Predicting Hospital Length of Stay of Neonates Admitted to the NICU Using Data Mining Techniques

Hospitals face many pressures, including limited budgets and resources. The Intensive Care Unit (ICU) mostly includes patients who are in critical condition and require costly sources of treatment and has attracted much attention from the medical community. The ability to predict the length of stay for newborns in the Neonatal Intensive Care Unit (NICU) can assist the health care system in allocating needed resources and also has clinical value as an indicator of newborn’s health status. This research utilized the Medical Information Mart for Intensive Care III database (MIMIC III), and the performance of different machine learning models on NICU patients was discussed. Data was filtered, extracted, and preprocessed from the database, and the preprocessing step included a vast amount of feature engineering. The performance of various regression models for predicting hospital length of stay for NICU patients was discussed and compared. Finally, high performing results with a high R2 score of 0.78 by exploiting only patients’ diagnoses data and demographics obtained at the first 24 hours of the admission was achieved.