Predicting poor neonatal outcome: a comparative study of noninvasive antenatal testing methods.

OBJECTIVE

The assessment of fetal well-being in the third trimester of pregnancy depends on many variables including fetal size, amniotic fluid volume, umbilical cord arterial Doppler waveforms, the nonstress test, and the biophysical profile, yet little has been written that directly compares these variables. In this study, we compared amniotic fluid indexes, umbilical cord arterial Doppler waveforms, nonstress tests, and biophysical profiles for predicting poor neonatal outcomes in fetuses who are small for gestational age (SGA).

SUBJECTS AND METHODS

From April 11, 1994, through August 1, 1995, a cohort of 108 SGA fetuses was identified. Follow-up was available in 97 of these cases. Receiver operating characteristic curves were constructed for Doppler systolic:diastolic ratios and for amniotic fluid indexes. Student's t test and logistic regression analysis were used to compare umbilical cord arterial Doppler imaging, amniotic fluid indexes, the nonstress test, and the biophysical profile for predicting poor neonatal outcome.

RESULTS

Of the 30 fetuses who had poor outcomes, five were emergency cesarean deliveries, three died, three had intracranial hemorrhages, one had a cerebral infarct, 12 had prolonged admission to the neonatal intensive care unit (NICU) (> 10 days), and six had NICU admissions at term. Of the variables we assessed, the sensitivities for predicting poor outcome were as follows: cord Doppler imaging, 64%; low amniotic fluid volume (oligohydramnios), 32%; biophysical profile, 18%; and nonstress test, 14%. Receiver operating characteristic curves showed that a systolic:diastolic ratio of 4.0 and an amniotic fluid index of 5 cm (independent of gestational age) were the most accurate cutoff values for predicting poor outcome. Logistic regression analysis showed that amniotic fluid indexes and umbilical cord arterial Doppler imaging were independent predictors of poor outcome and that the predictive value of the biophysical profile varied according to the amniotic fluid index.

CONCLUSION

Doppler waveform abnormalities were the most accurate predictor of poor neonatal outcome in a cohort of SGA fetuses. Umbilical cord arterial Doppler waveform analysis should be included in the surveillance of SGA fetuses.