Case Study: Kaiser Permanente improves newborn risk prediction with JMP

JMP software helps Kaiser Permanente develop predictive models

Kaiser Permanente’s research team, led by Dr. Gabriel Escobar, needed a better way to predict which newborns with respiratory distress were at risk for shock, respiratory failure, brain damage, or death. The challenge was turning large, complex clinical data into an objective tool that physicians and nurses could use quickly to identify at-risk babies and intervene early, using JMP statistical discovery software.

JMP helped Kaiser Permanente build predictive models using recursive partitioning and logistic regression, which led to the Richardson Score for assessing newborn risk. The score was validated and then disseminated across the Kaiser Permanente Medical Care Program, helping staff assign resources, decide on transfers to NICUs, and improve early treatment decisions. JMP also provided a scalable, cost-effective analytic platform that researchers could use for both advanced modeling and routine analyses.

Kaiser Permanente

Gabriel Escobar

Director, Perinatal Research Unit