Learning From the Adoption of a Readmission Clinical Decision Support Tool: Group Model Building Approach.

Area of Science:

• Health Informatics
• Health Services Research
• Systems Engineering

Background:

• Computerized clinical decision support (CDS) can enhance patient outcomes and resource planning.
• Adoption factors include perceived usefulness, ease of use, workflow integration, and support.
• CDS adoption is dynamic, influenced by evolving behaviors and workflows.

Purpose of the Study:

• To explore dynamic factors influencing the adoption of a specific CDS tool.
• To examine the implementation of Epic's "Unplanned readmission model version 1" at Duke University Health System.

Main Methods:

• Utilized group model building and system dynamics modeling.
• Conducted workshops with healthcare staff involved in patient discharge decisions.
• Developed causal loop diagrams from workshop transcripts and simulated adoption scenarios.

Main Results:

• Initial adoption was driven by external pressures (e.g., CMS penalties) and institutional incentives.
• Sustained use depended on perceived clinical utility and internal benefits, which were sometimes weak.
• Simulation showed initial success waned with workflow improvements and competing priorities; internal motivation sustained use.

Conclusions:

• External pressure motivates initial CDS tool adoption but may not ensure long-term use.
• Sustained and extensive use of CDS tools is more likely when users perceive direct internal benefits.