Development and validation of a tool to assess hypertension and diabetes management in older adults within the National Essential Public Health Services Program (NEPHSP) in China

Abstract

Abstract Purpose This study aimed to develop and refine an assessment tool for assessing chronic disease management outcomes in older adults based on the National Essential Public Health Services Program (NEPHSP) in China, and to validate its reliability and validity. Methods The development of the assessment tool involved generating 40 potential items through a comprehensive review and logical model analysis of the NEPHSP. Data were collected through a survey conducted as part of the 10-year evaluation project of the NEPHSP in China, which took place from November to December 2019. The assessment tool was evaluated using Cronbach's alpha coefficients and confirmatory factor analysis (CFA) to select and scale potential items, assess internal consistency, establish construct validity, evaluate convergent validity, and confirm discriminant validity. Results The assessment tool was structured around a second-order factor model comprising three domains: essential health services for all older adults, follow-up service for older adults with hypertension or diabetes, and self-assessment of patient experience satisfaction. After a two-step process, nine factors and twenty-nine items were selected, including elements such as 'health records and health education,' 'blood pressure/glucose monitoring,' and 'health coaching in follow-up,' among others. The Cronbach's alpha coefficients were excellent, measuring at 0.899 and 0.906. The 29-item instruments demonstrated strong fits in both hypertension and diabetes models. Fit indices for the hypertension model included Bollen-Stine bootstrap χ²/df = 1.78, RMSEA = 0.03, and GFI = 0.97. For the diabetes model, the indices were Bollen-Stine bootstrap χ²/df = 1.38, RMSEA = 0.02, and GFI = 0.97. Confirmatory factor analysis (CFA) revealed factor loadings ranging from 0.516 to 0.940 for the hypertension model and from 0.504 to 0.943 for the diabetes model. All three first-order factors were significantly correlated with each other (p < 0.01), and their correlation coefficients were lower than the square root of AVE. Models had favorable structural validity, convergent validity, and discriminant validity. Conclusions A valid and reliable assessment tool for evaluating hypertension and diabetes management in the older adults was successfully developed in China. This study provides robust evidence for the internal consistency reliability and structural validity of the tool. Furthermore, it serves as a foundational step for future tool refinement and holds promise for broader applications in essential public health and community healthcare program evaluations. However, further validation in these contexts is warranted.