Maternal serum screening for Down's syndrome: the effect of routine ultrasound scan determination of gestational age and adjustment for maternal weight

Abstract

Objective To investigate the effect of using a routine ultrasound estimate of gestational age and maternal weight adjustment on maternal serum alpha‐fetoprotein (AFP), unconjugated oestriol (uE3) and human chorionic gonadotrophin (hCG) levels in antenatal screening for Down's syndrome. Design Women with a singleton pregnancy without Down's syndrome were screened using the three serum markers and an estimate of gestational age based on ‘dates’ (time since first day of the last menstrual period) and one based on an ultrasound scan examination was recorded together with maternal weight. Setting Women attending the Homerton Hospital, Hackney, for their antenatal care between February 1989 and January 1990. Subjects 2113 women with a singleton pregnancy without Down's syndrome. Results The use of ultrasound to estimate gestational age (usually based on the biparietal diameter of the fetal skull) led to a significant reduction in the variance of each marker at a given week of pregnancy. The level of each marker was negatively associated with maternal weight, so that adjustment for weight also led to a reduction in variance. These data on gestational age and maternal weight, taken together with published data on pregnancies associated with Down's syndrome, indicate that the routine use of ultrasound to estimate gestational age will increase the detection rate from 58% to 67% while maintaining the false‐positive rate at 5%, or reduce the false‐positive rate from 5.7% to 3.1% while maintaining the detection rate at 60%. Routine maternal weight adjustment for the serum marker levels was much less useful, increasing the detection rate by about 0.5% for a given false‐positive rate, or reducing the false‐positive rate about 0.1% for a given detection rate. Conclusion An ultrasound gestational age estimate available at the time of Down's syndrome screening confers a substantial advantage to screening performance with a further small benefit resulting from maternal weight adjustment, which is worth adopting if it can be done without difficulty or extra cost.

Authors' Reply

“…The 10th centile and 90th centile for AFP and UE] and the data related to our Tables 4 and 51 show similarity with data published previously by Wald et al 15 as shown in Table 1. If, therefore, our population parameters in our earlier Table 6 are closer to the consensus estimates for AFP and The real test of any screening programme is not based on mathematical modelling of retrospective data and adjustments or estimates of the 'real world', but is based on true prospective experience.…”

“…When the actual data is examined, the median Free (3-hCG in Down's syndrome cases at the [14][15][16] weeks' gestation period was 2·52 and at greater than 16 weeks was 2' 32. The corresponding median for AFP in these two groups was o· 66 and 0'73 and the mean age was 30·9 and 30'2 years, respectively.…”

Assay precision of serum α fetoprotein in antenatal screening for neural tube defects and Down's syndrome

“…AFP concentrations were converted into multiples of the median (MoMs) using regression equations for gestational age based on either days from the last menstrual period (dates) or the biparietal diameter measurement from an ultrasound scan, as previously described. 7 Corrections were also made for maternal weight.…”

Effect of screening algorithm, parameter values and median smoothing on patient-specific risk estimates for Down's syndrome screening

“…Estimates of the parameters derived from these data are shown in Table 1 together with those derived by Wald. 4 The latter have been chosen for illustrative purposes as they are widely used. Others we could have used include those of Crossley, 2 Spencer 5 and Heyl.…”