Evaluation of general practice residency training programs in Chengdu, China: a mixed-methods study

China has implemented a nationwide standardized “5 + 3” residency training program to strengthen the general practitioner (GP) workforce for primary healthcare institutions, with tertiary hospitals designated as the primary teaching bases. However, residents are trained mainly in specialized tertiary hospitals focused on inpatient care, while many will later work in primary care settings that emphasize preventive medicine and outpatient consultations. The effectiveness of this training model and its alignment with actual primary care demands remain insufficiently examined. This study evaluated the GP training program across 10 tertiary teaching hospitals in Chengdu, China using a convergent mixed-methods design. Quantitative data were collected from 262 residents using the validated Competency Scale for General Practitioners in Primary Healthcare Institutions (Cronbach’s α = 0.979) and analysed using repeated-measures ANOVA, non-parametric tests, and Spearman’s rank correlation. Thematic analysis of 70 open-ended responses was used to explore residents’ training experiences and help interpret the quantitative findings. The mean overall self-reported competency score was 4.109 ± 0.695. No statistically significant difference was found across training hospitals. Scores differed across the five competency domains, F (3.14, 818.06) = 40.79, p < .001, partial η² = 0.135. Public Health Service had the lowest mean score (3.884 ± 0.849) and scored lower than each of the other domains in the adjusted pairwise comparisons. Training satisfaction was positively correlated with overall self-reported competency, Spearman’s ρ = 0.508, 95% CI [0.402, 0.610], Holm-adjusted p < .001. Qualitative analysis identified three themes: limited alignment between tertiary hospital training and outpatient and public health needs, non-educational duties and unclear trainee roles, and limited financial incentives. Residents reported relatively high self-reported competency scores, but the Public Health Service domain scored lower than the other domains. These scores reflect residents’ perceptions rather than objectively assessed competence. The association between training satisfaction and self-reported competency should not be interpreted as causal. Future reforms could consider standardizing primary care rotations, increasing supervised outpatient training, and reviewing remuneration mechanisms.