Timely Prognosis and Intervention Management
More than 200,000 hospitalized patients have cardiac arrests each year in the US, and 75% of those die. Many of those cardiac arrests could be prevented and many patients who did have cardiac arrests could be saved by timely risk assessment and transfer to an intensive care unit (ICU) followed by appropriate intervention/treatment. Because the time frames are very short and resources are extremely limited, it is crucial to know with high confidence which patients need to be sent to ICU immediately and what treatment/intervention they should receive. This life-saving project uses information from the electronic health record (EHR) to build a personalized Hidden Markov Model of a patient's physiological status over time, to provide a summary risk score for assisting with timely decisions on transfer of patients from hospital wards to the ICU and to recommend specific treatment(s)/intervention(s). The algorithms developed in this project outperform the previous state-of-the-art algorithms (e.g., the Rothman Index), and if implemented, have the potential to save thousands of lives each year.
Prof. Mihaela van der Schaar