Second-trimester ultrasound to detect fetuses with Down syndrome - A meta-analysis

Context Second-trimester prenatal ultrasound is widely used in an attempt t o detect Down syndrome in fetuses, but the accuracy of this method is unkno wn. Objective To determine the accuracy of second-trimester ultrasound in detec ting Down syndrome in fetuses. Data Sources English-language articles published between 1980 and February 1999 identified through MEDLINE and manual searches. Study Selection Studies were included if they recorded second-trimester fin dings of ultrasonographic markers, chromosomal abnormalities, and clinical outcomes for a well-described sample of women. A total of 56 articles descr ibing 1930 fetuses with Down syndrome and 130365 unaffected fetuses were in cluded. Data Extraction Articles were independently reviewed, selected, and abstrac ted by 2 reviewers. Discrepancies in data abstraction were resolved by cons ensus with a third reviewer. Overall estimates of sensitivity, specificity, and positive and negative likelihood ratios were calculated for the follow ing markers: choroid plexus cyst, thickened nuchal fold, echogenic intracar diac focus, echogenic bowel, renal pyelectasis, and humeral and femoral sho rtening. Results were stratified by whether markers were identified in isol ation or in conjunction with fetal structural malformations. Data Synthesis When ultrasonographic markers were observed without associat ed fetal structural malformations, sensitivity for each was low (range, 1%- 16%), and most fetuses with such markers had normal outcomes. A thickened n uchal fold was the most accurate marker for discriminating between unaffect ed and affected fetuses and was associated with an approximately 17-fold in creased risk of Down syndrome. If a thickened nuchal fold is used to screen for Down syndrome, 15 893 average-risk women or 6818 high-risk women would need to be screened for each case of Down syndrome identified. For each of the other 6 markers, when observed without associated structural malformat ions, the marker had marginal impact on the risk of Down syndrome. Because the markers were detected in only a small number of affected fetuses, the l ikelihood of Down syndrome did not decrease substantially after normal exam ination findings (none of the negative likelihood ratios were significant), Conclusions A thickened nuchal fold in the second trimester may be useful i n distinguishing unaffected fetuses from those with Down syndrome, but the overall sensitivity of this finding is too low for it to be a practical scr eening test for Down syndrome. When observed without associated structural malformations, the remaining ultrasonographic markers could not discriminat e well between unaffected fetuses and those with Down syndrome. Using these markers as a basis for deciding to offer amniocentesis will result in more fetal losses than cases of Down syndrome detected, and will read to a decr ease in the prenatal detection of fetuses with Down syndrome.