What Shapes a Radiation Oncologist? A Preliminary National Survey of Residents and Attendings on Perceived Determinants of Residency Training Quality

Introduction/Background The modifiable features of the radiation oncology residency environment that most strongly influence perceived training quality remain incompletely characterized. We sought to identify physician-reported determinants and to compare perceptions between residents and practicing attendings. Methods Cross-sectional, anonymous, electronic pilot survey distributed through program directors and chairs at 110 United States academic radiation oncology residency programs. A 15-item instrument covered training content, technology exposure, teaching methods, training environment, and satisfaction. Between-group differences were assessed with Fisher’s exact test (SAS 9.4). Results/Findings Eighty-three physicians responded (42 residents, 41 attendings). Education during outpatient consultations was ranked most important training component more often by attendings than residents (58.5% vs. 38.1%; p = 0.032). Mentorship was the most commonly cited positive influence (49.4% overall). The distribution of responses to the most negative training-environment factor differed significantly between groups (p = 0.003), driven by attendings’ greater endorsement of malignant leadership and a markedly higher non-response rate among residents on this single item (28.6% vs. 2.4%). One-on-one case discussion was the most valued teaching modality (combined 75.9%); only 12% of respondents rated remote didactic instruction as equal to or better than in-person (p = 0.85). Discussion/Implications Signals from this pilot are consistent with mentorship, faculty teaching engagement, outpatient clinical teaching, and one-on-one instruction as candidate pillars of high-quality training. Between-group discordance on negative-experience items underscores the need for confidential channels for trainee feedback. Findings are hypothesis-generating given a convenience sample drawn predominantly from academic practice and a non-validated forced-choice instrument. Conclusion A formally validated instrument with broader construct coverage, multi-site sampling, and inclusion of community-practice physicians is the appropriate next step.