Process mining, a fusion of data mining and process science, provides a powerful approach for optimising business processes using event log data. This technique is particularly relevant in the healthcare sector, where challenges such as increased costs, long waiting times, inefficient resource management, and poor patient outcomes persist. Using event logs that contain time-stamped clinical events, process mining can help identify system bottlenecks, anomalies, and inefficiency. This paper proposes a process mining methodology to study patient experiences in demographic groups. The methodology consists of four stages: data preprocessing, process discovery using specialized algorithms, and evaluation of process models through conformance checking. Applied to the MIMICEL ED case study that focuses on demographic attributes of gender and race, this approach reveals significant similarities and differences in patient journeys, offering actionable insights for the optimization of healthcare resources. Our findings demonstrate how our framework can effectively utilize process mining to compare patient experiences across demographics to reveal disparities and patterns critical to clinical process improvement. These results show the potential of this framework to inform health professionals of system inefficiencies and better resource management.
Jayawardane et al. (Thu,) studied this question.