Effect of Social Comparison Feedback on Laboratory Test Ordering for Hospitalized Patients: A Randomized Controlled Trial

Abstract [from journal]

Background: Social comparison feedback is an increasingly popular strategy that uses performance report cards to modify physician behavior. Our objective was to test the effect of such feedback on the ordering of routine laboratory tests for hospitalized patients, a practice considered overused.

Methods: This was a single-blinded randomized controlled trial. Between January and June 2016, physicians on six general medicine teams at the Hospital of the University of Pennsylvania were cluster randomized with equal allocation to two arms: (1) those e-mailed a summary of their routine laboratory test ordering vs. the service average for the prior week, linked to a continuously updated personalized dashboard containing patient-level details, and snapshot of the dashboard and (2) those who did not receive the intervention. The primary outcome was the count of routine laboratory test orders placed by a physician per patient-day. We modeled the count of orders by each physician per patient-day after the intervention as a function of trial arm and the physician’s order count before the intervention. The count outcome was modeled using negative binomial models with adjustment for clustering within teams.

Results: One hundred and fourteen interns and residents participated. We did not observe a statistically significant difference in adjusted reduction in routine laboratory ordering between the intervention and control physicians (physicians in the intervention group ordered 0.14 fewer tests per patient-day than physicians in the control group, 95% CI − 0.56 to 0.27, p = 0.50). Physicians whose absolute ordering rate deviated from the peer rate by more than 1.0 laboratory test per patient-day reduced their laboratory ordering by 0.80 orders per patient-day (95% CI − 1.58 to − 0.02, p = 0.04).

Conclusions: Personalized social comparison feedback on routine laboratory ordering did not change targeted behavior among physicians, although there was a significant decrease in orders among participants who deviated more from the peer rate.