Diagnostic performance of intracardiac echogenic foci for Down syndrome: a meta-analysis.

OBJECTIVE To synthesize the accumulated data on the diagnostic performance of intracardiac echogenic foci for Down syndrome, a meta-analysis was performed. DATA SOURCES We conducted MEDLINE and EMBASE searches (1985 to August 2002) using the key words "intracardiac (echogenic) focus/foci," "golfballs," "trisomy 21," and "Down syndrome." Bibliographies of retrieved articles were also screened, and experts were contacted. Both single and multiple intracardiac echogenic foci qualified, regardless of cardiac location. Eligible studies included and described both chromosomally normal and abnormal fetuses; the fetal karyotype was unknown at the time of sonographic examination; and chromosomal status was confirmed by karyotype or postnatal clinical examination. TABULATION, INTEGRATION AND RESULTS Sensitivity and specificity were recorded for intracardiac echogenic foci in a "combined" setting (regardless of the presence of other sonographic markers) and "isolated" setting (in the absence of other markers). Weighted estimates and summary receiver operating characteristic curves were calculated. Across 11 studies (51,831 pregnancies, 333 Down syndrome cases), random effects sensitivity and specificity were 26% (95% confidence interval 19, 34) and 95.8% (95% confidence interval 92.2, 97.8), respectively, with a positive likelihood ratio of 6.2 ("combined" setting, likelihood ratio 7.0; "isolated" setting, likelihood ratio 5.4). Summary receiver operating characteristic curves were also consistent with these values. With a 0.8% risk of amniocentesis-induced fetal loss, one fetus is lost per Down case detected when the background Down risk is 1:770. CONCLUSION Intracardiac echogenic foci increase the risk of Down syndrome five- to seven-fold. This information should be considered in the decision making for amniocentesis in conjunction with the woman's background risk.