This paper, an initiative of the Process-Oriented Data Science in Healthcare Alliance (PODS4H), offers a comprehensive overview of the application of process mining techniques within the healthcare domain. Process mining is a set of techniques used across various fields, including healthcare, to extract valuable insights from event logs – data recorded during the execution of processes. These event logs capture essential information like case IDs (e.g., a specific patient), activities performed, timestamps, resources involved, and other attributes. It acts as a bridge between process science and data science, enabling the analysis of real-life process behavior.
While healthcare organizations generate vast amounts of data from staff and machinery, there is currently no evidence of a systematic adoption of process mining beyond targeted research case studies. This is despite its significant potential to help manage and improve both clinical processes (e.g., care pathways) and organizational/administrative processes (e.g., billing). The paper highlights that process mining can contribute to the Quadruple Aim for healthcare improvement: enhancing population health, improving patient experience, reducing costs, and bettering the work-life balance of healthcare professionals.
The authors emphasize that the unique characteristics of healthcare processes – such as their inherent variability and patient-centered nature – demand specific attention when developing and applying process mining techniques. The paper identifies ten distinguishing characteristics of the healthcare domain, including the substantial variability of processes, the importance of infrequent behavior, the use of guidelines and protocols, the presence of “break the glass” situations, and the sensitivity and often low quality of data.
These distinguishing characteristics give rise to ten key challenges that the Process Mining for Healthcare (PM4H) community must address to establish process mining as a powerful tool for evidence-based process analysis and improvement. These challenges include the need to design dedicated methodologies, move beyond just process discovery to include conformance checking and enhancement techniques tailored for healthcare, manage concept drift, deal with real-world data, foster “Do It Yourself” adoption by healthcare professionals, pay close attention to data quality, ensure privacy and security, look at processes from the patient’s perspective, complement Health Information Systems (HISs) with a process perspective, and evolve in symbiosis with broader developments in healthcare. The paper serves as an inspiration for researchers and practitioners to contribute to this evolving field.
APA Reference for this article:
Munoz-Gama, J., Martin, N., Fernandez-Llatas, C., Johnson, O. A., Sepúlveda, M., Helm, E., Galvez-Yanjari, V., Rojas, E., Martinez-Millana, A., Aloini, D., Amantea, I. A., Andrews, R., Arias, M., Beerepoot, I., Benevento, E., Burattin, A., Capurro, D., Carmona, J., Comuzzi, M., Dalmas, B., de la Fuente, R., Di Francescomarino, C., Di Ciccio, C., Gatta, R., Ghidini, C., Gonzalez-Lopez, F., Ibanez-Sanchez, G., Klasky, H. B., Kurniati, A. P., Lu, X., Mannhardt, F., Mans, R., Marcos, M., de Carvalho, R. M., Pegoraro, M., Poon, S. K., Pufahl, L., Reijers, H. A., Remy, S., Rinderle-Ma, S., Sacchi, L., Seoane, F., Song, M., Stefanini, A., Sulis, E., ter Hofstede, A. H. M., Toussaint, P. J., Traver, V., Valero-Ramon, Z., van de Weerd, I., van der Aalst, W. M. P., Vanwersch, R., Weske, M., Wynn, M. T., & Zerbato, F. (2022). Process mining for healthcare: Characteristics and challenges. Journal of Biomedical Informatics, 127, 103994. https://doi.org/10.1016/j.jbi.2022.103994
