Abstract 1518 Poster Session III, Monday, 5/3 (poster 27)

Background/study objective: During the 1990's, survival increased for very low birth weight (VLBW) infants, due in large part to surfactant use. However, surfactant has not decreased the prevalence of chronic lung disease (CLD). The objective of this analysis was to study the possibility that in the 1990's (with frequent use of surfactant), the severity of CLD has decreased, improving the outcome for infants with CLD. Methods: We analyzed data for 2214 infants with: 1) birth weight 500 to 1500 gms, 2) no major congenital anomalies, and 3) a mother who resided in a 17-country region. CLD was defined as the need for supplemental oxygen at 36 weeks post-conception. Neonatal data were obtained from medical records. Data about the severity of CLD (as indicated by duration of oxygen use) and outcomes through one year adjusted age were obtained from our multidisciplinary follow-up clinic. For analyses of trends over time, infants were grouped into 4 three-year epochs, based on date of birth (7/1/84 to 6/30/96). For successive epochs, the strength of association between CLD and adverse outcomes was expressed as odds ratios (95% confidence limits), using VLBW infants without CLD as the referent group. Results: Surfactant and dexamethasone use increased in successive epochs [surfactant rates: 0, 10%, 59%, and 60%; dexamethasone rates: 0, 12%, 22%, and 44%]. Survival increased from 70% in epochs 1 and 2, to 82% (483/589) in epoch 3, and 87% (517/593) in epoch 4 [p<0.0001, test for trend]. In the first 3 epochs, the prevalences of CLD among surviving infants were 27%, 26%, and 25%; in epoch 4 (7/1/93-6/30/96) the prevalence decreased to 15% [p < 0.0001]. Among 399 infants with CLD who survived to discharge, duration of supplemental oxygen and home oxygen use did not decrease over time. The frequency of death after discharge decreased in successive epochs, from 10% (9/89) to 5% (5/108) to 2% (3/125) to 1% (1/77) [p=0.004, test for trend]. The frequency of weight < 10th percentile decreased from 64% in epoch 1 to 47% in epochs 2,3, and 4. Among the 352 infants with CLD who came for follow-up at 1 year adjusted age (92% of survivors), no change over time was found in the number of rehospitalizations, length or head circumference < 10th percentile, cerebral palsy, or Bayley Scales of Infant Development Mental Developmental Index (MDI) < 70. Between epoch 1 and subsequent epochs, a decrease was found in the odds ratio for mortality [4.5 (1.5-12.9) vs 1.5 (0.7-3.2)] and weight < 10th percentile [4.7 (2.6-8.3) vs 2.3 (1.8-3.1)]. No significant trends over time were found in the odds ratios for rehospitalization > once, length or head circumference < 10th percentile, MDI < 70, or cerebral palsy. Conclusion: Other than a decrease in post-discharge mortality, we found no evidence that the outcome of infants with CLD (through one year of age) has improved in the 1990's. CLD remains a useful predictor of adverse outcome in VLBW infants.