Psychometric evaluation of the problem areas in diabetes (PAID) scale among people with type 2 diabetes in Ethiopia: a tool validation study

The confirmatory factor analysis for the Ethiopian Problem Area in Diabetes (Eth-PAID) scales demonstrated good model fit to the one, two, three, and four factor structures. The one-factor model Eth-PAID-20 scale showed acceptable internal consistency (α = 0.82), but the "Diabetes distress" subscale of two, three, and four-factor structures partially satisfied the internal consistency (ranged α = 0.74-77). The Eth-PAID-20 scale correlated negatively with self-care efficacy (r = 0.131, P = 0.06) and positively with Fasting Blood Sugar (FBS) level (r = 0.02, P = 0.86), which implies poor convergent validity. Moreover, "lack of confidence," "food-related problem," and "support-related problem" subclasses of two, three, and four-factor models showed very weak convergent validity. Discriminative validity revealed that female patients (30.16 ± 13.11), t = - 2.73, p = 0.007, d = 0.4) and patients who lived alone (28.05 ± 12.98), t = 2.542, p = 0.021, d = 0.5) had significantly higher distress scores in Eth-PAID-20 as one factor model.