The Diabetes Treatment Satisfaction Questionnaire
A cross-cultural South African perspective
- 1South Africa Medical Research Council, Health and Development Research Group, Pretoria, Gauteng, South Africa
- 2School of Health Systems and Public Health, University of Pretoria, Pretoria, Gauteng, South Africa
- 3Faculty of Community and Health Sciences, University of the Western Cape, Cape Town, Western Cape, South Africa
- Address correspondence to Prof. Margaret Sandra Westaway, SA Medical Research Council, Health and Development Department, Private Bag X385, Pretoria, Gauteng, 0001 South Africa. E-mail:
Reliable and valid multicultural instruments are important in multicultural societies that are typical of modern cities, and clinicians, using psychosocial assessments, need to ensure that their diagnostic and screening tools are appropriate. This study was conducted with 176 diabetic outpatients from two culturally distinct groups (95 Bantu-speaking and 81 Afrikaans-speaking subjects) to 1) ascertain the underlying dimensions of treatment satisfaction as measured by the Diabetes Treatment Satisfaction Questionnaire (DTSQ status) (1), 2) determine the reliability (internal consistency) of the measures, and 3) investigate the effects of objective (HbA1c results) and subjective metabolic control, health (2), and well-being (3) on satisfaction with diabetes treatment.
Principal components analysis was conducted on the 8-item DTSQ (1). All communality estimates exceeded the criterion of 0.30 (4) for both Bantu-speaking and Afrikaans-speaking patients (range 0.62–0.79 and 0.55–0.76, respectively). Two factors explained 71% of the variance for Bantu-speaking patients and 68% of the variance for Afrikaans-speaking patients. The first factor consisted of the six treatment satisfaction items, and the second factor consisted of the two subjective metabolic control items. Reliability (internal consistency) coefficients were excellent (5) and very similar for both groups (>0.80 on all measures).
Treatment satisfaction was associated with fewer incidents of hyperglycemia (r = −0.58, P < 0.01) and hypoglycemia (r = −0.32, P < 0.01), higher general well-being (r = 0.56, P < 0.01), and better health (r = 0.44, P < 0.01) for Bantu-speaking patients. For Afrikaans-speaking patients, greater treatment satisfaction was associated with fewer incidents of hyperglycemia (r = −0.29, P < 0.01), higher general well-being (r = 0.54, P < 0.01), and better health (r = 0.50, P < 0.01). Language, sex, age, and employment status were not related to treatment satisfaction or general well-being (P > 0.05), confirming the construct validity of the measures. HbA1c results were not significantly related to treatment satisfaction, subjective metabolic control, general well-being, or general health for either group (P > 0.05).
For Bantu-speaking patients, fewer incidents of hyperglycemia significantly predicted 33% of the variance (P < 0.001) in treatment satisfaction; an additional 11% of the variance (P < 0.001) was explained by general well-being. For Afrikaans-speaking patients, general well-being predicted 29% of the variance (P < 0.001) in treatment satisfaction; an additional 7% of the variance (P = 0.001) was explained by general health.
In conclusion, the study demonstrated that the underlying dimensions of the DTSQ for both groups were treatment satisfaction and hyper- and hypoglycemia, all measures had excellent reliability (5), and well-being is an important predictor of treatment satisfaction for both groups of patients. These findings were consistent with those reported in the U.K. and Sweden (6–7) and support the idea that the DTSQ can be used in multicultural settings.
We thank Novo Nordisk (South Africa) for funding the interviewers’ salaries. We also thank Professor Clare Bradley for permission to use her measures and Rosalind Plowright for constructive comments on the application of the DTSQ in multicultural settings.