Measurement Properties of the Brazilian Version of the High-Activity Arthroplasty Score: An Item Response Theory Approach

OBJECTIVE: To evaluate the structural validity of the Brazilian version of High Activity Arthroplasty Score (HAAS-Brazil) using item response theory in patients after hip arthroplasty, in an outpatient setting. DESIGN: Cross-sectional clinimetric study. METHODS: Inclusion criteria were being at least 6 months post–hip arthroplasty, having an American Society of Anesthesiologists (ASA) classification of ≤ 2, undergoing elective surgery, and demonstrating fluent Brazilian Portuguese reading ability. Data were collected by self-administration. Exploratory factor analysis (EFA), confirmatory factor analysis (CFA), and Mokken scale analysis (MSA) were conducted. Samejima’s Graded Response Model (SGRM) was fitted to the data. Type I error α = .05 was considered for all tests (95% confidence interval). RESULTS: A sample of 112 patients, 51% female, with a mean age of 55.9 years, completed the HAAS-Brazil. EFA indicated a 1-factor model, corroborated by CFA and MSA (H = 0.42, Hi > 0.30, Hij > 0.30). Local item independence (no flagged W values) and monotonicity (V mon = 0) were confirmed by MSA. SGRM showed a good fit to the data. The instrument demonstrated an adequate precision index (F = 0.78) and reliability (C2 (2) = 1.16, P = .56). Items exhibited good overall discriminatory properties and covered different levels of physical function. CONCLUSION: The HAAS-Brazil had robust structural validity evidence for patients after hip arthroplasty in the Brazilian context. JOSPT Open 2025;3(3):296-301. Epub 4 March 2025. doi:10.2519/josptopen.2025.0084