Process Mining in Primary Care: Avoiding Adverse Events Due to Hazardous Prescribing

Area of Science:

• Health Informatics
• Clinical Pharmacy
• Data Science in Healthcare

Background:

• Process mining is underutilized in primary care despite its potential.
• Hazardous prescribing, particularly with NSAIDs and antiplatelets, leads to costly serious adverse events like gastro-intestinal bleeds (GiBs).
• Understanding primary care prescribing pathways is crucial for patient safety.

Purpose of the Study:

• To apply process mining techniques to primary care data.
• To analyze prescribing pathways for NSAIDs and antiplatelets.
• To identify challenges and solutions for process mining in primary care.

Main Methods:

• Extraction of prescribing events (drug start/stop) and gastro-intestinal bleeds (GiBs).
• Application of process mining to three distinct prescribing pathways.
• Analysis of event sequences to understand drug usage patterns and associated risks.

Main Results:

• NSAIDs were frequently prescribed short-term, while antiplatelets were often long-term prescriptions.
• Gastro-protection was more commonly co-prescribed with antiplatelets than NSAIDs.
• Identified specific challenges in processing primary care data for mining and proposed mitigation strategies.

Conclusions:

• Process mining of primary care data is achievable and offers significant utility.
• Findings can inform strategies to improve patient safety and reduce healthcare costs associated with adverse drug events.
• Further application of process mining in primary care settings is recommended.