Psychometric properties of the Diabetes Self-Management Questionnaire (DSMQ) in Urdu

Abstract

Background: Numerous study tools on diabetes self-care have been introduced; however, most existing tools do not show expectable and meaningful correlations with patients' glycaemic control. The Diabetes Self-Management Questionnaire (DSMQ) was designed to appraise self-care activities which can predict glycaemic control outcomes. However, this tool has not been validated in Pakistan. Therefore, the aim of this study was to translate and examine the psychometric properties of the Urdu version of DSMQ among type 2 diabetes patients. Method: Standard forward-backward translation was used to translate the DSMQ into Urdu language. A convenience sample of 130 patients with type 2 diabetes was collected to assess the Urdu version's psychometric properties. Reliability was assessed by Cronbach's coefficient α and validity was assessed using confirmatory factor analysis and criterion-related correlations. Results: High internal consistency was found for all DSMQ scales (Sum scale: α = 0.96, Glucose Management: 0.91; Dietary Control: 0.88; Physical Activity: 0.89; Health-Care Use: 0.73). The DSMQ subscales showed significant correlations with HbA1c (Glucose Management: -0.75; Dietary Control: -0.76; Physical Activity: -0.71; Health-Care Use: -0.64; Sum Scale: -0.78; all p < 0.001). However, when associations with HbA1c were assessed in one multiple linear regression model, only Glucose Management and Dietary Control were significantly associated with lower HbA1c values (Beta = -0.42, p = 0.004 and Beta = -0.30, p = 0.028, respectively), while Physical Activity and Health-Care Use were not (p > 0.05). Adequate fit to the data was achieved for single factor model after successively modelling all significant correlations between the items' error terms, with Chi2 = 106.6, df = 84, p = 0.049; TLI = 0.98, CFI = 0.99 and RMSEA = 0.05 (90% CI 0.01-0.07). Whereas a comparatively lower fit indices to data were observed in case of four factor model. Conclusion: The findings support the Urdu version of the DSMQ as a reliable and valid instrument for assessing self-care activities associated with glycaemic control in type 2 diabetes patients.