Single large study or meta-analysis parameters: choosing the most appropriate tool for Down syndrome screening in the first trimester

Aim The aim of the present study is to determine whether the use of population parameters derived from meta-analysis produces better test characteristics in the first-trimester combined screening than those obtained from large single studies, when screening for Down syndrome (DS) in singleton pregnancies. Methods Retrospective analysis was clone on a database of 4248 singleton pregnancies, including 13 cases of DS, for which the values of three markers used in the combined first-trimester test were available. The risk of DS was calculated for each mother from the different population parameters derived by meta-analysis or from two large single studies: the Fetal Medicine Foundation study (FMF) and the Serum, Urine and Ultrasound Screening Study (SURUSS). Results When the detection rate was fixed at about 85%, the false-positive rate was significantly higher (Chi-square=53.49, p<0.0001) when the parameters obtained by the meta-analysis (6.25%) were compared with the two single studies (FMF, 4.15% and SURUSS, 3.75%). Conclusions This study highlights the importance of selecting appropriate population parameters when insufficient DS cases are available in the study population. When screening for DS in singleton pregnancies during the first trimester, the use of population parameters derived from single large studies produced better test characteristics. Copyright (C) 2006 John Wiley & Sons, Ltd.