Peer comparison interventions increase physician burnout

Comparing physicians to their peers to boost job performance decreased job satisfaction and worsened the escalating issue of physician burnout, according to research published in Proceedings of the National Academy of Sciences.

“Publicizing data among all faculty feels inappropriate, as if we are all being ranked/valued according to this metric,” a primary care physician who participated in the study told researchers.

In health care settings, peer comparison interventions have been shown to curb antibiotic overprescribing, improve ED efficiency and increase adherence to best practices, according to Joseph S. Reiff, a PhD student in behavioral decision making at the University of California, Los Angeles, and colleagues. However, less is known about how these interventions affect the well-being of recipients.

To assess the effect of these interventions on physicians, Reiff and colleagues conducted a 5-month field experiment that included nearly 200 PCPs and 46,631 patients. The PCPs were cluster randomized at the clinic level to one of three study arms: a control group, a peer comparison intervention that aimed to improve preventive care performance, or the peer comparison intervention plus leadership training.

All PCPs in the study received monthly emails with feedback about their preventive care performance, which was summarized with a health maintenance completion rate reflecting the proportion of recommended preventive care measures that their patients received in the previous 3 months.

In the control group, the emails only included feedback about the PCP’s personal score. In the intervention groups, the emails also included a list of the month’s top 25 PCPs and information about where the PCP fell in the performance distribution. In the intervention plus leadership training group, the leaders at each clinic also participated in training on how to help physicians’ improve their preventive care performance.

Reiff and colleagues found no significant difference in order rates for preventive services among PCPs in the intervention groups compared with the control group.

When the researchers compared the control condition with the peer comparison intervention, they found a significant decrease in job satisfaction and an increase in burnout. Those who received a combination of peer comparison and leadership support training, though, experienced significantly higher job satisfaction and lower burnout.

One year after the intervention, the researchers presented PCPs with an example of the peer comparison email, regardless of which group they had been designated to for the study. When asked if they would want the emails to continue being sent, 54% of 150 PCPs who responded to the question said they would prefer if the emails did not continue.

The responses to an open-ended question further revealed the negative attitudes that PCPs held toward peer comparison, which were consistent with other research that suggests comparisons to higher-ranked peers “can be discouraging” and may result “in feelings of shame or stress,” Reiff and colleagues wrote.

“Frankly, I think it’s inappropriate,” one physician wrote in their response.

Reiff and colleagues wrote that “the intervention inadvertently signaled a lack of support” from their leaders, with some reporting that the intervention seemed “transgressive.”

“The use of peer comparison interventions can alter workers’ perceptions of and relationships with the leaders implementing the intervention as they try to make sense of how and why this information is being presented to them,” the researchers wrote. “Workers may perceive their leaders of the intervention as reflecting inadequate leadership support. Given that leadership support is key to work-related well-being, job satisfaction and burnout may be harmed by the use of peer comparison interventions.”

Training leaders in how to support physicians, however, mitigated the negative effects.

“These qualitative responses suggest that the manner in which the peer comparison intervention was administered in our context was seen as normatively inappropriate and reductionist and that adding leadership support training buffered against these perceptions by helping leaders contextualize the intervention,” Reiff and colleagues wrote.

Another option that might help is using an absolute criterion instead of evaluating performance in relative terms, which “may have seemed unjust,” according to the researchers. An absolute criterion, they added, would additionally allow “for the public list of top performers to potentially grow over time, which could have motivated people by highlighting a growing trend.”

“Our work suggests that if policymakers and organizational leaders only measure the behavioral outcomes of such interventions, they risk overlooking important effects on less visible outcomes, such as job satisfaction and burnout,” the researchers concluded. “These psychological outcomes need to be accounted for to estimate the aggregate impact of policies and to improve their design and implementation.”