Chris Schilling1, Michelle M Dowsey2, Philip M Clarke3, Peter F Choong2. 1. Centre for Health Policy, School of Population and Global Health, the University of Melbourne, Carlton, Victoria, Australia; The University of Melbourne Department of Surgery, St. Vincent's Hospital, Fitzroy, Victoria, Australia. Electronic address: chris.schilling@unimelb.edu.au. 2. The University of Melbourne Department of Surgery, St. Vincent's Hospital, Fitzroy, Victoria, Australia; Department of Orthopaedics, St. Vincent's Hospital Melbourne, Melbourne, Victoria, Australia. 3. Centre for Health Policy, School of Population and Global Health, the University of Melbourne, Carlton, Victoria, Australia.
Abstract
BACKGROUND: Patient-reported outcome measures (PROMs) are becoming increasingly popular in orthopedic surgery. Preoperative and postoperative follow-up often elicit PROMs in the form of generic quality-of-life instruments (e.g., Short Form health survey SF-12 [SF-12]) that can be used in economic evaluation to estimate quality-adjusted life-years (QALYs). However, the timing of postoperative measurement is still under debate. OBJECTIVES: To explore the timing of postoperative PROMs collection and the implications for bias in QALY estimation for economic evaluation. METHODS: We compared the accuracy of QALY estimation on the basis of utilities derived from the SF-12 at one of 6 weeks, 3 months, 6 months, and 12 months after total knee arthroplasty, under different methods of interpolation between points. Five years of follow-up data were extracted from the St. Vincent's Melbourne Arthroplasty Outcomes (SMART) registry (n = 484). The SMART registry collects follow-up PROMs annually and obtained more frequent outcomes on subset of patients (n = 133). RESULTS: Postoperative PROM collection at 6 weeks, 6 months, or 12 months biased the estimation of QALY gain from total knee arthroplasty by -41% (95% confidence interval [CI] -59% to -22%), 18% (95% CI 4%-32%), and -8% (95% CI -18% to -2%), respectively. This bias was minimized by collecting PROMs at 3 months postoperatively (6% error; 95% CI -9% to 21%). CONCLUSIONS: The timing of PROM collection and the interpolation assumptions between measurements can bias economic evaluation. In the case of total knee arthroplasty, we recommend a postoperative measurement at 3 months with linear interpolation between preoperative and postoperative measures. The design of economic evaluations should consider timing and interpolation issues.
BACKGROUND:Patient-reported outcome measures (PROMs) are becoming increasingly popular in orthopedic surgery. Preoperative and postoperative follow-up often elicit PROMs in the form of generic quality-of-life instruments (e.g., Short Form health survey SF-12 [SF-12]) that can be used in economic evaluation to estimate quality-adjusted life-years (QALYs). However, the timing of postoperative measurement is still under debate. OBJECTIVES: To explore the timing of postoperative PROMs collection and the implications for bias in QALY estimation for economic evaluation. METHODS: We compared the accuracy of QALY estimation on the basis of utilities derived from the SF-12 at one of 6 weeks, 3 months, 6 months, and 12 months after total knee arthroplasty, under different methods of interpolation between points. Five years of follow-up data were extracted from the St. Vincent's Melbourne Arthroplasty Outcomes (SMART) registry (n = 484). The SMART registry collects follow-up PROMs annually and obtained more frequent outcomes on subset of patients (n = 133). RESULTS: Postoperative PROM collection at 6 weeks, 6 months, or 12 months biased the estimation of QALY gain from total knee arthroplasty by -41% (95% confidence interval [CI] -59% to -22%), 18% (95% CI 4%-32%), and -8% (95% CI -18% to -2%), respectively. This bias was minimized by collecting PROMs at 3 months postoperatively (6% error; 95% CI -9% to 21%). CONCLUSIONS: The timing of PROM collection and the interpolation assumptions between measurements can bias economic evaluation. In the case of total knee arthroplasty, we recommend a postoperative measurement at 3 months with linear interpolation between preoperative and postoperative measures. The design of economic evaluations should consider timing and interpolation issues.
Authors: Edward Burn; Christopher J Edwards; David W Murray; Alan Silman; Cyrus Cooper; Nigel K Arden; Rafael Pinedo-Villanueva; Daniel Prieto-Alhambra Journal: Rheumatology (Oxford) Date: 2019-06-01 Impact factor: 7.580