RESUMO
OBJECTIVE: To assess the predictive value of early therapy for ventilated extremely low birth weight (ELBW) infants beyond information available at delivery. STUDY DESIGN: Prospective, single-center cohort analysis of 177 ventilated ELBW infants. We collected information known at delivery (gestational age, birth weight, singleton, sex, antenatal steroids) and additional information while infants were mechanically ventilated (head ultrasound scanning, clinician intuitions of death before discharge). An adverse outcome was defined as mortality or Bayley Mental Developmental Index or Psychomotor Developmental Index <70 at 2 years. We compared the predictive ability of clinical variables separately, in combination, and in addition to information available at delivery. RESULTS: A total of 77% of infants survived to follow-up; 56% of survivors had Bayley Mental Developmental Index and Psychomotor Developmental Index ≥ 70. A total of 95% of infants with both abnormal head ultrasound scanning results and predicted death before discharge had an adverse outcome, independent of gestational age. Conversely, 40% of infants with normal head ultrasound scanning results and no predicted death before discharge had an adverse outcome, independent of gestational age. After adjusting for variables known at birth, predicted death before discharge and abnormal head ultrasound scanning results added significantly to the ability to predict outcomes. CONCLUSION: Information gained early in the neonatal intensive care unit improves prediction of mortality or neurodevelopmental impairment in ventilated ELBW infants beyond information available in the delivery room.