Abstract

Mitigating preventable readmissions, where patients are readmitted for the same primary diagnosis within thirty days, is a significant challenge in delivery of high quality healthcare. Towards this end, it is imperative to understand the cause, risk propensity and timing associated with patient readmissions. This study develops a patient profiling model that can predict the propensity of readmission for a patient as well as the timing of future readmissions. We develop a new model termed as BG/EG Hurdle model that can simultaneously estimate both the propensity and timing of patient readmissions. We test this model using a unique dataset that tracks both patient demographic and clinical data for individual patients across 72 hospitals in North Texas. The results indicate that patient profiles derived from our model can serve as the building block for clinical decision support system to identify patients with high CHF readmission risk.

Download

Share

COinS
 

A Profiling Model for Readmission of Patients with Congestive Heart Failure: A multi-hospital study

Mitigating preventable readmissions, where patients are readmitted for the same primary diagnosis within thirty days, is a significant challenge in delivery of high quality healthcare. Towards this end, it is imperative to understand the cause, risk propensity and timing associated with patient readmissions. This study develops a patient profiling model that can predict the propensity of readmission for a patient as well as the timing of future readmissions. We develop a new model termed as BG/EG Hurdle model that can simultaneously estimate both the propensity and timing of patient readmissions. We test this model using a unique dataset that tracks both patient demographic and clinical data for individual patients across 72 hospitals in North Texas. The results indicate that patient profiles derived from our model can serve as the building block for clinical decision support system to identify patients with high CHF readmission risk.