Evaluating the impact of audit and feedback using a novel text and web‐based platform with unblinded peer comparison

Hospitalists vary in their ordering practices related to labs, imaging, and discharge order times, which can lead to inefficient care, excess utilization of diagnostic tests, and prolonged length of stay. Peer comparison can be an effective way to influence physician practices. We implemented an automated, text-based, unblinded peer comparison platform (Agathos) in a hospitalist group and measured: (1) hospitalist engagement with unblinded peer comparison and (2) the subsequent change in process and outcome metrics. We found a statistically significant decrease in computed tomography (CT) utilization as well as an increase in discharge orders by 10 a.m. The intervention achieved a 78% monthly engagement rate. We did not detect a difference in medical consult utilization, echocardiography utilization, repeat daily lab utilization, length of stay, or 30-day readmissions.