Neuroprotection Bundles and Intraventricular Hemorrhage Rates in Neonates <29 Weeks' Gestation.

Abstract

To evaluate the association of intraventricular hemorrhage-neuroprotection bundle (IVH-NB) implementation on intraventricular hemorrhage (IVH) rates in infants born at <29 weeks' gestation. This multicenter, retrospective, preimplantation and postimplementation cohort study included neonates of <29 weeks' gestation admitted to Canadian neonatal intensive care units (NICUs) participating in the Canadian Neonatal Network. Sites which implemented IVH-NB were identified via a survey. Data were collected for 36 months each for preimplantation and postimplementation periods per NICU, with 1-month washout. Infants with major anomalies or admitted >2 days after birth were excluded. Primary outcome was a composite of severe IVH (grade III/IV) or death ≤ first 7 days. Secondary outcomes included severe IVH, posthemorrhagic ventricular dilatation, and severe neurological injury (severe IVH or periventricular leukomalacia). Seventeen NICUs that implemented IVH-NB were included, contributing a total of 6522 infants (3216 preimplementation and 3306 postimplementation). Postimplementation, deferred cord clamping (56% vs 62%, P < .0001), prophylactic indomethacin (15% vs 19%, P < .0001), and normothermia (50% vs 56%, P < .0001) increased, and cardiopulmonary resuscitation (29% vs 25%, P < .0001) and endotracheal intubation at birth decreased (46% vs 41%, P < .0001). After implementation, aOR for the composite outcome (aOR 1.07, 95% CI 0.84, 1.35), severe IVH (aOR 1.02, 95% CI 0.80, 1.31), and posthemorrhagic ventricular dilatation (aOR 1.18 95% CI 0.78, 1.79) were unchanged. The only bundle component associated with improved outcomes was early patent ductus arteriosus treatment adopted by 2 sites (aOR 0.30, 95% CI 0.17, 0.55). No temporal trend in outcomes was identified. In this large multicenter cohort, IVH-NB implementation was associated with improved practices but not with improved outcomes.