Screening performance of first-trimester nuchal translucency for major cardiac defects: a meta-analysis

Objective The purpose of this study was to evaluate the screening performance of increased first-trimester nuchal translucency for the detection of major congenital heart defects. Study design A meta-analysis based on MEDLINE and EMBASE searches (up to June 2002) that assessed the diagnostic performance of increased nuchal translucency for congenital heart defect detection. Weighted sensitivity and specificity estimates (random effects) and summary receiver-operating characteristic curves were obtained. Results Eight independent studies with 58,492 pregnant women were analyzed. There was significant heterogeneity among the studies. Nuchal translucency above the 99th percentile had a sensitivity of 31% and specificity of 98.7% (random effects calculations), with a positive likelihood ratio of 24. Summary receiver-operating characteristic estimates were consistent with these values. The ability of nuchal translucency measurements above this threshold to detect cardiac malformations varied nonsignificantly (P = .64) for different congenital heart defects types (sensitivity range, 25%-55%). Conclusion Nuchal translucency screening is a modestly efficient strategy for congenital heart defect detection; the use of the 99th percentile threshold may capture approximately 30% of congenital heart defects