Introduction: Rapid advancements in artificial intelligence (AI) have generated uncertainty about the future of radiology among medical students. However, it is unclear whether this has affected radiology residency applications. The purpose of this study was to evaluate recent trends in the Canadian radiology residency match.
Methods: Canadian Resident Matching Service annual data reports from 2010-2020 were collected. Statistics were extracted for Canadian medical graduates applying to radiology in the R-1 main residency match and analyzed using linear regression.
Results: The number of available radiology residency positions decreased (P = .01); declining from 84 in 2010 to 81 in 2020 (mean = 83.1). The overall number of applicants did not change (P = .08, mean = 131.8). The proportion of applicants with radiology as their first choice decreased (P = .001); declining from 4.5% in 2010 to 3.1% in 2020 (mean = 3.4%). The number of applicants applying exclusively to radiology also decreased (P = .02); declining from 39 in 2010 to 16 in 2020 (mean = 23). Positions per applicant (P = 0.24, mean = 0.64), and positions per applicant with radiology as their first choice did not change (P = 0.07, mean = 0.91).
Conclusion: While the overall number of students applying to radiology did not change, the number of applicants ranking radiology as their first or only choice decreased sharply. This analysis corroborates recent reports of increased workload, burnout, and declining reimbursement as well as uncertainty about the future of radiology due to advances in AI.
Keywords: artificial intelligence; postgraduate medical education; radiology; residency training.