Multi-perspective process mining for emergency process

Abstract

Background: Multi-perspective process mining is an analytical approach that uses data to gain objective insights and uncover hidden problems in business processes from multiple perspectives.

Objective: In this paper, we apply multi-perspective process mining techniques in the emergency process through a goal-driven performance evaluation method in order to understand and diagnose the timeliness of the emergency process.

Methods: Unstructured and multi-disciplinary emergency data is analyzed by following Goal-Question-Feature-Indicator (GQFI) method. In this paper, the GQFI method is extended with perspectives, and the insights in the enriched event data are examined by a decision tree model. All of them are applied in a systematic way in relation to the goal of assessing and improving the emergency process in a university hospital.

Results: We detected the deviations (e.g., skipping the triage and consultation request steps) and two bottlenecks in the emergency process. Among the suggestions for improving the process, are performing defensive medicine in a harmless manner, classification of the emergency services, ensuring triage step is applied to all patients and effective usage of the call system application in consultation activities.

Conclusion: The results of this study showed that goal-oriented multi-perspective process mining is effective in identifying process improvements in emergency services.

Keywords: Process mining; emergency medicine; multi-perspective process mining; process improvement.