Triple serum tests screening for Down's syndrome in women at second trimester: a meta analysis / 中华围产医学杂志

ABSTRACT

Objective To explore the value of maternal serum triple markers screening,consisted of AFP (α-fetal protein),β-human chorionic gonadotropin (β-hCG) and free estriol (uE3),for Down's syndrome,in second trimester.Methods We searched published literatures from PubMed,MEDLINE,China National Knowledge Internet (CNKI) and Wanfang Database from January 1990 to August 2014 and articles met the following criteria were collected(1) The report was of a screening test;(2) The research purpose was to study the efficiency ofAFP,hCG and uE3 (triple markers) screening for Down's syndrome;(3) The research subjects were pregnant women in second trimester;(4) All of the studied cases were confirmed by amniocentesis and chromosome karyotyping;(5) The pregnant outcomes must be available;(6) Each report should have at least one case of Down's syndrome identified;(7) The literature could be retrieved by Science Citation Index or Chinese core periodicals;(8) When assessed by QUADAS (quality assessment of studies of diagnostic accuracy included in systematic reviews) Quality Assessment Scale,the score should be ≥ 8.Information were extracted,including name of the first author,publication time,sample size,sensitivity,specificity,maternal age,gestational age,cutoff value,β-hCG type and others.MetaDiSc 1.4 software was applied for meta-analysis.I2 was used for heterogeneity,and the fixed or random effects model for calculation of the combined sensitivity,specificity,diagnostic odds ratio (DOR) and the 95％CI.The summary receiver operating characteristic (SROC) curve was drawn.Deek's test in Stata Software was used to validate publication bias.Results A total of 49 literatures were recruited in this study with a total sample size of 960 245.Deck's funnel plot analysis showed that the correlation coefficient and the standard error of bias were-1.067 and 3.64 (t=-0.290,P=0.771).The correlation coefficient and the standard error of the slope were 3.578 and 0.26 (t=13.740,P=0.000).The random effects model showed the pooled sensitivity of the 49 literatures was 0.72 (95％CI0.70 0.74),the pooled specificity was 0.92 (95％CI0.92-0.93),and DOR was 33.80 (95％CI25.03-45.65).The area under the SROC was 0.900 7.DOR of younger age group (including three literatures) was 14.38 (95％CI3.67-56.42) and 26.64 (95％CI19.49-36.41) for the older age group (two literatures).For two literatures determining the gestational age based on ultrasonography,the DOR was 50.22 (95％CI26.91-93.71),and DOR was 33.09 (95％CI17.33-63.19) for those based on last menstrual period in these two literatures.For eight literatures applied the cutoff value of 1270,12 applied 1250 and four applied mixed cutoff value,the DOR was 10.94 (95％CI3.04-39.38),54.34 (95％CI42.19-70.01) and 36.37 (95％CI31.19-42.40),respectively.DOR of total β-hCG group (12 literatures) was 22.06 (95％CI16.46-29.58) and that of free β-hCG group (ten literatures) was 37.15 (95％CI30.00-46.02).Conclusions Triple regimen of second trimester maternal serum screening for Down's syndrome is much more efficient.The detection rate could be further improved by determination of the gestational age with ultrasound,and application of 1250 as the risk cutoff value and free β hCG as a screening marker.