Development of Health Literacy Assessment Tool for Thai Stroke Risk Group (HLSR)

Background: Stroke remains the leading cause of mortality in Thailand. Strengthening health literacy is essential for effective stroke prevention; however, existing assessment tools lack stroke-specific content for at-risk populations and often have incomplete validation. Objective: To develop and validate a health literacy assessment instrument for Thai adults at risk of stroke. Methods: Guided by Sørensen and Van den Broucke framework, this instrument development study followed five phases: literature review, qualitative inquiry through in-depth interviews and focus group discussions, expert assessment of content and face validity, and psychometric evaluation. A total of 300 adults at risk of stroke in Sukhothai Province were selected through multistage random sampling. Data were collected using an online questionnaire administered on-site at healthcare facilities with research assistant support and analyzed using descriptive statistics, exploratory and confirmatory factor analysis, and classical test theory. Results: Following Exploratory Factor Analysis (EFA) and first-order Confirmatory Factor Analysis (CFA) with iterative model refinement and content-based review, 13 items were removed to improve parsimony and reduce content overlap. The final Health Literacy for Stroke Risk (HLSR) instrument consists of 34 items across four dimensions: Access (5 items), Understand (12 items), Appraise (10 items), and Apply (7 items), with a maximum possible score of 170. EFA produced a four-factor structure explaining 83.10% of the total variance (KMO = 0.957; Bartlett p 2/df = 2.86; second-order CFA: χ2/df = 3.40; RMSEA = 0.04; CFI = TLI = 0.99; SRMR = 0.03), with first-order CFA standardized factor loadings ranging from 0.605 to 0.764, and second-order factor loadings of the four dimensions on the higher-order construct ranging from 0.520 to 0.966. Internal consistency was good (Cronbach's α = 0.922 overall; subscale α = 0.879–0.921). All corrected items–total correlations exceeded acceptable thresholds, supporting internal consistency. Conclusion: The HLSR demonstrates good construct validity and reliability for assessing health literacy among Thai adults at risk of stroke. As the current psychometric evaluation was conducted in a single regional sample, future cross-validation in independent, geographically diverse samples, together with evaluation of test–retest reliability, convergent, discriminant, and predictive validity, is planned. This instrument offers a practical tool for informing evidence-based interventions and public health strategies aimed at stroke prevention.