Background: Township hospitals in China face the challenge of shortage and turnover of healthcare workers. This study aims to evaluate the relationship between organisational support for career development (OSCD), organisational commitment, and turnover intentions among healthcare workers in township hospitals. Methods: The data in this study were obtained from the Health Service Capacity Survey of Primary Health Institutions (2020), one of the special surveys of rural health poverty alleviation in Henan, China. The sample comprised 298 healthcare workers. Three standardised instruments were used: the turnover intention scale, OSCD scale, and organisational commitment scale. Descriptive statistics, One-way Analysis of Variance, Pearson correlation analysis, and the simple mediation model were used for data analysis. Results: The results showed that the mean score of the healthcare workers’ turnover intentions was 2.21 ± 0.77, which was low. The healthcare workers’ turnover intentions had significant differences in gender, age, marital status, education, professional title, and organisational tenure. OSCD had no significant direct relationship with turnover intentions, while having a significant positive direct relationship with organisational commitment. Organisational commitment had a significant negative direct relationship with turnover intentions, and played a fully mediating role in the relationship between OSCD and turnover intentions. Conclusions: OSCD had an indirect negative effect on healthcare workers’ turnover intentions in township hospitals through the full mediation of organisational commitment. The findings of this study suggest the importance of policymakers and organisation managers to improve OSCD practice and pay attention to ‘humanistic management’. In addition, the formulation and implementation of talent retention policies should consider socio-demographic differences.

Background: The health management plays an important role in improving the quality of life of the elderly and relieving the pressure of health resource consumption. This study aims to assess the income-related inequality in utilisation of health management services (HMS) for the elderly and the contribution of the related factors to inequality in rural Henan China. Methods: The data from 2015 Henan Rural Residents Health Survey with 1403 elderly people as the final sample were used for analysis. The concentration index (CI) was used to measure inequality in HMS utilisation for the elderly (no HMS, health assessment, physical examination, auxiliary examination, and health guidance). The decomposition of CI was adopted to explain the contribution of various determinants to inequality in HMS utilisation for the elderly. Results: No HMS utilisation was disproportionately concentrated among the poor (CI = − 0.0730, p = 0.0155), utilisation of physical and auxiliary examination was disproportionately concentrated among the rich (CI = 0.0575, p = 0.0448; CI = 0.0811, p = 0.0044). In addition, the pro-poor effects of health assessment and guidance utilisation were not statistically significant (CI = − 0.0173, p = 0.4617; CI = − 0.0213, p = 0.3900). The results of CI decomposition revealed that household income and family size made positive contributions to inequality while social medical insurance, gender, marital status, and age made negative contributions to inequality. The improved service satisfaction with village clinics could reduce inequality in HMS utilisation, while the improved service satisfaction with township hospitals could increase inequality in HMS utilisation. Conclusions: Although HMS for the elderly is provided free of charge, its accessibility remains pro-rich due to various factors. Policy makers should adopt effective interventions to resolve the contradiction between these factors and the utilisation of HMS, and redress inequality in the utilisation of HMS.

This study investigated the association between living arrangements and psychological well-being (PWB) among older adults in rural China and whether activities of daily living (ADL) disability moderate this association. The study data were drawn from the sixth to eighth waves of the Chinese Longitudinal Healthy Longevity Survey. The baseline sample included 4,042 individuals aged 65 years and above. PWB was measured using seven mood-related and personality-related questions. ADL disability was measured using Katz’s 6 items of ADL. Living arrangements were categorized into six categories. Linear mixed-effect models were used to examine the association between older adults’ living arrangements and PWB as well as the moderating effect of ADL disability on this relationship. The results showed that compared with living alone, living with spouse only, and living with spouse and children were positively associated with PWB (β = .621, p < .01; β = .527, p < .05), and living with children (no spouse) was negatively associated with PWB (β = −.564, p < .001). ADL disability was negatively associated with PWB (β = −.457, p < .01). Living with children (no spouse) × ADL disability had a positive interaction effect (β = .278, p < .05). These findings suggest that living alone and living with children (no spouse) are not conducive to PWB among older adults in rural China, and ADL disability reduces the negative effects of living with children (no spouse) on PWB. Our findings provide a crucial direction for PWB interventions aimed at older adults in rural China.

Background: Integrated care for patients living with HIV in rural and economically disadvantaged regions remains a critical challenge. This study aims to develop a 3D evaluation index system to assess the effectiveness of integrated medical care and prevention for patients living with HIV in county-level regions of China. The framework encompasses disease prevention, clinical treatment, and the integration of medical services with public health interventions. Objective: This study aims to develop a scientifically validated, 3D evaluation index system for assessing the effectiveness of integrated medical care and prevention for patients living with HIV in county-level regions of China. Methods: A structured Delphi method will be used to establish consensus among experts on key evaluation indicators. The study will begin with focus group discussions guided by the Valentijn Rainbow Model, followed by multiple rounds of Delphi surveys to refine the indicator system. A total of 75 experts will include clinical professionals, public health specialists, health policy researchers, and representatives from nongovernmental organizations. Experts will participate in iterative rounds, ranking and validating indicators using statistical methods such as the analytic hierarchy process and Kendall W coefficient to assess consensus. Results: The Delphi study was initiated in April 2025, with 3 rounds of questionnaires distributed between May and September 2025. Data analysis is ongoing, and the finalized county-level HIV integration index system is expected to be published by spring 2026. The study is projected to be completed by January 2026. Conclusions: The developed evaluation system will offer a comprehensive and scientifically validated tool for assessing health care outcomes in resource-constrained settings, contributing to policy improvements for HIV service integration in China and other low-resource settings globally.

Background Primary care in China lacks structured mental-health assessment, and the machine-learning models that could support such screening are typically developed on heavily selected samples. Cumulative inclusion and exclusion criteria, though usually treated as neutral data-cleaning steps, can create heterogeneity in predictive reliability among retained participants. Using the China Health and Retirement Longitudinal Study (CHARLS) 2011 baseline, we quantified how selection funnels distort epidemiological associations and inflate machine-learning metrics, and tested selective prediction as mitigation. Methods Using the CHARLS 2011 baseline with temporal external validation in CHARLS-2018, we built a four-level selection funnel (L0-L3), evaluated five classifiers with nested cross-validation and SMOTE, and compared model-embedded uncertainty with a decoupled predictor-selector framework; XGBoost cross-validation residuals drove risk stratification and classification and regression tree (CART) rules. Results Sample sizes fell from L0 n=17,705 to L3 n=4,256 (24.0%). The cancer-depression odds ratio attenuated from 1.78 (95% CI 1.32-2.41) to 1.39 (0.74-2.63), losing significance. AUC rose with selection but not after multiple-comparison correction, whereas calibration error increased for four of five models. Model-embedded uncertainty succeeded only for XGBoost; with the decoupled XGBoost residual selector, all five models achieved selective prediction at approximately 20% coverage (test AUC 0.90, 95% CI 0.85-0.95), abstaining on approximately 80% of cases for individual safety. Risk stratification was stable (residual Spearman correlations >0.95; multi-seed Jaccard 0.88), and CART rules used self-rated health, education, pain, and marital status. Conclusions The findings support a deployable primary-care triage pathway: a four-variable rule identifies patients suitable for algorithm-assisted scoring (approximately 20% coverage) and routes the remainder to human evaluation. Methodologically, cumulative selection bias produces a dual distortion: epidemiological associations are compressed and machine-learning metrics inflated. Selective prediction is limited mainly by uncertainty-indicator design. Performance metrics should be reported with selection level, coverage, and calibration trajectory. Decoupled selective prediction with CART rule extraction provides an actionable framework for quality-controlled, tiered-care deployment. Keywords: selective prediction, selection bias, CHARLS, depression, predictor-selector decoupling, uncertainty quantification, classification and regression tree, triage, clinical decision support, health management.

Insulin resistance, indexed by the estimated glucose disposal rate (eGDR), may evolve over time, yet its longitudinal patterns and relevance for cardiovascular prevention remain unclear. Using data from a nationwide cohort of adults without diabetes, we derived eGDR from routinely collected variables and classified participants into data-driven trajectories across two follow-up visits. We then applied longitudinal targeted maximum likelihood estimation (LTMLE) with Super Learner to estimate the association between eGDR trajectories and incident cardiovascular disease (CVD) within a causal-inference framework, while accounting for time-varying confounding. Compared with the low-stable trajectory, participants in the high-stable and increasing trajectories had significantly lower risks of incident CVD. Restricted cubic splines confirmed an inverse, nonlinear relationship between eGDR and CVD. Benchmarking against the China-PAR model showed that adding eGDR trajectories significantly improved discrimination (AUC 0.554 → 0.609), reclassification (IDI 0.0153; continuous NRI 0.364, both P < 0.001), and net clinical benefit across threshold probabilities of 10–25% on decision curve analysis, whereas TyG trajectories conferred no meaningful incremental value. Findings were consistent in sensitivity and subgroup analyses and were replicated in an independent UK cohort. These results suggest that longitudinal monitoring of eGDR, derived from routine clinical measurements, may help identify adults at elevated cardiometabolic risk before diabetes onset and support early prevention strategies.