Validez y confiabilidad de un instrumento para evaluar la satisfacción del paciente en ortodoncia

Abstract

Objective: to assess the validity and reliability of the Colombian version of the instrument on satisfaction in orthodontic treatment proposed by Bos and collaborators in 2005. Methods: a diagnostic technology evaluation study was carried out using a questionnaire on satisfaction in the orthodontic treatment proposed by Bos and collaborators in 2005, and translated and adapted transculturally by Moreno and Rodríguez in 2015. As variables, the 58 questions of the instrument, the sociodemographic characteristics of the participants and some aspects related to the treatment were analyzed. of orthodontics. Reliability was assessed by analyzing the internal consistency with Cronbach's alpha (a value greater than 0.7 was considered satisfactory). The validation of the questionnaire was carried out using the Rasch analysis based on the guidelines proposed by Messick in 1995 and interpreted by Wolfe and Smith. Statistical software Stata I/C version 12 and Winsteps were used for the analysis. All participants signed an informed consent. Results: the sample consisted of 163 (57.4%) women, the average age was 27.9 ± 9.4 years [CI 95% 26.4 - 29.3], 121 (42.6%) men they presented an average age of 28.1 ± 8.5 years [95% CI 26.5-29.6]. An internal consistency of 0.9339 was obtained for the whole questionnaire, which was considered excellent. It was necessary to group categories 2 and 4 according to the values of reliability and separation of persons and items; not all the items were adjusted to the Rash model, so 15 items had to be withdrawn. The questions they adjusted had MNSQ values of infit between 0.70 and 1.17 and MNSQ of outfit between 0.69 and 1.40. A Differential Item Function was observed for questions 34, 35 and 55, so it was necessary to remove them. In relation to dimensionality, the measurement of the forty items explained 36.4% of the variance. Conclusions: it was necessary to reduce the Colombian version of the satisfaction questionnaire to forty items and reduce the number of categories from five to three. The measurement of internal consistency using the Cronbach alpha coefficient was excellent. Differential Function was presented in relation to the place where the participant was attended.