Paulsen, Aksel2; Roos, Ewa M.3; Pedersen, Alma B5; Overgaard, Søren4
1 Department of Clinical Medicine - Department of Clinical Epidemiology, Department of Clinical Medicine, Health, Aarhus University2 Ortopædkirurgi3 Institut for Idræt og Biomekanik4 The Department of Orthopaedics E, ?AS, Faculty of Health Sciences, Aarhus University, Aarhus University5 Department of Clinical Medicine - Department of Clinical Epidemiology, Department of Clinical Medicine, Health, Aarhus University
Background and purpose - The increased use of patient-reported outcomes (PROs) in orthopedics requires data on estimated minimal clinically important improvements (MCIIs) and patient-acceptable symptom states (PASSs). We wanted to find cut-points corresponding to minimal clinically important PRO change score and the acceptable postoperative PRO score, by estimating MCII and PASS 1 year after total hip arthroplasty (THA) for the Hip Dysfunction and Osteoarthritis Outcome Score (HOOS) and the EQ-5D. Patients and methods - THA patients from 16 different departments received 2 PROs and additional questions preoperatively and 1 year postoperatively. The PROs included were the HOOS subscales pain (HOOS Pain), physical function short form (HOOS-PS), and hip-related quality of life (HOOS QoL), and the EQ-5D. MCII and PASS were estimated using multiple anchor-based approaches. Results - Of 1,837 patients available, 1,335 answered the preoperative PROs, and 1,288 of them answered the 1-year follow-up. The MCIIs and PASSs were estimated to be: 24 and 91 (HOOS Pain), 23 and 88 (HOOS-PS), 17 and 83 (HOOS QoL), 0.31 and 0.92 (EQ-5D Index), and 23 and 85 (EQ-VAS), respectively. MCIIs corresponded to a 38-55% improvement from mean baseline PRO score and PASSs corresponded to absolute follow-up scores of 57-91% of the maximum score in THA patients 1 year after surgery. Interpretation - This study improves the interpretability of PRO scores. The different estimation approaches presented may serve as a guide for future MCII and PASS estimations in other contexts. The cutoff points may serve as reference values in registry settings.