Analysing workflows in radiation oncology: A process mining approach to identify treatment process challenges

Area of Science:

• Healthcare Operations Research
• Medical Informatics
• Process Management

Background:

• Radiation therapy (RT) workflows are complex, involving multiple steps and resources, making them suitable for process mining (PM) analysis.
• High-volume RT departments face challenges in workflow optimization and timely treatment delivery.
• Identifying workflow inefficiencies is crucial for improving patient outcomes in radiation therapy.

Purpose of the Study:

• To apply process mining (PM) to characterize the real-world workflow of a high-volume radiation therapy (RT) department.
• To identify critical transitions, bottlenecks, and factors influencing treatment suspensions and cancellations.
• To quantify the impact of workflow inefficiencies on RT treatment timelines.

Main Methods:

• Analysis of 43,183 patient-related events from 8608 RT treatments between January 2017 and December 2021.
• Utilized a first-order Markov model to analyze event sequences and identify workflow anomalies.
• Employed the Kruskal-Wallis test to compare median treatment completion times across different patient sub-cohorts.

Main Results:

• Key RT events including Prescription, Scheduling, CT simulation, RT start, RT cancelled, RT suspension, and RT end were mapped.
• Treatment suspensions significantly impacted the median time from Prescription to RT start across multiple cancer types (p<0.05).
• The study identified specific workflow transitions and delays impacting overall treatment duration.

Conclusions:

• Process mining (PM) serves as a valuable methodological framework for analyzing complex healthcare pathways, such as radiation therapy.
• PM effectively identifies delays, interruptions, and workflow inefficiencies within RT processes.
• The findings highlight the potential of PM for supporting process evaluation and optimization in clinical settings.