Identifying Gaps in Resuscitation Practices Across Level-IV Neonatal Intensive Care Units.

Abstract

This study aimed to describe resuscitation practices in level-IV neonatal intensive care units (NICUs) and identify possible areas of improvement. This study was a cross-sectional cohort survey and conducted at the Level-IV NICUs of Children's Hospital Neonatal Consortium (CHNC). The survey was developed with consensus from resuscitation and education experts in the CHNC and pilot tested. An electronic survey was sent to individual site sponsors to determine unit demographics, resuscitation team composition, and resuscitation-related clinical practices. Of the sites surveyed, 33 of 34 sites responded. Unit average daily census ranged from less than 30 to greater than 100, with the majority (72%) of the sites between 30 and 75 patients. A designated code response team was utilized in 18% of NICUs, only 30% assigned roles before or during codes. The Neonatal Resuscitation Program (NRP) was the exclusive algorithm used during codes in 61% of NICUs, and 34% used a combination of NRP and the Pediatric Advanced Life Support (PALS). Most (81%) of the sites required neonatal attendings to maintain NRP training. A third of sites (36%) lacked protocols for high-acuity events. A code review process existed in 76% of participating NICUs, but only 9% of centers enter code data into a national database. There is variability among units regarding designated code team presence and composition, resuscitation algorithm, protocols for high-acuity events, and event review. These inconsistencies in resuscitation teams and practices provide an opportunity for standardization and, ultimately, improved resuscitation performance. Resources, education, and efforts could be directed to these areas to potentially impact future neonatal outcomes of the complex patients cared for in level-IV NICUs. · Resuscitation practice is variable in level-IV NICUs.. · Resuscitation algorithm training is not uniform. · Standardized protocols for high-acuity low-occurrence (HALO) events are lacking.