A controlled trial of nurse practitioners in neonatal intensive care.

OBJECTIVE To compare a clinical nurse specialist/neonatal practitioner (CNS/NP) team with a pediatric resident team in the delivery of neonatal intensive care. DESIGN Randomized, controlled trial. SETTING A 33-bed tertiary-level neonatal intensive care unit. PATIENTS Of 821 infants admitted to the neonatal intensive care unit between September 1991 and September 1992, 414 were randomized to care by the CNS/NP team, and 407 were randomized to care by the pediatric resident team. INTERVENTION Infants assigned to the CNS/NPs team were cared for by CNS/NPs during the day and by pediatric residents during the night. Infants assigned to the pediatric resident team were cared for by pediatric residents around the clock. Neonatologists supervised both teams. MEASURES Outcome measures included mortality; number of neonatal complications; length of stay; quality of care, as assessed by a quantitative indicator condition approach; parent satisfaction with care, measured using the Neonatal Index of Parent Satisfaction; long-term outcomes, measured using the Minnesota Infant Development Inventory; and costs. RESULTS There were 19 (4.6%) deaths in the CNS/NP group and 24 (5.9%) in the resident group (relative risk [RR], 0.78; confidence interval [CI], 0.43 to 1.40). In the CNS/NP group, 230 (55.6%) neonates had complications, in comparison with 220 (54.1%) in the resident group (RR, 1.03; CI 0.91 to 1.16). Mean lengths of stay were 12.5 days in the CNS/NP group and 11.7 days in the resident group (difference in means, 0.8 days; CI, -1.1 to 2.7). The performance on the indicator conditions was comparable in the two groups except for two instances, jaundice and charting, both of which favored the CNS/NP group. Mean scores on the Neonatal Index of Parent Satisfaction were 140 in the CNS/NP group and 139 in the resident group (difference in means, 1.0; CI, -3.6 to 5.6). In the CNS/NP group, 6 (2.6%) infants performed 30% or more below their age level in the Minnesota Infant Development Inventory, in comparison with 2 (0.9%) in the resident group (RR, 2.87; CI, 0.59 to 14.06) The cost per infant in the CNS/NP group was $14,245 and in the resident group $13,267 (difference in means, $978; CI, -1303.18 to 3259.05). CONCLUSIONS CNS/NP and resident teams are similar with respect to all tested measures of performance. These results support the use of CNS/NPs as an alternative to pediatric residents in delivering care to critically ill neonates.