Impact of inpatient volume on residents’ In-training examination scores and burnout in Japanese community hospitals: a nationwide cross-sectional study

TL;DR

This study finds that residents in small Japanese hospitals perform as well on exams and have similar burnout rates as those in larger hospitals.

Contribution

The study provides new evidence on the relationship between hospital size, resident performance, and mental health in Japan's medical training system.

Findings

Residents in Very Low-Volume hospitals had higher GM-ITE® scores than Low-Volume hospitals but similar scores to Moderate and High-Volume hospitals.

Burnout symptoms were similarly prevalent across all hospital volume categories.

Very Low-Volume hospitals may be suitable for clinical training if they provide adequate educational environments.

Abstract

The optimal hospital size between adequate clinical training and resident well-being remains underexplored, particularly within Japan’s unique medical training environment. This study examined the relationship between residents’ nationwide General Medicine In-Training Examination (GM-ITE®) scores and their mental health to evaluate the validity of inpatient volume as a criterion for training hospital designation.

We conducted a nationwide cross-sectional study of 7,498 postgraduate year (PGY)-1 and PGY-2 residents from 608 community-based hospitals in Japan who participated in the GM-ITE® from January 17 to 30, 2024. The GM-ITE® is a highly reliable test for evaluating residents’ basic clinical skills. The 2023 GM-ITE® included 80 multiple-choice questions with a maximum score of 80. In the Mini Z 2.0 survey, a score of 3 or more was defined as the presence of burnout symptoms. Training facilities were categorized into four groups by yearly inpatient volume: Very Low (< 3,000), Low (3,000–5,999), Moderate (6,000–9,999), and High (≥ 10,000). We explored the relationships using multivariable analysis between GM-ITE® score, burnout symptoms, and training environment.

The GM-ITE® scores in the Very Low-Volume group were significantly higher than those of the Low-Volume group (45.8 ± 7.3 vs. 44.3 ± 7.1; adjusted coefficient (AC): 2.007; p = 0.021), but did not differ significantly from those of the Moderate- (45.8 ± 7.3 vs. 44.9 ± 6.7; AC: 1.637; p = 0.060) or High-Volume (45.8 ± 7.3 vs. 46.2 ± 6.9; AC: 1.638; p = 0.082) groups. The prevalence of burnout symptoms among residents in the Very Low-Volume group, did not differ significantly from that of residents in the Low- (13.5% vs. 13.1%; adjusted odds ratio (aOR): 0.974; p = 0.940), Moderate- (13.5% vs. 12.8%; aOR: 0.967; p = 0.923), and High-Volume groups (13.5% vs. 12.1%; aOR: 0.994; p = 0.986).

Residents at Very Low-Volume hospitals achieve comparable GM-ITE® scores and mental health outcomes to those of residents at larger hospitals. Even Very Low-Volume hospitals are individually assessed, and if recognized as having an educational environment where residents can acquire basic clinical skills, it may be appropriate to certify them as core clinical training hospitals. The educational environment in Very Low-Volume hospitals could be enhanced by supporting residents’ mental health and advancing international clinical training programs.

The online version contains supplementary material available at 10.1186/s12909-026-08664-3.