Comparison of antibiotic prescription indexes by patient and physician gender and age: a cross-sectional study based on administrative data

Few studies have examined the association between gender-differentiated patient–physician interactions and antimicrobial prescription (AP) patterns. We analyzed gender-specific differences in AP in relation to the sex and age of general practitioners (GPs), drawing on data from a comprehensive regional database in Tuscany, Italy. We extracted data on primary care APs for 2023 using a population sample of 3,022,332 patients and 2,311 GPs. We used defined daily doses per 1,000 population per day (DDD/1000 population/day), the percentage of subjects with at least one antimicrobial prescription (AP prevalence), and the percentage of Access AP according to Access, Watch and Reserve (AwaRE) classification as indicators. We fitted regression models for each indicator with patient-physician gender dyads and other covariates. Adjusted indicators and average marginal effects (AMEs) of having a female GP separately for the female and male patients are reported together with the difference between the above AMEs. A total of 1,583,893 female patients (52.4%) and 1,104 female GPs (54.1%) were included. Same-sex care was observed for 774,411 female patients (48.9%) and 815,887 male patients (56.7%). Female GPs prescribed fewer antibiotics—in terms of both DDD/1000 population/day and prevalence of AP—and were more likely to prescribe Access antibiotics. These differences were more pronounced among male patients. Among male patients, having a female GP was associated with an adjusted AME of -0.78 DDD/1000 population/day (95% CI -1.09 to -0.47), -0.21% in AP prevalence (95% CI -1.74 to -0.69), and + 0.73% in proportion of Access AP (95% CI -0.10 to 1.56). Among female patients with a female GP, the adjusted AME was − 0.51 DDD/1000 population/day (95% CI -0.86 to -0.15), -0.59% in AP prevalence (95% CI -1.18 to -0.004), and + 0.24% in proportion of Access AP (95% CI -0.52 to 0.99). When the patient-GP gender dyad and GP age were included, male GPs under 40 years of age showed consistently lower antibiotic prescription—both in DDD/1000 population/day and AP prevalence—and a higher proportion of Access AP compared with female GPs, particularly among female patients. In this subgroup, the adjusted AME of having a female GP was + 0.87 DDD/1000 population/day (95% CI 0.06 to 1.69), + 1.86% in AP prevalence (95% CI 0.52 to 3.2), and − 1.44% in Access AP (95% CI -3.01 to 0.13). In our analysis of a large administrative dataset, we observed that the gender-differentiated interaction between physician and patient was associated with antimicrobial prescription. Overall, female GPs showed lower AP rates. However, among physicians under 40 years of age, male GPs had lower prescription indicators than their female counterparts, especially for male patients. The differences highlighted in our study could be the target of stewardship interventions.