Evaluation of routine ultrasound in the prenatal diagnosis of structural anomalies of the fetus.

The objective of this study was to evaluate routine obstetric ultrasound in detecting fetal structural anomalies and the impact of changing clinical practice on success rates. A retrospective study of routine ultrasound in a population of 6,869 pregnancies was performed during 1985-86 (phase 1) to establish efficacy of ultrasound in the detection of fetal anomalies. Changes in ultrasound practice comprised timing, personnel and technique of fetal examination. After alterations in practice, a prospective study of 6,969 pregnancies during 1987-89 (phase 2) was performed. All abnormal fetuses (cases) were assessed in both studies. Random samples of normal infants (controls) were chosen from both populations to establish specificity. In 83 cases in phase 1,116 anomalies were diagnosed postnatally of which 11 were suspected by routine prenatal ultrasound at < 24 weeks (sensitivity 9%, 95% C.I. 4-15). In 72 cases in phase 2, 89 anomalies were identified postnatally of which 27 were suspected by routine prenatal ultrasound at < 24 weeks (sensitivity 30%, 95% C.I. 21-41). The change in sensitivity was statistically significant (p = 0.0003). Of the 382 control infants randomly selected in phase 1, false abnormal scans were not identified (specificity 100%, 95% C.I. 99-100). Of the 367 control infants in phase 2, 12 false abnormal scans were identified (specificity 96.7%, 95% C.I. 95-99). The decrease in specificity was statistically significant (p = 0.0013). The detection of urorenal and to a lesser degree central nervous system anomalies showed most improvement between the two phases. The alterations to routine ultrasound practice in timing, personnel and technique have significantly improved the detection of all fetal structural anomalies but at the cost of a small but significant loss of specificity.