Using standardised patients to assess the quality of primary care in rural China: a county-level mixed-methods study

Primary health care quality is central to recent reforms in China, yet real-world evidence remains limited. This study aimed to assess the quality of primary care delivered by village clinics in rural China after primary health care reforms and to identify system-level drivers of performance gaps. We conducted an explanatory sequential mixed-methods study combining unannounced standardised patient (SP) visits with qualitative interviews and non-participant observation (direct observation of routine clinic practice without researcher involvement, included to contextualise clinician behaviour). Sixty-one unannounced SP visits were completed (influenza n = 30; tuberculosis n = 31). Fifteen participants were interviewed and 42 h of non-participant observation were conducted in six clinics. Outcomes included SP-assessed process quality, diagnostic accuracy and correct case management benchmarked to clinical guidelines. Qualitative analysis of determinants drew on Structure–Process–Outcome and the Consolidated Framework for Implementation Research. Process quality was low: clinicians completed 14.6% of essential history-taking items and 13.3% of indicated examinations; mean consultation time was 3.2 min. Correct diagnosis was 26.2% overall (influenza 13.3% vs. tuberculosis 38.7%) and correct management 31.2% overall (influenza 16.7% vs. tuberculosis 45.2%). Antibiotics were avoided in 63.9% of encounters. Interviews and observation attributed gaps to premature diagnostic closure and limited differential diagnosis skills, reinforced by weak performance incentives, limited township-level managerial authority and accountability, and low-yield training. Tuberculosis care performed relatively better due to long-standing programme support and clearer referral pathways. Despite major investment, primary care quality in village clinics in the study county remained low. Deficits appear driven less by material scarcity than by organisational, incentive and capability constraints. Strengthening county-level, performance-linked purchasing, enabling meaningful patient choice, and embedding workplace-based, technology-supported supervision and decision support may improve process quality and narrow condition-specific gaps.