Objective The objective was to conduct a pilot randomized controlled trial to assess the feasibility, logistics, and potential effect of monthly provider funnel plot feedback reports from Press Ganey data and semiannual face-to-face coaching sessions to improve patient satisfaction scores. Methods This was a pilot randomized controlled trial of 25 emergency medicine faculty providers in one urban academic emergency department. We enrolled full-time clinical faculty with at least 12 months of baseline Press Ganey data, who anticipated working in the ED for at least 12 additional months. Providers were randomized into intervention or control groups in a 1:1 ratio. The intervention group had an initial 20-minute meeting to introduce the funnel plot feedback tool and standardized feedback based on their baseline Press Ganey scores and then received a monthly e-mail with their individualized funnel plot depicting cumulative Press Ganey scores (compared to their baseline score and the mean score of all providers) for 12 months. The primary outcome was the difference in Press Ganey “doctor-overall” scores between treatment groups at 12 months. We used a weighted analysis of covariance model to analyze the study groups, accounting for variation in the number of surveys by provider and baseline scores. Results Of 36 eligible faculty, we enrolled 25 providers, 13 of whom were randomized to the intervention group and 12 to the control group. During the study period, there were 815 Press Ganey surveys returned, ranging from four to 71 surveys per provider. For the standardized overall doctor score over 12 months (primary outcome), there was no difference between the intervention and control groups (difference = 1.3 points, 95% confidence interval = –2.4 to 5.9, p = 0.47). Similarly, there was no difference between groups when evaluating the four categories of doctor-specific patient satisfaction scores from the Press Ganey survey (all p > 0.05). Conclusions In this pilot trial of monthly provider funnel plot Press Ganey feedback reports, there was no difference in patient satisfaction scores between the intervention and control groups after 12 months. While this study was not powered to detect outcome differences, we demonstrate the feasibility, logistics, and effect sizes that could be used to inform future definitive trials.
