Evidências de validade baseada na estrutura interna do questionário de segurança de medicação fundamentado na teoria do comportamento planejado (TPB-MSQ-B)

Abstract

Medication safety has gained prominence in the context of patient safety, driven by actions aimed at improving the quality of care at all stages of the process, from prescription to administration, with the goal of preventing errors and avoiding harm to patients. This study aims to evaluate evidence of validity based on the internal structure and reliability of the Medication Safety Questionnaire, grounded in the Theory of Planned Behavior (TPB-MSQ-B), for use in Brazil. This is a methodological study, with data collection conducted from November 2023 to October 2024 in two phases: the first application included 743 students, and the second included 401 students from nursing, pharmacy, and medical programs. Participants were from public universities in Ceará, including the State University of Ceará (UECE), Federal University of Ceará (UFC), University for International Integration of the Afro-Brazilian Lusophony (UNILAB), Regional University of Cariri (URCA), and State University of Vale do Acaraú (UVA). Psychometric validation of the questionnaire utilized confirmatory factor analysis, Cronbach’s alpha, McDonald’s omega, the graded response model of Item Response Theory (IRT). Analyses were performed using the open-source software R and Jasper. The study was approved by the Research Ethics Committee of the State University of Ceará (approval number 6.902.733). In the first application, the confirmatory factor analysis of the initial 48 items in the questionnaires revealed that the factorial structure was composed of 29 items for nursing students, 28 for pharmacy students, and 23 for medical students. The reliability of the factors attitude, behavioral control, and subjective norms ranged from 0.324 to 0.635 for nursing students; from 0.220 to 0.560 for pharmacy students; and from 0.492 to 0.634 for medical students. The IRT analysis indicated satisfactory parameters for the evaluated factors and suggested reducing the Likert scale to five points. After adjustments to the questionnaire, the second application demonstrated good internal consistency. The factor loadings of the items ranged from 0.201 to 0.969, and reliability values ranged from 0.686 to 0.879. The final version of the questionnaires included 26 items for nursing students, 28 items for pharmacy students, and 22 items for medical students. Thus, the study provides questionnaires with evidence of validity based on internal structure and reliability, in a reduced factorial structure, applicable to nursing, pharmacy, and medical students in the Brazilian context.