Abstract
OBJECTIVES: Measurement of treatment satisfaction in diabetes is important as it has been shown to be associated with positive outcomes, reduced disease cost and better health. The aim of this study was to assess the construct validity and internal consistency reliability of the Greek version of the Diabetes Treatment Satisfaction Questionnaire (DTSQ). METHODS: A sample of type II diabetes patients (N = 172) completed the DTSQ status version, the SF-36 health survey and also provided data regarding treatment method, clinical and socio-demographic status. Instrument structure, reliability (Cronbach's a) and construct validity (convergent, discriminative, concurrent and known-groups) were assessed. RESULTS: The DTSQ measurement properties were confirmed in the Greek version with confirmatory factor analysis (CFA). Scale reliability was high (Cronbach's a = 0.92). Item-scale internal consistency and discriminant validity were also good, exceeding the designated success criteria. Significant correlations were observed between DTSQ items/overall score and SF-36 scales/component scores, which were hypothesized to measure similar dimensions. Known groups' comparisons yielded consistent support of the construct validity of the instrument. CONCLUSIONS: The instrument was well-accepted by the patients and its psychometric properties were similar to those reported in validation studies of other language versions. Further research, incorporating a longitudinal study design, is required for examining test-retest reliability and responsiveness of the instrument, which were not addressed in this study. Overall, the present results confirm that the DTSQ status version is a reasonable choice for measuring diabetes treatment satisfaction in Greece.