Residency Competency Assessments and Early-Career Primary Care Physicians' Antibiotic Prescribing

Antibiotic prescribing varies across primary care physicians (PCPs) and disproportionately harms older adults, yet whether residency training is a factor in this variation is unknown.

To evaluate whether PCPs' Accreditation Council for Graduate Medical Education (ACGME) competency ratings toward the end of residency are associated with their early-career pattern of antibiotic prescribing to older adults.

This cohort study involved graduates of US ACGME-accredited internal medicine and family medicine residency programs between January 1, 2015, and December 31, 2020, who were in their first 5 years of primary care practice. Their ACGME Milestone ratings (hereafter competency ratings) were linked to Medicare Parts B and D public claims from January 1, 2015, to December 31, 2022. Data analysis was conducted between March 2024 and May 2026.

Overall competency rating in the middle of the final year of residency. Ratings for 22 subcompetencies (5-point developmental scale) were averaged within 6 core competency domains (patient care, medical knowledge, professionalism, interpersonal and communication skills, systems-based practice, and practice-based learning and improvement), averaged across domains, and then standardized at the residency program level.

Annual antibiotic prescription rate. This primary outcome was expressed as an incidence rate ratio (IRR) adjusted for physician, residency program, practice location, and patient-panel factors. Residency training program was examined as a secondary factor.

The analysis included 31 800 PCPs (mean [SD] age on practice entry, 32.8 [4.2] years; 15 990 females [50%]) who graduated from 1090 residency programs, representing 129 225 physician-years of prescribing data. Among these physicians, 58% graduated from a US medical school, 90% earned an MD degree, and 51% specialized in internal medicine and 49% specialized in family medicine. Each physician panel had a mean (SD) of 169 (144) Medicare beneficiaries per year (mean [SD] age, 70.3 [3.5] years; 59% female patients). Physicians prescribed a mean (SD) of 54.1 (52.2) antibiotics per year, and the full cohort wrote 7 664 372 antibiotic prescriptions across the study period. In unadjusted analyses, graduates from internal medicine and family medicine residency programs with highest vs lowest quintile prescribing rates prescribed a difference of 14.9 (95% CI, 13.8-15.9; P < .001) and 18.8 (95% CI, 17.9-19.7; P < .001) additional antibiotics per 100 beneficiaries per year, respectively. In adjusted analysis, each 1-SD increase in overall competency rating was associated with 0.7 (95% CI, 0.5-0.9) fewer antibiotic prescriptions per 100 beneficiaries per year (adjusted IRR, 0.98; 95% CI, 0.97-0.99; P < .001). Each year of experience was associated with higher antibiotic prescribing (IRR, 1.01; 95% CI, 1.01-1.02; P < .001), even accounting for the reduction in prescribing that occurred across study years.

This study found that higher assessed competence near the end of residency was associated with lower antibiotic prescribing in early primary care practice. Competency assessments may inform prescribing variation-focused efforts during training and early practice.