Study subjects
This study targeted 247 PHC institutions in Shanghai, systematically evaluating the current status of community-based CHD management from two dimensions: institutional management and general practitioners’ (GPs’) practices. A census method was applied to community health centres, covering all 16 administrative districts in Shanghai, with 247 PHC institutions included as a full-sample survey (details in Supplementary file 1). For community GPs, a multi-stage sampling method was adopted: Stage 1 (Cluster Random Sampling): 50% of PHC institutions were randomly selected from the 16 administrative districts. Stage 2 (Stratified Random Sampling): Licensed physicians within the sampled institutions were stratified into four levels based on professional titles—resident physicians, attending physicians, associate chief physicians, and chief physicians—with 50% of GPs randomly selected from each stratum. In China, GPs are classified into three primary professional title levels based on years of service, qualifications, and professional assessment: junior (e.g., Resident Physician), intermediate (e.g., Attending Physician), and senior (e.g., Associate Chief Physician or Chief Physician). In this study, we used these standard categories to describe the professional composition of participants.
Inclusion Criteria
① PHC institutions in Shanghai are willing to participate.
② Community GPs currently employed and on duty.
③ Participants voluntarily signed informed consent after fully understanding the study’s purpose and content, committing to provide necessary cooperation.
This study was approved by the Ethics Committee of Yangpu Hospital Affiliated to Tongji University (Approval No.: LL-2025-LW-001).
Methods
Survey tools
Based on the two levels of investigation targets (institutions and physicians), the survey included institution-specific and physician-specific questionnaires. Both questionnaires were developed in accordance with the Primary Care Guidelines for Stable Coronary Heart Disease (2020) [19] and the Primary Care Guidelines for Cardiac Rehabilitation in CHD (2020) [20], and were structured using the Donabedian quality assessment framework (Structure-Process-Outcome, SPO model) [21]. This study employs Donabedian’s Structure–Process–Outcome (SPO) model, which provides substantial practical value through its three-dimensional framework for systematically evaluating all aspects of PHC quality [22]. By examining the interrelationships among resource allocation (structure), diagnostic and treatment protocols (process), and core knowledge of CHD (as a proxy outcome indicator), the model offers a comprehensive analytical approach [23]. Given the limited accessibility of PHC outcome data, this study indirectly assesses diagnostic and treatment quality through the CHD knowledge levels of GPs. In contrast to the United Kingdom’s Quality and Outcomes Framework (QOF) [24], which emphasises performance incentives and outcome metrics, the SPO model is better aligned with the quality improvement needs of Shanghai’s PHC system, enabling a more nuanced analysis of current bottlenecks and potential development pathways in CHD management.
The present study encompasses the following dimensions: (1) Structure: Focused on institutions, assessing staffing for CHD speciality clinics, availability of diagnostic/therapeutic equipment, and medication supply. (2) Process: Targeted institutions, covering implementation programs, quality control, and IT infrastructure (e.g., collaboration with referral institutions, high-risk population screening, standardised CHD training programs, and chronic disease information systems). (3) Outcome: Evaluated community GPs’ mastery of CHD core knowledge (e.g., disease concepts, diagnostic criteria, treatment principles, management goals, and drug mechanisms). (4) Challenges and Needs: Surveyed GPs’ perceived barriers and priorities in CHD management (details in Supplementary file 2).
This study adopts the accuracy of GPs’ knowledge, rather than patient treatment outcomes, as the outcome variable, based on two primary considerations. First, obtaining reliable CHD treatment outcome data from PHC institutions in Shanghai is highly challenging: (a) electronic health records are fragmented, and standardised cardiovascular disease endpoint indicators are lacking [25, 26]; and (b) high patient mobility leads to considerable loss-to-follow-up bias [27]. Second, GPs’ knowledge directly influences the standardisation of care, serves as a foundational component of PHC quality, and is closely associated with CHD management outcomes. Therefore, it represents a critical indicator for improving healthcare quality.
Regional classification criteria and quantitative evaluation indicators
In this study, the classification of PHC institutions into urban and suburban areas was primarily based on Shanghai municipal regulations governing the identification of suburban PHC institutions [28]. Specifically, the suburban scope included all PHC institutions in Chongming, Fengxian, Qingpu, and Jinshan, Songjiang districts, as well as selected areas of Pudong New district (details in Supplementary file 3). According to these regulations, the study included 154 urban PHC institutions and 93 suburban institutions.
To evaluate medical resource allocation and diagnostic capability at the PHC level, two indicators were employed: mention rate and correct response rate. The mention rate is defined as the proportion of institutions that reported being equipped with specific medications or diagnostic tools, calculated as: Number of institutions equipped with specific medications or devices / Total number of surveyed institutions × 100%. This indicator reflects the availability and distribution of essential resources, similar to facility readiness metrics used in service capacity assessments [29].The correct response rate, defined as the proportion of institutions correctly identifying the recommended diagnostic or treatment option in knowledge-based scenarios, serves as a proxy for diagnostic capacity and provider knowledge, calculated as: Number of respondents providing accurate answers / Total number of surveyed personnel× 100%.This approach aligns with previous health system evaluations that assess institutional knowledge and readiness through simulated case responses or structured questionnaires [30]. Additionally, to determine the ranking of resource indicators, we calculated the overall mention rate for each specific item within its category. This mention rate was defined as the proportion of institutions reporting that they met the predefined standard for that item. Items were then ranked in descending order of their mention rates to highlight the most and least widely available resources across surveyed institutions.
These quantitative indicators objectively reflect the resource allocation status and diagnostic service capabilities of PHC institutions. They provide scientific evidence for interregional comparisons and offer research support for optimising regional health resource allocation.
Survey methodology and quality control
From April to May 2024, a questionnaire survey targeting PHC institutions and their GPs was conducted. To ensure scientific validity and accuracy, domain experts were invited to review the questionnaire content prior to formal implementation. A pilot survey was then carried out following the convenience sampling principle, involving two PHC institutions and ten GPs. Ambiguous questions identified during the pilot phase were revised to enhance clarity, resulting in the finalised version of the questionnaire.
The formal questionnaires were distributed electronically via Wenjuanxing (a professional online survey platform), utilising the quality control network of the Shanghai Municipal Clinical Quality Control Centre for General Practice. During pre-testing, we observed that respondents who completed the full questionnaire in under two minutes frequently provided low-quality or patterned responses (e.g., selecting the same option throughout), indicating inattentiveness. This observation aligns with prior studies that have set comparable thresholds to identify insufficient effort responding [31]. Based on these findings, responses with completion times under two minutes or containing missing entries were classified as invalid and excluded from analysis. This rigorous quality control protocol guaranteed the integrity and representativeness of the collected data.
Statistical analysis
Survey data collected via the Wenjuanxing platform were exported to Excel and subsequently imported into Statistical Package for Social Sciences (SPSS) for Mac (Version 26.0, SPSS, Inc., Chicago, IL, USA) for statistical analysis. During data processing, continuous variables such as years of professional experience were assessed for normality using standard tests. Variables conforming to a normal distribution were described using mean ± standard deviation. Categorical data, including frequencies and proportions of classification variables, were expressed as rates or percentages. Intergroup comparisons were performed using the chi-square (χ²) test to evaluate differences between subgroups, with a P-value < 0.05 considered statistically significant. This analytical framework ensured rigorous evaluation of associations and trends within the dataset, aligning with established methodologies for healthcare resource and service capability research.
link