Design and validation of the Disaster Health Literacy Questionnaire for diabetes patients in Iran: a mixed-methods study

Abstract

Objectives To develop and psychometrically evaluate a multidimensional Disaster Health Literacy Questionnaire (DHLQ) for diabetic patients in Iran, using advanced item response theory approaches. The questionnaire was designed in the Persian (Farsi) language. Design A sequential mixed-methods study incorporating qualitative (scoping review and interviews) and quantitative (psychometric validation) phases. Setting Diabetes clinics and healthcare centres across Iran (2022-2023). Participants The study enrolled 570 patients with diabetes (56 female, mean age 45.57 +/- 16.33 years) for quantitative validation; 15 experts and 15 patients for qualitative validation. Outcome measures The psychometric properties evaluated included content validity (using content validity ratio (CVR) and content validity index (CVI)), construct validity (assessed via multidimensional item response theory (MIRT)), and reliability (measured by Cronbach's alpha and test-retest Kappa). Additionally, item parameters (multidimensional difficulty (MDIFF) and multidimensional discrimination (MDISC)) and model fit indices (RMSEA, CFI and TLI) were examined. Results The final 30-item DHLQ demonstrated excellent content validity (scale-level CVI=1; item-level CVI>0.79; CVR>0.49). Cronbach's alpha for the total scale was 0.606; test-retest reliability showed significant agreement (Kappa=0.35-1, p<0.05). MIRT confirmed a three-factor structure: Disaster Perception Risk (14 items), Medication-Nutritional Literacy (11 items), and Self-help and Emergency Literacy (five items). Model fit was excellent (RMSEA=0.016, CFI=0.96, TLI=0.95). Item analysis revealed that 73 of items had moderate-to-high discrimination (MDISC >= 0.65), and 83 had medium-to-low difficulty (MDIFF <0.5). Conclusion The DHLQ is a rigorously validated tool for assessing disaster health literacy in diabetic populations. Its multidimensional structure and strong psychometric properties support its use in clinical and emergency planning contexts to identify literacy gaps and tailor interventions.