Antibiotic prescribing audit with peer benchmarking does not reduce prescriptions

Among primary care physicians, quarterly personalized antibiotic prescribing audits and feedback with peer benchmarking did not reduce antibiotic prescriptions, according to results of a randomized clinical trial in JAMA Internal Medicine.

Heiner C. Bucher, MD, MPH, director of the Basel Institute for Clinical Epidemiology and Biostatistics and a professor at the University of Basel in Switzerland, told Healio that “in absolute terms, most antibiotics are prescribed in primary care.”

“Overuse of antibiotics is directly correlated with antibiotic resistance in the community,” he said. “To preserve these precious drugs, routine monitoring and feedback in primary care should become a cornerstone of antibiotic stewardship programs in any health care system.”

The randomized clinical trial included 3,426 primary care physicians and pediatricians in Switzerland who were among the top 75% prescribers of antibiotics. The researchers conducted the trial from Jan. 1, 2018, to Dec. 31, 2019.

The physicians were randomly assigned in a 1:1 ratio to undergo quarterly antibiotic prescribing audit and feedback with peer benchmarking or no intervention for 2 years. Those in the intervention group received evidence-based guidelines for UTI and respiratory tract infection management and information about community antibiotic resistance. For audit and feedback, the researchers used anonymized patient-level claims data from the health insurers that serve roughly 50% of people insured in Switzerland.

The researchers noted that those in the intervention group were “blinded regarding the nature of the trial,” and those in the control group did not know of the trial.

Compared with 2017, the baseline year, Bucher and colleagues reported a 4.2% (95% CI, 3.9-4.6) relative increase in the antibiotic prescribing rate during the second year of the intervention. Overall, the intervention group had a median annual antibiotic prescribing rate of 8.2 per 100 consultations vs. 8.4 per 100 consultations in the control group.

The researchers noted that, relative to the overall increase, they found a 0.1% (95% CI, –1.2 to 1) lower antibiotic prescribing rate per 100 consultations compared with the control group. Also, the researchers found no relevant reductions in specific antibiotic prescribing rates between groups, “except for quinolones in the second year of the intervention” (–0.9%; 95% CI, –1.5 to –0.4), they wrote.

“Routine nationwide audit and feedback for antibiotic prescribing in primary care based on claims data does not reduce antibiotic prescribing in a setting with already low antibiotic use,” Bucher said.

He also said that, due to the processing of claims data, there was a delay in feedback that “might have been of less interest in the current clinical situation,” so providing feedback in a shorter time could improve prescribing.

Bucher also noted that the research team “did not have any diagnostic data, making it impossible to give feedback on the appropriateness of antibiotic prescriptions.”

“Thus, providing more personalized feedback or incentives like credit points might be an option [to reduce antibiotic consumption],” he said. “Another option is to provide more point of care tests like PCR panels for respiratory infections with results available within 24 hours.”