RESUMO
OBJECTIVES: To compare the performance of double screening (measuring maternal serum levels of alpha-fetoprotein [AFP] and total beta-human chorionic gonadotrophin [hCG] as markers for Down syndrome) with that of triple screening (also measuring levels of unconjugated estriol [uE3]) in the second trimester of pregnancy, and to compare ethnic variance between Chinese and Caucasian populations. METHODS: The study investigated 15096 normal singleton pregnancies and 24 pregnancies affected with Down syndrome. Frequency distributions of AFP, hCG, and uE3 levels were analyzed. Likelihood ratios (LRs) were calculated using the multiple of median value (MoM) of AFP, hCG, and uE3 as variables. After multiplying maternal age risk by the LR values for the markers used in double and triple screening, the specific risks obtained with double and triple screening were estimated. The detection rate (DR) and false-positive rate (FPR) were calculated at different cut-off points. The serum markers' levels were also compared with those of Caucasian women. RESULTS: The median MoM value of hCG was higher in women with affected pregnancies (1.40) than those with unaffected pregnancies (1.00). However, the median MoMs of AFP and uE3 (0.79 and 0.68) were lower in affected than in unaffected pregnancies. At a FPR of 5%, the detection rates reached with double and triple screening were 50% and 66.7%, respectively. Ratios of the 3 serum markers' medians to those in a study with Caucasian women were 1.06 (range=1.04-1.09) for AFP, 1.14 (range=1.10-1.17) for hCG, and 1.28 (range=1.23-1.41) for uE3 for the relevant gestational weeks. CONCLUSION: Triple screening performed better than double screening in the second trimester. Ethnic variance should be taken into account in Down syndrome screening.