The Prognostic Respiratory Intensity Scoring Metric (PRISM)

Invasive respiratory interventions save infant lives yet, have negative consequences for their general health and neurodevelopment. Tools that can both accurately summarize the intensity of ventilation needed during an infants stay in the NICU and be used to make predictions about their future health and neurodevelopmental outcomes are currently lacking. Here we present the Prognostic Respiratory Intensity Scoring Metric (PRISM) as an accurate, summative tool of respiratory support needs in hospitalized neonates well suited for predicting an infants future neurodevelopmental outcomes. 218 infants were included in this study. We trained a classification and regression tree model using PRISM scores to classify respiratory diagnoses and compared performance to the Neonatal Sequential Organ Failure Assessment Respiratory subscore (nSOFA-r). PRISM outperforms the nSOFA-r in classifying infants with respiratory distress syndrome (PRISM AUC = 0.71, nSOFA-r AUC = 0.64) and chronic lung disease (PRISM AUC = 0.86, nSOFA-r AUC = 0.76) according to Delongs Test (p=0.002, p=0.001). PRISM had significant associations both with an infants length of hospital stay ({beta} = 0.03 (95% CI [0.02, 0.03], p < 0.001) and gestational age ({beta} =-0.03 (95% CI [-0.03,-0.02], p <0.001). PRISM successfully identifies which infants will develop an intraventricular hemorrhage (83% accuracy) and is highly successful at determining which infants will develop retinopathy of prematurity (95% accuracy). PRISM is a tool designed for clinicians and researchers alike to summarize the invasive nature of neonatal respiratory support.