Paper Number
1520
Abstract
Unplanned readmission entails unnecessary risks for patients and avoidable waste of medical resources, especially intensive care unit (ICU) readmissions, which increases likelihood of length of stay and more severely mortality. Identifying patients who are likely to suffer unplanned ICU readmission can benefit both patients and hospitals. Readmission is typically predicted by statistical features extracted from completed ICU stays. The development of predictive process monitoring (PPM) technique aims to learn from complete traces and predict the outcome of ongoing ones. In this paper, we adopt PPM to view ICU stay from electronic health record (EHR) as a continuous process trace to enable us to discover clinical and also process features to predict likelihood of readmission. Using events logs extracted from MIMIC-IV database, the results show that our approach can achieve up to 65% accuracy during the early stage of each ICU stay and demonstrate the feasibility of applying PPM to unplanned ICU readmission prediction.
Recommended Citation
CHEN, QIFAN; Lu, Yang; Tam, Charmaine; and Poon, Simon, "Outcome-Oriented Predictive Process Monitoring to Predict Unplanned ICU Readmission in MIMIC-IV Database" (2022). ECIS 2022 Research-in-Progress Papers. 31.
https://aisel.aisnet.org/ecis2022_rip/31
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.