OBJECTIVES: The study aims to develop and validate scales (direct and indirect) based on a modified Theory of Planned Behavior (TPB) to measure factors associated with the provision of PCare for HDS users by Thai CPs.
METHOD: Item generation for the scales was based on the theoretical constructs of the modified TPB framework, literature review, and authors' previous qualitative study. Draft items were then subjected to content validity and face validity. Psychometric testing was carried out among CPs in Bangkok, Thailand. Refinement of the scales utilized factor analysis and validity was assessed using factor analysis and Rasch analysis. Internal consistency reliability and construct reliability were used to assess the scales' reliability.
RESULTS: Initially, the direct and indirect scales contained 15 and 28 items, respectively and were reduced to 12 and 16 items, after experts' review. Factor analysis further reduced the number of items of the indirect scale to 13. For both scales, confirmatory factor analysis showed model-data fit. Each construct of the direct scale was significant predictors of intention. Moreover, each construct of the direct scale correlated positively and significantly with the respective construct of the indirect scale, signifying concurrent validity. No misfit item was identified in the Rasch analysis and the majority of items were invariant across gender. Internal consistency reliability and construct reliability of the scales were acceptable.
CONCLUSION: This study presents the development and validation of theoretically-grounded scales to measure the factors associated with the provision of PCare for HDS users by Thai CPs.
METHODS: This was a cross-sectional validation study among 159 T2DM patients attending a public primary care clinic in Selangor. The DMOQ English version underwent adaptation, translation, face validation and field testing to produce the Malay version. Psychometric analysis was performed using Exploratory Factor Analysis, internal consistency and testretest reliability.
RESULTS: The DMOQ domains were conceptually equivalent between English and Malay language. A total of 13 items and two domains were removed during the validation process (three items during the content validation, three items due to poor factor loadings, five items as they loaded onto two domains which were not interpretable, one item as it did not fit conceptually into the factor it loaded onto and one openended question as it did not fit into the retained domains). Therefore, the final DMOQ Malay version consisted of 21- items within five domains. The Cronbach alpha was 0.714 and the intraclass-correlation coefficient was 0.868.
CONCLUSION: The DMOQ Malay version is a valid and reliable tool which is consistent over time. It can be used to examine the perception of T2DM patients towards the risk of their offspring developing diabetes and possibility of intervention in Malay-speaking patients.