PEPCOR — A risk prediction model for pediatric intensive care units utilizing ventilator days and length of stay

Abstract

Great sources of concern for pediatric and neonatal intensive care units are the total resource utilization and cost of caring for very sick children. This paper attempts to create a usable clinical decision support system (PEPCOR) that would help improve personalized health care and avoid unnecessary secondary, and costly, procedures that do not improve the care of a critically ill patient. The system uses the Children's Healthcare of Atlanta ICU database to analyze the effects on the risk, (quantified as the risk of invasive ventilation and associated complications) when combining two procedures versus administering one procedure without the other. The risk is calculated by computing the ratio of ventilator days to the length of stay, with a higher risk score indicating the second procedure did not improve the overall health of the patient. 82/90 of the procedure combinations showed a statistically significant difference with a p-value <; 0.05 when the stand-alone procedure was compared to the combination of two procedures.

DOI: 10.1109/BHI.2016.7455841

3 Figures and Tables

Cite this paper

@article{Pethiyagoda2016PEPCORA, title={PEPCOR — A risk prediction model for pediatric intensive care units utilizing ventilator days and length of stay}, author={Theruni Pethiyagoda and Nikhil Chanani and Chihwen Cheng and May D. Wang}, journal={2016 IEEE-EMBS International Conference on Biomedical and Health Informatics (BHI)}, year={2016}, pages={86-89} }