Korean translation, cultural adaptation and psychometric validation of the revised patients’ attitudes towards deprescribing questionnaire for caregivers

OBJECTIVES: To translate, cross-culturally adapt and validate the Korean version of the revised Patients' Attitudes Towards Deprescribing (rPATD) questionnaire for caregivers. DESIGN: Cross-sectional validation study. SETTING: Long-term care facilities (LTCFs) in South Korea. PARTICIPANTS: 1000 caregivers of older adult residents in LTCFs. PRIMARY AND SECONDARY OUTCOME MEASURES: Internal consistency was assessed using Cronbach's alpha and corrected item-total correlations. An alpha-if-item-deleted analysis was conducted to evaluate individual item contributions. Construct validity was evaluated using confirmatory factor analysis (CFA). RESULTS: The overall Cronbach's alpha was 0.881, indicating high internal consistency. Factor-specific reliability was acceptable for the burden (0.781) and appropriateness (0.766) factors but relatively lower for the involvement (0.637) and concerns about stopping (0.563) factors. Despite lower reliability in some subscales and a borderline corrected item-total correlation for question 17 (0.244), all items were retained based on their conceptual importance, statistically significant CFA loading (β=0.283, p<0.001) and consistency with international validation studies. CFA of the four-factor structure yielded marginal but partially acceptable fit indices (Comparative Fit Index=0.830, Tucker-Lewis Index=0.795, Root Mean Square Error of Approximation=0.094 and Standardised Root Mean Square Residual=0.074), with all factor loadings statistically significant (p<0.001), providing partial support for the construct validity of the four-factor structure. CONCLUSIONS: The Korean version of the rPATD questionnaire for caregivers demonstrated good overall internal consistency and acceptable structural validity, though performance varied across subscales. Specifically, the concerns about stopping and involvement subscales showed relatively lower reliability, and the CFA indicated marginal model fit, suggesting that further psychometric refinement may be warranted. Nevertheless, this tool may be valuable for identifying specific barriers and facilitators to deprescribing in this critical population, thereby contributing to medication optimisation and shared decision-making in Korean LTCF settings.