Adaptation into Basque language and validation of the Barthel Index, EQ-5D-5L and PHQ-9 questionnaires in adult patients

Patient Reported Outcome Measures (PROMs) tools are limited resources in Basque, co-official language in Basque Autonomous Community, Spain. The aim of this study was to translate and adapt Basque Barthel Index (BI) and Patient Health Questionnaire-9 (PHQ-9) and to validate them along with Basque EQ-5D-5L. A cross-sectional observational study was carried out to assess the psychometric properties of Basque questionnaires. We recruited Basque speaking adults suffering from diabetes mellitus, heart failure, respiratory or musculoskeletal diseases and a convenience sample (overall n = 347) via post mail, all of them covered by the Galdakao-Usansolo University Hospital. Participants were classified depending on the number of comorbidities and Charlson Comorbidity Index (CCI). BI and PHQ-9 were translated to Basque and back-translated to Spanish and English, respectively. Basque EQ-5D-5L was already available. The comprehensibility and feasibility of instruments were evaluated by cognitive debriefing interviews. Reliability (Cronbach’s alpha), factor analysis, convergent validity and known-groups validity were assessed. Cronbach alpha coefficient was 0.93 for BI, 0.85 for EQ-5D-5L and 0.89 for PHQ-9. Each tool fit into its own single dimension. Three questionnaires correlated independently (each one’s ρ > 0.4, p < .0001) with the physical component (PCS) of Short Form-36 Health Survey (SF-36) and the PHQ-9 with the mental component (MCS) of SF-36, confirming the convergent validity. All the instruments detected differences depending on the number of comorbidities or CCI, demonstrating the known-groups validity. Basque BI, EQ-5D-5L and PHQ-9 are reliable and valid tools that satisfy the psychometric quality requirements for use in clinical practice and research studies.