Psychometric properties of the general self-efficacy scale among Thais with type 2 diabetes: a multicenter study

Abstract

Background. Type 2 diabetes (T2D) is one of the most common chronic diseases in the world. In recent decades the prevalence of this disease has increased alarmingly in lower to middle income countries, where their resource-limited health care systems have struggled to meet this increased burden. Improving patient self-care by improving diabetes knowledge and diabetes management self-efficacy represents a feasible way of ameliorating the impact of T2D on the patient, and the health care system. Unfortunately, the relationships between self-efficacy, diabetes self-management, and thereafter, patient outcomes, are still far from well understood. Although a domain-specific measure of diabetes management self-efficacy, the Diabetes Management Self-Efficacy Scale (DMSES), has been validated in the Thai T2D population, more general measures of self-efficacy, such as the General Self-Efficacy scale (GSE) have not been validated in this population. In this paper we translate and examine the psychometric properties of the GSE in Thais living with T2D. Methods. In this nation-wide study we examined the psychometric properties of the GSE in 749 Thais diagnosed with T2D within the last five years, and evaluated its relationship with the DMSES along with other patient characteristics. Reliability of GSE was assessed using Cronbach’s alpha, and the construct validity was examined using confirmatory factor analysis, along with GSE’s convergence and discrimination from DMSES. Results. The Thai version of the GSE was shown to have good psychometric properties in Thais living with T2D. Cronbach’s alpha was shown to be 0.87 (95% CI [0.86, 0.88]). We also demonstrated the structural validity of the GSE (Tucker-Lewis Index = 0.994, Cumulative Fit Index = 0.995, Adjusted Goodness of Fit Index = 0.998, Root Mean Square Error of Approximations = 0.025, 95% CI [0.06–0.039]), and that this instrument has a similar structure in Thais as in other populations. GSE was also shown to have some overlap with the DMSES with correlations among GSE and the DMSES domains ranging from 0.18 to 0.26, but also the GSE has substantial discrimination from DMSES (Disattenuated correlation coefficient = 0.283, 95% CI [0.214–0.352], p < 0.001). This suggests that while general and diabetes management self-efficacy are somewhat associated, there are aspects of diabetes management self-efficacy not captured by the more stable general self-efficacy. Conclusions. We demonstrate that the Thai GSE is a reliable and valid measure. We believe the GSE may represent a useful tool to examine the efficacy of proposed and existing diabetes self-management, and management self-efficacy interventions.