Leveraging workflow control patterns in the domain of clinical practice guidelines

Area of Science:

• Computer Science
• Medical Informatics

Background:

• Clinical practice guidelines (CPGs) are essential for patient care.
• Authoring executable CPG models is labor-intensive, despite tool assistance.
• Facilitating early CPG modeling stages is a key challenge.

Purpose of the Study:

• To propose a pattern-based approach for CPG model authoring.
• To semi-automatically transform CPG patterns into executable languages.
• To streamline the development of clinical practice guidelines.

Main Methods:

• Utilized workflow control patterns from workflow systems and business process management.
• Performed qualitative analysis of CPG texts to assess pattern suitability.
• Implemented selected patterns in Asbru and PROforma CPG languages using BPMN 2.0 notation.
• Developed XSLT transformations for converting BPMN 2.0 patterns to Asbru and PROforma.

Main Results:

• Workflow patterns are largely suitable for CPG procedural knowledge, with some limitations for single-patient care.
• CPGs may necessitate additional patterns beyond standard workflow patterns.
• CPG-suitable patterns are implementable in Asbru and PROforma.
• Individual patterns can be semi-automatically transformed from BPMN 2.0 to executable CPG languages.

Conclusions:

• A pattern and transformation-based approach offers a framework for CPG model development.
• Identifying adequate patterns and implementing transformations are crucial first steps.
• This approach facilitates the efficient authoring of clinical practice guideline models.