Data Science in Clinical Care (2025-2026)
The Data Science in Clinical Care team investigated how clinical decision support (CDS) tools can be more effectively designed, implemented and sustained in real-world healthcare settings.
CDS tools are computerized algorithms that typically draw from electronic health record data to help providers make informed decisions and provide the right treatment to the right patient at the right time. However, CDS tools don’t always perform as well in real-world situations as they do in the development stage, in part because healthcare providers may not use the tools as expected. While some reasons have to do with the tools themselves (e.g., providers not trusting the scores), others have to do with systemic factors (e.g., annual staff turnover can result in gaps in training).
The team partnered with Duke Health staff to explore factors that affect their adoption of CDS tools. Their case study focused on surgical site infections (SSIs) following lower extremity bypass surgery. SSIs remain a significant complication after this procedure, and CDS tools can help identify high-risk patients earlier, but only if the tools reflect how clinicians actually understand and navigate risk.
The team used an approach called Group Model Building, where stakeholders collaboratively map the dynamic factors underlying a complex problem, before building the CDS tool algorithm for SSI risk.
Through workshops conducted with Duke Health surgical residents and nursing staff, the team produced a preliminary Causal Loop Diagram depicting how interacting variables shape SSI risk. The model showed how social determinants of health, like housing insecurity, job instability and limited financial resources amplify SSI risk at every stage. These findings revealed potential intervention points at both the patient and hospital level that a future CDS tool could be designed to address.
Timing
Summer 2025- Spring 2026
Team Outputs
Data Science in Clinical Care (Poster presentation at the Fortin Foundation Bass Connections Showcase, April 15, 2026)
Mapping the Complexity around Surgical Infection Risk (Team profile)
Developed a Causal Loop Diagram that shows how interacting variables shape SSI risk
See related teams, Data Science in Clinical Care (2024-2025) and Data Science in Clinical Care (2026-2027).