A Probabilistic Model for Constructing Healthcare Episodes

Episode creation describes the task of classifying medical events and related clinical data to a high-level concept, such as a disease, illness or care. Traditional approaches to the problem have been limited to feature-poor claims records utilizing simplistic strategies such as rules, filters and code transformations. However, these approaches have suffered a number of shortcomings including: inconsistencies in defining episodes; lack of sufficient information to infer episodes; and differences in methods for diagnosing and resolving episodes. With the advent of the electronic medical record, which contains multiple sources of patient-related information, data is now accessible to construct more accurate and refined episodes. A probabilistic model is described that utilizes features extracted from different medical repositories (e.g., claims records, structured medical reports) to guide a context-sensitive combinatorial approach for associating medical data contained in a patient’s record with their underlying episodes. Results from a preliminary evaluation of our prototype on a set of knee pain episodes is presented.