Hospital Presence and Reported Working Hours Among Specialty Residents in Japan: A Nationwide Cross-Sectional Survey Conducted in 2024

Background: Accurate assessment of physicians' working hours is essential for evaluating working conditions, as reported working hours may not fully reflect the time spent in hospitals.We examined discordance between reported working hours and hospital presence among physicians undergoing specialty training.Methods: We analyzed 996 responses from 198 hospitals in a voluntary nationwide specialty-resident survey distributed through training supervisors at 955 hospitals during July 20-August 10, 2024 (198/955 hospitals; 20.7% hospital participation).Both working-hour and hospital-presence questions referred to the same most recent week and included outside clinical work, including moonlighting.Categories were cross-tabulated at 60 and 80 hours/week using exact McNemar tests.An exploratory reduced multivariable model accounted for hospital clustering.Results: Reported working hours were ≥60 hours/week for 415/996 (41.7%) and ≥80 hours/week for 78/996 (7.8%), compared with 613 (61.5%) and 187 (18.8%) for hospital presence.At the 60-hour threshold, 210 reported ≥60 hours of presence but <60 hours of work, whereas 12 showed the reverse pattern (222 (22.3%) discordant; p<0.001).At the 80-hour threshold, discordance was observed in 133 (13.4%; p<0.001).In exploratory adjusted analysis, working ≥60 hours/week was associated with male sex (adjusted odds ratio (aOR) 1.65, 95% CI 1.26-2.16),surgery versus internal medicine (2.20, 1.40-3.46),and urban setting (1.57, 1.15-2.14),while psychiatry was inversely associated (0.16, 0.09-0.29).Conclusions: Among participating specialty residents, prolonged weekly hospital presence was reported substantially more frequently than prolonged weekly working hours.Neither indicator directly establishes legally defined overtime, unpaid work, patient harm, or a causal effect of the reform.