Engaging Hospital Staff to Identify Levers for Adoption of Clinical Decision Support: Protocol for a Single-Site Case...

Area of Science:

• Healthcare systems engineering
• Health informatics
• System dynamics modeling

Background:

• Clinical decision support (CDS) tools, including machine learning-based predictions, face adoption challenges due to perceived lack of transparency and utility.
• Integration gaps exist between data scientists and clinical end-users, hindering effective CDS implementation.

Purpose of the Study:

• To utilize group model building (GMB) and system dynamics to identify facilitators and barriers for implementing early warning scores (EWSs) in medical-surgical wards.
• To develop a causal model reflecting health staff insights on EWS implementation.
• To evaluate GMB as a strategy for enhancing CDS adoption.

Main Methods:

• Conducted 3 sequential GMB sessions with health system staff to elicit variables and understand adoption dynamics.
• Developed a causal loop diagram (CLD) to represent system behavior.
• Created a stock-and-flow simulation model from the CLD to quantify feedback loop influences on EWS adoption.
• Assessed perceived acceptability, appropriateness, and feasibility of EWS through pre- and postsession questionnaires.

Main Results:

• Completed 3 GMB sessions and qualitative causal loop diagrams.
• Ongoing data analysis involves translating the CLD into a stock-and-flow simulation model.
• Results will present a CLD, narrative, transcripts, simulation model test results, and participant feedback on EWS and the GMB process.

Conclusions:

• The GMB approach provides a framework for understanding and addressing barriers to CDS implementation.
• Findings from the EWS model will inform broader applications of system dynamics for CDS adoption.
• Future work will extend the model to multiple clinical use cases to capture multilevel adoption factors.