Improving the prediction of surfactant deficiency in very low-birthweight infants with respiratory distress.

With the availability of exogenous surfactant therapy, distinguishing between surfactant deficiency and other causes of respiratory distress soon after birth is critical. Of 149 very low-birthweight (VLBW) infants of 24 to 30 weeks' gestation, 107 (72%) had both clinical and radiographic features of respiratory distress syndrome (RDS). Analysis of amniotic fluid obtained within 24 hours before birth or of the initial tracheal aspirate after birth for phosphatidylglycerol (PG) showed that absence of PG had a positive predictive value (PPV) of 80% for predicting RDS, while an immature lecithin/sphingomyelin (L/S) ratio had a PPV of 86% with a 68% specificity. When PG was absent and the L/S ratio was immature, the PPV for RDS increased to 89% (97% specificity). Phospholipid analysis was superior to gestational age alone in predicting RDS (P less than or equal to .02). Receiver operating characteristic (ROC) analysis detected different thresholds for immature L/S ratios for amniotic fluid (2.0) and initial tracheal aspirate (3.0) for predicting RDS. A tracheal aspirate L/S ratio less than or equal to 3.0 predicted RDS with 91% accuracy, while an amniotic fluid L/S ratio less than or equal to 2.0 predicted RDS with 90% accuracy. ROC curve thresholds for either tracheal aspirate or amniotic fluid permit selection of VLBW infants most likely to benefit from surfactant treatment. These analyses improve the accuracy of diagnosing surfactant-deficiency-associated RDS using only clinical and radiographic diagnoses, and may permit a more focused approach for rescue surfactant therapy in infants presenting with clinical symptoms of respiratory distress.(ABSTRACT TRUNCATED AT 250 WORDS)