What is the minimum response rate on patient-reported outcome measures needed to adequately evaluate total hip arthroplasties?

Abstract

Abstract Background Unknown is which response rate on patient-reported outcome measures (PROMs) is needed to both obtain an accurate outcome and ensure generalizability in evaluating total hip arthroplasty (THA) procedures. Without an evidence based minimum response rate (MRR) on THA PROMs, it is possible that hospitals report invalid patient-reported outcomes (PROs) due to a too low response rate. Alternatively, hospitals may invest too much in achieving an unnecessary high response rate. The aim of this study is to gain an insight into the MRR on PROMs needed to adequately evaluate THA procedures from a clinical perspective. Methods Retrospective study on prospective collected data of primary, elective THA procedures was performed. MRR was investigated for each PROM (NRS pain at rest, NRS pain during activity, EQ-5D-3L, HOOS-PS, anchor function, OHS, anchor pain and NRS satisfaction) separately to calculate the primary outcome: MRR for the THA PROMs set. MRR on a PROM needed to have (condition 1.) similar PRO change score (3 month score minus preoperative score) including confidence interval, (condition 2.) maintaining the influence of each change score predictor and (condition 3.) equal distribution of each predictor, as those of a 100% PROM response rate group. Per PROM, a 100%-group was identified with all patients having the PRO change score. Randomly assessed groups of 90% till 10% response rate (in total 90 groups) were compared with the 100%-group. Linear mixed model analyses and linear regressions were executed. Results The MRR for the THA PROMs set was 100% (range: 70–100% per PROM). The first condition resulted in a MRR of 60%, the second condition in a MRR of 100% and the third condition in a MRR of 10%. Conclusions A 100% response rate on PROMs is needed in order to adequately evaluate THA procedures from a clinical perspective. All stakeholders using THA PROs should be aware that 100% of the THA patients should respond on both preoperative and 3 month postoperative PROMs. For now, taking the first step in improving evaluation of THA for quality control by achieving at least two of the three conditions of MRR, advised is to require a response rate on PROMs of 60% as the lower limit.

Keywords

• Patient-reported outcome measures
• Response rate
• Total hip arthroplasty