Patient-reported outcome measures offer little additional value two years after arthroplasty : a systematic review and meta-analysis.

Abstract

Aims: The use of patient-reported outcome measures (PROMs) to assess the outcome after total knee (TKA) and total hip arthroplasty (THA) is increasing, with associated regulatory mandates. However, the robustness and clinical relevance of long-term data are often questionable. It is important to determine whether using long-term PROMs data justify the resources, costs, and difficulties associated with their collection. The aim of this study was to assess studies involving TKA and THA to determine which PROMs are most commonly reported, how complete PROMs data are at ≥ five years postoperatively, and the extent to which the scores change between early and long-term follow-up.

Methods: We conducted a systematic review of the literature. Randomized controlled trials (RCTs) with sufficient reporting of PROMs were included. The mean difference in scores from the preoperative condition to early follow-up times (between one and two years), and from early to final follow-up, were calculated. The mean rates of change in the scores were calculated from representative studies. Meta-analyses were also performed on the most frequently reported PROMs.

Results: A total of 24 studies were assessed. The most frequently reported PROMs were the Oxford Knee Score (OKS) for TKA and the University of California, Los Angeles activity scale for THA. The mean rate of follow-up based on the number of patients available at final follow-up was 70.5% (39.2% to 91.0%) for knees and 82.1% (63.2% to 92.3%) for hips. The actual rates of collection of PROM scores were lower. For TKA, the mean OKS, Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), and transformed WOMAC changes were -16.3 (95% CI -17.5 to -15.2), 23.2 (95% CI 17.2 to 29.2), and -29.7 (95% CI -32.4 to -27.0) points for short-term follow-up. These decreased to 1.3 (95% CI -0.8 to 3.3), -3.4 (95% CI -7.0 to 0.3), and 4.7 (95% CI -1.5 to 10.9) points for the remaining follow-up. A similar meta-analysis was not possible for studies involving THA. We commonly observed that the scores plateaued after between one and two years, and that there was little or no change beyond this time.

Conclusion: The long-term PROMs for TKA and THA beyond one or two years are often incomplete and lose sensitivity at this time. Given the considerable resources, costs, and challenges associated with the collection of these scores, their clinical value is questionable. Therefore, consideration should be given to abandoning the requirement for the collection of long-term PROMs in favour of more robust and reliable measures of success that offer more clinical relevance and use.