Peer effects on appropriate antibiotic prescribing within a patient-sharing physician network in China

Resource limitations frequently impede efforts to enhance antibiotic prescribing practices. Previous studies have suggested that physicians' behaviors can be influenced by their peers, potentially amplifying the effect of interventions through peer effects. This study aimed to examine the peer effects on physicians' appropriateness of antibiotic prescribing and to identify key physicians within their networks.

We extracted outpatient records from a regional electronic medical records database in Yinzhou, China. Physicians who prescribed more than 100 outpatient antibiotics prescription annually were included. We constructed physician networks by connecting physicians who shared at least ten patients in the same year and calculated networks' descriptive indicators. We then estimated the magnitude of peer effects on antibiotic prescribing by regressing the rate of appropriate antibiotic prescription for each physician in the current year on the average rate of their peers in the previous year.

We included 2,586 physicians, generating 13,856 physician-year observations, with an overall annual appropriate antibiotic prescribing rate of 60.7%. After adjustment, an 1% increase in the appropriate antibiotic prescribing rate among peers in the previous year is significantly associated with a 0.62% (95% CI 0.58, 0.66) increase in an individual physician's rate in the current year. Physicians with higher network centrality associate with greater effects on their peers, compared with those with lower centrality (stratified by degree centrality: 0.57[95% CI = 0.53, 0.61] vs 0.05 [0.04, 0.06]).

Peer effect provides a promising solution to improve intervention efficiency by targeting influential opinion leaders and facilitate the diffusion of appropriate antibiotic prescribing.