1 Poissy-Saint Germain Hospital, Centre hospitalier Poissy-Saint Germain, 10, rue de Champ Gaillard, 78303 Poissy Cédex, France, 2 Reproductive Epidemiology, University of Leeds, UK, 3 Public health evaluation and study unit, Henri Mondor Hospital, 51 avenue Maréchal de Lattre de Tassigny, 94000 Creteil, University Paris XII, 4 Antoine Béclère Hospital, 157 rue Porte de Trivaux, 92140 Clamart, 5 Dreux Hospital, 92 rue Rieuville, 28100 Dreux, 6 Metz-Thionville Hospital, 28 rue XXème Corps Américain, 57000 Metz, 7 Franco-Britannic Hospital,3 rue Barbès, 92300 Levallois Perret 8 Evreux Hospital, 17 rue St Louis, 27000 Evreux, Centre hospitalier Poissy-Saint Germain, 10 rue de Champ Gaillord, Poissy Cédex, France
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: alpha-fetoprotein/Down's syndrome/free ß-HCG/screening/ultrasound
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Sonographic and biochemical screening methods have developed simultaneously and independently. As a consequence, many women undergo a sequential two-stage screening with NT measurement followed by MSM at 1418 weeks when NT screening is negative. Although each method was developed to generate a 5% false-positive rate, a two-stage screening has a cumulative false-positive rate. Thus, the proportion undergoing invasive prenatal diagnosis can be as high as 10%, or even higher in countries offering invasive prenatal diagnosis to women aged >35 years, regardless of any screening result. This leads to an increase in iatrogenic loss of normal pregnancies and in all related costs. Therefore, the rate of invasive procedures for fetal karyotyping has become an important public health issue which needs to be controlled. One approach is to combine NT and MSM results into a single risk assessment and not to consider their results independently.
Thus, a multicentre prospective study was carried to assess the performance of a combined screening. The objective was to: (i) evaluate a potential increase in the detection rate with a combined screening programme; and (ii) examine the difference in related cost (Hagard and Carter, 1976; Ganiats, 1996
).
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Ultrasound examination was carried out between 12+0 weeks and 14+0 weeks of gestation. A 7.5 MHz transvaginal probe was used primarily, and a 5 MHz transabdominal probe was then used when the fetal position prevented a satisfactory examination. The aims were to confirm gestational age by crownrump length measurement, to look for major fetal abnormalities, and to measure NT using a previously described technique (Nicolaides et al., 1992).
Maternal venous blood was taken at between 14+1 weeks and 17+0 weeks of gestation. A separating gel was used in dry glass bottles (Beckton-Dickinson). Centrifugation was performed within 30 min, and serum was kept at 4°C to be assayed within 3 days. Free ß-HCG and AFP were assayed by immunoradiometric assay (ELISA-FBHCG and ELISA-AFP; CISBio International SA). Marker levels were expressed in multiples of the gestation-specific median (MoMs), and risks were estimated from maternal age and MoMs using commercial software (Free-ß-screen; CISBio International SA) that uses normal median values derived from a population of French pregnant women tested at 1417 weeks.
Amniocentesis was advised in two circumstances: (i) when NT was 3 mm and the procedure was delayed until after maternal blood was taken; and (ii) when the risk based on MSM results was
1 in 250 at term.
Some women had only ultrasound or only MSM but have nevertheless been included in the study statistical analysis, as the prevalence of Down's syndrome would otherwise be flawed. Pregnancies lost to follow-up or ending in spontaneous abortions without karyotyping were excluded from the analysis.
Retrospective statistical analysis of the data collected in the study was performed. NT measurements and MSM levels were expressed in MoMs using our own dataset, and risks were calculated from parameters based on meta-analyses and tailored to the study population as described previously (Cuckle, 1995; Cuckle and van Lith, 1999
). MSM MoMs were calculated using a regression equation based on median levels observed in the study for each completed week of gestation, and regressing median level on median days weighted for the number of women tested. Maternal weight was available for all pregnancies, and MoM values were divided by the expected weight-specific MoM obtained by regression analysis. Log Gaussian modelling was also carried out using standard methods (Royston and Thompson, 1992
) to estimate the predicted detection and false-positive rates from our parameters when applied to a population with a mean maternal age of 28 years and a coefficient of variation of 16%.
Economic evaluation
With regard to the cost of each of the three strategies, ultrasound alone, serum markers combined with maternal age alone, and the combination of ultrasound and serum markers were estimated from the viewpoint of the healthcare system. A cost per case diagnosed was computed, but this did not include the outcome of the pregnancy with an abnormal karyotype. The costs were computed from the second trimester through to the end of the pregnancy. Only empirical data were used for the economic analysis, and hence the lifetime cost of a person with Down's syndrome was not modelled. Societal costs were not included, and neither were intangible costs such as the loss of a normal baby or the unwanted birth of a child with Down's syndrome. The costs included were: the ultrasound performed in the hospital; maternal serum markers; one genetic counselling clinic; amniocentesis and karyotyping for patients with a positive test result; and termination of pregnancy for patients with abnormal fetal karyotype.
The costs of hospital procedures were derived from the national hospital cost database which estimated a cost per diagnosis-related group from accounting data. The costs of tests and genetic counselling were estimated from the cost program used by the national public insurer. Both are available on the Internet at www.le-pmsi.fr and www.cnam.fr. In order to estimate the cost of a given strategy, the cost of each event was multiplied by the probability of its occurrence. For any of the three strategies tested, the average cost per patient was: the cost of the initial test or tests; the costs of amniocentesis and karyotype for all patients with a positive test result (i.e. both true-positive and false-positive results); and the cost of termination of pregnancy for patients with an abnormal fetal karyotype. True- and false-positive rates were computed from the present patient population. Costs in French francs (2001) were converted into UK£ using the OECD purchasing power parity index (1 UK£ = 10 FF).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The median (25th; 75th centiles) maternal age was 30.5 (27.8; 33.9) years.
Ultrasound examination was carried out in 6327 women, and NT was successfully measured in 6234 of these (98.6%) by 41 operators. The mean duration of the scan was 12 min (range 526 min). NT was measured at 3 mm in 187 women (3.0%). Second-trimester MSM screening was carried out in 8297 women, and was positive in 486 (5.8%) who subsequently underwent amniocentesis. A total of 5506 women had both NT and serum screening, of whom only 13 (0.23%) had both raised NT and positive MSM. However, the proportion of invasive testing generated by this two-stage screening was 8.6%.
NT was 3 mm in 13/21 Down's syndrome pregnancies, giving a detection rate of 62%. Second-trimester MSM screening was carried out in 20 cases and was positive in 11, giving a detection rate of 55%. Seventeen cases (81%) had either raised NT or positive MSM screening, or both (7/20, 35%). Thus, the overall detection rate was 81% (17/21), for an invasive testing rate of 8.6%. One patient with a 4.3 mm NT at 12 weeks underwent chorionic villus sampling at 12 weeks and requested a termination of pregnancy before the MSM screening could be performed.
The median, 25th centile, 75th centile and standard deviation of each marker in both unaffected and Down's syndromes cases are listed in Table I.
|
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Using both ultrasound and MSM screening programmes had the advantage of confirming the independence of ultrasound and biochemical tests, and it was possible to assess the sensitivity and false-positive rates of both strategies for the same patients. In the present study, 8.6% of patients underwent an amniocentesis because of an increased risk with any of the tests. This was generated by the addition of the false-positive rates induced by each method. Because ultrasound and biochemical tests are independent risk markers, they select different populations. Indeed, only 0.2% of the patients were at high risk for both screening programmes. The present results are in agreement with those of others (Herman et al., 2000), who showed that in 508 normal pregnancies the screen-positive rates differed significantly in NT and second-trimester MSM screening programmes; moreover, when either test showed an increased risk for Down's syndrome, the probability of the other test predicting the outcome same was negligible.
A cumulative effect on the false-positive rate was also found by using a combination of ultrasound and biochemical screening (Audibert et al., 2001; Schuchter et al., 2001
). Although the methodology of one of these studies (Audibert et al., 2001
) was similar to the present investigation, the sensitivity for a 5% false-positive rate of combined screening (NT and MSM) was lower in the present study (80 and 90% respectively). This difference might be explained by the small size of Audibert et al.'s series, however.
Others (Kadir and Economides, 1997) also examined the effect of introducing first-trimester NT measurement on the results of second-trimester MSM. The likelihood ratio for a positive result and the positive predictive value of biochemical screening decreased from 9.1 to 5, and from 2.7% to 0.45% respectively. It was also found that implementation of NT screening significantly reduced the positive predictive value of second-trimester maternal serum biochemistry testing (Thilaganathan et al., 1997
).
Sequential screening programmes also increase the invasive procedure-related pregnancy loss rate. The present study involved women among whom the prevalence of Down's syndrome was below 1 in 1000 births, and an invasive testing rate of 8.6% might therefore be considered unacceptably high in this population who were at relatively low risk.
It was possible, however, to take advantage of the independence of ultrasound and biochemical markers to evaluate a single risk assessment that integrated results of NT measurement, free ß-HCG and AFP assays together with maternal age. The present results suggest that, with such an algorithm, it would be possible to keep the sensitivity as high as that obtained with sequential screening while maintaining an invasive testing rate at 5%. However, this would imply that a 25% increase in detection rate would delay risk calculation and invasive testing by 24 weeks. When termination of pregnancy is an issue, this has potential serious societal, medical and economic implications on screening policies.
Only women <38 years were included in the present study for two reasons: (i) the aim was to avoid an inclusion bias due to maternal age; and (ii) in France, invasive testing for fetal karyotype is routinely offered to women >38 years. Thus, it did not seem ethical to delay the invasive procedure once this had been planned and requested.
In the present study, NT measurement was reproducible: it could be performed in 98.6% of the cases, and the median value did not differ significantly between centres and between operators at centres where there were between three and 12 operators. Consequently, the present results varied from those reported elsewhere (Bewley et al., 1995; Roberts et al., 1995
; Kornman et al., 1996
; Haddow et al., 1998
), with the latter authors finding NT measurement to be unsuitable for screening due to its great inter-centre variability.
The results of the present study using ultrasound screening were also different from those reported in the three largest recent trials on the subject (Taipale et al., 1997; Snijders et al., 1998
; Schwarzler et al., 1999
). In the present study, sensitivity (66%) was higher than that published by others (54%; Taipale et al., 1997
), using the same threshold of 3 mm. However the present false-positive rate was much higher (3.0 versus 0.8%). It was suggested by others (Taipale et al., 1997
) that different equipment may explain differences in false-positive rates. Although in the present study high-resolution transvaginal ultrasonography was used primarily, transabdominal ultrasonography was also used when necessary. The difference in gestational age might also explain the difference in test performances; whilst other groups performed ultrasound screening until up to the 16th week of gestation, NT was not taken into account until after 14 weeks in the present study. Indeed, NT screening has been validated up to 14 weeks only (Nicolaides et al., 1992
; Snijders et al., 1998
).
Sensitivity in the present study, when based on a 95th percentile threshold (54.6%), was much lower than that of others (Snijders et al., 1998; Schwarzler et al., 1999
), who reported sensitivities of 77 and 78% respectively. In the present study, a threshold of 3 mm was used primarily to increase the risk of a patient on the basis of NT measurement. This had initially been reported by others (Nicolaides et al., 1992
) at the time this study had started. The same authors moved towards a continuous risk assessment by integrating maternal age, gestational age and previous history, as well as NT measurement. Although the present data were corrected for continuous assessment of NT with maternal age and gestational age by using multiples of the median, the data were recorded prospectively, with the aim of identifying cases with NT
3 mm. It is therefore reasonable to assume that our NT measurements distribution might be less accurate below 3 mm, and therefore may have led to a lower sensitivity. Another explanation might be that the present report relates to a low-risk population with only 21 cases of trisomy 21, though conclusions drawn from a rather small population must be considered with caution.
The most interesting aspect of the present study was to evaluate prospectively two screening strategies for trisomy 21. Such a comparison has been attempted previously only once (Haddow et al., 1998), but that study was hindered by poor reproducibility of the ultrasound measurements between centres. A combination of maternal serum pregnancyassociated plasma protein-A (MS PAPP-A) and ß-HCG has been shown to be promising, however. Others (Orlandi et al., 1997
; Spencer et al., 1999
; Krantz et al., 2000
) have confirmed the value of PAPP-A and free ß-HCG in combination with NT measurement, though this approach must be evaluated prospectively in a large series.
The present study confirms that ultrasound examination has an important role in prenatal screening for Down's syndrome. Ultrasound has the advantage over biochemical screening in that it may detect several severe malformations which can be recognized in early pregnancy.
The economic aspects of screening for Down's syndrome have been studied previously both in empirical investigations and models (Goldstein and Philip, 1990; Moatti et al., 1990
; Seror et al., 1992
; Wald et al., 1992
; Shackley et al., 1993
; Fletcher et al., 1995
; Torgerson, 1996
). Results ranged from beneficial (where savings exceed costs) to £52 000 per case averted, depending on the hypothesis and the methods of cost computation employed.
An economic evaluation of the present study showed the use of ultrasound alone to be more cost-effective than the use of both tests. However, whilst the additional cost per case diagnosed following screening with a combined method is modest, a combination of ultrasound and serum markers reduces the false-positive rate dramatically and in turn also reduces the risk of performing iatrogenic amniocentesis.
Recently, the cost effectiveness of various antenatal screening strategies for Down's syndrome was compared (Gilbert et al., 2001). These authors concluded that the choice of screening strategy should be between integrated test, first-trimester combined test, quadruple test or NT measurement, depending on how much service providers are willing to pay, the total budget available, and safety values. Screening based on either maternal age, second-trimester double test or first-trimester serum test was less effective, less safe, and more expensive.
In conclusion, the results of the present study suggest that, for a 5% false-positive rate, a 25% increase in detection rate could be obtained by a single risk assessment at 16 weeks using the results of NT and MSM screening at 12+014+0 weeks and 14+117+0 weeks respectively. However, this would delay risk assessment and invasive testing by 24 weeks when compared with NT alone. Such information should form part of genetic counselling, especially when the termination of a pregnancy for a fetal abnormality is an issue. Societal implications of antenatal screening policies were beyond the scope of this report. The combination of NT and first-trimester MSM screening appears feasible on the basis of only one small prospective series (Krantz et al., 2000). This combination remains to be established as an effective screening strategy in a large, population-based series as it may overcome any delay in risk assessment.
![]() |
Notes |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Association ABA (1999) Dépistage de la trisomie 21 par les marqueurs sériques maternels: Bilan de l'année 1997. Médecine Ftale, 37, 1114.
Bewley, S., Roberts, L.J., Mackinson, A.M. and Rodeck, C.H. (1995) First trimester fetal nuchal translucency: problems with screening the general population 2. Br. J. Obstet. Gynaecol., 102, 386388.[ISI][Medline]
Cuckle, H.S. (1995) Improved parameters for risk estimation in Down's syndrome screening. Prenat. Diagn., 15, 10571065.[ISI][Medline]
Cuckle, H.S. (1996) Established markers in second trimester maternal serum. Early Hum. Dev., 47 (Suppl.), S27S29.[ISI][Medline]
Cuckle, H.S. and van Lith, J.M.M. (1999) Appropriate biochemical parameters in first trimester screening for Down's syndrome. Prenat. Diagn., 19, 505512.[ISI][Medline]
Fletcher, J., Hicks, N.R., Kay, J.D.S. and Boyd, P.A. (1995) Using decision analysis to compare policies for antenatal screening for Down's syndrome. Br. Med. J., 311, 351356.
Ganiats, T.G. (1996) Justifying prenatal screening and genetic amniocentesis programs by cost-effectiveness analyses: a re-evaluation. Med. Decision Making, 16, 4550.[ISI][Medline]
Gilbert, R.E., Augood, C., Gupta, R., Ades, A.E., Logan, S., Sculpher, M. and van Der Meulen, J.H. (2001) Screening for Down's syndrome: effects, safety, and cost effectiveness of first and second trimester strategies. Br. Med. J., 323, 123125.
Goldstein, H. and Philip, J. (1990) A cost-benefit analysis of prenatal diagnosis by amniocentesis in Denmark. Clin. Genet., 37, 241263.[ISI][Medline]
Haddow, J.E., Palomaki, G.E., Knight, G.J., Williams, J., Miller, W.A. and Johnson, A. (1998) Screening of maternal serum for fetal Down's syndrome in the first trimester. N. Engl. J. Med., 338, 955961.
Hagard, S. and Carter, A.C. (1976) Preventing the birth of infants with Down's syndrome: a cost-benefit analysis. Br. Med. J., 1, 753756.[Medline]
Herman, A., Weinraub, Z., Dreazen, E., Arieli, S., Rozansky, S., Bukovsky, I. and Maymon, R. (2000) Combined first trimester nuchal translucency and second trimester biochemical screening tests among normal pregnancies. Prenat. Diagn., 20, 781784.[ISI][Medline]
Kadir, R.A. and Economides, D.L. (1997) The effect of nuchal translucency measurement on second-trimester biochemical screening for Down's syndrome. Ultrasound Obstet. Gynecol., 9, 244247.[ISI][Medline]
Kornman, L.H., Morssink, L.P., Beekhuis, J.R., De Wolf, B.T.H.M., Heringa, M.P. and Mantingh, A. (1996) Nuchal translucency cannot be used as a screening test for chromosomal abnormalities in the first trimester of pregnancy in a routine ultrasound practice. Prenat. Diagn., 16, 797805.[ISI][Medline]
Krantz, D.A., Hallahan, T.W., Orlandi, F., Buchanan, P., Larsen, J.W. and Macri, J.N. (2000) First trimester Down syndrome screening using dried blood biochemistry and nuchal translucency. Obstet. Gynecol., 96, 207213.
Moatti, J.P., Le Galès, C., Julian, C., Lanoe, J.L. and Ayme, S. (1990) Analyse coût-bénéfice du diagnostic prénatal des anomalies chromosomiques par amniocentèse. Rev. Epidémiol. Santé Publique, 38, 309321.[ISI][Medline]
Nicolaides, K.H., Azar, G., Byrne, D., Mansur, C. and Marks, K. (1992) Fetal nuchal translucency: ultrasound screening for chromosomal defects in first trimester of pregnancy. Br. Med. J., 304, 867869.[ISI][Medline]
Orlandi, F., Damiani, G., Hallahan, T.W., Krantz, D.A. and Macri, J.N. (1997) First-trimester screening for fetal aneuploidy: biochemistry and nuchal translucency. Ultrasound Obstet. Gynecol., 10, 381386.[ISI][Medline]
Roberts, L.J., Bewley, S., Mackinson, A.M. and Rodeck, C.H. (1995) First trimester fetal nuchal translucency: problems with screening the general population 1. Br. J. Obstet. Gynaecol., 102, 381385.[ISI][Medline]
Royston, P. and Thompson, S.G. (1992) Model-based screening by risk with application to Down's syndrome. Statist. Med., 11, 257268.[ISI]
Schuchter, K., Hafner, E., Stangl, G., Ogris, E. and Philipp, K. (2001) Sequential screening for trisomy 21 by nuchal translucency measurement in the first trimester and maternal serum biochemistry in the second trimester in a low-risk population. Ultrasound Obstet. Gynecol., 18, 2325.[ISI][Medline]
Schwarzler, P., Carvalho, J.S., Senat, M.V., Masroor, T., Campbell, S. and Ville, Y. (1999) Screening for fetal aneuploidies and fetal cardiac abnormalities by nuchal translucency thickness measurement at 1014 weeks of gestation as part of routine antenatal care in an unselected population. Br. J. Obstet. Gynaecol., 106, 10291034.[ISI][Medline]
Seror, V., Moatti, J.P., Muller, F. Le Gales, C. and Boue, A. (1992) Analyse coût-efficacité du dépistage prénatal de la trisomie 21 par marqueur sérique maternel (le cas de l'hCG). J. Gynecol. Obstet. Biol. Reprod., 21, 915922.[Medline]
Shackley, P., McGuire, A., Boyd, P.A., Dennis, J., Fitchett, M., Kay, J., Roche, M. and Wood, P. (1993) An economic appraisal of alternative pre-natal screening programmes for Down's syndrome. J. Public Health Med., 15, 175184.[Abstract]
Snijders, R.J.M., Noble, P., Sebire, N., Souka, A., Nicolaides, K.H., for the Fetal Medicine Foundation First Trimester Screening Group (1998) UK multicentre project on assessment of risk of trisomy 21 by maternal age and fetal nuchal translucency thickness at 1014 weeks of gestation. Lancet, 352, 343346.[ISI][Medline]
Spencer, K., Souter, V., Tul, N., Snijders, R. and Nicolaides, K.H. (1999) A screening program for trisomy 21 at 1014 weeks using fetal nuchal translucency, maternal serum free ß-human chorionic gonadotropin and pregnancy-associated plasma protein-A. Ultrasound Obstet. Gynecol., 13, 231237.[ISI][Medline]
Taipale, P., Hiilesmaa, V., Salonen, R. and Ylostalo, P. (1997) Increased nuchal translucency as a marker for fetal chromosomal defects. N. Engl. J. Med., 337, 16541658.
Thilaganathan, B., Slack, A. and Wathen, N.C. (1997) Effect of first-trimester nuchal translucency on second-trimester maternal serum biochemical screening for Down's syndrome. Ultrasound Obstet. Gynecol., 10, 261264.[ISI][Medline]
Torgerson, D.J. (1996) The impact of maternal age on the cost effectiveness of Down's syndrome screening. Br. J. Obstet. Gynaecol., 103, 581583.[ISI][Medline]
Wald, N.J., Kennard, A., Densem, J.W., Cuckle, H.S., Chard, T. and Butler, L. (1992) Antenatal maternal serum screening for Down's syndrome: results of a demonstration project. Br. Med. J., 305, 391394.[ISI][Medline]
Submitted on March 5, 2001; accepted on December 20, 2001.