Preterm Clinical Comorbidity Index: A Novel Scoring System and Modeling for Predicting Length of Stay among Preterm Infants during Birth Hospitalization

Abstract

Abstract Background: A scoring system that accounts for comorbidities will provide a novel model to predict the length of stay for preterm infant. Objective: To develop a novel scoring system to predict the length of stay in neonatal intensive care units for all admissions of preterm babies based on their comorbidity developed during the stay. Methods: A total of 25 comorbidities were identified. The Cox regression model was used to develop a standardized Preterm Clinical Comorbidity Index (PCCI) score for each comorbidity. The final length of the stay prediction model was developed using PCCI scores by gestational age. Results: Assigned comorbidity scores for each co-morbidity ranged from -1 to 6. Least square regression using index score to predict LOS had nearly comparable R2 in the main sample (0.76) and in a random sample (0.79). Conclusion: This novel scoring system can be used in the counseling of parents, evaluating healthcare policies, and insurance billing.