An Evidence-Based Decision Support Framework for Clinician Medical Scheduling

In healthcare management, waiting time for consultation is an important measure that has strong associations with patient's satisfaction (i.e., the longer patients wait for consultation, the less satisfied they are). To this end, it is required to optimize medical scheduling for clinicians. A typical approach for deriving the optimized schedules is to perform experiments using discrete event simulation. The existing work has developed how to build a simulation model based on process mining techniques. However, applying this method for outpatient processes straightforwardly, in particular medical scheduling, is challenging: 1) the collected data from electronic health record system requires a series of processes to acquire simulation parameters from the raw data; and 2) even if the derived simulation model fully reflects the reality, there is no systematic approach to deriving effective improvements for simulation analysis, i.e., experimental scenarios. To overcome these challenges, this paper proposes a novel decision support framework for a clinician's schedule using simulation analysis. In the proposed framework, a data-driven simulation model is constructed based on process mining analysis, which includes process discovery, patient arrival rate analysis, and service time analysis. Also, a series of steps to derive the optimal improvement method from the simulation analysis is included in the framework. To demonstrate the usefulness of our approach, we present the case study results with real-world data in a hospital.