Short version of the Problem Areas in Diabetes scale(PAID-13) in Brazilian patients with diabetes: a structural and criterion validity study.

Abstract

Background: The Problem Areas in Diabetes scale (PAID) is used to measure emotional distress levels related to diabetes mellitus (DM). However, consensus on its internal structure is lacking.

Objectives: To compare the different internal structures of the PAID and propose a shortened version for Brazilian patients with diabetes.

Design and setting: Structural and criterion validity study.

Methods: We included Brazilian patients with type 1 DM (DM1) and type 2 DM (DM2) in this study. In accordance with the international consensus recommendations, we assessed the structural validity using confirmatory factor analysis (CFA) and used the following indices to evaluate model fit: root mean square error of approximation (RMSEA), comparative fit index (CFI), Tucker-Lewis index (TLI), standardized root mean square residual (SRMR), chi-square/degrees of freedom (DF), Akaike information criterion (AIC) and sample-size adjusted Bayesian information criterion (SABIC). Modification indices and factor loadings were used to reduce the number of items.

Results: One hundred eighty-five patients, most of whom included women with DM2, participated in the study. The reduction in the PAID generated a unidimensional structure with 13 items (PAID-13). The PAID-13 presented the best-fit indices (chi-square/DF = 2.15, CFI = 0.989, TLI = 0.986, RMSEA = 0.079, and SRMR = 0.049). When the PAID versions with 13 and 20 items (original version) were correlated, a strong correlation was observed (rho = 0.941, P < 0.001).

Conclusion: The short version of the PAID scale with 13 items presented a more appropriate internal structure for Brazilian patients with diabetes.