The outcome of multiple pregnancy
Pat Doyle
Epidemiology Unit, Department of Epidemiology and Population Sciences,
London School of Hygiene and Tropical Medicine, Keppel Street, London
WC IE 7HT, UK
The incidence of multiple pregnancy and delivery has increased dramatically
over the past 10-15 years in many developed countries of the world. Data
for England and Wales show that between 1980 and 1993 the twin maternity
rate increased by ~25% and the triplet and higher order maternity rate
more than doubled. Similar trends have been reported elsewhere. The
majority of these increases have been linked to the use of ovarian stimulants
and assisted reproduction techniques, and multiple pregnancy must be
considered to be one of the most important adverse outcomes in current
methods of infertility treatment. Obstetric complications associated with
multiple pregnancy include prenatal screening problems and increased
incidence of pregnancy-induced hypertension, antepartum haemorrhage,
preterm labour and assisted or surgical delivery. Neonatal problems include
low birthweight and increased prevalence of congenital malformations.
Compared with singletons, neonatal mortality was seven times higher in
twins and >20 times higher in triplets and higher order births in England
and Wales in 1991. Survivors also suffer higher rates of cerebral palsy and
other neurological impairments. Most studies of pregnancies and babies
resulting from assisted reproduction have demonstrated similar, if not higher,
risks of adverse obstetric and neonatal outcomes for multiple births compared
with national expectations. A poorer outcome in multiple pregnancy, especially in triplet and higher order pregnancy, supports the replacement of two
good quality embryos in assisted reproduction treatment cycles.
Key words: assisted reproduction/multiple maternity/multiple pregnancy/trends
Introduction
In the treatment of infertility, ovarian stimulation and the replacement of multiple
embryos increase the probability of pregnancy. These techniques also increase
the risk of multiple pregnancy. Although many infertile couples may welcome
multiple pregnancy as a way of achieving their desired family, many are unaware
of the complications and poorer outcomes associated with them. Data from
throughout the world now show that multiple pregnancy, especially triplet and
110
O European Society for Human Reproduction and Embryology
Human Reproduction Volume 11 Supplement 4 1996
Outcome of multiple pregnancy
1986
1991
•5
3
8
o
o"
1
Figure 1. Trends in multiple maternities in England and Wales, 1971-1993. (A) All multiple maternities. (B)
Twin maternities. (C) Triplet and higher order maternities. (D) Numbers of twin, triplet and quadruplet plus
maternities. These data include maternities where still births occurred. (Office of Population Censuses and
Surveys, FM1.)
higher order pregnancy, is one of the most important adverse effects of ovarian
stimulation and assisted reproduction treatment. This paper discusses the impact
of infertility treatments on national trends in multiple births, and considers the
major short- and long-term health consequences of multiple birth.
National trends in multiple maternities
The incidence of deliveries with more than one baby (multiple maternities) has
increased dramatically over the past decade in many countries. There have been
reports of increasing multiple maternities in the USA (Luke, 1994; Jewell and
Yip, 1995), Canada (Millar et al, 1994), Belgium (Derom et al, 1993), The
Netherlands (van Duivenboden et al, 1991), France (Tuppin et al, 1993) and
Taiwan (Chen et al, 1992). In England and Wales, the downward trend evident
from the early 1950s slowed and began an upward course in the early 1980s in
all maternal age groups except the under 20s (Botting et al, 1987). This
increasing trend has continued into the 1990s (Figure 1A) (Office of Population
Censuses and Surveys, Series FM1). The proportion of multiple maternities
increased from 9.8 per 1000 in 1980 to 12.7 per 1000 in 1993. Most of this
increase is made up of twin maternities, because they are the most common
multiple pregnancy (Figure IB). Triplets and higher order maternities remain
111
RDoyle
relatively rare, but they showed an even more dramatic change within the time
period, increasing from 1.5 per 10 000 in 1980 to 3.7 per 10 000 in 1993 (Figure
1C). Changes in the numbers, rather than rates, of multiple maternities are shown
in Figure ID. In 1980 there were 6308 twin, 91 triplet and five quadruplet or
higher order maternities in England and Wales. This increased to 8302 twin, 234
triplet and 13 quadruplet or higher order maternities by 1993.
Contribution of ovarian stimulation and assisted reproduction
treatment to national multiple birth rates
Although some of the increase in multiple births can be attributed to national
increases in age at childbirth, the majority is associated with treatments for
infertility. An accurate assessment of the impact of infertility treatment is difficult
in the absence of complete national treatment data, but estimates have been made
using hospital birth series, drug sales records and multiple birth registers. A UK
study of 156 triplet, 12 quadruplet and one quintuplet delivery occurring in 1989
found that 31% were conceived naturally, 34% were the result of ovarian
stimulation and 35% of the mothers had undergone in-vitro fertilization (IVF)
or gamete intra-Fallopian transfer (Levine et al., 1992). Using national data on
multiple maternities and fertility drug sales in France, Tuppin et al. (1993)
estimated that between 1985 and 1989 50% of all triplet deliveries resulted
from treatments with ovulation-inducing agents and 26% from other assisted
reproduction techniques such as IVF. Similarly, data from the East Flanders
Prospective Twin Study have confirmed the close association between fertilityenhancing drugs, other reproductive technologies and multiple pregnancy (Derom
et al, 1993).
National registers of assisted reproduction have reported proportions of multiple
pregnancies ranging between 15 and 27% (Lancaster, 1992; Logerot-Lebrum
et al., 1995). Correspondingly, the proportions of multiple birth babies lie between
30 and 40% of all babies born following assisted reproduction treatment.
Obstetric complications
Multiple pregnancy carries extra risk for both mother and babies, and greater
monitoring of the pregnancy is required. Prenatal screening poses particular
difficulties. Apart from the technical problems of invasive procedures such as
amniocentesis, the couple may have to cope with severe dilemmas when faced
with discordancy for abnormality (Nielson, 1992). Recent work has shown that
serum-free a-human chorionic gonadotrophin concentrations can be successfully
adjusted to produce standards for use in twin pregnancies (Wald and Densem,
1994), but difficulties remain for prenatal screening in higher order pregnancies.
The incidence of pregnancy-induced hypertension is greatly increased in
multiple pregnancies, and both pre-eclampsia and eclampsia are more common
112
*•»•
Outcome of multiple pregnancy
(Chamberlain, 1991; Douglas and Redman, 1994). In a study of births following
assisted conception in the UK, 23% of women with a multiple pregnancy were
admitted to hospital because of hypertension compared with 13% of mothers
with singleton pregnancies (Beral et al, 1990). Similarly, bleeding during
pregnancy is more common in multiple births: 22% of the women with a multiple
pregnancy in the above study were admitted to hospital because of bleeding at
some time during their pregnancy compared with 17% of mothers of singletons
(Beral et al, 1990). These rates are generally higher than would be expected in
the general population (Tan et al, 1992). Growth retardation can occur at any
time during a multiple pregnancy, but is more likely to appear in the third trimester
when fetal demands on the placenta are greatly increased (Chamberlain, 1991).
It is of very great significance that preterm labour resulting in low birthweight
is the most important determinant of perinatal and neonatal mortality. National
data show median gestational durations of 40 weeks for singletons, 37 weeks
for twins and 33 weeks for triplets (Chamberlain, 1991), and the incidence of
preterm delivery (<37 completed weeks of gestation) follows a clear upward
trend with increasing plurality. The UK register of babies resulting from assisted
conception found preterm delivery rates of 13% for singletons, 57% for twins
and 95% for triplets and higher order pregnancies (Beral and Doyle, 1990).
These figures are significantly higher than expected using national rates, which
were 6% for singletons and 38% for twins (Office of Population Censuses and
Surveys). Similar results have been reported from other registers of assisted
reproduction (Lancaster, 1992).
Surgical and assisted delivery is common in all multiple pregnancies, but there
is evidence that rates are higher in pregnancies following assisted reproduction
treatment. Only 13% of twin and 2% of triplet and higher order pregnancies
experienced a normal vaginal delivery in the UK series (Beral et al, 1990). In
a comparative study of IVF versus naturally conceived pregnancies in Finland,
62% of IVF multiple pregnancies were delivered by Caesarean section compared
with 41% of naturally conceived multiple pregnancies (Gissler et al, 1995).
Neonatal outcome
Congenital malformations
There is good evidence that babies from multiple pregnancies have a higher
prevalence of reported malformations at birth than singletons. Neural tube defects
and structural malformations of the gastro-intestinal tract are increased in twins
compared with singletons (Doyle et al, 1991). The study of malformations in
assisted reproduction treatment babies is hampered by low statistical power to
assess relatively rare outcomes and the incomplete nature of national congenital
malformation registry data for comparison. Meta-analysis of data from several
assisted reproduction registers is required to investigate whether assisted reproduction and other infertility treatments are associated with risks of congenital
malformation over and above those expected, in both singleton and multiple births.
113
RDoyle
Table I. Mortality multiplicity of birth, England and Wales, 1991"
Mortality
Singleton Twin
babies
babies
Triplet and higher
order babies
Stillbirth rate (late fetal deaths per 1000 total
births)
Early neonatal mortality rate (deaths in first 6 days
per 1000 live births)
Late neonatal mortality rate (deaths at ages 7-27
completed days per 1000 live births)
Post-natal mortality rate (deaths at ages >28 days
but < 1 year per 1000 live births)
Infant mortality rate (deaths at age < 1 year per
1000 live births)
4.4
14.2
19.3
2.9
22.8
75.6
0.8
3.9
10.6
2.4
6.3
15.1
6.1
33.0
101.4
"Office of Population Censuses and Surveys, Series DH3, No. 26.
Low birthweight
Median birthweights in the UK are ~3300 g for singletons, 2500 g for twins,
1800 g for triplets and 1500 g for quadruplets and above (Botting et al, 1990),
and the proportion of low birthweight babies (<2500 g) rises with increasing
plurality. In the UK MRC/IVF register data, 12% of singletons, 55% of twins
and 94% of triplets or higher order babies were of low birthweight. After making
allowance for differences in maternal age and parity, there is some evidence that
these proportions are higher than expected compared with national data (Beral
and Doyle, 1990; Tan et al, 1992).
Mortality
Early fetal death in the form of 'vanishing sacs' is a fairly common occurrence
in a multiple pregnancy following assisted reproduction treatment. An ultrasound
study of 38 triplet pregnancies found that 50% experienced spontaneous reduction
of at least one embryo (Manzur et al, 1995). It is likely that a similar phenomenon
occurs in naturally conceived pregnancies, and multiple conception rates are
probably much higher than the detected multiple pregnancy and corresponding
multiple maternity rates. Late fetal death is increased in multiple pregnancy, with
the stillbirth rate being over three times higher in twins (14.2/4.4 = 3.2) than in
singletons and over four times higher in triplets and higher order births (19.3/
4.4 = 4.4) m England and Wales in 1991 (see Table I and Figure 2, Office of
Population Censuses and Surveys, Series DH3). Neonatal deaths (0-27 days)
show the greatest disparity by multiplicity, the twin rate being seven times and
the triplet plus rate >20 times (86.2/3.7 = 23.3) the singleton rate. Despite their
relative rarity, multiple births make a large contribution to overall mortality rates.
In England and Wales, multiple births made up 2.5% of all births, but 8% of all
stillbirths, 19% of all neonatal deaths and 7% of all postneonatal deaths in 1991
(Office of Population Censuses and Surveys, Series DH3). Mortality of multiple
114
Outcome of multiple pregnancy
Figure 2. Mortality by multiplicity of birth: England and Wales, 1991.
births resulting from assisted reproduction treatment generally follows this pattern
of increasing risk with increasing plurality (Lancaster, 1992), although there is
a tendency for the rates to be non-significantly higher than expected on the basis
of national rates (Beral and Doyle, 1990).
Morbidity
The vast majority of excess mortality in multiple births is attributable to a
low birthweight resulting from premature delivery (Chamberlain, 1991). Low
birthweight is also a major risk factor for infant and childhood morbidities such
as cerebral palsy, mental retardation and cataract. Two recent studies of the
incidence of cerebral palsy reported dramatically increased risks in multiple
births: twins had risks approximately five times higher and triplets 17 times
higher that in singletons (Petterson et ai, 1993), and the risks of producing at
least one child with cerebral palsy was estimated to be 1.5% for twin, 8% for
triplet and almost 50% for quadruplet pregnancies (Yokoyama et al, 1995). The
prevalence of cerebral palsy in babies resulting from infertility treatments is not
known, but results such as this make long-term follow-up studies imperative.
Conclusion
A dramatic increase in the numbers of multiple births in developed countries has
stimulated interest in the progress and outcome of such pregnancies. As well as
considerable health risks, the social, psychological and financial impact of
115
RDoyle
multiple births, especially higher order multiple births, has been highlighted
(Botting et al, 1990; Levine et al, 1992; Gissler et al, 1995). Treatment for
infertility has improved dramatically over the past 15 years, but it is now well
recognized that a balance has to be struck between offering a technique which
has a reasonable chance of success whilst keeping the proportion of multiple
pregnancies low. Methods of fetal reduction have been shown to be safe and
effective in reducing higher order pregnancies (Evans et al, 1994), but the
psychological and ethical issues surrounding the technique remain problematic
and the avoidance of these pregnancies must be a priority. The Human Fertilization
and Embryology Authority of the UK recommend the replacement of up to three
embryos per IVF cycle (four in exceptional circumstances), and the Royal College
of Obstetricians and Gynaecologists (RCOG; London, UK) has recommended
abandonment of ovarian stimulation cycles if multiple mature follicles are present
(RCOG, 1990). Recent trials on the replacement of two good quality embryos
for specific groups of women reduced triplet pregnancy rates but did not
significantly depress pregnancy rates (Nijs et al, 1993; Staessen et al, 1993;
Kodama et al, 1995), and there is recent evidence that the replacement of fewer
embryos in assisted reproduction treatment cycles in the USA has been followed
by a general decline in the number of higher multiple pregnancies (Evans et al,
1995). The availability of large amounts of patient and treatment data would
allow a statistical investigation of risk factors for multiple birth in assisted
reproduction treatment, with the ultimate aim of tailoring treatment regimes to
different types of patient. Pooling of data both nationally and internationally for
this purpose is to be recommended.
References
Beral, V. and Doyle, P. (1990) Births in Great Britain resulting from assisted conception, 1978—
87. Br. Med. J., 300, 1229-1233.
Beral, V., Doyle, P., Tan, S.L. et al, (1990) Outcome of pregnancies resulting from assisted
conception. Br. Med. Bull., 46, 753-768.
Botting, B.J., MacDonald-Davies, I. and MacFarlane, A. (1987) Recent trends in the incidence of
multiple births and associated mortality. Arch. Dis. Childhood, 62, 941-950.
Botting, B., MacFarlane, A. and Price, F. (1990) Three, Four or More: A Study of Triplet and
Higher Order Births. HMSO, London, UK.
Chamberlain, G. (1991) Multiple pregnancy. Br. Med. J., 303, 111-115.
Chen, C.J., Lee, T.K., Wang, C.J. et al., (1992) Secular trend and associated factors of twinning
in Taiwan. Ada Genet. Med. Gemelloi, 41, 205-213.
Derom, C , Derom, R., Vlietinck, R. et al., (1993) Iatrogenic multiple pregnancies in East Flanders,
Belgium. Fertil. Steril., 60, 493-496.
Douglas, K.A. and Redman, C.W. (1994) Eclampsia in the United Kingdom. Br. Med. J., 309,
1395-1400.
Doyle, P., Beral, V., Botting, B. and Wale, C.J. (1991) Congenital malformations in twins in
England and Wales. J. Epidemiol. Community Health, 45, 43-48.
Evans, M.I., Dommergues, M., Timor-Tritsch, I. et al, (1994) Transabdominal versus transcervical
and transvaginal multifetal pregnancy reduction: international collaborative experience of more
than one thousand cases. Am. J. Obstet. Gynecol., 170, 902-909.
Evans, M.I., Littman, L., Louis, L.S. et al, (1995) Evolving patterns of iatrogenic multifetal
116
Outcome of multiple pregnancy
pregnancy generation: implications for aggressiveness of infertility treatments. Am. J. Obstet.
GynecoL, 172, 1750-1755.
Gissler, M., Silvero, M. and Hemminki, E. (1995) In-vitro fertilization pregnancies and perinatal
health in Finland 1991-1993. Hum. Reprod., 10, 1856-1861.
Jewell, S.E. and Yip, R. (1995) Increasing trends in plural births in the United States. Obstet.
GynecoL, 85, 229-232.
Kodama, H., Fukuda, J., Kambe, H. et al. (1995) Prospective evaluation of simple morphological
criteria for embryo selection in double embryo transfer cycles. Hum. Reprod., 10, 2999-3003.
Lancaster, P. (1992) International comparisons of assisted reproduction. Assist. Reprod. Rev., 2,
212-221.
Levine, M.I., Wild, J. and Steer, P. (1992) Higher multiple births and the modern management of
infertility in Britain. Br. J. Obstet. Gynaecol, 99, 607-613.
Logerot-Lebrum, H., De Mouzon, J., Hatchelot, A. and Spira, A. (1995) Pregnancies and births
resulting form in vitro fertilization: French national registry, analysis of data 1986-90. Fertil.
Steril., 64, 747-756.
Luke, B. (1994) The changing pattern of multiple births in the United States: maternal and infant
characteristics, 1973 and 1990. Obstet. GynecoL, 84, 101-106.
Manzur, A., Goldsman, M.P., Stone, S.C. et al., (1995) Outcome of triplet pregnancies after
assisted reproductive techniques: how frequent are the vanishing embryos. Fertil. Steril., 63,
252-257.
Millar, W.J., Wadhera, S. and Nimrod, C. (1994) Multiple births: trends and patterns in Canada,
1974-1990. Health Rep., 4, 223-250.
Neilson, J.P. (1992) Prenatal diagnosis in multiple pregnancies. Cum Opin. Obstet. GynecoL, 4,
280-285.
Nijs, M., Geerts, L., Van Roosendaal, E. et al., (1993) Prevention of multiple pregnancies in an
in-vitro fertilization program. Fertil. Steril., 59, 1245-1250.
Office of Population Censuses and Surveys. Birth Statistics. HMSO, London, UK, Series FM1,
Nos. 1-22.
Office of Population Censuses and Surveys. Mortality Statistics, Perinatal and Infant: Social and
Biological Factors. HMSO, London, UK, Series DH3, No. 26.
Petterson, B., Nelson, K.B., Watson, L. and Stanley, F. (1993) Twins, triplets, and cerebral palsy
in births in Western Australia in the 1980s. Br. Med. J., 307, 1239-1243.
Royal College of Obstetricians and Gynaecologists (1990) Guidelines on Assisted Reproduction
Involving Superovulation. RCOG, London, UK.
Staessen, C , Janssenswillen, C , Van den Abbeel, E. et al., (1993) Avoidance of triplet pregnancies
by elective transfer of two good quality embryos. Hum. Reprod., 8, 1650-1653.
Tan, S.L., Doyle, P., Campbell, S. et al., (1992) Obstetric outcome of in vitro fertilization
pregnancies compared with normally conceived pregnancies. Am. J. Obstet. GynecoL, 167,
778-784.
Tuppin, P., Blonde], B. and Kaminski, M. (1993) Trends in multiple deliveries and infertility
treatments in France. Br. J. Obstet. Gynaecol., 100, 383-385.
van Duivenboden, Y.A., Merkus, J.M. and Verloove-Vanhorick, S.P. (1991) Infertility treatment:
implications for perinatology. Eur. J. Obstet. GynecoL Reprod. Biol., 42, 201-204.
Wald, N.J. and Densem, J.W. (1994) Maternal serum free a-human chorionic gonadotrophin levels
in twin pregnancies: implications for screening for Down's syndrome. Prenat. Diagn., 14,
717-719.
Yokoyama, Y, Shimizu, T. and Hayakawa, K. (1995) Prevalence of cerebral palsy in twins, triplets
and quadruplets. Int. J. Epidemiol., 24, 943-948.
117
Discussion
Simpson: I have a comment just for the record. I think the difficulty in screening
in multiple gestations is less that of being able to perform amniocentesis, which
is now fairly simple with good ultrasound, but the lack of efficiency. It is difficult
if not impossible to screen for Down syndrome because algorithms were not
prepared for the detection of trisomies and neural tube defects in multiple
gestations. For example, if a 5% rate in the detection of a neural tube defect in
multiple gestation is accepted, then only 40% of cases are detected. In order to
achieve the 80% obtained in singleton gestations, we would have to perform
amniocentesis on 20% of cases. Thus, the greater problem is lack of detection
efficiency.
My specific question relates to your thoughts about the US, where most patients
are paying for IVF out of their own pockets at a rate of about US $5000 to
$10000 per treatment. Another $2000 or $3000 is added for ICSI. Given this, it
is difficult to demand restriction on the number of embryos that are transferred,
even though one can argue that if there are multiple order gestations, the family
would eventually never be able to pay all the bills. Thus, society would be
involved.
Perhaps a middle ground could be devised. Because pregnancy success is a
function of maternal age, one might transfer more embryos in, say a 35 to 38
year old woman than in a younger woman. In the absence of a flexible policy,
i.e. with a policy of restricted transfer, preimplantation genetic diagnosis to verify
the normalcy of embryos would probably be instituted. Thus, at least in the
USA, there would be a potential financial downside to restricting embryos. How
would you respond to our side of the Atlantic?
Doyle: I think that is a very good point about infertile couples desiring multiple
births. But they have to be made aware, and most clinics do make people aware,
of the long-term costs. The monetary costs of the treatment are an important
issue of course, but what are the long-term health costs? Having a child with
severe prematurity with a so called normal child is very costly to that family
and they should know the risk. My talk was based largely on national data.
Whether risks are higher in ART is not known, but the idea about cost has to be
long-term cost and not just financial cost. The other issue about changing
treatment algorithms according to the type of woman: her age, her diagnostic
problem, and so on to maximise the probability of producing a live birth, but
minimising the chance of multiplicity, seems entirely appropriate. With good
enough data, clinicians could work out suitable schedules for specific groups of
patients.
Zeilmaker: I very much appreciated one of your last remarks, namely that triplets
should be avoided at all costs. In this regard, I would like to make a small
comment on the table you presented. The percentage of triplets of all pregnancies
was about four in your table, and it was a bit lower in another table presented
earlier. If we could calculate the number of triplets on the basis of pregnancies
only originating from transfers of three or more embryos, that percentage would
118
Discussion
be much higher. In Rotterdam, we calculated that about 10% of all pregnancies
originating after transfer of three embryos were triplet pregnancies, which led to
the decision that never more than two embryos can be transferred. So our doctors
have to tell the women that if you get a pregnancy after transfer of 3 embryos
you have a 10% chance of getting a triplet. This is of course totally unacceptable.
We have analysed the effect of age in Rotterdam, and up to the age of 38 the
risk of producting a triplet is the same when three embryos or more are transferred
and the patient becomes pregnant. We have to be careful with transfers even
with patients of advanced age.
Doyle: I appreciate your comment and I think the figures you quote illustrate
how treatments have improved. I have data only from 1978 to 1987, when
success was much lower. We had data on numbers of embryos transferred and
the plurality of pregnancy, showing a clear link, but nowhere near as high as
you have just quoted. However, the data are rather old, and you are referring to
treatments using new ovarian stimulation methods.
Diedrich: Is fetal reduction a solution to avoid triplet deliveries?
Doyle: This is out of my field, but I have read the literature. Data that have been
published show low risks for both mother and remaining fetus, but there is much
literature about the psychology and the short- and long-term psychological aspects
of the procedure.
Camus: With a fetal reduction in a triplet pregnancy, the outcome to expect is
that of a twin pregnancy. The obstetrical outcome is similar to the obstetrical
outcome of control twin pregnancies.
Devroey: We have reported that perinatal morbidity is lower after reduction of
triplet pregnancies to twin pregnancies.
Anonymous: How many pregnancies are lost when fetal reduction is performed?
Camus: Several studies report that pregnancy loss after embryo reduction of
triplets to twins is about 10%. There is a similar take-home baby rate in reduced
and non-reduced triplet pregnancies but the perinatal morbidity is significantly
less after embryo reduction.
Nygren: I have three comments. First, the question of fetal reduction is very
much dependent on which society you live in. In Scandinavia where we have
been traditionally very liberal to abortion as part of a treatment schedule, fetal
reduction is not acceptable. We do it occasionally, but it would never be 'part of
a system'. That would endanger the whole abortion law in my country.
Secondly, yes, the women would like to have two eggs. My question is: would
it be possible to say, yes you will have two eggs, but one at a time? If the other
one is frozen, how many are lost?
Thirdly, most multiple pregnancies in my country arise not from IVF anymore,
but originate from ovarian stimulation alone. We have issued guidelines now
that ovulation must not be induced if you have more than three follicles of 18
mm or more present.
Devroey: With women less than 37 years of age, the implantation rate will be
20% when one embryo is replaced. Secondly, we must make guidelines about
ovulation induction: any gynaecologist can convert from ovulation induction to
119
Discussion
IVF when he sees that there are too many follicles. I think that could be a
solution to that problem.
Ron-El: Vanishing twins are accepted to occur with a frequency of 18 to 20%.
So we are lucky that triplets go spontaneously to twins in around 50% of the
cases. In our programme, we are hardly trying to convince the patients to reduce
the triplets to twins. The overall pregnancy loss in recent reports is between 3
and 5% and this is also the range in our series.
Tarlatzis: An interesting presentation at the 1995 IFFS meeting in Montpellier
from Salat-Baroux showed a prospective study comparing perinatal outcome in
non-reduced triplets and reduced triplets. Perinatal outcome was significantly
better when the triplets were reduced in this prospective study.
Diedrich: Is a lower success rate more acceptable than to reduce a triplet
pregnancy?
Doyle: If you choose embryos carefully, there is no reduction in success rate.
There are ways around the problem. Again, I am not a practitioner, but recent
reports, using only good-quality embryos, look very promising. Two-good quality
embryos, or even one good-quality embryo, has quite a good chance of
implantation.
Liebaers: Transferring only two good-quality embryos does not reduce the rate
of twins, but does reduce the rate of higher multiples. If we want to reduce the
rate of twins, we should transfer one. Another issue is whether we get the same
success rate by using frozen embryos in subsequent cycles. That question was
not answered.
120