Screening for Down’s syndrome: effects, safety, and cost
effectiveness of first and second trimester strategies
R E Gilbert, C Augood, R Gupta, A E Ades, S Logan, M Sculpher, J H P van der Meulen
Objective To compare the effects, safety, and cost
effectiveness of antenatal screening strategies for
Design Analysis of incremental cost effectiveness.
Setting United Kingdom.
Main outcome measures Number of liveborn babies
with Down’s syndrome, miscarriages due to chorionic
villus sampling or amniocentesis, healthcare costs of
screening programme, and additional costs and
additional miscarriages per additional affected live
birth prevented by adopting a more effective strategy.
Results Compared with no screening, the additional
cost per additional liveborn baby with Down’s
syndrome prevented was £22 000 for measurement of
nuchal translucency. The cost of the integrated test
was £51 000 compared with measurement of nuchal
translucency. All other strategies were more costly and
less effective, or cost more per additional affected baby
prevented. Depending on the cost of the screening
test, the first trimester combined test and the
quadruple test would also be cost effective options.
Conclusions The choice of screening strategy should
be between the integrated test, first trimester
combined test, quadruple test, or nuchal translucency
measurement depending on how much service
providers are willing to pay, the total budget available,
and values on safety. Screening based on maternal
age, the second trimester double test, and the first
trimester serum test was less effective, less safe, and
more costly than these four options.
The provision of screening services in the NHS lags far
behind advances in performance of screening tests
over the past decade.1In 1998, 57% of antenatal care
providers offered the second trimester double test for
Down’s syndrome and 8% offered screening based only
on maternal age.2Few NHS providers offered the more
effective nuchal translucency measurement (7%) or
quadruple test (3%). The integrated test, which is the
most effective screening test and involves testing in the
first and second trimesters,3is available only privately.
The main considerations for providers of screening
for Down’s syndrome should be minimising the risk of
babies with Down’s syndrome being missed by the test,
reducing miscarriage due to amniocentesis or chori-
onic villus sampling,and costs.We therefore compared
the effects, safety, and cost effectiveness of nine
strategies currently available for screening for Down’s
syndrome in the United Kingdom.
We compared no screening with nine screening
strategies offered in the first or second trimesters (table
1). Prenatal diagnosis included abdominal chorionic
amniocentesis thereafter if screening indicated a
greater than 1 in 300 risk of Down’s syndrome. We
determined the number of liveborn children with and
without Down’s syndrome,pregnancy losses (including
terminations, spontaneous losses, and miscarriages
due to chorionic villus sampling or amniocentesis),and
the healthcare cost of the screening programme in
10 000 women with the age distribution of women
delivering in England and Wales in 1995.
Estimates used in the model are based on a systematic
review,1other published sources, and discussions with a
reference group of UK service providers and users (see
acknowledgements). We assumed that 100% of women
attend antenatal clinic between 10 and 14 completed
weeks of gestation (median 12 weeks gestation) and are
offered tests in the first trimester or between 15 and 19
weeks (median 15 weeks) for tests in the second trimes-
ter. Table 1 gives the time required for counselling and
for processing of screening and diagnostic tests and the
proportion of women accepting procedures.
Based on the age specific prevalence of pregnan-
cies affected by Down’s syndrome at 10 and 16 weeks’
gestation and at term,6 10we took account of the weekly
risk of spontaneous fetal loss in affected and unaffected
pregnancies. Overall, we estimated that 45% of affected
fetuses would be spontaneously lost between 10 weeks’
gestation and term. The test characteristics for nuchal
translucency measurements were derived from a study
fetuses,4 11adjusted for verification bias due to termina-
tion of affected fetuses who would otherwise have been
The test characteristics for all serum tests were
derived from the analysis of stored serum samples
from 77 women with affected pregnancies and 385
Institute of Child
R E Gilbert
senior lecturer in
research fellow in
research assistant in
senior lecturer in
Institute of Child
A E Ades
reader in biostatistics
Centre for Health
University of York,
senior research fellow
Public Health and
School of Hygiene
J H P van der
senior lecturer in
R E Gilbert
BMJ VOLUME 32325 AUGUST 2001 bmj.com
amniocentesis.6We selected combinations of serum
analytes that, according to the systematic review,6
performed best.Alternative combinations of analytes—
for example, substitution of free ? human chorionic
gonadotrophin for total human chorionic gonado-
trophin or inhibin A for unconjugated oestriol, would
not substantially affect test performance. We assumed
that all women would have an ultrasound dating scan
and that, for second trimester serum tests, maternal
weight would be taken into account.
We estimated the risk of giving birth to a Down’s
syndrome baby for each woman, based on the screen-
ing test results and maternal age. Firstly, we calculated
the distribution of likelihood ratios for the screening
test separately for pregnancies affected and unaffected
by Down’s syndrome using Monte Carlo simulation12
and assumed independence between the nuchal trans-
lucency measurement,first trimester serum test (PAPP-
A), and quadruple test. We then calculated the
distributions of risk of an affected pregnancy for each
year of maternal age by multiplying likelihood ratios by
the age specific risk of an affected live birth. Finally, we
calculated a single, age specific distribution of risk after
unaffectedpregnancies, all ofwhom had
screening by taking account of the proportion of preg-
nant women at each maternal age and the age specific
risk of Down’s syndrome.
The costs of screening tests included laboratory
expenses (consumables and staff), informing the
women of the results (by telephone if positive, by post
if negative), service costs (including processing results
and monitoring the service), overheads, and (for
nuchal translucency measurement) training (table 1).
The costs of chorionic villus sampling and amniocen-
tesis included counselling before the procedure,equip-
ment and staff to do the procedure, laboratory
expenses (consumables and staff, non-reagent and
labour costs), and overheads. For all these costs, we
assumed an existing infrastructure for antenatal
screening and diagnosis of Down’s syndrome. We also
estimated costs of events arising from screening (table
2) All costs were adjusted to June 1998 prices (for full
details see www.ich.ucl.ac.uk/srtu).
The sensitivity analyses examined the effect of a 50%
increase and decrease in the costs of chorionic villus
sampling,amniocentesis,the screening test,and nuchal
translucency measurement. We also tested the effect of
varying the cut-off point for a positive result.
Effects and costs
Table 3 and the figure show the effects and costs of no
screening and of the nine screening strategies.
Measurement of nuchal translucency would result in
7.6 fewer births of babies with Down’s syndrome com-
pared with no antenatal screening, at a total cost of
£171 000 per 10 000 pregnant women. The incremen-
tal cost effectiveness ratio for nuchal translucency
compared with no screening is £22 000 (171 000/7.6)
Table 1 Detection rate and estimated uptake, process time, and cost of screening strategies for Down’s syndrome
First trimester screening (10 to 14 weeks)‡:
Nuchal translucency measurement
First trimester double test (PAPP-A, HCG)
First trimester combined test (nuchal translucency, PAPP-A, HCG)
Second trimester screening (15 to 19 weeks)‡:
Second trimester double test (AFP, HCG)
Triple test (AFP, HCG, uE3)
Quadruple test (AFP, HCG, uE3, inhibin A)
Integrated test (first trimester: nuchal translucency, PAPP-A;
second trimester: quadruple test)
Amniocentesis (>15 weeks)
Chorionic villus sampling (11-14 weeks)‡
Surgical dilatation, evacuation (11 to 13 weeks)‡
Medical with mifepristone (>14 weeks)
Detection rateReported rateUptake
AFP=? fetoprotein, HCG=? human chorionic gonadotrophin, PAPP-A=pregnancy associated plasma protein A, uE3=unconjugated oestriol.
*Time between procedure and date results given to the woman.
†For details of derivation see technical report.
‡Refers to completed weeks.
§Detection rate for a 5% false positive rate.
¶1 week allowed before procedure for administration and counselling.
**Culture of amniotic fluid takes 2 weeks, preparation of chorionic villus sample takes 2 days.
Table 2 Risk and costs of events after screening for Down’s syndrome
Miscarriage due to the procedure
Chorionic villus sampling
Sample failure and repeat procedure in all
Chorionic villus sampling
Spontaneous loss in unaffected pregnancies (10 weeks
Spontaneous loss in Down’s syndrome pregnancies (10
weeks to term)†
RiskReported rangeUnit cost
0.0 to 1.9%1
−2.2% to 2.0%*1
0.5 to 1.0%1
1.0 to 1.6%1
*Difference compared with second trimester amniocentesis.
†Costs incurred because dilatation and evacuation assumed for all losses.
BMJ VOLUME 32325 AUGUST 2001 bmj.com
per prevented birth of a baby with Down’s syndrome.
The integrated test results in 2.0 fewer births of babies
with Down’s syndrome than the nuchal translucency
measurement at a total cost of £276 000. The
incremental cost effectiveness ratio of the integrated
test compared with nuchal translucency is £52 000
(276 000− 171 000)/2.0) per affected baby prevented.
Service providers with an intermediate total budget
should consider the first trimester combined test or the
quadruple test. Compared with measuring nuchal
translucency, the first trimester combined test costs an
extra £57 000 per affected liveborn baby prevented
(total cost £238 000). At a similar total cost (£241 000),
the quadruple test costs an additional £75 000 per
affected baby prevented compared with measuring
Effects and safety
The integrated test is the most effective and safest strat-
egy. All other strategies result in more liveborn babies
with Down’s syndrome and more miscarriages of unaf-
fected pregnancies due to amniocentesis or chorionic
villus sampling. Compared with no screening, the inte-
grated test results in 0.14 miscarriages due to chorionic
villus sampling or amniocentesis per birth of a Down’s
syndrome baby prevented. The next safest strategies,
compared with no screening, are the first trimester
combined test (0.22 miscarriages), nuchal translucency
measurement (0.34), and the quadruple test (0.42).
A 50% increase or decrease in the unit costs of chorionic
villus sampling and amniocentesis would hardly alter
the cost effectiveness. If the cost of amniocentesis fell by
50%, the quadruple test would enter the efficiency fron-
tier (cost effectiveness ratio compared with nuchal trans-
lucency £39 000 and for the integrated test compared
with the quadruple test £51 000).
The cost effectiveness of the screening strategies is
most susceptible to variation in the costs of the screen-
ing test. A 50% decrease in the unit costs of all screen-
ing tests would increase the cost effectiveness ratio of
all strategies above that of the integrated test (£19 000
compared with no screening), although the cost effec-
tiveness ratios would be close for nuchal translucency
measurement and the first trimester combined test. A
50% increase in the unit costs of measuring nuchal
translucency would increase the cost effectiveness ratio
for nuchal translucency compared with no screening
to £25 000. The cost effectiveness ratios for the first tri-
mester combined, quadruple, and integrated tests
would be similar (£51 000-£56 000). The second
trimester double test, triple test, maternal age, and first
trimester serum test were not cost effective at any of the
Varying the cut-off point for doing amniocentesis
or chorionic villus sampling between risks of 1 in 100
and 1 in 400 did not alter the cost effectiveness ranking
of the screening strategies. Similarly, varying the risk
cut-off point corresponding to a 5% false positive rate
for the screening test did not alter the ranking (see
The nuchal translucency measurement,quadruple test,
first trimester combined, and integrated tests represent
the best options in terms of effectiveness, cost effective-
ness, and safety. All other strategies would be less
Table 3 Number of live births affected by Down’s syndrome, miscarriages due to chorionic villus sampling or amniocentesis, and total
costs per 10 000 pregnancies and incremental cost effectiveness ratios assuming risk of 1 in 300 as cut-off point for positive result
Maternal age and amniocentesis
Maternal age and chorionic villus sampling
First trimester double test
Second trimester double test
Nuchal translucency measurement§
First trimester combined test
Total No of
affected live births
Miscarriages due to
Incremental cost per affected birth
*Ranked in order of effectiveness (number of Down’s syndrome live births).
†Above cost of no screening.
‡Compared with no screening
§Most efficient strategies.
¶Compared with nuchal translucency.
D=dominance—that is, there is a cheaper and more effective strategy; ED=extended dominance—that is, a more effective option has a lower incremental cost
Age + amniocentesis
Age + chorionic villus sampling
First trimester double
Second trimester double
Total cost for 10 000 women (£1000s)
No of live born babies with Down's syndrome
Cost effectiveness of screening strategies for Down’s syndrome with
a risk of 1 in 300 as cut-off point for positive result. The continuous
line is the efficiency frontier. Gradient represents the additional cost
incurred per additional birth prevented of an affected baby by
adopting a more effective strategy (compared with next cheapest
strategy). Strategies above the efficiency frontier are dominated
(more costly and less effective than nuchal translucency
measurement or integrated test) or ruled out by extended dominance
(more costly or less effective and resulted in higher costs per
prevented birth than a more effective option).13
BMJ VOLUME 32325 AUGUST 2001bmj.com
effective, cost more per additional birth of an affected
baby prevented, and be less safe. This finding was
robust in the sensitivity analyses.
The choice between the four options depends on
how much service providers are willing to pay to
prevent one affected liveborn baby, on the total budget
available for antenatal screening, and on how much
service providers value safety. We would expect service
providers to be willing to spend at least £30 000 to
£40 000 per additional affected baby prevented, as this
reflects the incremental costs paid by most service pro-
viders that offer screening to all women.14
Implications for practice
These findings contrast with current practice in the
United Kingdom, where the second trimester double
test is most commonly offered.2Moving from the
double test to the first trimester combined test or quad-
ruple test would not cost any more and would result in
1.5 (for the first trimester combined) or 1.2 (for the
quadruple test) fewer affected liveborn babies for every
10 000 pregnancies (table 2). Alternatively, the nuchal
translucency measurement would be more effective
than the double test (0.3 fewer affected babies) at a total
cost saving of £70 000. Finally, moving from the double
test to the integrated test would result in 2.3 fewer
affected babies and cost £13 000 per additional affected
Limited research on women’s preferences suggests
that the choice of screening strategy should be based
primarily on minimising the number of affected preg-
nancies that are missed and miscarriages due to chori-
onic villus sampling or amniocentesis.15 16Timing of
test results and termination seem to be a secondary
consideration.Nevertheless,given the realistic distribu-
tion of gestation at attendance for antenatal screening
that we used in our analysis,only 31% of women would
be in time for a surgical termination (before 13 weeks
gestation) after nuchal translucency or age based
screening and none after the first trimester combined
test (see www.ich.ucl.ac.uk/srtu).
Our results are susceptible to variation in performance
pregnancies affected by Down’s syndrome is highly con-
sistent across different studies,1whereas performance of
nuchal translucency measurement varies widely.17 18
Such variation may be explained by verification bias,
chance,referral patterns,variation in test reliability,19and
the use of repeated measurements. We used perform-
ance characteristics for nuchal translucency measure-
ment derived from routine screening of 95 000
pregnancies, which were then adjusted for verification
bias.4However,we do not know how much test perform-
ance was optimised by repeated measurements or refer-
ral for a second opinion.In addition,the performance of
the integrated test, which was derived by combining test
characteristics for serum tests and the nuchal translu-
cency measurement from different datasets, needs to be
evaluated in a study of pregnant women.
A further concern, examined in the sensitivity
analysis,is the possibility of an increased relative risk of
spontaneous miscarriage in women with a positive test
result ( > 1 in 300 risk of Down’s syndrome) compared
with those with a negative result. Reports of an
increased relative risk range from 1.6 to 2.8 for nuchal
translucency measurements >3mm and would dimin-
ish the detection rate of nuchal translucency measure-
ment.20 21In our analysis, a relative risk of 2.8 would
result in 1.3 more liveborn babies with Down’s
syndrome per 10 000 pregnancies after measuring
nuchal translucency, but the strategy would still be cost
The choice of screening strategy may be affected by
several factors that were beyond the scope of our
analysis. Firstly, we did not adopt a societal perspec-
tive,13which would seek to maximise health gain for a
given cost. Such analyses are problematic in terms of
measures of effect—for example, should the outcome
measure be prevented liveborn baby with Down’s syn-
drome or quality adjusted life years.22There are also
difficult decisions about which costs to include. Some
previous cost effectiveness analyses included direct and
indirect lifetime incremental costs per liveborn baby
with Down’s syndrome (including education, health,
and lost productivity). These ranged from £85 000 in
199023to £340 000 in 1996.24The studies avoided
including other indirect costs attributable to fetal losses
by assuming a replacement pregnancy.
If the lowest estimate (£85 000) for the incremen-
tal lifetime costs associated with a Down’s syndrome
child is included in our analysis the net saving
compared with no screening would be largest for the
integrated test. The net saving would be £548 000
(costs avoided (9.7 Down’s syndrome babies prevented
× £85 000) − screening programme costs (£276 000)).
Factors that could affect results
Three factors that we did not consider may affect our
results. Firstly, capital costs for additional laboratory or
clinical capacity may be incurred for use of the enzyme
linked immunosorbent assay (ELISA) test for inhibin A
(the fourth serum marker in the quadruple test), imple-
mentation of routine nuchal translucency measurement,
or the creation of additional facilities for chorionic villus
sampling.Secondly,uptake of all screening tests and ter-
mination may not be the same for all strategies and may
be lower for the integrated test,which requires two visits.
Decreased uptake of termination in later pregnancy is
likely and would mean that our analysis overestimates
the effectiveness of the integrated and quadruple tests.
What is already known on this topic
Screening strategies that combine nuchal translucency measurement
with serum testing perform better than either of these tests used alone
Serum testing in the second trimester using the triple test is cost
effective compared with screening based on maternal age
What this study adds
The integrated test is the most effective, safest, and most expensive
The choice of screening strategy should be between the integrated test,
first trimester combined test, quadruple test, or measurement of nuchal
Screening based on maternal age, the second trimester double test, and
the first trimester serum test is less effective, less safe, and more costly
than the above options
BMJ VOLUME 32325 AUGUST 2001 bmj.com
Thirdly, service providers need to provide for women
who attend antenatal booking clinic after the first
trimester (up to 35% based on results from 21 000
women in six inner London
Alternative strategies for late attenders include no
screening, the quadruple test, or strategies to encourage
earlier attendance at antenatal booking clinic—for
example,by speeding up referrals from primary care.
The choice of screening strategy should be between the
integrated test, first trimester combined test, quadruple
test, or nuchal translucency measurement. Screening
based on maternal age,the second trimester double test,
and the first trimester serum test would be less effective,
less safe,and more costly than these four options.
John Kingdom and Susan Bewley provided advice on the devel-
opment of the project and commented on drafts of the report.
Maureen Dalziel chaired the reference group and commented
on the study design. Roxanne Chamberlain, Jean Chapple, Nick
Fisk,Mike Gill,David Highton,M L Ko,Mike Lobb,Lucy Moore,
Tracey Reeves, and Tracy Stein were members of the reference
group and commented on the study design and results. Susan
Bewley, Elizabeth Dormandy, Ross Hastings, Wayne Huttly,
Linda Mulhair, Kypros Nicolaides, and Pran Pandya provided
information on screening practices. The report may not reflect
the views of the funding body.
Funding: The project was commissioned by Maureen
Dalziel, Sally Davies, and Ron Kerr from the London regional
office of the NHS Executive.
Contributors: REG, AEA, SL, MS, and JHPvdM contributed
to the study design. CA reviewed the literature. RG did the com-
puter programming, which was supervised by AEA. MS advised
on the economic analysis. JHPvdM devised the analytical
strategy for the screening tests. All authors contributed to
writing the manuscript. REG wrote the final version, and acts as
guarantor for the study.
Competing interests: None declared.
1Wald NJ, Kennard A, Hackshaw A, McGuire A. Antenatal screening for
Down’s syndrome. J Med Screen 1997;4:181-246.
Wald NJ, Huttly WJ, Hennessy CF. Down’s syndrome screening in the UK
in 1998. Lancet 1999;354:1264.
Wald NJ, Watt HC, Hackshaw AK. Integrated screening for Down’s
syndrome based on tests performed during the first and second
trimesters. N Engl J Med 1999;341:461-7.
Nicolaides KH,Snijders RJ,Cuckle HS.Correct estimation of parameters
for ultrasound nuchal translucency screening. Prenat Diagn 1998;18:519-
Wald NJ, Hackshaw AK. Combining ultrasound and biochemistry in
first-trimester screening for Down’s syndrome. Prenat Diagn 1997;17:821-9.
6Hecht CA, Hook EB. The imprecision in rates of Down syndrome by
1-year maternal age intervals:a critical analysis of rates used in biochemi-
cal screening. Prenat Diagn 1994;14:729-38.
Wald NJ, Cuckle HS, Densem JW, Nanchahal K, Royston P, Chard T, et al.
Maternal serum screening for Down’s syndrome in early pregnancy. BMJ
Wald NJ, Cuckle HS, Densem JW, Kennard A, Smith D. Maternal serum
screening for Down’s syndrome: the effect of routine ultrasound scan
determination of gestational age and adjustment for maternal weight. Br
J Obstet Gynaecol 1992;99:144-9.
Wald NJ, Densem JW, George L, Muttukrishna S, Knight PG. Prenatal
screening for Down’s syndrome using inhibin-A as a serum marker. Pre-
nat Diagn 1996;16:143-53.
10 Macintosh MC,Wald NJ,Chard T,Hansen J,Mikkelsen M,Therkelsen AJ,
et al.Selective miscarriage of Down’s syndrome fetuses in women aged 35
years and older. Br J Obstet Gynaecol 1995;102:798-801.
11 Snijders RJ, Noble P, Sebire N, Souka A, Nicolaides KH. UK multicentre
project on assessment of risk of trisomy 21 by maternal age and fetal
nuchal-translucency thickness at 10-14 weeks of gestation.Fetal Medicine
Foundation First Trimester Screening Group. Lancet 1998;352:343-6.
12 Van der Meulen JHP, Mol BW, Pajkrt E, van Lith JM, Voorn W. Use of the
disutility ratio in prenatal screening for Down’s syndrome. Br J Obstet
13 Siegel JE,Weinstein MC,Torrance GW.Reporting cost-effectiveness stud-
ies and results.In:Gold MR,Siegel JE,Russell LB,Weinstein MC,eds. Cost
effectiveness in health and medicine. New York: Oxford University Press,
14 Torgerson DJ, Spencer A. Marginal costs and benefits. BMJ 1996;312:35-6.
15 Heckerling PS, Verp MS, Hadro TA. Preferences of pregnant women for
amniocentesis or chorionic villus sampling for prenatal testing: compari-
son of patients’ choices and those of a decision-analytic model. J Clin Epi-
16 Kuppermann M, Feeny D, Gates E, Posner SF, Blumberg B, Washington
AE. Preferences for women facing a prenatal diagnostic choice:
long-term outcomes matter most. Prenat Diagn 1999;19:711-6.
17 Palomaki GE, Knight GJ, McCarthy JE, Haddow JE, Donhowe JM.
TI-Screening of maternal serum for fetal Down’s syndrome in the first
trimester. N Engl J Med 1998;338:955-61.
18 Mol BW,Lijmer JG,van der Meulen JHP,Pajkrt E,Bilardo CM,Bossuy PM.
Effect of study design on the association between nuchal translucency
measurement and Down syndrome. Obstet Gynecol 1999;94:864-9.
19 Pajkrt E, Mol BW, Boer K, Drogtrop AP, Bossuyt PM, Bilardo CM. Intra-
and interoperator repeatability of the nuchal translucency measurement.
Ultrasound Obstet Gynecol 2000;15:297-301.
20 Hyett J, Moscoso G, Papapanagiotou G, Perdu M, Nicolaides KH. Abnor-
malities of the heart and great arteries in chromosomally normal fetuses
with increased nuchal translucency thickness at 11-13 weeks of gestation.
Ultrasound Obstet Gynecol 1996;7:245-50.
21 Bewley S, Roberts LJ, Mackinson AM, Rodeck C. First trimester fetal
nuchal translucency: problems with screening the general population.2.
Br J Obstet Gynaecol 1995;102:386-8.
22 Ganiats TG.Justifying prenatal screening and genetic amniocentesis pro-
grams by cost-effectiveness analyses: a re-evaluation. Med Decis Making
23 Shackley P,McGuire A,Boyd PA,Dennis J,Fitchett M,Kay J, et al.An eco-
nomic appraisal of alternative prenatal screening programmes for
Down’s syndrome. J Public Health Med 1993;15:175-84.
24 Beazoglou T, Heffley D, Kyriopoulos J, Vintzileos A, Benn P. Economic
evaluation of prenatal screening for Down syndrome in the USA. Prenat
25 Gibb DM, MacDonagh SE, Gupta R, Tookey PA, Peckham CS, Ades AE.
Factors affecting uptake of antenatal HIV testing in London: results of a
multicentre study. BMJ 1998;316:259-61.
(Accepted 21 May 2001)
Commentary: Results may not be widely applicable
Euan M Wallace, Sheila Mulvey
The average age at which women in Western countries
choose to have children continues to increase.Although
this trend has several important implications for the
outcome of pregnancy,one of the most obvious effects is
an increase in the incidence of Down’s syndrome.1
Prenatal screening for Down’s syndrome has become an
established part of antenatal care in many centres,
resulting in a reduction in the number of babies born
with Down’s syndrome in the populations screened.1
Several serum markers of Down’s syndrome have
been described, and different screening programmes
use different combinations of markers, making it diffi-
cult to compare their results.2The recent introduction
of first trimester serum and ultrasound screening has
further complicated such comparisons. Therefore, it is
often unclear which screening strategy is the most
effective, the most cost effective, or the most appropri-
ate for a given setting. Gilbert and her colleagues sum-
marise various approaches to screening and describe
the most cost effective strategies. Such information will
be invaluable to all those providing antenatal care.
Some authorities are likely to disagree with the conclu-
sions reached by Gilbert et al. There is evidence that
simple two marker protocols (such as ? fetoprotein and
free ß human chorionic gonadotrophin) can achieve
higher detection rates than those estimated by Gilbert
et al for more complex combinations, even after
Centre for Women’s
Euan M Wallace
clinical research fellow
E M Wallace
BMJ VOLUME 32325 AUGUST 2001bmj.com
correction for the high rate of spontaneous loss of
pregnancies affected by Down’s syndrome.2 3How can
A potential limitation of the analyses is that the
effectiveness modelling is based on an archived set of
serum samples derived from 77 women with pregnan-
cies affected by Down’s syndrome and 385 with
unaffected pregnancies, rather than an analysis of pub-
lished detection rates from prospective series, as was
done for nuchal translucency. If the archived data set
differs from the population being screened then the
performance characteristics may not be the same in
practice. This emphasises the importance of evaluating
screening tests locally,where the expertise exists,rather
than relying on a single archived serum bank.
It is also possible that marker combinations not ana-
lysed in this report will offer highly effective screening.
For example, second trimester serum screening has bet-
ter detection rates than first trimester tests.4The combi-
nation of nuchal translucency with second trimester
serum markers may prove more effective than the first
trimester test that Gilbert et al recommend.Several pro-
spective comparative studies are underway worldwide
that will investigate this possibility.
The cost comparisons of the various strategies
highlight that more expensive approaches may be the
most cost effective. Such insights will be extremely
helpful to health planners. However, the costings used
in the analyses are for a programme within a large
public service such as the NHS and will not easily
translate across healthcare systems. For example, the
£4.40 allowed for measuring nuchal translucency is
much less than the $200 (£143) estimated in the
United States.5Again, it will be important for local pro-
grammes to derive local costs before planning their
What is not covered by these analyses, although the
authors mention it, is patient preference. We have little
information about what women prefer in screening,
but there is some evidence that they want testing as
early in pregnancy as possible, even if a later test
performs slightly better.6Such information is surely
fundamental to the development of future screening
programmes. It would be unfortunate if the technical
advances in screening were assessed and compared
without considering women’s wishes. Similarly, it is
unacceptable that access to screening for Down’s
syndrome varies across the United Kingdom. Contri-
butions such as that by Gilbert et al will inform relevant
authorities, such as the UK National Screening
Committee,and help ensure that appropriate and high
quality screening is available to all women who want it.
Competing interests:EMW holds the patent for use of inhibin A
as a marker of Down’s syndrome.
1Cheffins T, Chan A, Haan EA, Ranieri E, Ryall RG, Keane RJ, et al. The
impact of maternal serum screening on the birth prevalence of Down’s
syndrome and the use of amniocentesis and chorionic villus sampling in
South Australia. Br J Obstet Gynaecol 2000;107:1453-9.
Wald NJ, Kennard A, Hackshaw A, McGuire A. Antenatal screening for
Down’s syndrome. J Med Screen 1997;4:181-246.
Spencer K. Second trimester prenatal screening for Down’s syndrome
using alpha-fetoprotein and free beta hCG:a seven year review. Br J Obstet
Berry E, Aitken DA, Crossley JA, Macri JN, Connor JM. Screening for
Down’s syndrome: changes in marker levels and detection rates between
first and second trimesters. Br J Obstet Gynaecol 1997;104:811-7.
Vintzileos AM, Ananth CV, Smulian JC, Day-Salvatore DL, Beazoglou T,
Knuppel RA. Cost-benefit analysis of prenatal diagnosis for Down
syndrome using the British or the American approach. Obstet Gynecol
Mulvey S, Wallace EM. Women’s knowledge of and attitudes to first and
second trimester screening for Down’s syndrome. Br J Obstet Gynaecol
Wallace and Mulvey criticise our study for relying on a
single archived serum bank. Instead, they advocate
comparing detection rates from “prospective studies,”
by which they mean routinely collected hospital data.
We disagree. It is well known that data from routine
practice may overestimate the performance of screen
tests because positive results lead to interventions
whereas affected pregnancies with negative results may
not be counted (termed verification bias).1Such bias
depends on the gestational age at testing and can be
allowed for only indirectly. Differences in detection
rates (assuming a constant false positive rate) may also
relate to uptake of interventions and referral patterns.
Finally, we required the test performance characteris-
tics rather than the detection rate in order to standard-
ise for age and cut-off point for positivity. The study
based on an archived serum bank avoided biased
detection of affected fetuses (all were karyotyped), took
account of ultrasound pregnancy dating, and reported
performance characteristics for all the serum tests.
1Mol BW, Lijmer JG, van der Meulen J, Pajkrt E, Bilardo CM, Bossuyt PM.
Effect of study design on the association between nuchal translucency
measurement and Down syndrome. Gynecol 1999;94:864-9.
BMJ VOLUME 323 25 AUGUST 2001 bmj.com