Adaptação transcultural e análise das propriedades psicométricas da versão brasileira da escala de catastrofização da dor para idosos com dor lombar

Abstract

Low back pain is common in the elderly population. Studies have showed pain catastrophizing as an important factor in the etiology and chronicity of this condition in young adults. Instruments for measuring specific pain catastrophizing for seniors are needed to understand the impact of this factor in low back pain in this population. The aim of the study was to conduct cross-cultural adaptation of the Pain Catastrophizing Scale (PCS) into Portuguese of Brazil, to evaluate the intra and interrater reliability examiner and assess the construct validity of the instrument for individuals with low back pain. PCS was translated and culturally adapted following a standardized protocol, we assessed the intra and inter-rater (50 seniors) by Quadratic Weighted Kappa Coefficient (Kp) and Intraclass Correlation Coefficient (ICC). Psychometric properties were investigated using Rasch analysis in 131 elderly. The mean age of the 131 patients was 66.6 ± 7.1 years, and 92.3% were women, 40.7% were married and 32.3% had incomplete primary education. Theintra-rater reliability was Kp = 0.80 ± 0.01 (95% CI = 0.72 to 0.78) and ICC = 0.88 (0.87 to 0.90) and inter-rater, Kp = 0 , 75 ± 0.01 (95% CI 0.72 to 0.78) and ICC = 0.77 (0.73 to 0.80). Were detected calibration stability coefficients of 0.95 and 0.90 for items and individuals respectively, which was considered adequate. The separation index for the elderly was 2.95, indicating that the items divides people in approximately three levels of catastrophizing, and for items, 4.59, indicating approximately five levels of catastrophizing. According to Rasch analysis, the pain catastrophizing was considered a one-dimensional construct. Of the 13 items, onedid not fit the model (item 11), which was considered acceptable, according to the assumptions of Rasch analysis. This erratic item was justified between individuals evaluated, considering its complex grammatical construction, low level of formal education and recall bias. The pattern of use of the scores indicates that the "nothing" was used in 18% of responses, "mild" in 14%, "moderate" in 24%, "intense" and 17% "always" 27%. Whereas the expectation of the Rasch model is that there is increasing use of ordination scores, there was a slight distortion in the use of category 3 (severe), which seems to be harder to use than the category 2 (moderate) and 4 (always) . The first hypothesis is that individuals found more familiar and easier to answer "always" that "intense" and the second is that the participants were indoubt because the answers "intense" and "always" express different word classes.Looking at the map of the distribution of items and individuals in the same linear continuum, there is no effect ceiling or floor effect, and most people stood at moderate to high levels of catastrophizing. The instrument proved to be clinically useful in the sample. Moreover, its application was considered fast and easy and was well understood by most individuals. Subsequent studies should carry out the sameanalyzes in populations with different characteristics.