Predicting hospital admissions and emergency room visits using remote home monitoring data

© 2017 IEEE. The costs of lengthy hospital admissions (HA) and multiple emergency room visits (ER Visits) from patients with conditions such as heart failure (HF) and chronic obstructive pulmonary disease (COPD) can place a significant burden on healthcare systems. Understanding the various factors contributing to hospitalization and ER visits could aid cost-effective management in the delivery of services leading to potential improvement on quality of life for patients. This can be facilitated by collecting data using remoting patient monitoring (RPM) services and using analytics to discover important factors about patients. This paper presents our research that utilizes predictive modeling to determine key factors that are significant determinant to hospitalization and multiple ER Visits. The results shows that gender, past medical history and vital status are key factors to hospital admissions and ER Visits. Additionally, when a factor to indicate the period before, during and after an ER Visits was included, the resulting model shows a very high likelihood ratio and improved p values on all vital status. Our results shows that more research is needed to fully understand the temporal patterns among variables during hospitalization or ER visit