Identifying and investigating empirical patient-pathways

Identifying and investigating empirical patient-pathways the healthcare sector, large volumes of data are now being accumulated, including health insurance claims, electronic health records, hospital records, imaging data, as well as information from wearables and fitness devices [25].In Germany, health insurance claims data have increasingly been used for research [26].These datasets typically provide basic demographic information, diagnoses, medication prescriptions, information on laboratory tests, medical procedures, physician visits, and hospitalizations [26, 27].Statutory regulations ensure data standardization across different insurance providers [28].The main drawback, however, stems from the fact that the data were collected for billing purposes rather than for research.This necessitates careful operationalization of populations, events, and diseases, often relying on surrogate parameters [26].Moreover, the data offer only limited clinical detail: while a laboratory test is recorded in the data, the result of the test remains unknown.Medical information such as blood pressure, weight, height, test results or X-rays are not included in claims data, to name a few important examples [26].Yet, claims data are well-suited for patient pathway analysis because of their comprehensiveness, inclusion of large case numbers over extended periods of time, and relatively low selection bias [26]. Analytical methods to identify empirical patient pathwaysResearch on empirical patient pathways requires not only appropriate data but also robust analytical methods.Tracing a single patient's pathway based on their medical history, visualizing it, and comparing it to treatment recommendations may seem intuitive.However, extending these steps to entire populations implies handling numerous, possibly diverging pathways simultaneously.Extracting meaningful insights from this analysis is therefore both complex and challenging. Knowledge discovery in databasesParallel to the data explosion, interest and efforts grew to make this data usable and to develop methods that could extract meaningful information: The field of Knowledge Discovery in Databases (KDD) can be defined as the 'process of identifying valid, novel, potentially useful, and ultimately understandable structure in data' [29, 30,