Psychometric evaluation of the Diabetes Self-Management Instrument Short Form (DSMI-20)

Accurately assessing the self-management of patients with diabetes requires a valid and reliable assessment instrument. The purpose of this study was to revise and simplify the original assessment instrument (DSMI-35), and to evaluate the psychometric properties of the new short form (DSMI-20) in order to enhance its practicality in clinical settings.

The short form instrument (DSMI-20) was created by revising our previous longer instrument (DSMI-35) through having in-depth small group discussions and validation by experts. The draft (DSMI-20) was pilot tested, and its items were analyzed. We verified the draft (DSMI-20) by conducting an exploratory factor analysis (EFA) and correlating the (DSMI-20) with the diabetes empowerment scale (DES). We tested its reliability including internal consistency and established its test-retest reliability using Cronbach's alpha and Pearson correlation coefficients.

EFA identified four factors with loadings ranging from -0.727 to -0.535 and 0.547 to 0.907: communication with HCPs, self-integration, self-monitoring of blood glucose, and problem solving. These factors accounted for 57.110% of the total variance. The Cronbach's alpha coefficient of the DSMI-20 total scale was 0.925 and of subscales ranged from 0.838 to 0.892. The test-retest correlation for the DSMI-20 total scale was acceptable (r=0.790, p<0.001). Correlation with the DES was r=0.552 (p<0.001).

The DSMI-20 is a rapid, viable assessment tool for identifying self-management problems among patients with diabetes.