Since the observation of pulmonary hemorrhage (PH) in VLBW neonates during controlled exogenous surfactant trials, debate as to the relative contributions of surfactant or patent ductus arteriosus (PDA) to the development of PH has existed. To add to our studies of the relationships of PH, PDA, and surfactant therapy, we performed a retrospective cohort study of all consecutively ventilated VLBW infants (<1500g) in the period immediately before surfactant was introduced (Sept '89). PH was defined as any non-traumatic episode of blood-tinged pulmonary secretions from the endotracheal tube in the first 28 days; severity of PH was assessed by # of episodes, character of hemorrhage, and radiographic/clinical deterioration. 22 of 96 infants (23%) had PH (med. age=3 d); 19 infants (86%) had multiple and 10 (45%) had frank blood episodes. Logistic regression (modelP=0.02, controlled for RDS) indicated that predictors of developing PH were hemodynamically important PDA (O.R. 4.2, 95% C.I. 1.2-15.1) and BW(O.R. 1.3, 95% C.I. 1.01-1.7 per -100g). Predictors and outcomes of PH:Table PH infants were more likely to need ventilation>30d (41% vs 10%) and O2>50d (41% vs 13%)(P<0.01). We conclude that before the introduction of surfactant (1) PDA was a strong predictor of PH and (2) PH was associated with adverse pulmonary outcome.

Table 1
Full size table