needed,
but how much such information may move patient attitudes remains to be evaluated.
57
6 Cost-effectiveness of SET versus DET strategies
6.1
Introduction
The dilemma between SET and DET is interesting from an economic point of view. Multiple
pregnancies are more expensive than singleton pregnancies due to increased need for antenatal and
neonatal care. However, SET reduces the twin pregnancy rate resulting in lower costs in the health
care sector due to a reduced need for extra monitoring during pregnancy, less complicated deliveries
and less need for neonatal intensive care, as well as a lower risk of having neurological sequelae, i.e.
cerebral palsy. A SET-policy will expectedly result in a lower pregnancy rate per transfer of fresh or
frozen embryo(s) and possibly also cumulatively, as well as fewer children born.
This trade-off between effectiveness and costs will be evaluated in the present cost-effectiveness
analysis.
6.2 Purpose
The purpose of the study was two-fold: The first aim was to calculate the costs of singleton and
twin IVF pregnancies, as well as the three first months after delivery of the child. Secondly, the aim
was to investigate the cost-effectiveness of single-embryo transfer (SET) compared with doubleembryo transfer (DET) in IVF, i.e. to elucidate the hypothesis whether a SET-policy is cost-effective
compared with the traditional DET-policy used most frequently today.
The perspective of the health economic analysis was societal as days absent from work during
pregnancy (production lost), as well as the out-of-pocket expenditures for the couples was recorded.
However, the result is reported both with and without the inclusion of the production lost.
6.3 Method – type of analysis
The health economic analysis of SET compared with DET was carried out as cost-effectiveness analysis. Three measures of effectiveness were chosen in the cost-effectiveness analysis – clinical pregnancies,
births, and children born – with the two of them being final endpoints, whereas clinical pregnancy rate
(defined after eight weeks of pregnancy) can be considered as an intermediate measure of effectiveness.
However, prior to the cost-effectiveness analysis a cost analysis was carried out estimating the costs
of a singleton IVF child/pregnancy compared to the costs of a twin born IVF child/pregnancy.
These cost result were used in the cost-effectiveness analysis comparing SET and DET IVF.
The design of the health economic analysis was a mix of a prospective collection of resource use in
the antenatal and neonatal phases (three months after delivery) followed by a modelling (decision
tree) of the cost-effectiveness of the patient flow in the SET versus DET IVF strategies. Data were
analysed using SPSS 13.0 statistical package and Microsoft Excel.
In this decision tree model information of the cost of the IVF treatment part came from a previous
study on IVF by Ingerslev et al. (2001). Furthermore, data on the cost of having either a singleton
or twins following IVF came from the prospective collections described below. Finally, future costs,
related to the IVF treatment and the increased risk following IVF twins, such as cerebral palsy, was
included in the model based on evidence from the literature (Strømberg et al. 2002, Pinborg et al.
2004). This last cost component was, however, only added in sensitivity analyses.
Resource use and cost in other sectors, e.g. social care, were not included.
58
6.4 Material
The inclusion criteria for the women (couples) invited to participate in the prospective collection
study was that they have had IVF or intra-uterine insemination (IUI) treatment at the fertility
clinic at Aarhus University Hospital, Skejby resulting in a live clinical pregnancy as diagnosed by
vaginal ultrasound in week 8. All women (couples) fulfilling these criteria were invited to participate
in the study. Both IVF and IUI pregnancies were included in order to generate enough participants
within the study period. It was considered hard to identify reasons to believe that significant differences may exist between the two groups in terms of pregnancy complications and outcome. A posthoc analysis was planned to elucidate this.
As specific risk conditions can influence the economic result, selective SET due to medical indications such as previous preterm delivery, pre-eclampsia, and cervical incompetence were criteria
for exclusion from the study. This group of patients was expected to be small.
The pregnant women (and couples) were invited to participate in the study from April 1st 2002,
and the last woman was included in the study 12 months later (March 31st 2003). With a 12month data collection for each woman (couple) the data for the last woman was collected by April
1st 2004.
In total, 277 couples, including 14 FER, obtained a clinical pregnancy by IVF or ICSI during the
study period.
6.4.1 Data collection
Data on resource use during the antenatal, delivery and neonatal phases were collected with the use
of cost diaries, which were filled in prospectively by the women (couples). This registration was
initiated at the fertility clinic at the ultrasound scan visit at week 8, where the first cost diary was
given to the couple together with information of the project. After three months a new diary was
mailed to the couple for registration of the activities during the next three months. In total the
one-year data collection consisted of four cost diaries, which were mailed to the woman (couple)
every 3-month. A 3-month period of each cost diary was chosen to minimize problems with recall
bias.
The data collection period in the neonatal phase covered as the rest of the cost diary period only
three months. This was, however, assessed to be adequate, as a previous study in the county of
Funen had revealed that the length of inpatient stay for IVF twins in the neonatal intensive care
unit was 20 days (median) (Westergaard 1998).
Women who were not pregnant anymore due to spontaneous abortion, or simply did not want to
continue in the project, did not receive further cost diaries.
With the information on resource use from the cost diaries it was possible to calculate the cost in
the antenatal, delivery and neonatal phases of having either a singleton or twins following the IVF
treatment.
The difference in the total costs of the SET and DET policies is explained by the differences in
the pregnancy rates and number of multiple pregnancies (twins) and deliveries, experienced in the
two groups.
59
6.4.2 Data on resource use
To be able to estimate the costs of singleton and twin IVF births/children, and in the end the costeffectiveness of SET and DET strategies in IVF, focus was upon resource use during the IVF
treatment, resource use in the antenatal phase as well as resource use in the neonatal phase. The
two last phases are interesting as the resource use in these phases may be related indirectly to the
IVF treatment. Furthermore, the couples own expenditures as well as days absent from work was
included as a resource use. Only resource use where a difference is expected was included.
Below are the specific resource use collected listed.
H
Resource use for the IVF treatment.
H
Resource use in the antenatal phase.
H
Visits at the midwife.
H
Visits at the general practitioner (GP) or emergency doctor.
H
Outpatient visits at the hospital.
H
Inpatient stay at the hospital.
H
Pharmaceutical drugs.
H
Abortions.
H
Resource use for the delivery.
H
Type of delivery (normal or caesarian section).
H
Neonatal phase.
H
Visits by the health visitor.
H
Visits at the GP or emergency doctor.
H
Outpatient visits at the hospital.
H
Inpatients stay at the hospital (neonatal intensive care unit).
H
H
Resource use by the couple.
H
Transportation to visits.
H
H
Production lost.
H
Days absent from work.
Following the IVF or IUI treatment and a positive pregnancy test these resource uses were systematically collected for each woman during the pregnancy period as well as three months after the
delivery of the child/children. This means that the resource use prospectively collected in total
covered one year. Pregnancy was verified with an ultrasound scanning five weeks after embryo
transfer at the fertility clinic.
6.5 Results
6.5.1 Cost of an IVF singleton versus IVF twins
In total 213 women (couples) – 190 IVF and 23 IUI – fulfilled the criteria for inclusion and
accepted to participate in the study with a participation rate of 77%.
Among the 213 women 14 women withdrew from the study without returning any cost diary,
which left 195 women (couples) as active participants in the study (70%). All four cost diaries were
returned by 164 women during the 1-year of data collection (59%). Most of the dropouts happened
after cost diary 1 for the first three-month period (21 women), 12 women returned the first two
diaries, and 2 women returned three cost diaries.
Four women had a spontaneous abortion, and two women delivered a stillborn child. The reason(s)
for dropping out of the study were not stated explicitly by the rest (43 women).
60
All cost diaries returned were used in the analysis of the cost of an IVF singleton versus IVF twins.
The cost diaries were fulfilled during the period of April 1st 2002 to April 29th 2004.
6.5.1.1 Background data on participants, obstetric outcome and the children
The average age of the women with singleton pregnancy was 32.0 years compared with 30.9 years
for women with twin pregnancies. As seen from Table 21 146 women had a singleton pregnancy
and 49 women were pregnant with twins (25%). Two of these were a result of IUI. For 82% of
the women with a singleton pregnancy this was their first child, whereas this was the case for 90%
of the women with a twin pregnancy.
TABLE 21
Previous pregnancies
First child
Singleton pregnancy (NΩ146)
Twin pregnancy (NΩ49)
121
44
Not the
first child
25
5
Total
146
49
1. Test of the difference in previous pregnancy: c2 Ω1.349; dfΩ1; pΩ0.245.
Table 22 shows the gestational age at delivery and the mean birth weight. As expected twin pregnancies had a significant shorter mean duration of gestation (week 36) than singletons (week 39) and
a significant lower mean birth weight (2,470g vs. 3,400 g).
TABLE 22
Average week of delivery and average birth weight
Week of delivery
Week 39
Week 36
Birth weight of
child/twin child 1
3,412 g
2,470 g
Birth weight of
twin child 2
–
2,432 g
1. Test of the difference in week of delivery: tΩ6.540; dfΩ140; p∞0.001.
Test of the difference in birth weight of child 1: tΩ8.577; dfΩ141; p∞0.001.
2.
6.5.1.2 Resource use
Antenatal, delivery-associated and neonatal resource use is described in the following section. Some
of the various types of resource use are different in twin and singleton pregnancies probably because
the monitoring of two types of pregnancies is organized differently. For example, the number of
visits to the midwife and the general practitioner (GP) during a twin pregnancy were fewer in twin
than in singleton pregnancies (Table 23 and 24) – although only significantly different for midwife
visits. The difference between singleton pregnancies and twin pregnancies in terms of visits is
explained by a more intensive monitoring of twin pregnancies – especially during the last half of
pregnancy, which is done in the antenatal clinic in the obstetric departments. These more visits in
the outpatient clinic for twin pregnancies are seen in Table 25. These outpatient visits substitute
some of the midwife and GP visits in twin pregnancies.
Fewer midwife visits (3 vs. 5), especially during the last three months may also be explained by the
fact that women with twin pregnancies deliver earlier (Table 2).
After delivery the women visit their GP for the first examination of the newborn child as well on
the basis of need.
61
TABLE 23
Number of midwife visits
Period (months)
1-3
4-6
7-9
Total period
1-3
4-6
7-9
Total period
Total number
of visits
Average
per woman
66
205
476
747
0.45
1.40
3.26
5.12
27
67
66
160
0.55
1.37
1.35
3.27
1. Test of the difference in mean number of midwife visits (total):
tΩ5.243; dfΩ193; p∞0.001.
TABLE 24
Number of visits at the GP during pregnancy and
until three months after delivery
Period (months)
1-3
4-6
7-9
After delivery
Total period
1-3
4-6
7-9
After delivery
Total period
Total number
of visits
Average
per woman
198
209
212
275
894
1.36
1.43
1.45
1.88
6.12
73
83
39
67
262
1.49
1.69
0.8
1.37
5.34
1. Test of the difference in mean number of GP visits (total): tΩ
1.681; dfΩ193; pΩ0.094.
On average the women with singleton pregnancies visited the GP six times on average in the 12months period for the reason of their pregnancy or their newborn child. For women with a twin
pregnancy this was five visits on average.
The women in the study reported use of around 100 different types of drugs. Most of these were
over-the-counter pharmaceutical drugs, such as vitamins and iron products routinely recommended
during pregnancy. All types of pharmaceutical drugs were included in the cost analysis.
The total and average number of outpatient visits at the hospital during pregnancy and three
months after delivery is presented in Table 25.
62
TABLE 25
Number of outpatient visits at the hospital during
pregnancy and three months after delivery
Period (months)
1-3
4-6
7-9
After delivery
Total period
1-3
4-6
7-9
After delivery
Total period
Total number
of visits
Average
per woman
196
183
177
42
598
1.34
1.25
1.21
0.29
4.10
92
125
133
18
368
1.88
2.55
2.71
0.37
7.51
1. Test of the difference in mean number of outpatient visits (total):
tΩª5.837; dfΩ193; p∞0.001.
As expected the average number of visits was significantly higher for women with a twin pregnancy
than for women with singleton pregnancy (7.5 vs. 4.1 visits). The largest difference between the
two groups is especially found from the fourth to the ninth month of pregnancy. Some of the visits
during this period at the midwife and GP are normally substituted by visits at the antenatal clinic
in the hospital.
The frequency of hospitalization during pregnancy, in association with delivery, post partum or
during the neonatal period of the woman or child(ren) is reported in Table 26.
TABLE 26
Average number of inpatient stays at the hospital during
pregnancy and three months after delivery (including the
child’s inpatient stay1)
Period (months)
Pregnancy period
After delivery
Total period
Pregnancy period
After delivery
Total period
Average number
Average number of
of inpatient stays inpatient days in hospital
0.29
0.14
0.43
1.59
0.77
2.36
0.45
0.24
0.69
3.16
6.20
9.37
1. When twins require neonatal care both children are assumed to be hospitalized together with their mother.
2. Test of the difference in mean number of inpatient stays (total): tΩ
ª1.828; dfΩ193; pΩ0.069.
3. Test of the difference in mean number of inpatient days (total): tΩª3.221;
dfΩ193; pΩ0.001.
Both during pregnancy as well as post partum women with twins had more inpatient stays compared with women with singletons. After delivery, where hospitalization was primarily due to need
for neonatal intensive care, a twin had eight times more inpatient days in hospital than a singleton,
as shown in Table 26 (pΩ0.001).
The proportion of vaginal and caesarian deliveries in the two groups is presented in Table 27.
63
TABLE 27
Proportion of vaginal and caesarian delivery
Type of delivery
Vaginal delivery
Caesarian section
1Vaginal delivery
Caesarian section
Total number
Percent
96
31
76%
24%
11
23
32%
68%
1. Information on delivery type is missing for 19 women with
singleton pregnancy.
2. Information on delivery type is missing for 15 women with
twin pregnancy.
3. Test of the difference in type of delivery: c2Ω22.493; dfΩ1;
p∞0.001.
Caesarian section is significantly frequent in twin pregnancies than in singleton pregnancies. Caesarian section was performed in 68% of twin deliveries compared to only 24% of singletons. This
is higher than previously found by Henriksen et al. (1994) and Sperling et al. (1994) (28% vs. 9%)
in the general population, but the national caesarean section rates have increased significantly since
1994.
After delivery the health visitor visits the woman and her newborn baby during the first period.
These visits were recorded by the participants in cost diary number 4, as seen in Table 28.
TABLE 28
Number of visits by the health visitor
Period (months)
After delivery
After delivery
Total number of visits
by the health visitor
Average number of
visits per woman
527
3.61
177
3.61
1. Test of the difference in mean number of visits by the health visitor
(total): tΩª0.007; dfΩ193; pΩ0.994.
As the Table shows there was no difference in the average number of visits by the health visitor in
the two groups.
Finally, days of absence from work due to illness or necessary visits to during the pregnancy period –
before maternity leave – was recorded in both groups of pregnancy. The results are presented in
Table 29.
TABLE 29
Number of days of absence from work due to the pregnancy
Period (months)
1-3
4-6
7-9
Total period
1-3
4-6
7-9
Total period
Total number of days
of absence from work
Average per
pregnant woman
358
916
500
1774
2.45
6.27
3.43
12.15
494
590
154
1238
10.08
12.04
3.14
25.27
1. Test of the difference in mean number of days of absence from work
(total): tΩª2.405; dfΩ193; pΩ0.017.
64
Women with a twin pregnancy were found to be absent from work 25 days on average compared
to 12 days on average for women with a singleton pregnancy (pΩ0.017). The largest difference
was found during the first three months of pregnancy, where absence from work was more than
five times higher in the twin group, but also during the second trimester, where absence from work
was two times higher in the twin group.
6.5.1.3 Unit costs
Relevant unit costs were identified to value the resource use. These unit costs, as well as their
sources are presented in Table 30.
TABLE 30
Unit costs (Danish kroner (DKK))
Unit cost
IVF treatment
FER (IVF from frozen egg)
Pregnancy check-up at the GP
(1st visit, 2nd-3rd visit, following visits)
Source
Ingerslev et al. (2001)
Ingerslev et al. (2001)
www.plo.dk
Check-ups at the GP after delivery (child examination)
(1st visit, following visits)
Cost per midwife visit
Ultrasound scan
Pharmaceutical drugs
www.plo.dk
Outpatient visit at the hospital
Inpatient stay in the hospital
Uncomplicated vaginal delivery (normal delivery)
Caesarian section, normal degree of complication
Neonatal Intensive Care Unit: Inpatient stay for children with a birth weight
between 1,000-2,499 g (uncomplicated)3 DRG2 nr. 1508
Neonatal Intensive Care Unit: Inpatient stay for children with a birth weight
above 2,499 g (uncomplicated)3 DRG2 nr. 1511
Cost per health visitor visit
Transportation
Wage rate per hour (production lost)
www.sundhedsinformation.dk and www.drs.dk
DAGS1 nr. BG50A
Pharmacy (Århus Jernbane Apotek) d. 27/7 2004.
Over-the-counter pharmaceutical drugs: www.apotek.dk
Prescription pharmaceutical drugs: www.medicinpriser.dk
DAGS1 nr. BG50A
DRG2 nr. 1410
DRG2 nr. 1407
DRG2 nr. 1405
DKK 73,449.00
Price
DKK 15,215.00
DKK 5,877.00
DKK 296.84
DKK 150.28
DKK 106.81
DKK 181.47
DKK 106.81
DKK 62.72
DKK 1,393.00
Various
DKK 1,393.00
DKK 10,272.00
DKK 11,147.00
DKK 23,372.00
DKK 19,590.00
MPH Dissertation nr. 59 and www.drs.dk
www.told&skat.dk
Own calculations and Danish Statistics
DKK 177.15
DKK 1.62 per km.
DKK 215.61
1. DAGS (Dansk Ambulant Grupperingssystem) – a casemix payment system for outpatients.
2. DRG – Diagnose Related Groups – a casemix payment system for inpatients.
3. In case of twins both children will be hospitalized in the neonatal care intensive unit.
The price of an IVF treatment was taken from a previous HTA comparing standard IVF with
clomiphene stimulated IVF (Ingerslev et al. 2001). Labour costs were adjusted to a 2004 level with
an annual increase of 3.2 percent, and with respect to drugs, updated 2004 prices were found.
However, costs for various articles, overhead and depreciation were kept constant in the updated
calculation of the cost of an IVF treatment.
With respect to pregnancy visits at the general practitioner the charge of the first visit was DKK
296.84 according to the agreement between the organisation of the general practitioners (PLO) and
the county council association. The price of the second and third visit was DKK 150.28, and for
additional visits the price is DKK 106.81. With respect to visits after the delivery focusing on
examination of the newborn child the charge for the first visit is DKK 181.47 and DKK 106.81
for additional visits.
The cost per visit at the midwife was based on the assumption that the visit last 20 minutes and
that the hourly wage rate for a midwife is DKK 188.26.4 Using these assumptions the cost per
midwife visit becomes DKK 62.72. This cost estimate is probably lower than the actual cost, as it
only includes salary, and on the other hand excludes running costs, overheads and building expenditures (capital).
4
The hourly wage rate for a midwife was based on a monthly wage of DKK 27,846 and 1,775 hours of work annually.
65
The cost per visit by the health visitor was based on the assumption that the visit last 1.25 hours,
according to a time study from the county of Ringkjøbing, and that the hourly wage rate for a
health visitor was DKK 155. The cost per visit by the health visitor then amounts to DKK 177.15.5
For inpatient stay at the hospital before and after the pregnancy the calculation of the cost was
based on the actual number of days the women (and child(ren)) was hospitalized, recorded in the
cost diaries. If hospitalization was longer than the average days in the relevant DRG charge, then a
day charge of DKK 1,522 was added to the specific DRG charge.
To evaluate the production lost in society due to absence from work an average wage rate was
calculated as a weighted average of the salary for women and men in 2003 prices.
6.5.1.4 Average costs per pregnancy/child
The total average costs of a singleton or twin IVF pregnancy and child(ren) were estimated on the
basis of the prospectively collected resource consumption data (diaries), and the associated unit
costs. The cost of the IVF treatment itself was not included in this calculation, which should cover
costs related to pregnancy and the three-month period post partum only. Table 31 shows the average
costs per pregnancy or per child for singletons and twin pregnancies according to phase (antenatal,
delivery, neonatal).
TABLE 31
Average costs per pregnancy/child for a singleton and twin pregnancy/child (DKK)
Singleton pregnancy/child
(NΩ146)
Average cost per
pregnancy/per child (DKK)
Antenatal phase
Pregnancy control at the GP
Visit at midwife
Ultrasound scan
Inpatient and outpatient stay
Delivery phase
Delivery (normal/caesarian section)
Neonatal phase
Neonatal Intensive Care Unit
Outpatient visits at the hospital
Visits by the health visitor
Visit at the GP (child examination)
Women’s own expenditures2
Pharmaceutical drugs (OTC and prescription)
Transportation to visits
All costs (excl. production lost)
Production lost
All costs (incl. production lost)
Twin pregnancy/child (NΩ49)
Average cost per
Average cost
pregnancy (DKK)
per child (DKK)
721
321
5,305
3,851
686
205
9,950
5,893
343
103
4,975
2,947
12,292
13,473
6,737
3,558
401
639
266
28,0061
512
640
193
14,003
256
320
97
114
434
27,903
19,884
47,787
303
588
60,449
40,311
100,760
152
294
30,225
20,156
50,380
1
Assumes that both children are hospitalized at the neonatal intensive care unit.
2. Covers drugs bought by the women and transportation costs to the visits.
3. Result in parentheses is average costs, excluding production lost.
The average cost per pregnancy or delivery is more than twice as high, for a twin as compared with
a singleton pregnancy (p∞0.001). The major reasons to this difference are significantly higher
antenatal and neonatal costs per pregnancy, as well as more production lost in the twin pregnancy
group. It is noteworthy that the average cost per pregnancy for neonatal care intensive hospitalization is DKK 28,006 for twins (both hospitalized) compared with DKK 3,558 for singletons.
5
The hourly wage rate for a health visitor was based on a monthly wage rate of DKK 24,295.13 and 1,775 hours of work annually.
66
The need for and costs associated with neonatal care for IVF neonates found in the present study
seems somewhat lower than in a previous Danish estimate (Westergaard et al. 1998). The impact
of neonatal care costs calculated in that study was therefore further investigated in a separate
sensitivity analysis. Ingerslev et al. (2001) used the data estimated by Westergaard et al. (1998) in
a previous Danish HTA concerning IVF treatment.
Production lost per pregnancy was twice as high in the twin compared with the singleton pregnancy
group. Omitting production lost from the calculation of the total costs then limits the difference
in average costs between the two groups only little, since the average cost per pregnancy/delivery
after a twin pregnancy is still more than 1.5 times higher than after a singleton pregnancy.
When considering the average pregnancy associated costs per child born, the costs of the twin
pregnancy group comes close to the cost per child born from a singleton IVF pregnancy. Actually,
the cost per child born from a twin IVF pregnancy is marginally lower. In case the wish of the
parents is to have two children from the IVF treatment one could argue that this average cost per
child born is the relevant one for comparison.
6.5.2 Cost-effectiveness of SET- versus DET-policies
The main purpose of the health economic analysis was to investigate the cost-effectiveness of a
SET-policy with a DET-practice, which is presently predominant in Denmark. In the previous
section, we have calculated the costs of a singleton and twin pregnancy and child(ren). Combining
these with updated cost estimates concerning IVF treatment from a previous HTA (Ingerslev et al.
2001), and with evidence on the effectiveness of SET- and DET-policies (see the technology section), it is possible to model the cost-effectiveness of the two transfer policies.
The evidence on the effectiveness of SET- and DET-policies available in the literature is presented
in the following section. Afterwards a decision analytic model to analyze the cost-effectiveness of
the two transfer policies is presented.
6.5.2.1 Clinical evidence
As referred to in the Technology chapter a Cochrane review by Pandian et al. (2004) systematically
reviewed the available evidence of the effectiveness of SET in IVF compared with elective-DET.
The RCT’s included in the Cochrane review is presented below in Table 32.
TABLE 32
Sample size of the RCT’s included in the
Cochrane review (Pandian et al. 2004)
Gerris et al. (1999)
Martikainen et al. (2001)
Lukassen et al. (2002)
Total
Sample size (women (cycles))
SET
DET
26
27
74
70
26
25
126
123
The pooled effectiveness results of the three RCT’s as presented in the Cochrane review is summarized below in Table 33 (Pandian et al. 2004).
67
TABLE 33
Pooled effectiveness results from Pandian et
al. (2004) in the comparison of SET versus DET
in IVF/ICSI
Clinical pregnancy rate
Livebirth rate
Multiple pregnancy rate
Single pregnancy rate
Pandian et al. (2004)
SET
DET
37.3%
54.9%
31.0%
45.9%
1.6%
16.4%
35.7%
38.5%
The clinical pregnancy rate was significantly higher for DET compared with SET, which is also the
case with the livebirth rate. On the other hand, the twin pregnancy rate was significantly lower
with the SET strategy than in case two embryos were transferred.
However, an overall problem with the three clinical trials reviewed in the Cochrane review – especially Gerris et al. (1999) and Lukassen et al. (2002) – is that the methodological quality of the
studies are inadequate for a robust conclusion from the results to catalyse a change in clinical
practice (Pandian et al. 2004). More research and evidence is needed.
A newer and larger multicenter randomized controlled clinical trial carried out in Sweden, Norway
and Denmark has been published after the literature search in the Cochrane review (June 2003).
In this study by Thurin et al. (2004) compared elective SET with DET in 661 women less than
36 years, and with at least two good-quality embryos, underwent randomization. The SET group
had subsequent transfer of one fresh and one frozen embryo (if any). The effectiveness result of
this trial is shown below in Table 34.
TABLE 34
Effectiveness result from Thurin et al. (2004) in the
comparison of eSET versus DET in IVF/ICSI (intention
to treat)
Clinical pregnancy rate, cumulative
(fresh and frozen)
Livebirth rate
Multiple pregnancy rate1
Single pregnancy
Thurin et al. (2004)
eSET
DET
47.9%
52.6%
38.8%
0.8%
47.1%
42.9%
33.1%
19.5%
1. Covered delivery of 46 twins and one triplet.
In accordance with the previously mentioned three randomized controlled studies, Thurin et al.
(2004) found a significantly higher clinical pregnancy rate in the DET group compared to the
elective SET group when only fresh embryos were considered (52.6% vs. 33.6%). However, when
both fresh and frozen were included the difference was not significant (52.6% vs. 47.9%). The
same was the case for livebirth rate in the two alternatives. However, the number of multiple
pregnancies was significantly lower with the eSET-policy as compared with the DET-policy (0.8%
vs. 33%).
In the health economic analysis it was chosen to include and use both the effectiveness results from
the clinical study by Thurin et al. (2004) and the combined data from the Cochrane review (Pandian et al. 2004).
The different cost-effectiveness results are presented in section 6.5.2.3.
68
6.5.2.2 Decision analytic model
To analyse the cost-effectiveness of SET- versus DET-policies in IVF simple decision analytic models
in the form of decision trees was designed. These decision trees were based on the clinical evidence
as referred to above with respect to the probabilities for having a singleton, twins or no delivery
from either a SET-policy or a DET-policy. The costs of the IVF treatment using a standard IVF
protocol, the costs of FER-IVF (Ingerslev et al. 2001) and the costs of having either a singleton or
a twin child after IVF, as calculated in result section 1 in this health economic analysis, was entered
in the decision tree as cost inputs. As examples the decision trees based on clinical pregnancy rates
as measure of effectiveness in the cost-effectiveness analysis are illustrated below in figure 3 and 4.
The two sources for the clinical evidence on the clinical pregnancy rate obtained are Thurin et al.
(2004) as well as the Cochrane review by Pandian et al. (2004).
FIGURE 3 Decision tree based on evidence on the clinical pregnancy rate with SET and DET from Thurin et al.
(2004).
FIGURE 4 Decision tree based on evidence on the clinical pregnancy rate with SET and DET from Cochrane review
by Pandian et al. (2004).
Similar models and decision trees, not shown here, were designed with respect to deliveries and
child born as measures of effectiveness in the cost-effectiveness analysis.
6.5.2.3 Cost-effectiveness result
Using the decision tree the cost-effectiveness of SET versus DET was analysed for the three measures
of effectiveness: cost per clinical pregnancy, cost per delivery and cost per child born from IVF.
Tables 35 and 36 show the cost-effectiveness results using the different effectiveness data.
69
The first cost-effectiveness result as presented in Table 35 is based on the evidence from the newly
Scandinavian study by Thurin et al. (2004).
TABLE 35
Cost-effectiveness of eSET versus DET – effectiveness based on
Thurin et al. (2004)1 (DKK)
Cost per clinical pregnancy
Cost per clinical pregnancy, including frozen embryos (FER)2
Cost per delivery3
Cost per delivery3, including frozen embryos (FER)
Cost per child born
Cost-effectiveness ratio
eSET
DET
131,446
115,321
114,842
115,321
149,833
120,324
125,172
120,324
148,204
93,265
1. Results in parentheses are excluding production lost during the IVF procedure.
2. Cost of IVF using frozen embryos are included.
3. The cost of the one triplet in Thurin et al. (2004) is set equal to the cost of a twin in
the analysis.
As can be seen from the Table the cost per clinical pregnancy is lower for the DET strategy
compared with the elective SET strategy. Including the clinical pregnancies from frozen embryos
in Thurin et al. (2004) made the cost per clinical pregnancy almost equal. The same result is found
for cost per delivery with the DET strategy having the lowest costs per delivery. The cost-effectiveness ratio decreases measured in terms of children born. The cost per child born with DET is only
two-thirds of the costs of having a child born with eSET. However, the DET-policy is both more
effective and involves higher total costs than the eSET-policy.
Table 36 below show the cost-effectiveness based upon effectiveness data from Pandian et al. (2004).
TABLE 36
Cost-effectiveness of SET versus DET – effectiveness based on Pandian et al. (2004)1 (DKK)
Cost per clinical pregnancy
Cost per delivery
Cost per child born
Cost-effectiveness ratio
SET (DKK)
DET (DKK)
125,050
114,548
140,899
127,661
134,026
94,066
1. Results in parentheses are excluding production lost during
the IVF procedure.
Using the effectiveness data from this study in the decision tree does not change the cost-effectiveness result with the DET-policy as having a lower cost-effectiveness ratio than the SET-policy. It is
still the case that a DET-policy is more effective in terms of clinical pregnancies and deliveries per
woman treated, but that the costs per treated woman are higher.
Overall the cost-effectiveness ratios, no matter which clinical source is used, are rather close to each
other for SET and DET. The largest difference is found for cost per child born. The cost per
delivery with DET is around DKK 120,324-DKK 127,661 compared with DKK 140,899-DKK
149,833 for SET, which means that in all cases the cost per delivery with DET is lower than with
SET. Furthermore, it illustrates that the results are rather robust, as the cost results using the three
clinical sources of evidence are close to each other.
The extra costs paid today in the health care sector and in the society due to the higher total costs
per woman treated, including costs for antenatal care, delivery, neonatal care and production lost,
with the use of DET in routine IVF practice is therefore around DKK 69,313 and DKK 82,502
per extra delivery obtained or DKK 47,257 and DKK 50,065 per extra IVF child born using DET
instead of SET in IVF.
70
6.6 Sensitivity analyses
Uncertainty with respect to parameters, assumptions made, etc. is always present in health economic
analyses. Sensitivity analyses should be performed in order to handle the uncertainty and to assess its
impact on the result as well as the robustness of the analysis and its conclusion (Briggs et al. 1999).
In the present analysis four parameters and assumptions were changed and their consequences upon
the conclusions regarding cost-effectiveness were investigated. First, all the assumption that the costs
of IVF pregnancies and children are equal to IUI pregnancies and children, as assumed in the
baseline analysis with the inclusion of the 23 IUI couples as part of the 213 enrolled couples in
the study, was tested. Secondly, the impact of neonatal intensive care unit upon the total costs and
cost-effectiveness was investigated. Thirdly, the impact of including cost of the risk of neurological
sequelae like cerebral palsy, which is increased in IVF children. Finally, the consequences of increasing or decreasing the production lost, as well as the direct costs with both 30% were investigated
in separate sensitivity analyses. Each of these sensitivity analyses is presented in sections below.
6.6.1 IUI couples excluded from the analysis
For the baseline analysis some IUI couples where enrolled in the study together with the IVF
couples at the fertility clinic. In total 23 IUI couples were enrolled as part of the 213 couples
enrolled in the study in total. Most of them did result in a singleton pregnancy and only four
resulted in twin pregnancies. In sensitivity analysis it was therefore investigated how it influenced
the result, if only the cost-effectiveness analysis was based on the 190 IVF couples.
The total cost per pregnancy increases a bit for singleton pregnancies, if IUI pregnancies are left
out, whereas the cost per pregnancy decreases about one-fifth for twin pregnancies (from DKK
50,380 to DKK 42,005). The reasons to this is especially a decrease in neonatal care costs and
production lost, as one of the IUI deliveries did involve high neonatal intensive care costs and an
associated high rate of production lost. As there were only four IUI couples with twin pregnancies,
this tendency to higher IUI costs should, however, not be generalised.
The cost-effectiveness result does not change, although the cost per delivery and cost per child
decreases a bit for DET (Appendix 3).
6.6.2 Neonatal Intensive Care
In case her child was hospitalized at the neonatal intensive care unit the women was asked in her
diary to report for how many days. The data showed that a twin child had a higher risk of needing
hospitalization after delivery than a singleton IVF child, and that the number of inpatient days at
the neonatal intensive care unit was higher for the twin child than for the singleton. For the average
twin child the number of inpatient days at the neonatal intensive care unit during the first three
months after delivery was 6.2 days compared with only 0.77 days on average in case of a singleton.
Therefore days at the neonatal intensive care unit explained some of the (extra) costs of being born
as a twin IVF child.
In a previous study in the county of Funen in Denmark the number of inpatient days at the
neonatal intensive care unit for IVF children was recorded for a period of six months in 1996
(Westergaard 1998). He found that the median inpatient days at the neonatal intensive care unit
for a singleton IVF child was 8 days, whereas it was 20 days for a twin IVF child. These figures
were used in a previous retrospective and model-based health economic analysis of IVF, where the
average cost due to hospitalization at the neonatal intensive care unit for a singleton IVF child was
DKK 9,493 and DKK 36,513 for a twin IVF child (Ingerslev et al. 2001).
As the inpatient data found by Westergaard (1998) are higher than the similar ones prospectively
collected in the present study, the consequences on the cost-effectiveness of using these inpatient
days for the child were investigated in a sensitivity analysis.
71
The total cost per delivery and the total cost per child increases both for singletons and for twins.
However, the increase is largest for a twin IVF delivery, which increases from DKK 100,760 to
DKK 144,436, which shows the impact of the higher number of inpatient days at the neonatal
intensive care unit (from 6 days on average to 20 days (median)).
The cost-effectiveness ratios do also increase a bit – both with respect to cost per delivery and cost
per child, when using these higher neonatal intensive care costs (Appendix 3). However, the costeffectiveness of the DET-policy with a lower cost per child born does not change in this sensitivity
analysis.
6.6.3 Cost of cerebral palsy
It is known that IVF children have a higher risk of having neurological sequelae, i.e. cerebral palsy,
compared with natural controls, and that the risk is slightly higher for IVF twins than for IVF
children born as singletons. Strömberg et al. (2002) found in a Swedish long-term study his to be
the case for 12 IVF singleton children out of a total of 3,183 IVF singleton children (0.0038%)
compared with 15 IVF twin children out of a total of 2,014 IVF twin children (0.0074%). In the
natural control group the similar risk figures was 0.0016% for singletons and 0.0069% for twins,
which for both groups represents lower risks than following IVF, although still with the highest risk
for twins. In a more recent Danish study Pinborg et al. (2004) found the risk of cerebral palsy to
be 0.0032% for IVF twins and 0.0025% for IVF singletons, which was a bit lower than in the
study by Strömberg et al. (2002).
To investigate the effect on the cost-effectiveness result of including the risk of cerebral palsy in the
calculations the risk proportions found by Pinborg et al. (2004) was used. The cost of having a
cerebral palsy was set equal to the Danish DRG charge (DRG-nr. 2640) – neurological rehabilitation of severely brain-damaged children (without complications) of DKK 376,809. The extra cost
per average child is, however, low due to the relatively low risk of having cerebral palsy.
The total cost per delivery and the cost per child almost doubles for both types of pregnancy
(singleton or twin), when the cost of cerebral palsy is included. Similarly the cost-effectiveness
results increases. The cost per child born does also increase, especially in the DET strategy, where
the cost per child born doubles compared with the base-case analysis, where the cost of cerebral
palsy was left out from the calculations. In SET the cost per child born on increases with onefourth, when the cost of cerebral palsy is included.
Based on the results of the sensitivity analysis the conclusion of the cost-effectiveness analysis does
therefore not change, despite inclusion of the costs of cerebral palsy, but the cost per delivery or
per child born increases.
6.6.4 Production lost changed
The women’s days away from work due to illness in the pregnancy were recorded in the diaries.
However, to investigate the influence of the production lost upon the cost-effectiveness result this
was evaluated in sensitivity analyses by either decreasing or increasing the production lost with
30%. The cost-effectiveness ratios are shown in Appendix 3.
As expected the cost per delivery and the cost per child falls when the production lost is decreased
by 30% – around DKK 7,000 below baseline in SET and around DKK 8,000 in DET. As also
expected the cost per delivery and cost per child increases above the baseline results, when production lost is increased with 30%. The relation between SET and DET does, however, not change,
and DET has still the lowest cost-effectiveness ratios.
6.6.5 Direct costs changed
Similarly as with production lost the influence of changes in the direct costs upon the cost-effectiveness ratios was investigated by either decreasing or increasing the direct costs by 30% respectively.
72
These scenarios (∫30% in direct costs) are shown in Appendix 3. The cost-effectiveness result
decreases, when the direct costs are decreased by 30% and increases, when the direct costs are
increased with 30%. However, this change in the direct costs does not influence the overall conclusion of the cost-effectiveness analysis.
6.7 Discussion
This chapter has evaluated the health economic consequences of using either a SET-policy or the
traditional DET-policy. This was done in a modelling study and the prospective collection of the
cost of having either an IVF-singleton or an IVF-twin child based on the inclusion of 190 IVF
couples and 23 IUI couples visiting the fertilization clinic at Aarhus University Hospital, Skejby
between April 2002 and March 2003. These couples were followed for one year to collect data on
the costs in the pregnancy period as well as three months after delivery.
As expected the costs of having IVF twins were higher than the costs of having an IVF-singleton.
However, in terms of children the costs were similar. Focusing on the two transfer policies the
advantage then of SET is a low rate of the more expensive multiple pregnancies opposed to DET,
where the twin rate is around 25%.
Some IUI couples were included in the present analysis together with the IVF couples. This might
have increased the cost estimates for singletons and twins a bit, but a sensitivity analysis showed
that the cost-effectiveness result did not change with their inclusion, why they were kept in the
baseline analysis.
The present prospective estimation of the cost of singleton pregnancies and twin pregnancies has
only revealed a difference of a factor two. Lukassen et al. (2004) in their study found a five times
difference between singleton and twin pregnancies, i.e. s 2,549 per singleton pregnancy and
s 13,469 per twin pregnancy. The costs per twin pregnancy was around the same as found in the
present analysis, but the cost per singleton pregnancy was less than half of that found in the present
study. However, the study by Lukassen et al. (2004) did only include antenatal costs, delivery and
neonatal costs, and not productivity costs. In an older American study Callahan et al. (1994)
predicted that the total charges to the family in 1991 for a singleton delivery was $ 9,845 as
compared with $ 37,947 per twin delivery. This study covers all types of singletons and twins, and
not only singletons and twins from assisted reproduction, but comes closer to the result of the
present study.
However, in the present study the SET-policy did not show to be more cost-effective compared with
the DET-policy, which is both more effective (higher clinical pregnancy rate, higher rate of delivery
and children), but also more expensive (higher delivery cost and neonatal intensive care costs). However, the cost per delivery using SET was not significantly higher than the cost using DET.
Measured in terms of cost per child born there is a larger difference between the two transfer
policies. The reason to this is the extra children born as twin with the DET-policy, which more
than compensate for the extra costs following delivery and neonatal care. The sensitivity analyses
showed that the result of the analysis was robust.
In the literature a few health economic studies have compared SET with DET (or multiple-embryo
transfer). Sutter et al. (2002) found in a modelling study that the cost per child using SET and
DET to be rather equal (s 10,563 versus s 11,297), and marginally lower for SET. In a recent
prospective study Gerris et al. (2004) found the cost per child of SET was only half of the cost of
DET (s 4,700 versus s 8,613) due to lower neonatal costs with single-embryo transfer. These
results in favour of SET compared with DET differs from those found in the present analysis,
where the cost per child using SET was one-third higher than the cost per child using DET.
73
One reason to the lower cost per child with the DET-policy in the present study might have been
the size of the neonatal intensive care costs estimated on the basis of the prospective data collection
among the couples participating. However, this was further investigated in a sensitivity analysis,
where higher cost figures from a previous estimation of neonatal intensive care in the County of
Funen associated with IVF singletons and IVF twins was used instead. However, the use of these
higher cost figures for neonatal intensive care for twins did not change the cost-effectiveness result,
as the DET-policy was still cost-effective. Another limitations of the study is that only three months
resource use after delivery are included. Neonatal intensive care might last longer, as well as the
impact from severe diseases. A few of the estimated unit costs, such as midwife visits might be kept
at an unreasonable low level, as only labour costs are included.
IVF children, and especially IVF twins, has an increased risk of having neurological sequelae, i.e.
cerebral palsy, compared with natural controls. Cerebral palsy affects their quality of life and involves
high costs for the health care sector for those with the disease. In the present cost-effectiveness
analysis the costs of cerebral palsy was included in a sensitivity analysis. This analysis documented,
however, that the economic consequences of cerebral palsy have only minor influence upon the
cost-effectiveness of the average IVF child born. The reason to this is that very few children get
cerebral palsy.
The conclusion from this Danish cost-effectiveness analysis comparing a SET-policy with the traditional DET-policy used today is that the cost per delivery or the cost per child is higher using
SET compared with DET. However, DET involves higher total costs per woman treated, i.e. costs
for antenatal care, delivery, neonatal care and production lost due to the higher frequency of twin
pregnancies and deliveries, but at the same time also more effective. In the end it will be up to the
decision-maker and the patients to decide, which policy to be used.
74
7 Organisational consequences of SET
The aim of an organisational analysis is to identify various organisational dimensions of importance
for evaluation of the particular technology, i.e. the influence of the change in technology upon the
organisational processes, e.g. workload and patient flow, staffing, education. Furthermore, the analysis can be expanded to structure and management (Vrangbæk 2001).
The present organisational analysis was performed to elucidate consequences for process and structure following SET or eSET replacing or supplementing a DET policy in Denmark. Different
scenarios may be relevant in terms of public funding to IVF treatment in Denmark following an
obligatory SET or a more restricted policy of eSET.
The present analysis aims at elucidation of changes in organisational processes in terms of workload
and flow of patients (e.g. referrals) and secondly effects on staffing (resources and education).
This was done by using data from our own randomized study, previously published randomized
studies, clinical practice and empirical data. Secondly, possible effects of SET on the structure of
fertility treatment in Denmark are discussed based upon the present conditions for fertility treatment in the public health care system and in private clinics, respectively. Possible changes in the
national funding policy that could neutralize deterioration of the conditions for treatment for
patients are outlined, i.e. more reimbursed treatments to compensate for a lower average pregnancy
rate due to SET, possibility to have treatment for the second child to compensate for lack of
possibility to have two children by twinning. However, such considerations are hypothetical and
based upon assumptions.
7.1
The process
7.1.1
Treatment numbers
As previously described obligatory SET yield a lower pregnancy per embryo transfer than DET.
Elective SET may reduce the difference.
By measuring success rates accumulated as clinical pregnancies/deliveries coming from all fresh
transfers and frozen single embryo transfers (if any), derived from one hormone stimulation/oocyte
pick-up, the final result for the individual patient could be acceptable, but the pregnancy rate will
still be lower than by SET than by DET. Accordingly, the difference in cumulative pregnancy rate
between the two strategies would be equivalent to the difference between pregnancy rates by transferring one or two fresh embryos for those, who do not have frozen embryos. For those, who have
frozen embryo(s) the difference will be less, i.e. between pregnancy rates obtained following transferring a fresh embryo and a frozen, since in DET the two best embryos are both transferred freshly,
whereas in SET one is transferred fresh and the second frozen with a lower implantation potential.
Subsequent transfers of frozen embryos would give rise to an equal number of gestations. Accordingly, cumulatively SET of frozen thawed embryos should theoretically give rise to the same number
of children as DET, but fewer deliveries due to twinning. The overall result is that more treatments
should be offered if the average cumulative pregnancy rate for patients undergoing IVF should be
maintained.
More straws are to be frozen in SET than in DET, since only one embryo is to be used at a time.
Accordingly, more thaws are to be performed since only 50-60% of all embryos survive thaw, and
following thaw of one embryo, you have to continue with next until a surviving embryo is achieved
for transfer, whereas fewer straws are to be thawed to obtain at least one surviving embryo if two
embryos are frozen and thawed at a time. Finally, more transfers of frozen-thawed embryos are
necessary in case of SET.
75
It is difficult precisely to calculate what the specific consequences of these changes will be in respect
to logistics and need for resources, but estimates can be done.
The data from the RCT between elective SET and DET by Thurin et al. (2004) showed an
approximately 5% decrease in cumulative pregnancy rate between the eSET (one fresh and one
frozen and thawed embryo) and the DET group (two fresh embryos). It is reasonable to assume
that subsequent transfers of frozen and thawed embryos contribute only little to change this difference. However, the thawing and transfer of a single embryo at a time may result in slightly more
pregnant women and deliveries than thaw and transfer of two embryos, but the difference is small.
On the other hand the eSET group in the Thurin study was specifically defined by having two
high quality embryos. In case obligatory SET is implemented, the difference between SET and
DET may be closer to the success rate following transfer of one (33.6%) and two (52.6%) fresh
embryos (i.e. 19%), since only approximately 40% of all IVF cycles yield surplus embryos of
sufficient quality for freezing. Accordingly, taking results of freezable embryos into consideration, a
conservative estimate of the likely difference in pregnancy rates between DET and obligatory SET
could be 10%.
Thus to compensate for the lower cumulative pregnancy rate following SET a 5-10% higher number of hormone stimulations and oocyte pick-ups is probably needed, depending upon the kind of
SET policy implemented (SET to all, eSET following strict rules, recommended SET).
Since the Danish yearly number of hormone stimulations and oocyte pick-ups is around 10.000
(9598 in 2004) (http://www.fertilitetsselskab.dk/), the extra load will be in the order of 500-1000
more hormone stimulations and oocyte pick-ups, and fresh embryo transfers per year in Denmark.
Furthermore more embryos will be frozen each in one straw and consequently more transfers of
fresh and frozen embryos have to be performed. An estimate could be that the work load will be
doubled in respect to freezing, thawing and transfer of frozen embryos. In 2004 1853 transfers of
frozen embryos was performed in Denmark, i.e. 19% (1853/9598) of all treatment cycles result in
a frozen embryo transfer. Since only 50-60% of all embryos survive thawing, this figure represents
at least 3700 frozen embryos – and single straws in a SET strategy.
Elective SET of frozen embryos increase the number of transfers of frozen embryos with 50-100%
corresponding to 900-1800 per year, assuming that between 50 to 100% of frozen embryo transfers
is presently with two embryos. Again, these represent a qualified estimate.
A single child per treatment is not necessarily optimal for the infertile couple, as it may seem from
a medical point of view. Presently, only one child is offered in the public health care system in
Denmark, besides in the Counties of Storstrøm, Roskilde and Vestsjælland were treatment to a
second child is allowed. From the survey previously mentioned we know that only 3.8% of the
infertile patients would like to have only one biological child, 52.5% would prefer to have two
biological children, while 43.7% would want to be the mother or father of three or more biological
children. Since some of the infertile patients were secondarily infertile and had given birth to one
or more children previously (17%), it can be concluded that 79.2% would need treatment for a
second or more children.
In 2004, 5999 IVF/ICSI treatments were performed in the public health care system in Denmark
(http://www.fertilitetsselskab.dk/). A clinical pregnancy was obtained in 28% corresponding to 1719
couples achieving pregnancy. No data specific for the public health care are available in terms of
twin rates and results of frozen embryo transfers. The twinning rate in the national data was in
2004 was 22% following fresh transfers. Accordingly around 378 ongoing twin pregnancies were
initiated. In addition, around 26 twin pregnancies were the result of transfer of frozen and thawed
embryos. Thus, the total number of couples with a twin pregnancy following treatment in the
public health care system in 2003 was approximately 404. The rest (1315) achieved one child only.
It can be assumed that 79.2% or 1041 of these wants a second child. A minority of these (a guess
76
could be 10% or 132) have frozen embryos to transfer for a second child, in most cases for one or
two transfers only (1.5). With a pregnancy rate of 20% per transfer and 1.5 embryos available
approximately 36 couples get pregnant in a DET or a SET situation. Eighteen extra transfers of
frozen and thawed embryos can be expected in a SET situation. Under these circumstances 1005
couples having one child only following treatment in the public system may seek renewed IVF
treatment in the private sector in Denmark for a second child in a DET situation. In an obligatory
SET strategy, this number will increase to 1383 with those 378, who would have had twins with
DET, corresponding to an increase of 27% of couples wanting child number two.
An increase of 378 couples wanting a second child is not easy to convert to a precise number of
cycles, since no hard data exists on how many treatments these couples need to achieve a second
child or to refrain from further treatment. A qualified guess could be two treatments. Accordingly,
the increase in number of IVF treatments could be 756 per year in Denmark to obtain a second
child if SET is practised instead of DET. This would be a significant increase (8%) of the yearly
IVF/ICSI cycle number (9,598) in Denmark. According to the present funding policy, treatment
for child number two is not allowed in the public health care system. The 756 treatments represent
an increase of 22% of the yearly number of cycles in the private sector (756/3,398).
In conclusion, introduction of a SET policy is estimated to have the following consequences in
Denmark, supposing maintenance of an unchanged cumulative pregnancy rate by increased number
of cycles offered to compensate for a lower pregnancy rate in SET and still only treatment for the
first child:
a.
b.
c.
d.
e.
500-1000 more hormone stimulations, oocyte pick-ups, and fresh embryo transfers per year
Freezing of 900-1800 extra straws containing a single embryo
900-1800 more transfers of frozen embryos per year
18 more transfers of frozen embryos per year to obtain a second child
The distribution of this increased activity between the public and private sector can be assessed to
be in proportion with the present distribution of IVF/ICSI performed yearly in the two sectors in
2004, i.e. 64.6% and 35.4% (http://www.fertilitetsselskab.dk/) except for (d) which is a public
health care service and point (e) which is exclusively performed in private clinics.
7.2
Estimates of economic consequences of SET
7.2.1 Public health care clinics
In Denmark there are 9 fertility clinics in the public health care system and 12 private. The
proportion of treatments performed yearly in the two sectors in 2004 was 64.6% and 35.4%,
respectively. This means that corresponding to the impact of a SET policy outlined above (a-e),
between 323 and 646 extra full IVF treatments are needed in the public health care system with
handling of 581-1162 extra straws and a similar number of transfers of frozen and thawed embryos.
This increase in activity corresponds to around 3% and 6% in terms of IVF cycles and 50-100%
increase of freezing and thawing procedures with derived transfers. Such an increase in activity is
probably not possible without extra staffing, and funding of medication and other articles. Using
the cost estimated of a IVF cycle used in the health economic analysis (DKK 15,215, Ingerslev et
al. (2001)) 323-646 extra IVF cycles cost between DKK 4.9-9.8 mio.. Freezing, thawing and
transfer of 900-1800 extra frozen and thawed embryos at a price of (DKK 5,877, Ingerslev et al.
(2001)) each can similarly be estimated to cost between DKK 5.3-10.6 mio.. In total, the extra
expenses in the public health care system can be estimated to be in the order of DKK 10.2-20.4
mio.
77
7.2.2 Private fertility clinics
Corresponding to the impact of a SET policy outlined above (a) between 177 and 342 extra IVF
treatments are needed in the private sector with handling of 319-638 extra straws and a similar
number of transfers of frozen and thawed embryos.
The estimate was that 378 couples would obtain a single child by SET instead of twins by DET.
They would need 756 more hormone stimulations and oocyte pick-ups, and fresh embryo transfers
per year to obtain a second child in private fertility clinics. It was estimated from the national data
that 19% of fresh cycles result in a frozen embryo transfer. Accordingly, with a DET policy 144
frozen embryo transfers would be derived from these 756 treatment cycles, but with a SET policy
for frozen embryos this figure is 50-100% higher or 216-288.
In the private sector the total increase associated with an obligatory SET policy will be in the order
of 933-1098 extra IVF cycles and 535-926 extra frozen embryo transfers.
The average price of IVF treatment per cycle in private clinics in Denmark is in 2005 around
DKK 18,000, excluding hormone medication, which are paid by the National Health Insurance
(reimbursement) and the patient them selves. The price does neither include ICSI, which has an
average price in the private market of DKK 3,300 in 2005-prices. The average price per IVF-cycle,
including medication, but excluding ICSI, in private IVF clinics is therefore around DKK 24,000
(compared with the DKK 15,215 for a public clinic). The average price for a frozen embryo transfer
(including freezing) in private clinics in Denmark is in the order of DKK 7,000.
Using this estimate of a price for a IVF treatment in the private fertility clinics, the cost for 9331098 extra IVF cycles will be DKK 22.4-26.4 mio. per year and for 535-926 extra frozen embryo
transfers DKK 3.7-6.5 mio.
In the private clinics the patient pays for most of these expenses. However, expenses for hormone
stimulation used in IVF treatments in are estimated to represent DKK 6,000/DKK 24,000 or 25%
of these expenses, corresponding to DKK 5.6-6.6 mio. In 2004 the reimbursement rate by the
National Health Insurance in 2004 for the whole group of IVF hormone pharmaceutical drugs
(G03G) was 84% (Source: Register of Medicinal Product Statistics, the Danish Medicines Agency
(personal communication with Jesper Winther Koch, Danish Medicines Agency)). This does, however, not include any reimbursement made at the municipality level. With a reimbursement rate of
84% for IVF hormone pharmaceutical drugs by the National Health Insurance then the public
health care sector will pay DKK 4.7-5.5 mio. extra for medication to child number two in private
clinics in case of SET is introduced in Denmark. Frozen embryo transfers are performed either
without hormone treatment or with negligible expenses.
Accordingly, an estimate of economic consequences in case SET is introduced in Denmark is that
the public health care system will carry an extra burden of around DKK 14.9-25.9 mio. in case an
unchanged outcome of IVF treatments should be maintained in terms of chance of achieving one
child in the public health care system and access to treatment for a second child in the private
system. However, these estimated should be evaluated with care. An increase in number of treatment
cycles of 10% in established clinics may very well represent a marginal increase without consequences for staffing, buildings etc. In such case the extra expenses to the public health care sector
associated with SET may be halved to DKK 7-13 mio.
7.2.3 Savings in expenses to twin pregnancy, delivery and neonatal care following SET
SET may in principle reduce the twinning problem associated with IVF to zero if practised consequently to all patients. Elective SET may reduce the twinning rate by a proportion which depends
upon selection criteria for those who receive SET and especially the proportion of the total population this includes. Denmark had a SET frequency of 18.4% in 2002 (21.8% in 2003) and a
subsequent twinning rate in IVF pregnancies of 23.1%, whereas in Finland in 2002 one embryo
78
was transferred in 38.7% with a twinning rate of 15.2%, and in Sweden in 2003, 58.5% had SET
with a twinning rate of 11.7. Accordingly, the twinning rate can be reduced by at least 50% using
elective SET.
In Denmark, 450 twins were born in 2003 following IVF/ICSI, FER or oocyte donation.
Previously, we have calculated the extra expenses associated with pregnancy, delivery and neonatal
care in twins (DKK 100,760) compared to singletons (DKK 47,787) to DKK 52,973. In case an
obligatory SET policy is introduced the twin rate following these procedures is falling towards zero.
In such case the resulting saving amounts DKK 23.8 mio. per year. If eSET is used, it may reduce
the twinning rate by 50% and savings will be less, DKK 11.9 mio. per year.
7.2.4 Economic balance in case of SET
Based upon the calculations above, implementation of SET will cost the public health care system
between DKK 14.9 and 25.9 mio. yearly, depending upon what the conditions will be for introduction of SET. The savings in respect to reduced obstetrical and neonatal care represent between DKK
12 and 24 mio. Based upon these assumptions, introduction of SET seems to represent a change
in the public expenses associated with IVF between a saving of DKK 9.1 mio. (best case) and
increase in expenses of DKK 13.9 mio. (worst case).
7.2.5 Considerations on public funding of IVF in Denmark
Presently public funding covers IVF treatment to couples fulfilling the following criteria: 1. Female
age below 40 years, 2. No previous biologically shared children, and 3. Indication for treatment.
Three embryo transfers are funded. Cancelled cycles do not count, unless the number of cycles started
exceeds 5. Transfer of frozen and thawed embryos derived from the oocyte pickups is not limited.
In a situation where the cumulative chance of delivery within a limited number of treatment cycles
is decreased by regulations as to number of embryos in order to avoid twins to reduce morbidity
and mortality and to save the expenses related to this, one may find arguments to expand the public
funding to more cycles to keep the cumulative probability for delivery unchanged.
Another point is, that today two children (twins) are provided to 25-30% of the couples covered
by public funding. These couples could feel a significant deterioration of the conditions for treatment in case they are imposed SET. An expansion of public funding to treatments for child number
two could represent a compensation for that – and may help to decide on SET in a situation,
where the couples are responsible for the choice between SET or DET.
However, the savings associated with a reduction in twin rates may not fully finance either of these
suggestions.
7.3
Staffing and education
In case SET is introduced, extra staffing is necessary, and specialized education of new staff members
will be necessary. This can be realized within six to twelve months.
7.4
Communication and culture
SET may be introduced as a voluntary possibility with a specified obligation to inform the patients
about twin risks. Another possibility is that the Government or National Board of Health defines
more or less specific criteria for DET and SET. A third strategy could be obligatory SET, prohibiting
DET.
79
Only the last situation will not give rise to challenges in terms of communicating risk data concerning twin pregnancy and delivery to the patient, leaving to her/them to decide if one or two embryos
should be replaced to the uterus with a risk/chance of twins of 25% on average.
Parents’ decision-making about the number of embryos to be replaced is often influenced by their
desire to have more than one child, by their appraisal of the potential risks and, to a lesser extent,
by the financial pressures created through restricted access to public funding of infertility treatment.
It is the physician’s responsibility to inform parents fully about the appropriate treatment for individual cases and the risks associated with multiple pregnancies, minimizing the potential for conflict,
for example about the number of embryos to transfer. Thus, the physician should act with responsibility to the potential child rather than through a paternalistic attitude towards the potential parents
(Shenfield 2003).
Communicating risk data represent a challenge to the medical profession (Edwards 2003, Godolphin 2003). The information on risks associated with twin pregnancy and delivery should be
communicated nondirectively. Knowing and understanding the frequency of an event in a population provides no certainty for individuals – only a guide to be used according to their own circumstances, values, and preferences (Thornton 2003). The business of enabling patients to understand
risk so that they might incorporate it into their decision-making process is fraught with difficulties.
It goes without saying that health practitioners need the knowledge, skills, confidence, communication skills, and the decision aids to provide this essential component of shared decision making
(Thornton 2003).
Nevertheless, data from several sources may indicate a reluctance to SET, despite type of information
given or compensatory reimbursed cycles. A randomized study between ‘‘standard clinic pack about
IVF’’, a leaflet informing on twin pregnancies and a leaflet group having a short discussion with a
nurse revealed a surprisingly lack of influence on views on SET. Neither content or method of
information encouraged patients to favour limiting the number of embryos transferred (Murray et
al. 2004). However, Child et al. (2004) found that patients recognizing increased risk of multiple
pregnancy were significantly less likely to want this outcome than those who did not. The fact that
only around 25% of patients invited to participate in a randomized study between DET and SET,
despite the SET group were offered an additional reimbursed cycle, indicates that even limited
compensation does not make SET an attractive option.
Specific risk estimates means clarity, not odds ratios or similar indications of a relative proportion
of increased risk. Nationally standardized written information to patients combined with some type
of tutorial meeting with either the patient couple individually or in groups could be helpful in this
process. Accordingly, any non-obligatory SET policy will impose new or enhanced tasks of informing the infertile couple on risks associated with twin pregnancy and delivery against consequences for their chances to obtain pregnancy to help them decide to have one or two embryos
transferred. This represents undoubtedly a change in culture and priority of the professionals treating the patients.
A draft to written information to patients is shown in appendix 1 and in Danish in appendix 2.
7.5
Discussion and conclusions on organisational perspectives
The present analysis was based on facts in many aspects, but assumptions were involved in many
of the estimates. Therefore, the conclusions should be taken with some caution. The various RCTs
and observational studies indicate SET strategy of any kind will have an inevitable negative impact
on the chance of pregnancy for the individual patient. The reduced success rate will increase number
of treatments necessary to reach an unchanged cumulative pregnancy rate. However, the recent
national Swedish data are against such an expectation (Bergh et al. Läkartidningen 2005, in press).
80
Nevertheless, handling of single embryos in individual straws for freezing, thawing and transfer
increases the overall workload. These consequences will lead to increased costs for infertility treatment. Furthermore, the fact that most couples aim at more than one child and 25% achieved two
children at a time (twins) by SET, more treatments will be necessary for child number two (and
more). A n estimate of the economic consequences in case SET is introduced in Denmark was that
the public health care system will carry an extra burden of up to DKK 14.9-25.9 mio. in case an
unchanged outcome of IVF treatments should be maintained in terms of chance of achieving one
child in the public health care system and access to treatment for a second child in the private
system. The increased costs to IVF treatments following SET can be counter-balanced by savings
caused by reduced health care expenses associated with twin pregnancy, delivery and neonatal care –
and long term costs. The proportion of these savings depend upon the reduction in twin rates
following IVF, which depend upon the SET policy chosen. Denmark had a SET frequency of
18.4% in 2002 (21.8% in 2003) and a subsequent twinning rate in IVF pregnancies of 23.1%,
whereas in Finland in 2002 one embryo was transferred in 38.7% with a twinning rate of 15.2%,
and in Sweden in 2003, 58.5% had SET with a twinning rate of 11.7. Accordingly, the twinning
rate can be reduced by at least 50% using elective SET.
In case an obligatory SET policy is introduced the twin rate following IVF is falling towards zero.
In such case the resulting saving amounts DKK 23.8 mio. per year. If eSET is used, it may reduce
the twinning rate by 50% and savings will be less, DKK 11.9 mio. per year. Based upon these
assumptions, introduction of SET seems to represent a change in the public expenses associated
with IVF between a saving of DKK 9.1 mio. (best case) and increase in expenses of DKK 13.9
mio. (worst case). However, expansion of public funding to treatments for child number two as a
compensation for the lost chance of having two children in one reimbursed treatment was not
considered in this context.
Any eSET policy will create increased challenges to information of patients about the perspectives
of the choice between one and two embryos.
81
8 Discussion and conclusion
The issue of this report was to throw light on a question which has several shareholders. The answer
to the question whether one or two embryos should be transferred in IVF treatments is relevant to
the infertile couples wanting to decide on their treatment to obtain the outcome they prefer and
with the best possible success rate and to the doctors responsible for the treatments. Obstetricians
taking care of the pregnant woman and the paediatricians responsible for the newborn are involved,
and finally administrators and politicians are interested because of the implications the decision on
this question has for allocation of resources.
The HTA represents an overview that integrates evaluation of the technical factors, patient-related
perspectives, and economic and organisational consequences of choosing between two techniques.
The background for the present HTA was that the twin rate in Denmark has increased significantly
(2.4 fold) during the last decades as in most European countries. The focus of the present analysis
was IVF and ICSI treatments which seem to be responsible for one third of the increase, while
other types of infertility treatment have caused another third, and increasing age of women at
establishment of family explaining yet another third part (Bergh et al. 1999). Accordingly, intervention against the practise of two embryo transfers may reduce the increase by one third at the
most.
In Denmark the 2004 figures for IVF/ICSI show that 27.4%, 67.5% and 5.2% were single, double
and triple embryo transfers, respectively. A total of 2056 clinical pregnancies treatments were established with 72.7% singletons, 26.4% twins, and 18 (0.9%) triplets. In terms of expected delivered
children, these figures correspond to a total of 3176 children with 1084 singletons, 1084 twins and
36 triplets (http://www.fertilitetsselskab.dk/).
Elective single embryo transfer (eSET) means that SET is offered to a selected group of patients with
a better than average chance of achieving pregnancy – and twin pregnancy in case of DET. All published trials comparing SET with DET have been studying selected groups of ‘‘good prognosis’’ patients. However, it seems difficult by means of prediction models to identify a group of patients with a
predictable high chance of pregnancy and risk of twins without negative consequences for pregnancy
rates on one side and limited effects on total twin rates on the other. Thus the published randomized
studies (RCT) showed all a significantly lower pregnancy rate following eSET compared to DET, but
subsequent transfer of a frozen embryo (if any) may reduce this difference to an insignificant level
(Pandian et al. 2004, Thurin et al. 2004). Thus, in patients without a freezable surplus embryo the
pregnancy rate could be reduced from a level of 48% to 31%. Somewhat surprisingly, the RCT presented in this HTA involving randomisation between SET and DET of an unselected group of infertile
patients at their first treatment cycle did not show a difference in clinical pregnancy rates between the
two groups. The best possible explanations for this could be random variation, or an unquantifiable
difference in embryo quality in the two groups.
Observational data from Belgium and Finland indicate that very satisfying pregnancy rates can be
maintained despite a high rate of single embryo transfers (Tiitinen et al. 2003, De Sutter et al.
2003, Gerris et al. 2002, 2004). Recent national data from Sweden where legal rules have been
introduced, have imposed a SET rate of 60% with a twin rate of less than 10% and an unchanged
delivery rate (Bergh 2005). Improved identification of the best embryo or lack of otherwise expected
increase in pregnancy rates may explain such data.
Denmark had a SET frequency of 18.4% in 2002 (21.8% in 2003) and a subsequent twinning
rate in IVF pregnancies of 23.1%, whereas in Finland in 2002 one embryo was transferred in
38.7% with a twinning rate of 15.2%, and in Sweden in 2003, 58.5% had SET with a twinning
rate of 11.7%. Accordingly, the twinning rate can be reduced by at least 50% using elective SET.
82
Twin pregnancy and delivery is associated with increased risks to the mother and the foetus. Preterm
delivery and complications associated to that is the dominating problem, but also other disorders
such as preeclampsia (OR 2.4) complicate twin pregnancies. These complications results in an
increased consumption of health resources in the antenatal period, associated with delivery and
during the neonatal period.
Caesarean section (CS) rates are considerably higher in IVF twin than singleton pregnancies with
2-3-fold increased relative risks from about 50% to 20% in singletons with considerable variations
between the countries (Dhont et al. 1999, Westergaard et al. 1999, Klemetti et al. 2002, Koivurova
et al. 2002b, Pinborg et al. 2004c).
Overall, the risk of preterm delivery and low birth weight in IVF pregnancies is higher than in the
general population (Bergh et al. 1999, Dhont et al. 1999, Westergaard et al. 1999, Schieve et al.
2002). Meta-analyses have shown that IVF singletons carry a higher risk of preterm delivery and
low birth weight than spontaneously conceived singletons (Bergh et al. 1999, Schieve et al. 2002,
Helmerhorst et al. 2004, Jackson et al. 2004). However, the predominant contributor is a higher
twin birth rate. Nevertheless, the obstetric outcome is similar in IVF vs. spontaneously conceived
twins (Dhont et al. 1999, Helmerhorst et al. 2004, Pinborg et al. 2004b). A lower morbidity
explained by a lower rate of monochorionic twins following IVF (1-2%) compared to spontaneous
conceptions (20%) is possibly at play (Sebire et al. 1997, Loos et al. 1998, Derom et al. 2001).
The occurrence of singleton deliveries starting as a twin pregnancy (‘‘vanishing twins’’) (10.4% of
live-born IVF singletons) may contribute to the increased risk of IVF singleton pregnancies compared to naturally occurring pregnancies (Pinborg et al. 2004e).
IVF twins are born with an average gestational age three weeks earlier than IVF singletons and with
a mean birth weight about 1000 g lower (Pinborg et al. 2004c). In the Danish register study the
age- and parity adjusted odds ratio of birth ∞37 completed weeks was 10-fold increased (OR 9.9,
95%CI 8.7-11.3). The crude percentages of children born ∞37 weeks of gestational age were
43.9% in IVF twins and 7.3% in IVF singletons and odds ratio of birth ∞32 completed weeks
was increased 7-fold (OR 7.4, 95%CI 5.6-9.8) crude frequencies being 8.5% in IVF twins vs. 1.3%
in IVF singeletons (Pinborg et al. 2004c). The increased prematurity rate results in need for intensive care. The Danish national twin cohort study revealed that IVF/ICSI twins had a 3.8-fold
increased risk of admittance to neonatal intensive care unit (NICU) compared with IVF singletons
(56.4% vs. 25.4) and IVF twins spent on average 9 days more in NICU than singletons (Pinborg
et al. 2004c). Moreover, perinatal mortality in IVF twins was twice as high as in IVF singletons;
20.7 vs. 11.0 per 1000 (Pinborg et al. 2004c).
The prevalence rates of neurological sequelae and cerebral palsy seem to be similar in IVF twins,
control twins and IVF singletons (Pinborg et al. 2004d), but higher than in spontaneously conceived
singletons (Strömberg et al. 2002, Lidegaard et al. 2005). The ratio of cerebral palsy in IVF vs.
non-IVF singletons was 1.8 (1.2-2.8) (0.33% vs. 0.19%) (Lidegaard et al. 2005). Again the phenomenon of vanishing twins may explain the increased risk in IVF singletons compared with
spontaneously conceived singletons. Pinborg et al. (2003a) found that special needs (ergo or physiotherapy, speech therapy or a special remedial teacher) were present in significantly more IVF twins
than singletons (9.9% vs. 6.1%) and speech therapy was provided to 6.4% vs. 3.2%. Furthermore,
Strömberg et al. (2002) revealed an increased requirement of treatment in childhood disability
centres in children born after IVF/ICSI compared to controls, even among singletons.
Overall, these figures show increased risks of serious immediate complications associated with twin
pregnancy and delivery, but also in the long term. As a minimum these data should be considered
together with the infertile couple when planning infertility treatment. The practise and level of
information presently offered to the infertile patients concerning the twin question is hardly in
concordance with the principle of informed choice. Although most patients in the present qualitative study and survey on patient’s attitudes to the twin question indicated that they were happy
83
with the information given, there were strong indications that more information is wanted by the
patients and that the information given is rather unspecific.
Formal guidelines describing the type and extent of information and necessity of counselling by a
doctor or a nurse may ensure a proper and qualified informed choice.
Both patient studies in the present report showed that the majority of the patients and their partners
(58.4%) prefer twins to having one child at a time, while only a small fraction claimed to be
indifferent. A larger proportion (78.5%) had decided to have to embryos replaced in the next
treatment, only small fraction (6.2%) wanted SET. This means that around 20% use DET to
optimise their chance to get pregnant and accepts the twin risk. Perceived physical and psychological
stress was important for the decision between one or two embryos. The present proportion of
couples preferring twins seems high. Other studies have shown lower proportions of fertility patients
preferrering twins (14%, Kalra et al. 2003, 20.3%, Ryan et al. 2004, 38.9%, Child et al. 2004 and
32%, Murray et al. 2004, respectively).
Despite the widespread desire for twins there are reasons to believe that more specific information
may change the decisions. Both patient studies indicated that the extent of information about twins
could be better. Nearly half of those who preferred to have one child at a time were worried
about maternal or foetal risks or pregnancy complications, indicating that risks have importance
for preferences. It is likely, that more specific risk information may increase worries among the
couples concerned. The respondents’ response to three different risk scenarios showed that the
desire for twins decreased with increasing risk and differently depending upon wording of risk vs.
chance. Attitudes towards eSET seemed independent of methods of information provision in a
small randomized study by Murray et al. (2004), but the present survey showed that the large
proportions of patient preferred to have oral information by a nurse or a doctor at the Fertility
Clinic or through a leaflet. In the present qualitative study there were indications that risk information on twins was received, but subsequently interpreted in such a way that it was contained
and did not pose a risk for the overall wish and motivation of the couples.
Another question is if a reduction of the national twin rate is a goal in itself. The costs associated
with twin pregnancy, deliveries and neonatal care may urge such a policy. A paternalistic attitude
based upon a view that patients should not – or cannot – decide upon the choice between one or
two embryos themselves may be incentive to an obligatory SET policy or eSET to strictly defined
groups of patients. However, the present results strongly indicate that an obligatory single embryo
policy would be in conflict with patient interests and wishes.
The dilemma between SET and DET represents a trade-off between effectiveness and costs. A SETpolicy will expectedly result in a lower pregnancy rate per transfer of fresh or frozen embryo(s) and
possibly also cumulatively, as well as fewer children born. However, SET reduces the twin pregnancy
rate resulting in lower costs in the health care sector due to a reduced need for extra monitoring
during pregnancy, less complicated deliveries and less need for neonatal intensive care.
The total costs of a singleton or twin IVF pregnancy and the three-month period post partum and
child(ren) were estimated on the basis of the prospectively collected resource consumption data
(cost diaries). As expected the costs of having IVF twins were higher than the costs of having an
IVF-singleton. The total cost per singleton pregnancy was DKK 47,787, whereas the total cost per
twin pregnancy was twice as high – DKK 100,760. However, in terms of children the costs equals.
The major reasons to this difference were significantly higher antenatal and neonatal costs per
pregnancy, as well as more production lost in the twin pregnancy group. The average cost per
pregnancy for neonatal care intensive hospitalization was four times higher for twins compared with
singletons. When considering the average pregnancy costs per child born, the costs of a twin came
close to the cost (actually marginally lower) of a child from a singleton IVF pregnancy.
84
DET-policy, which is both more effective (higher clinical pregnancy rate, higher rate of delivery
and children), but also more expensive (higher delivery cost and neonatal intensive care costs). The
extra cost paid today using the more effective DET-policy is around DKK 69,000 and DKK 82,000
per extra delivery obtained. However, the cost per delivery using SET was not significantly higher
than the cost using DET.
Measured in terms of cost per child born there was a larger difference between the two transfer
policies. The cost per child born using DET was around DKK 93,000-94,000 compared with DKK
134,000-148,000 using SET. The reason to this is the extra children born as twin with the DETpolicy, which more than compensated for the extra costs following delivery and neonatal care. The
sensitivity analyses showed that the result of the analysis was robust. One sensitivity analysis furthermore revealed that the inclusion of the risk and cost of cerebral palsy did not have an impact upon
the costs of the average cost per pregnancy or child born.
In the literature results of a few health economic studies have been in favour of SET compared
with DET, which differ from those found in the present analysis. The present Danish cost-effectiveness analysis comparing a SET-policy with the traditional DET-policy was that the cost per delivery
or the cost per child is higher using SET compared with DET. One reason to the lower cost per
child with the DET-policy in the present study might have been the size of the neonatal intensive
care costs estimated on the basis of the prospective data collection among the couples participating.
However, DET involves higher total costs per woman treated, i.e. costs for antenatal care, delivery,
neonatal care and production lost due to the higher frequency of twin pregnancies and deliveries,
but at the same time also more effective. DET does not result in total costs that are much higher
than SET – around DKK 58,000 per woman treated with DET compared with DKK 41,00043,000 per woman treated with SET.
Introduction of a SET policy was estimated to have the following organisational consequences in
Denmark, supposing maintenance of an unchanged cumulative pregnancy rate by increased number
of cycles offered to compensate for a lower pregnancy rate in SET and still only treatment for the
first child:
a.
b.
c.
d.
e.
500-1000 more hormone stimulations, oocyte pick-ups, and fresh embryo transfers per year
Freezing of 900-1800 extra straws containing a single embryo
900-1800 more transfers of frozen embryos per year
18 more transfers of frozen embryos per year to obtain a second child
Based upon these estimates the economic consequences in case SET is introduced in Denmark were
that the public health care system will carry an extra burden of around DKK 14.9-25.9 mio. However,
these estimates should be evaluated with care. An increase in number of treatment cycles of 10% in
established clinics may very well represent a marginal increase without consequences for staffing, buildings etc. In such case the extra expenses to the public health care sector associated with SET may be
half. These extra expenses to IVF treatment may be balanced by savings due to fewer twin pregnancies.
An obligatory SET policy resulting the twin rate following IVF to fall close to zero may result in savings
of DKK 23.8 mio. per year. If eSET is used, it may reduce the twinning rate by 50% and savings will be
less, DKK 11.9 mio. per year. Based upon these assumptions, introduction of SET seems to represent a
change in the public expenses associated with IVF between a saving of DKK 9.1 mio. (best case) and
increase in expenses of DKK 13.9 mio. (worst case).
A second child was not contained in these considerations. Today two children (twins) are provided
to 25-30% of the couples covered by public funding. These couples could feel a significant deterio-
85
ration of the conditions for treatment in case they are imposed SET. An expansion of public funding
to treatments for child number two could represent a compensation for that – and may help to
decide on SET in a situation, where the couple is responsible for the choice between SET and
DET.
There seem to be five different options concerning policy on the question of transferring one or
two embryos:
a. An unchanged strategy, using two embryos when possible, unless contraindicated for medical
reasons, or because the patient chooses otherwise. The existing literature on risks associated with
twin pregnancies may contradict such a strategy. Two embryos are in accordance with a widespread wish among patients for more than one child.
b. Strict rules defining the frame and content of information of the risks of twin pregnancy and
delivery, and describing recommendations to the patient, but leaving the decision to the couple.
This option is in agreement with the principle of informed choice and is respecting patient
autonomy. Ultimately it is the couple that carries the consequences of choosing either one or
two embryos, i.e. chance of pregnancy against risk of twins. The type of information to the
patient is exemplified in Appendix 1, which is a slightly modified version of information on
SET written by Helle Ejdrup (personal communication) and discussed in the Danish Fertility
Society. Such a strategy will change patient choices towards SET, but we do not know to what
extent this will reduce the twin rate following IVF. It is likely, that more specific risk information
may increase worries among the couples concerned.
c. An elective SET policy with strictly defined criteria for SET, e.g. by age and embryo number
and quality. This somewhat like the Swedish model, but more workable if criteria are strictly
defined. In such case little is left for discussion between the patient and the doctor. The problem
is that although many prognostic factors have been identified in various studies, few of these are
strong predictors. In more general terms, few of the prognostic factors perform helpfully in
identifying specific groups at high risk of achieving pregnancy/twin pregnancy. In such case
rather strictly defined criteria are preferable to minimize uncertainty for the patients. An age
limit similar to or lower than that practised by Thurin et al. (2004) seems reasonable to avoid a
too substantial effect on pregnancy rates. Two or more high quality embryos is another criterion.
However, despite these selection criteria Thurin et al. (2004) found a lower pregnancy rate when
transferring one compared with two embryos. Further, it seems reasonable to apply SET to the
first and/or second treatment cycle only since there is a small decrease in pregnancy rates according to cycle number, and since funding in Denmark only covers three embryo transfers only.
d. An obligatory SET policy allowing SET only to all patients. This option carries an inherent risk
of potentially reduced cumulative pregnancy rates, especially for the oldest females without possibility of a significant number of compensatory reimbursed treatments.
e. A soft stimulation protocol with clomiphene citrate has been demonstrated to result in very few
twin pregnancies since few oocytes are harvested and few embryos are available for transfer
(Ingerslev et al. 2001b). However, this protocol is hampered by a lower clinical pregnancy rate
measured per started cycle compared with the long down regulation protocol due to a higher
cycle cancellation rate, which can only be compensated for by increasing the number of cycles
offered to the patient. Today, a sufficient hormone stimulation with harvest of a proper number
of oocytes (8-10) yielding possibility for embryo freezing and later transfer in a natural cycle
seems a more efficient strategy if a good freezing protocol is at hand.
It is beyond doubt that even a carefully designed eSET policy, but especially obligatory SET to all
patients will have a significant impact on the ultimate chance of a child for the individual couple.
Compensatory cycles are necessary to keep chances even compared to the present DET situation.
Although the results in present survey do not indicate that more cycles is an attractive compensation
for the infertile patients they will maintain an unchanged cumulative pregnancy rate. In terms of
costs for this versus savings by a reduction in the twin rate there seemed to be a likely balance.
86
The purpose of the present health technology assessment (HTA) report was to elucidate the consequences of obligatory single embryo transfer versus optional two embryo transfer in Denmark
Overall, the present analysis allows the following answers to the HTA questions:
All previous randomized studies have shown that elective single embryo transfer reduces the pregnancy rate per fresh cycle significantly. However, the present randomized study did not reveal any
difference, possibly due to random variation or an unquantifiable difference in embryo quality.
Observational data have indicated that it is possible to maintain unchanged pregnancy rates following introduction of single embryo transfer to selected groups of patients.
2. What is the basis for decisions of the couples to decide about the twin question, and how would an obligatory
single embryo transfer policy be in keeping with the interests of the infertile couple?
The present studies of patient attitudes revealed a strong desire for twins among couples undergoing
fertility treatment. Patients formed this decision on the basis of an evaluation of the social, psychological and physical discomfort related with IVF treatment combined with the wish to have more
than one biological child. Accordingly, an enforced selective embryo transfer policy would be in
conflict with patient interests and wishes.
3. What organizational consequences are expected in case of introduction of an obligatory single embryo transfer
policy?
Introduction of SET seem to represent a change in the public expenses associated with IVF between
a saving of DKK 9.1 mio. (best case) and increase in expenses of DKK 13,9 mio. (worst case). Any
eSET policy will create increased challenges to information of patients about the perspectives of the
choice between one and two embryos.
4. What are the expected health economic consequences for the society following an obligatory single embryo
transfer as judged from an expected lower pregnancy rate and a reduced consumption of resources with
respect to delivery, neonatal service etc.?
DET-policy, which is both more effective (higher clinical pregnancy rate, higher rate of delivery
and children), but also more expensive (higher delivery cost and neonatal intensive care costs). In
terms of cost per child born the cost using DET was around DKK 93,000-94,000 compared with
DKK 134,000-148,000 using SET.
Thus, the question ‘‘Should one or two embryos be transferred in IVF?’’ is not easy to answer.
Respect for patient autonomy should be considered against economic aspects including the effectiveness of these rather physically and psychologically stressful treatments and complications and against
long term sequelae associated with preterm delivery derived from twin pregnancies resulting from
transfer of two embryos. The ultimate answer to this question is to be given by the decision-makers.
87
References
Andersen AN, Gianaroli L, Nygren KG. European monitoring programme; European Society of Human Reproduction and
Embryology. Assisted reproductive technology in Europe, 2000. Results generated from European registers by ESHRE.
Hum Reprod 2004; 19:490-503.
Bergh C. Single embryo transfer: a mini-review. Hum Reprod 2005; 20:323-327.
Bergh T, Ericson A, Hillensjö T, Nygren K-G, Wennerholm U-B. Deliveries and children born after in vitro fertilisation in
Sweden 1982-95: a retrospective cohort study. Lancet 1999; 354:1579-1585.
Bernard H, Russel. Research Methods in Anthropology. London and New York: Sage, 1994.
Briggs AH, Gray AM. Handling uncertainty when performing economic evaluation of healthcare interventions. Health
Technology Assessment 1999; 3(2).
Callahan TL, Hall JE, Ettner SL, Christiansen CL, Greene MF, Crowley WF. The Economic Impact of Multiple-Gestation
Pregnancies and the Contribution of Assisted-Reproduction Techniques to Their Incidence. The New England Journal of
Medicine 1994; 331(4):244-249.
Campbell DM, MacGillivray I. Preeclampsia in twin pregnancies: incidence and outcome. Hypertens Pregnancy 1999; 18:197207.
Child TJ, Henderson AM, Tan LS. The desire for multiple pregnancy in male and female infertility patients. Hum Reprod
2004; 3:558-561.
Coonrod DV, Hickok DE, Zhu K, Easterling TR, Daling JR. Risk factors for preeclampsia in twin pregnancies: a populationbased cohort study. Obstet Gynecol 1995; 85:645-650.
Dàlton M. Infertility and the desire for multiple births. Fertil Steril 2004; 3:523-525.
De Neubourg D, Gerris J, Mangelschots K et al. Single top quality embryo transfer as a model for prediction of early
pregnancy outcome. Hum Reprod 2004; 19:1476-1479.
De Sutter P, Van der Elst J, Coetsier T, Dhont M. Single embryo transfer and multiple pregnancy rate reduction in IVF/
ICSI: a 5-year appraisal. Reproductive BioMedicine Online 2003; 6:464-469.
Derom R, Derom C, Vlietinck R. The risk of monozygotic twinning. In: Blickstein I, Keith LG (eds.) Iatrogenic Multiple
Pregnancy: Clinical Implications. Pathenon Publishing, New York, NY, USA 2001; 9-19.
Dhont M, Sutter PD, Ruyssinck G, Martens G, Bekaert A. Perinatal outcome of pregnancies after assisted reproduction:
a case-control study. Am J Obstet Gynecol 1999; 181:688-695.
Dickey RP, Taylor SN, Lu PY, Sartor BM, Storment JM, Rye PH, Pelletier WD, Zender JL, Matulich EM. Spontaneous
reduction of multiple pregnancy: Incidence and effect on outcome. Am J Obstet Gynecol 2002; 186:77-83.
Doyle P, Beral V, Botting B, Wale C. Congenital malformations in twins in England and Wales. J Epidemiol Community
Health 1991; 45:43-48.
Edwards A. Communicating risks. BMJ 2003; 327:691-692.
ESHRE Campus Course Report. Prevention of twin pregnancies after IVF/ICSI by single embryo transfer. Hum Reprod
2001; 16:790-800.
Fertilitetsklinikken. Ægtransplantationsbehandling (IVF og ICSI). Århus Universitetshospital 2004. Available from
http://www.auh.dk/sks/afd/afdy/dk/Aegtrans.htm
Gardner DK, Surrey E, Minjarez D, Leitz A, Stevens J, Schoolcraft W. Single blastocyst transfer: a prospective randomized
trial. Fertil Steril 2004; 81:551-555.
Gerris J, De Neuborg D, Mangelschots K, Van Royen E, Van de Meerssche M, Valkenburg M. Prevention of twin pregnancy
after in-vitro fertilization or intracytoplasmic sperm injection based on strict embryo criteria: a prospective randomized clinical trial. Hum Reprod 1999; 14:2581-2587.
Gerris J, De Neubourg D, Mangelschots K, Van Royen E, Vercruyssen M, Barudy-Vasquez J et al. Elective single day 3
embryo transfer halves the twinning rate without decrease in the ongoing pregnancy rate of an IVF/ICSI programme. Hum
Reprod 2002; 17:2626-2631.
88
Gerris J, De Neubourg D, De Sutter P, Van Royen E, Mangelschots K, Vercruyssen M. Cryopreservation as a tool to reduce
multiple birth. Reprod Biomed Online 2003; 7:286-294.
Gerris J, De Sutter P, De Neubourg D, Van Royen E, Van der Els J, Mangelschots K et al. A real-life prospective health
economic study of elective single embryo transfer versus two-embryo transfer in first IVF/ICSI cycles. Hum Reprod 2004;
19:917-923.
Gerris JMR. Single embryo transfer and IVF/ICSI outcome: a balanced appraisal. Hum Reprod Upd 2005; 11:105-121.
Godolphin W. The role of risk communication in shared decision making. BMJ 2003; 327:692-693.
Grobman WA, Milad MP, Stout J, Klock SC. Patient perceptions of multiple gestations;: An assessment of knowledge and
risk aversion. Am J Obstet Gynecol 2001; 4:920-924.
Hansen M, Bower C, Milne E, de Klerk N, Kurinczuk J. Assisted reproductive technologies and the risk of birth defects –
a systematic review. Hum Reprod, advance access, November 26, 2004.
Hassan MAM, Killick SR. Negative lifestyle is associated with a significant reduction in fecundity. Fertility and Sterility
2004; 81:384-392.
Helmenhorst FM, Perquin DA, Donker D, Keirse MJ. Perinatal outcome of singletons and twins after assisted conception:
a systematic review of controlled studies. Br Med J 2004; 328:261-265.
Henriksen TB, Sperling L, Hedegaard M et al. Cesarean section in twin pregnancies in two Danish counties with different
cesarean section rates. Acta Obstet Gynecol Scand 1994; 73:123-128.
Højgaard A, Ingerslev HJ, Dinesen J. Friendly IVF: patient opinions. Hum Reprod 2001; 7:1391-1396.
Hoozemans AD, Schats R, Lambalk CB et al. Human embryo implantation: current knowledge and clinical implications in
assisted reproductive technology. RBM online 2004; 9:692-715.
Houghton FD, Hawkhead JA, Humpherson PG et al. Non-invasive amino acid turnover predicts human embryo developmental capacity. Hum Reprod 2002; 17:999-1005.
Houghton FD, Leese HJ. Metabolism and developmental competence of the preimplantation embryo. European Journal
of Obstetrics & Gynecology and Reproductive Biology 2004; 115:S92-S96.
Hunault CC, Eijkemans MJC, Pieters MHEC et al. A prediction model for selecting patients undergoing in vitro fertilization
for elective single embryo transfer. Fertility and Sterility 2002; 77:725-732.
Imaizumi Y. A comparative study of Twinning and Triplet Rates in 17 Countries, 1972-1996. Acta Genet Med Gemellol
1998; 47:101-114.
Ingerslev HJ, Højgaard A, Poulsen PB, Kesmodel U, Dinesen J. Medicinsk teknologivurdering af lavstimulations IVF sammenlignet med standard IVF. Medicinsk Teknologivurdering – puljeprojekter; 1(1) Center for Evaluering og Medicinsk Teknologivurdering, 2001a.
Ingerslev HJ, Højgaard A, Hindkjær J et al. Friendly IVF? A randomized study comparing IVF in the unstimulated cycle and
following clomiphene citrate. Hum Reprod 2001b; 16:696-702.
Jackson RA, Gibson KA, Wu YW, Croughan MS. Perinatal outcomes in singletons following in vitro fertilization: a metaanalysis. Obstet Gynecol 2004; 103:551-563.
Kalra SK, Milad MP, Klock SC, Grobman WA. Infertility Patients and Their Partners: Differences in the Desire for Twin
Gestations. Obstet Gynecol 2003; 102:152-155.
Källén B. Congenital malformations in Twins: A population study. Acta Genet Med Gemellol 1986; 35:167-178.
Klemetti R, Gissler M, Hemminki E. Comparison of perinatal health of children born from IVF in Finland in the early and
late 1990s. Hum Reprod 2002; 17:2192-2198.
Koivurova S, Hartikainen AL, Karinen L, Gissler M, Hemminki E, Martikainen H et al. The course of pregnancy and delivery
and the use of maternal healthcare services after standard IVF in Northern Finland 1990-1995. Hum Reprod 2002;
17:2897-2903.
Krogh R. Performance in IVF and ICSI treatment – consequences of smoking. Research Year Thesis. 2004.
Lambalk CB, Van Hoof M. Natural versus induced twinning and pregnancy outcome: a Dutch nationwide survey of
primiparous dizygotic twin deliveries. Fertil Steril 2001; 75:731-736.
89
Lancaster PAL. Number of gestational sacs and singleton IVF preterm delivery. Hum Reprod 19 suppl. 1: Abstract book:
O-245 2004; i85.
Lidegaard Ø, Pinborg A, Nyboe Andersen A. Imprinting diseases and in vitro fertilisation. Danish National IVF cohort
study. Hum Reprod 2005; 20:950-954.
Loos R, Derom C, Vlietinck R et al. The East Flanders Prospective Twin Survey (Belgium). Twin Research 1998; 1:167-175.
Lukassen HGM, Braat DD, Zielhuis GA, Adang EMM, Kremer JAM. 2X1 versus 1X2, a randomized study. Hum Reprod,
Abstracts from the 18th Annual Meeting of ESHRE. Vienna, July 1-3, 2002. 2002; 17 (Abstract book 1):V-0064.
Lukassen HGM, Schönbeck Y, Adang EMM, Braat DDM, Zielhuis GA, Kremer JAM. Cost analysis of singleton versus twin
pregnancies after in vitro fertilization. Fertility and Sterility 2004; 81:1240-1246.
Lukassen HGM, Braat DD, Wetzels AMM, Zielhuis GA, Adang EMM, Scheenjes E, Kremer JAM. Two cycles with single
embryo transfer versus one cycle with double embryo transfer: a randomized controlled trial. Hum Reprod 2005; 20:702708.
Malone FD, Kaufman GE, Chelmow D, Athanassiou A, Nores JA, D’Alton ME. Maternal morbidity associated with triplet
pregnancy. Am J Perinatol Jan 1998; 15(1):73-77.
Martikainen H, Tiitinen A, Tomás C, Tapanainen J, Orava M, Tuomivaara L et al. One versus two embryo transfer after
IVF and ICSI: a randomized study. Hum Reprod 2001; 16:1900-1903.
Martikainen H, Orava M, Lakkakorpi J et al. Day 2 elective single embryo transfer in clinical practice: better outcome in
ICSI cycles. Hum Reprod 2004; 19:1364-1366.
Mattingly C. Healing dramas and clinical plots. Cambridge: Cambridge University Press, 1998.
Michaelsen P. Førstegangsfødendes barsel på sygehuse: samfundsøkonomisk evaluering. Master of Public Health, Aarhus
Universitet, udgivelse nr. 59. 2002.
Milki AA, Sunny HJ, Hinckley MD, Behr B, Giudice LC, Westphal LM. Incidence of monozygotic twinning with blastocyst
transfer compared to cleavage-stage transfer. Fertil Steril 2003; 79:503-506.
Montfoort APA, Dumoulin JCM, Land JA et al. Elective single embryo transfer (eSET) policy in the first three IVF/ICSI
treatment cycles. Hum Reprod 2005; 20:433-436.
Murray S, Shetty A, Rattray A, Taylor V, Bhattacharya. A randomized comparison of alternative methods of information
provision on the acceptability of elective single embryo transfer. Hum Reprod 2004; 4:911-916.
National Vital Statistics Report. Births: Final data for 2001. Vol 51. No. 2.
Nygren KG, Andersen AN. Assisted reproductive technology in Europe, 1998. Results generated from European registers
by ESHRE. Hum Reprod 2001; 16:2459-2471.
Official statistics of Denmark. The Danish Society of Gynecology and Obstetrics. http://www.DSOG.dk
Official Statistics of Finland. Perinatal Statistics. http://stakes.info/files/pdf/Tilastotiedotteet/Tt15_04.pdf
Official statistics of Norway. University of Bergen. http://www.uib.no/mfr/html
Official statistics of Sweden. The National Board of Health and Welfare. 2003; http://www.sos.se
Ozturk O, Bhattacharya S, Saridogan E et al. Role of utero-ovarian vascular impedance: predictor of ongoing pregnancy
in an IVF-embryo programme. Reproductive BioMedicine Online 2004; 9:299-305.
Pandian Z, Bhattacharya S, Ozturk O, Serour GI, Templeton A. Number of embryos for transfer following in-vitro fertilisation or intra-cytoplasmic sperm injection. The Cochrane Library, 2004. Issue 4.
Pharoah POD, Adi Y. Consequences of in-utero death in a twin pregnancy. Lancet 2000; 355:1597-1602.
Pinborg A, Loft A, Schmidt L, Nyboe Andersen A. Morbidity in a Danish National cohort of 472 IVF/ICSI twins, 1132 nonIVF/ICSI twins and 634 IVF/ICSI singletons: health-related and social implications for the children and their families.
Hum Reprod 2003a; 18:1234-1243.
Pinborg A, Loft A, Schmidt L, Andersen AN. Attitudes of IVF/ICSI-twin mothers towards twins and single embryo transfer.
Hum Reprod 2003b; 3:621-627.
90
Pinborg A, Loft A, Schmidt L, Langhoff-Roos J, Nyboe Andersen A. Maternal risks and perinatal outcome in a Danish
National cohort of 802 twin pregnancies: the role of IVF/ICSI treatment. Acta Obstet Gynecol Scand 2004a; 83:75-84.
Pinborg A, Loft A, Rasmussen S, Schmidt L, Jens Langhoff-Roos, Greisen G, Nyboe Andersen A. Neonatal outcome in a
Danish national cohort of 3438 IVF/ICSI and 10362 non-IVF/ICSI twins born in 1995 to 2000. Hum Reprod 2004b; 19:435-441.
Pinborg A, Loft A, Nyboe Andersen A. Neonatal outcome in a Danish national cohort of 8602 children born after in
vitro fertilization (IVF) or microinsemination (ICSI): the role of twin pregnancy. Acta Obstet Gynecol Scand 2004c;
83:1071-1078.
Pinborg A, Loft A, Schmidt L, Greisen G, Rasmussen S, Nyboe Andersen A. Neurological sequelae in twins born after
assisted conception: controlled national cohort study. Br Med J 2004d; 329:311-314.
Pinborg A, Lidegaard Ø, La Cour Freiesleben N, Nyboe Andersen A. Consequences of vanishing twin in IVF/ICSI pregnancies. Submitted to Hum Reprod 2004e; 2005.
Ryan GL, Zhang SH, Dokras A, Syrop CH, Van Voorhis BJ. The desire of infertile patients for multiple births. Fertil Steril
2004; 3:500-504.
Ryan GL, Van Voorhis BJ. The desire of infertile patients for multiple births do they know the risks? Fertil Steril 2004;
3:526.
Santema JG, Koppelaar I, Wallenburg HC. Hypertensive disorders in twin pregnancy. Eur J Obstet Gynecol Reprod Biol
1995; 58:9-13.
Staessen C, Van den Abbeel E, Carle M, Devroey P, Van Steirteghem AC. Comparison between human serum and
Albuminar-20TM supplement for in-vitro fertilization. Hum Reprod 1990; 5:336-41.
Schachter M, Raziel A, Shevach F et al. Monozygotic twinning after assisted reproductive techniques: a phenomenon
independent of micromanipulation. Hum Reprod 2001; 16:1264-1269.
Scher AI, Petterson B, Blair E, Ellenberg JH, Grether JK, Haan E, Reddihough DS, Yeargin-Allsopp M, Nelson KB. The Risk
of Mortality or Cerebral Palsy in Twins: A Collaborative Population-Based Study. Pediatr Res 2002; 52:671-681.
Schieve LA, Meikle SF, Ferre C, Peterson HB, Jeng G, Wilcox LS. Low and very low birth weight in infants conceived with
use of assisted reproductive technology. N Engl J Med 2002; 346:731-737.
Sebire NJ, Snijders RJ, Hughes K, Sepulveda W, Nicolaides KH. The hidden mortality of monochorionic twin pregnancies.
Br J Obstet Gyneacol 1997; 104:1203-1207.
Seeberg J. Barnløshedsbehandling og det uønskedes non-narrativitet. Sygdom og Samfund 1, 2. 2005.
Seeberg J. IVF behandling med oplægning af ét eller to befrugtede æg? Patientperspektiver. Rapport for antropologisk
delstudie. Århus: HRA Consult. 2003.
Senat MV, Ancel PY, Bouvier-Colle MH, Breart G. Clin Obstet Gynecol 1998; 41:78-83.
Shenfield F. The responsibility of the fertility specialist in preventing multiple pregnancies in ART. In Adashi EY, Barri PN,
Berkowitz R et al. Report of the meeting Infertility therapy-associated multiple pregnancies (births): an ongoing epidemic. RBMOnline 2003; 7:515-542.
Shiku H, Shiraishi T, Ohya H et al. Oxygen consumption of single bovine embryos probed by scanning electrochemical
microscopy. Analytical Chemistry 2001; 73:3751-3758.
Sperling LS, Henriksen TB, Ulrichsen H et al. Indications for cesarean section in singleton pregnancies in two Danish
counties with different cesarean section rate. Acta Obstet Gynecol Scand 1994; 73:129-135.
Statens Institut for Medicinsk Teknologivurdering. Medicinsk Teknologivurdering. Hvorfor? Hvad? Hvornår? Hvordan?
Sundhedsstyrelsen, København 2000.
Strandell A, Bergh C, Lundin K. Selection of patients suiTable for one-embryo transfer may reduce the rate of multiple
births by half without impairment of overall birth rates. Hum Reprod 2000; 15:2520-2525.
Strömberg B, Dahlquist G, Ericson A, Finnström O, Köster M, Stjernqvist K. Neurological sequelae in children born after
in vitro fertilisation: a population-based study. Lancet 2002; 359:461-465.
Sutter PD, Gerris J, Dhont M. A health-economic decision-analytic model comparing double with single embryo transfer
in IVF/ICSI. Human Reproduction 2002; 17(11):2891-2896.
91
Templeton A, Morris JK. Reducing the risk of multiple births by transfer of two embryos after in vitro fertilization. N
Engl J Med 1998; 339:573-577.
Thornton H. Patients’ understanding of risk. BMJ 2003; 327:693-694.
Thurin A, Hausken J, Hillensjö T, Jablonowska B, Pinborg A, Strandell A, Bergh C. Elective Single-Embryo Transfer versus
Double-Embryo Transfer in in Vitro Fertilization. N Eng J Med 2004; 351:2392-2402.
Thurin A, Hardarson T, Hausken J et al. Predictors of ongoing implantation in IVF in a good prognosis group of patients.
Hum Reprod 2005; deh872.
Tiitinen A, Halttunen M, Härkki P, Vouristo P, Hydén-Granskog C. Elective single embryo transfer: the value of cryopreservation. Hum Reprod 2001; 6:1140-1144.
Tiitinen A, Unikila-Kallio L, Halttunen M, Hyden-Granskog C. Impact of elective single embryo transfer on the twin
pregnancy rate. Hum Reprod 2003; 18:1449-1453.
Tiitinen A, Hydén-Granskog C, Gissler M. What is the most relevant standard of success in assisted reproduction? The
value of cryopreservation on cumulative pregnancy rates per single oocyte retrieval should not be forgotten. Hum
Reprod 2004; 19:2439-2441.
Tiitinen A, Gissler M. Effect of in vitro fertilization practices on multiple pregnancy rates in Finland. Fertil Steril 2004;
82:1689-1690.
Tong S, Short RV. Dizygotic twinning as a measure of human fertility. Hum Reprod 1998; 13:95-98.
Trimarchi JR. A mathematical model for predicting which embryos to transfer – an illusion of control or a powerful tool?
Fertility and Sterility 2001; 76:1286-1287.
van Montfoort APA, Dumoulin JCM, Land JA, Coonen E, Derhaag JG, Evers JLH. Elective single embryo transfer (eSET9
policy in the first three IVF/ICSI treatment cycles. Hum Reprod 2005; 20:433-436.
van Montfoort APA, Fiddelers AAA, Janssen JM, Derhaag JG, Dirksen CD, Dunselman GAJ, Land JA, Geraedts JPM, Evers
JHL, Dumoulin JCM. Single embryo transfer (SET) in unselected patients: a randomised controlled trial (RCT). Abstract O150, ESHRE Copenhagen. 2005.
Vilska S, Tiitinen A, Hydén-Granskog C, Hovatta O. Elective transfer of one embryo results in accepTable pregnancy rate
and eliminates the risk of multiple birth. Hum Reprod 1999; 14:2392-2395.
Wennerholm UB, Bergh C, Hamberger L, Lundin K, Nilsson L, Wikland M, Källén B. Incidence of congenital malformations
in children born after ICSI. Hum Reprod 2000a; 15:944-948.
Westergaard HB, Johansen AM, Erb K, Andersen AN. Danish National in-vitro fertilization registry 1994 and 1995: a
controlled study of births, malformations and cytogenetic findings. Hum Reprod 1999; 14:1896-902.
Westergaard LG. Upubliceret appendix i Vejle Amt (1998): Redegørelse vedrørende IVF behandling på de offentlige
klinikker i Danmark. 1998.
Westergaard T, Wohlfahrt J, Aaby P. Population based study of rates of multiple pregnancies in Denmark, 1980-94. Br
Med J 1997; 314:775-779.
Wheeler CA, Cole BF, Frishman GN et al. Predicitng probabilities of pregnancy and multiple gestation from in vitro
fertilization-A new model. Obstet Gynecol 1998; 91:696-700.
Wolff NJ. Ovulation induction. Clinical Obstetrics and Gynecology 2000; 43:902-915.
Wright VC, Schieve LA, Reynolds MA, Jeng G, Kissin D. Assisted reproductive technology surveillance – United States,
2001. MMWR Surveill Summ 2004; 53:1-20.
92
Appendix 1
Information on the choice between one or
two embryos and on twin pregnancies
1. How is my probability to get pregnant influenced by choosing one embryo in stead of two?
The largest study to answer this question (1) shows that you transfer one embryo to one half and
two embryos to another half of women below 36 years of age and who have at least two high
quality embryos, the one-embryo group will have 30% chance of delivery, the two-embryo group
43%. However, freezing and subsequent thawing and transfer of the other high quality embryo
increases the delivery rate in the one-embryo group to 39%. Accordingly, if you include the pregnancies from freezing and thawing of spare embryos the chance of pregnancy is not very different in
the two groups.
2. What is the probability of twins in case I choose two embryos?
In the two-embryo group mentioned above 33% got pregnant with twins.
3. What are the risks of a twin pregnancy and delivery?
Two recent large studies from Denmark (2, 3), comprising among others 3000 IVF twins revealed
that twins were born on average three weeks earlier than IVF singletons. Other differences were:
Preterm delivery
Mean birth weight
Caesarean section
Still born
Admission to neonatal intensive care unit (NICU)
Number of days in NICU
Mental disability
IVF twins
43.9%
2500 g
53%
1,3%
56.4%
19.8
0.9%
IVF singletons
7.3%
3500 g
20%
0,7%
25.4%
11.0
0.8%
To be pregnant and to deliver a baby is unavoidably associated with a small risk. With twins,
however, the risk is increased. The twin pregnancy represents more strain to the mother. She has
to leave her job for childbirth earlier and there are more check-ups during pregnancy.
4. How do I choose?
We can recommend that you consider the question of one or two embryos carefully before the
embryo transfer.
However, it is not until that day we can give you information on how many oocytes fertilized,
cleaved and how the quality is.
One embryo
Normally, we offer you one embryo if
H
The female is below 37 years and
H
You are in your first or second cycle and
H
Your embryos have cleaved to four cells and are of good quality
H
There are spare embryo(s) for freezing
Moreover, one embryo is possible if
H
You are treated to obtain a second child
H
You have delivered twins previously
H
If there is risk of hyperstimulation syndrome
93
Two embryos
Normally you could have two embryos if
H
The female is 37 or more or
H
It is your second or third treatment or
H
If your embryos are of less optimal quality and
H
If there are no spare embryos for freezing
H
Other circumstances
References
1
2
Pinborg A, Loft A, Schmidt L, Nyboe Andersen A. Morbidity in a Danish National cohort of 472 IVF/ICSI twins, 1132 nonIVF/ICSI twins and 634 IVF/ICSI singletons: health-related and social implications for the children and their families. Hum
Reprod 2003; 18:1234-1243.
3
94
Appendix 2
Information omkring valget mellem et eller to æg
og om tvillingegraviditet
1. Hvordan påvirkes min graviditetschance, hvis jeg kun får oplagt ét æg i stedet for to?
Den største af de undersøgelser (1), der indtil nu er lavet for at belyse dette spørgsmål viser, at hvis
man tager en gruppe kvinder, som er under 36 år og som ved IVF (ægtransplantation) ender med
at have mindst to flotte æg og lægger ét æg op på den ene halvdel og to på den anden, vil
étægsgruppen have 30% chance for at føde et barn, mens den var 43% i toægsgruppen. Men når
det andet gode æg i etægsgruppen blev frosset ned og lagt op, hvis kvinden ikke blev gravid i første
omgang, så blev chancen for fødsel i étægsgruppen 39%. Når man medregner de graviditeter, der
opstår efter oplægning af frosne æg, vil chancen således være tæt på hinanden i de to grupper.
2. Hvor stor er risikoen/chancen for tvillinger, hvis man får lagt to æg op?
I toægsgruppen i den undersøgelse, vi beskriver ovenfor var det en chance/risiko for tvillinger på 33%.
3. Hvad er risikoen ved tvillingegraviditet og -fødsel?
I to nylige undersøgelser (2, 3), der er gennemført i Danmark, og som bl.a. omfattede 3000 børn
født som tvillinger, fandt man ved at sammenligne med enkeltfødte børn, at IVF tvillinger fødes i
gennemsnit tre uger før IVF enkeltbørn. Andre væsentlige forskelle var:
For tidlig fødsel
Gennemsnitlig fødselsvægt
Kejsersnit
Dødfødte
Indlæggelse på afdeling for for tidligt fødte
Indlæggelsesdage
Mentalt handicap
IVF tvillinger
43.9%
2500 g
53%
1,3%
56.4%
19.8
0.9%
IVF enkeltfødte
7.3%
3500 g
20%
0,7%
25.4%
11.0
0.8%
Der vil altid som det kan ses, være en lille risiko, når man føder børn for, at der sker noget.
Men risikoen er større, hvis man føder tvillinger. Graviditeten er også en hårdere belastning for
kvinden, hun skal gå på barsel tidligere, og der er flere kontroller i graviditeten.
4. Hvordan kunne man vælge?
Det er klogt, at I gør jer nogle overvejelser inden selve ægoplægningen.
I får dog først ved selve ægoplægningen besked om, hvor mange befrugtede og delte æg der er, samt
besked om deres kvalitet.
Et befrugtet æg
Vi vil normalt tilbyde tilbagelægning af et befrugtet æg hvis:
H
Kvinden er under 37 år og det er
H
Første eller anden behandling og
H
Æggene har delt sig til 4 celler og er af god kvalitet og
H
Der er overskydende æg til nedfrysning
Vi kan ligeledes lægge et befrugtet æg tilbage efter aftale med jer hvis:
H
Det er med henblik på andet barn.
H
Kvinden tidligere har født tvillinger.
H
Der er risiko for overstimulation.
95
To befrugtede æg
Vi vil normalt tilbyde oplægning af to befrugtede æg hvis:
H
Kvinden er 37 år eller derover og
H
Det er anden eller tredje behandling.
H
Hvis de befrugtede æg ikke er topkvalitet og
H
Der ikke er befrugtede æg til nedfrysning
H
Særlige forhold.
Referencer
1
2
Pinborg A, Loft A, Schmidt L, Nyboe Andersen A. Morbidity in a Danish National cohort of 472 IVF/ICSI twins, 1132 nonIVF/ICSI twins and 634 IVF/ICSI singletons: health-related and social implications for the children and their families. Hum
Reprod 2003; 18:1234-1243.
3
96
Appendix 3
Sensitivity analyses – cost-effectiveness ratios
IUI couples excluded from the analysis
Cost per delivery (Pandian et al. 2004)
Cost per delivery (Thurin et al. 2004)
Cost per child (Pandian et al. 2004)
Cost per child (Thurin et al. 2004)
Changed estimate of days in neonatal intensive care unit (Westergaard (1998)
Cost per delivery (Pandian et al, 2004)
Inclusion of risk of cerebral palsy (Pinborg et al. 2004)
Production lost decreased by 30%
Production lost increased by 30%
Direct costs decreased by 30%
Direct costs increased by 30%
SET (DKK)
DET (DKK)
141,747
151,429
134,832
149,783
122,836
116,742
90,511
90,489
147,530
154,891
140,333
153,207
146,235
136,281
107,752
105,634
191,729
199,012
182,376
196,848
191,026
180,942
140,756
140,252
134,619
143,801
128,053
142,238
119,507
112,581
88,058
87,263
147,178
155,866
139,999
154,171
135,815
128,067
100,074
99,267
132,027
141,355
125,587
139,818
115,803
109,120
85,329
84,581
149,770
158,311
142,465
156,591
139,519
131,527
102,804
101,949
97
Ordforklaring
Summary
HTA: health technology assessment, medicinsk teknologivurdering
Randomise: trække lod
Survey: spørgeskemaundersøgelse
1. Introduction
IVF: in vitro fertilisation, ægtransplantation, ‘‘reagensglasbefrugtning’’
ICSI: intracytoplasmatisk sædcelleinjektion
Intrauterine: i livmoderen
Ovarium: æggestok
Follicle: ægblære
Embryo: befrugtet æg, tidligt foster
Transfer: oplægning (af æg)
Multiple birth: flerfoldsfødsel
Single embryo transfer (SET): oplægning af ét æg ad gangen
Double embryo transfer (DET): oplægning af to æg ad gangen
2. Multiple birth rates
Dizygotic twins: tveæggede tvillinger
Monozygotic twins: énæggede tvillinger
Paediatrician: børnelæge
Hypertensive disorder: sygdom med forhøjet blodtryk
Thrombo embolism: blodpropssygdom
Urinary tract infection: urinvejsinfektion
Anaemia: blodmangel
Vaginal-uterine haemorrhage: blødning fra skede eller livmoder
Placental abruption: for tidlig løsning af moderkagen
Placenta praevia: forliggende moderkage
3. Risks associated with multiple pregnancy
Corticosteroids: binyrebarkhormoner
Caesarian section: kejsersnit
Preterm delivery: for tidlig fødsel
Maternal mortality: mødredødelighed
Preeclampsia: svangerskabsforgiftning
Intrahepatic cholestasis: gulsot betinget af reduceret afløb af galde fra leveren
Metaanalysis: statistisk analyse af resultater fra en række studier inden for et særligt område med
det formål at integrere resultaterne i en samlet analyse alternativt statistisk beregning, hvor man
samler resultaterne fra flere undersøgelser omkring et specifikt emne, så man får en samlet større
analyse
Conceive: undfange
Monochoric twins: tvillinger, der ligger i samme ydre graviditetshinde
Paritet: fødsels nummer
Odds ratio: risikoberegning, hvor odds for et udfald i en bestemt gruppe individer divideres med
odds for det samme udfald i en anden gruppe af kontrol individer.
Malformation: misdannelse
Patent ductus arteriosus: ductus arteriosus er et kar som hos fosteret forbinder venstre lungepulsåre
og hovedpulsåren. Ved patent ductus arteriosus forbliver dette kar åbentstående og lukker ikke
som på normal vis indenfor barnets første levedøgn.
Undescended testes: testikler, som ikke er kommet ned i pungen
Neural tube defects: manglende lukning af det rør, som nervesystemet udgør i fostertilstanden
98
Hydrocephaly: vand i hovedet
Alimentary tract: fordøjelseskanalen
Neonatal intensive care unit: afdeling for for tidligt fødte
Perinatal mortality: i dette arbejde defineret som antal dødfødte efter udgangen af 28 fulde svangerskabsuger og dødsfald blandt levendefødte i første leveuge per 1000 fødte
Cerebral palsy: cerebral parese, spastisk lammelse
Gestational sac: gestationssæk, graviditetsanlæg
4. Factors influencing the twin birth rate
Anovulatory: manglende ægløsning
Triplet: trilling
Primipara: førstegangsfødende
Ovulation: ægløsning
Ovary: æggestok
Gonadotrophin: overordnet hormon, som stimulerer æggestok
Tuba: æggeleder
Multivariate analysis: statistisk beregning med det formål at bestemme hvilke faktorer, der er af
betydning for et bestemt udfald ex. hvilke faktorer, der har betydning for udvikling af cerebral
parese.
Cleavage stage: delingsstadie (normalt 4-8 cellestadie i fosterudviklingen)
Blastocyst: udviklingsstadie af fostret lige før det sætter sig fast i livmoderen
Implantation: den proces, hvorved ægget sætter sig fast i livmoderslimhinden
Elective: planlagt
Cochrane review: systematisk litteraturgennemgang
Frozen-thawed: nedfrosset og optøet
Ongoing pregnancy: igangværende graviditet
Oocyte pick-up: ægudtagning
Testicular retrieval: fremskaffelse (af sædceller) fra testiklen
Cervical incompetence: cervixinsufficiens, eftergivelig livmoderhals
Spermatozoa: sædceller
Reimbursed: betalt (af det offentlige)
Endometriosis: endometriose; sygdom med livmoderslimhindevæv i f.eks. bughulen
Miscarriage: abort
Clinical pregnancy: ultralydspåviselig graviditet
Fragmentation: sønderdeling af celler i det tidlige befrugtede æg
Logistic regression analysis: statistisk beregning med det formål at bestemme hvilke faktorer, der
er af betydning for et bestemt udfald ex. hvilke faktorer, der har beydning for udvikling af
cerebral parese.
BMI: body mass index, beskriver relation mellem vægt i kg og højde
Basal FSH: niveau af hormonet FSH målt i dagene efter menstruationen
5. What do the infertile couples prefer: Single or double embryo transfer – a single child or twins – and why?
Gender: køn
G.P: General Practitioner, alment praktiserende læge (egen læge)
Likert scale: en type spørgsmål, hvor respondenten anmodes om at angive i hvilken grad han eller
hun er enig eller uenig med et udsagn
6. Cost-effectiveness of SET versus DET strategies
Cost-effectiveness analysis (CEA): en analyseform der måler omkostninger i forhold til effekter
ved et program eller en teknologi – omkostningseffektiviteten. Effekterne er sundhedsoutcome
og måles i naturlige enheder, eksempelvis vundne leveår eller antal opnåede fødsler. CEA sammenligner det relative forhold mellem omkostninger og effekter for forskellige alternativer.
Cost-effectiveness ratio: de gennemsnitlige omkostninger opnået ved en enhed af en sundhedseffekt (eks. kroner pr. leveår) ved én intervention sammenlignet med en alternativ intervention.
99
DAGS: Dansk Ambulant Grupperingssystem – et case-mix system til afregning i forbindelse med
ambulante besøg på sygehuset for somatiske ambulante patienter.
DRG: Diagnose Relaterede Grupper – et case-mix system, bl.a. til afregning pr. udskrivning fra
hospitalet.
Measure of effectiveness: måden hvorpå effekten af en intervention opgøres. Effektmålet kan have
form af et endeligt effektmål (eks. vundne leveår) eller et mellemliggende effektmål (eks. korrekt
diagnosticerede patienter).
Unit costs: omkostningen ved én enhed af en procedure.
Sensitivity analysis: en matematisk beregning hvormed usikkerhed omkring konkrete parametre
og sammenhænge kan håndteres systematisk og kvantificeres, med det formål at undersøge hvor
robust resultatet af den sundhedsøkonomiske analyse er.
Incremental cost-effectiveness ratio (ICER): ratioen der måler forskellen i omkostninger mellem
to alternativer i forhold til forskellen i effektivitet mellem samme alternativer, og derved tilnærmelsesvist udtrykker ekstra omkostningen ved at udvide et program med én ekstra enhed produceret.
Overhead costs: omkostninger fra hjælpeafdelinger, eksempelvis teknisk- eller vedligeholdelsesafdelinger, der ikke direkte relaterer sig til den specifikke produktion i en behandlende afdeling.
Perspective: synsvinklen ud fra hvilket en cost-effectiveness analyse udføres, eksempelvis et samfundsmæssigt perspektiv eller et hospitalsperspektiv. Perspektivet har betydning for omfanget af
omkostningsmålingen.
Production lost: den produktion der mistes i samfundet på grund af en persons mistede eller
reducerede arbejdsevne
7. Organisational consequences of SET
Straw: strå, i hvilke befrugtede æg nedfryses
100

should one or two embryos be transferred in ivf?

Transcription

Similar documents

Medical Intake Form - Reproductive Medicine Associates of New

of Having a Baby

READ MORE - Reproductive Medicine Associates of New Jersey

Have a baby at new low, low price, says fertility doctor Joel Batzofin

In Vitro Fertilization Orientation

Egg Thaw Cycle Orientation - NYU Langone Medical Center

Am I pregnant? Signs and symptoms

TWI NS - University of Houston

Risky Business

IVF-Treatment-Fertility-Medications-and-Their-Side