Prediction of the severity of meconium aspiration syndrome

Meconium aspiration syndrome is associated with increased neonatal morbidity and mortality. We sought to determine whether the need for neonatal mechanical ventilation or prolonged ventilation (> or = 3 days) was predictable from antepartum, intrapartum, and immediate neonatal events.

Between 1987 and 1989, 8003 of 43,906 (18%) live infants had meconium-stained amniotic fluid; 82 of these infants had clinical and radiographic evidence of meconium aspiration, and of these 82, 39 (48%) required mechanical ventilation. Predictors of ventilation or prolonged ventilation were determined by means of stepwise logistic regression.

Predictors of neonatal ventilation included fetal tachycardia, interval from meconium detection to delivery, low 5-minute Apgar score, respiratory distress necessitating intubation in the delivery suite, and delivery by cesarean section. Sensitivity was 72%, and specificity was 64%; the model was incorrect in 32% of the cases. Predictors of prolonged ventilation were ominous fetal heart rate tracing, umbilical arterial pH < 7.20, birth weight > 90th percentile, nulliparity, and Apgar score > 4 at 1 minute and < or = 6 at 5 minutes. Sensitivity was 67%, and specificity was 91%; prolonged ventilation was incorrectly predicted in 17% of the cases.

Use of these models to determine the need for mechanical ventilation or prolonged ventilatory assistance will enhance identification of infants at risk of severe meconium aspiration and will lead to early transfer to the neonatal intensive care unit for intense observation and management.