Prenatal sonography for the detection of fetal anomalies: results of a prospective study and comparison with prior series.

OBJECTIVE

Reported rates of sonographic detection of fetal anomaly vary widely. The purpose of our study was to determine how well we were detecting fetal anomaly in our region and to compare our results with published series using standardized criteria.

SUBJECTS AND METHODS

Over a 2-year period, we compared the reports of sonographic studies done in 7880 pregnancies at 16-20 weeks' gestation with all pregnancy outcomes established by medical records and pathologic results of the infants, whether born alive or dead. Fetal anomaly was defined as any structural anomaly except for those specifically excluded. We compared our results, expressed as percentage of anomalies found and anomalies found per 1000 pregnancies screened, with those of published studies, modifying published figures where necessary to compensate for our defined exclusion criteria.

RESULTS

At postnatal examination, we found 157 anomalies in 144 babies who had prenatal sonography at 16-20 weeks' gestation. We detected 93 anomalies (60%) in 84 fetuses, leading to termination of pregnancy in 42 cases (50% of anomalous fetuses). The prevalence of fetal anomaly in our population was 19.8 anomalies per 1000 pregnancies screened. We detected 11.8 anomalies per 1000 screened, leading to termination of 5.3 pregnancies per 1000 screened. Our success rate for fetal anomaly detection ranged from 92% for CNS anomalies to 31% for cardiac anomalies and 25% for craniofacial anomalies. There were fewer babies born with a major malformation in the group whose anomalies were detected prenatally (21%) than in the group whose anomalies were not detected (58%). By comparison, in prior series, prevalence of anomalies varied from 5.67 to 25.95 per-11000 screened; anomaly detection varied from 16.6% to 74.4%; detection rate varied from 3.08 to 11.03 per 1000 screened. There was no correlation between the percentage and per 1000 detection rates. Thus, the RADIUS and Helsinki trials had similar detection rates per 1000, but more than a twofold difference in percentage detection rate. Similarly, the percentage success rate in studies could be the same, but the detection rate per 1000 different threshold. There was a similar lack of correlation for rates of pregnancy termination. All series except for the RADIUS study reported a reduction in the rate of adverse outcome if the anomalies were detected prenatally.

CONCLUSION

Our rate of detection of fetal anomaly is satisfactory for most organ systems but remains poor for cardiac, skeletal, and craniofacial anomalies. The relative ranking of the success of fetal anomaly detection in prior studies depends on whether the detection rate is reported as a percentage or as the number of anomalies detected per 1000 screened. Our results compare favorably with those reported in other series.