Using Participatory Design to Engage Physicians in the Development of a Provider-Level Performance Dashboard and Feedback System

Abstract

Problem definition: Performance feedback, in which clinicians are given data on select metrics, is widely used in the context of quality improvement. However, there is a lack of practical guidance describing the process of developing performance feedback systems.

Initial approach: This study took place at the University of California, San Francisco (UCSF) with hospitalist physicians. Participatory design methodology was used to develop a performance dashboard and feedback system. Twenty hospitalist physicians participated in a series of six design sessions and two surveys. Each design session and survey systematically addressed key components of the feedback system, including design, metric selection, data delivery, and incentives. The Capability Opportunity Motivation and Behavior (COM-B) model was then used to identify behavior change interventions to facilitate engagement with the dashboard during a pilot implementation.

Key insights, lessons learned: In regard to performance improvement, physicians preferred collaboration over competition and internal motivation over external incentives. Physicians preferred that the dashboard be used as a tool to aid in clinical practice improvement and not punitively by leadership. Metrics that were clinical or patient-centered were perceived as more meaningful and more likely to motivate behavior change.

Next steps: The performance dashboard has been introduced to the entire hospitalist group, and evaluation of implementation continues by monitoring engagement and physician attitudes. This will be followed by targeted feedback interventions to attempt to improve performance.

Figure 1:

Shown here is a prototype data dashboard that inspired our dashboard. Provider performance metrics are displayed on a radar graph (or “spiderweb plot”) in yellow, while desired targets are in blue. Data are scaled such that higher values indicate better performance. LOS, length of stay; PCP, primary care physician; VTE, venous thromboembolism.

Figure 2:

Shown here is the final performance dashboard based on the participatory design process. Top: Radar graph depicting performance of provider in yellow, compared to divisional targets in blue. Data are scaled such that higher values indicate better performance, and font color indicates if provider is performing better than (green) or worse than (red) desired targets. Bottom: Individual user’s performance for each metric is ranked in percentile groups; user’s individual performance is highlighted in yellow. DHM, Division of Hospital Medicine; LOS, length of stay; ACP, advanced care planning; H&P, history and physical.