977 CORRELATION OF FETAL OUTCOME WITH ANTENATAL TESTING IN ISOIMMUNIZED PATIENTS

Abstract

Sixty-seven pregnant women at risk for fetal isoimmunization were followed prospectively from 6/78–6/84 and scored by nonstress testing (NST) (reactive =2, nonreactive =0), amniotic fluid Δ 450 (slope: increasing =2, stable =1, decreasing =0), and fetal ultrasound (U/S) (erythroblastosis fetalis: none =2, minimal-moderate =1, moderate-severe =0). Postnatally, the 67 infants (GA 36.6±3.1 wks, mean±SD, BW 2850±656 grams) were independently scored without knowledge of the maternal score (living =2, dead =0; treatment: none =2, phototherapy =1, exchange transfusion =0; problems of prematurity: none =2, mild =1, moderate-severe =0). The mothers and infants were each assigned a rank from 1–10 according to their scores. A sign test (non-parametric) between the maternal and infant scores revealed no significant difference indicating agreement between rankings. The chi square goodness of fit for the distribution of scores when grouped as 1/2/3, 4/5/6/7, and 8/9/10 also indicated no significant differences in rankings. The correlation coefficient for infants rank against maternal rank was .76 at p<.001. No single antenatal test has thus far been shown to be predictive of neonatal outcome. We have shown prospectively that, in combination, NST, Δ450 and U/S can reliably predict severity of isoimmunization. The composite score can therefore be used to optimize timing of delivery with respect to the risks of prematurity and erythroblastosis fetalis.