Delivery Room of the Future Research Overview

Immediately after birth, all newborns experience rapid cardiopulmonary adaptations. Time-critical physiologic targets for vital signs are used globally to guide resuscitation. This physiology is inherently disrupted in infants with congenital anomalies, but the extent and nature of these differences are not well known for distinct physiologic phenotypes. "Normal" vs "worrisome" vital sign trajectories are not yet determined. Without data-driven physiologic targets, providers operate "blindly" during resuscitation for newborns with anomalies. Our goal is to characterize the minute-to-minute values and variation of physiologic vital signs after birth for specific phenotypes of congenital anomalies. We will then identify time critical targets in vital signs that are associated with short-term outcomes. We have established a CHOP Delivery Databasewith essential clinical variables for all infants born in the SDU at CHOP. We will integrate streaming physiologic data acquired during resuscitation, a significant advancement over a clinical database alone, and we will pair these with registry data to support observational and epidemiologic studies of DR resuscitation. We will use these data to identify opportunities for therapeutic interventions to be assessed in DR clinical trials and integrate phenotype-specific physiologic reference ranges to establish goal-directed resuscitation targets in our Digital Coach platform.

Errors during neonatal resuscitation are common, and these are more frequent during complex resuscitations. The incidence of intensive resuscitation is up to 100-fold higher among infants with major anomalies compared with low-risk infants. We previously demonstrated that care disruptions are common during preterm resuscitation, with an average of 53 disruptions per resuscitation. Team performance during resuscitation for infants with anomalies has not been systematically assessed. We will employ our previously established technology and tools (video recording, eye tracking, and work systems analyses) to characterize the current state of delivery room resuscitation team performance for infants with anomalies at CHOP. We are applying a Human Factors framework to identify targets to improve the delivery room system to optimize provider performance and clinical outcomes. In addition, we have implemented a multidisciplinary video review program to systematically map and review delivery room resuscitations and identify opportunities for improvement through rapid cycle interventions.

Decision aids can help prevent errors, but most current decision aids are static instruments to disseminate knowledge that providers must synthesize in the context of a rapidly evolving physiologic state. We previously developed a prototype 'Critical Knowledge' platform that displays care guidelines in the resuscitation rooms on large bedside touchscreens. We deployed our existing prototype platform in the Special Delivery Unit clinical space in 2018. This electronic cognitive aid provides a visual reference for indexed content on large touch screens at bedside. In its current state, the platform is used by clinicians to access key disease-specific information before delivery while preparing for resuscitation. We will advance this prototype to the ideal state of a voice-guided, interactive, hands-free decision support that can be used throughout resuscitation. We will connect multi-channel Knowledge Streams that integrate with clinical devices such as patient monitors and ventilators. We will engage with software developers to develop clinical decision support tool that follows algorithms incorporating the patient's diagnosis, expected physiologic phenotype, current age (minutes after birth), interventions in place, and patient response (integrated physiologic data from vital signs). These algorithms will generate individualized, context-specific, and real-time clinical support to clinicians during DR resuscitation of high-risk newborns.