Tradução, adaptação transcultural e validação da versão portuguesa do Banco de Itens Comportamento de Dor do Patient-Reported Outcomes Measurement Information System – PROMIS®

• Ano de defesa: 2018
• Autor(a) principal: Xavier, Douglas Ataniel Alves
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Ciências da Saúde
• Palavras-chave em Português: Dor; Pain; Medidas de resultados relatados pelo paciente; Patient reported outcome measures; Psicometria; Psychometrics; Qualidade de vida; Quality of life; Teoria de resposta ao item; Item response theory; CNPQ::CIENCIAS DA SAUDE::MEDICINA; CNPQ::CIENCIAS DA SAUDE; CNPQ::CIENCIAS DA SAUDE::ENFERMAGEM
• Link de acesso: https://repositorio.ufu.br/handle/123456789/28665; http://dx.doi.org/10.14393/ufu.di.2019.1225

Introduction: Chronic pain is an important manifestation in various health conditions and is a relevant clinical, social and economic problem. It affects different age groups and its economic cost is high for both patients and the health system. It is estimated that the cost for your treatment exceeds that of heart disease or cancer. Objective: This study aims at cross-culturally adapt and validate the universal Portuguese version of the PROMIS Pain Behavior Item Bank. Methods: Universal Portuguese version of the PROMIS Pain Behavior Item Bank was cross-culturally adapted and validate in 370 patients with chronic pain and 350 healthy individuals. Reliability (internal consistency and test-retest reliability) was assessed. Item Response Theory (IRT) model assumptions were evaluated using Confirmatory (CFA) and Exploratory (CFE) Factor Analysis. Items were calibrated using a Samejima’s Graded Response Model (GRM). Differential Item Functioning (DIF) was examined in respect to age, gender, scholarship and language. Results: No major cultural adaptations were needed. Internal consistency (Cronbach’s alpha = 0.966) and test-retest reliability (intraclass correlation coefficient = 0.963) were good. CFA (CFI = 0.97, TLI = 0.97, RMSEA = 0.088) and CFE analysis supported sufficient unidimensionality. The data also fit the GRM and demonstrated good coverage across fatigue construct (threshold parameters range from -1.12 to 3.23 and standard factor loads (λ) from 0.54 to 0.88). No items demonstrated significant DIF. Conclusion: The universal Portuguese version of the PROMIS® Pain Behavior Item Bank provides a reliable, precise, and valid measure as assessed by classical psychometric theory and item response theory (IRT). Stability of the measurement properties of the item bank allows its use to assess pain behavior in clinical research in Portuguese-speaking adults patients. Short forms and computerized adaptive tests may now be developed.