Who gets referred for knee or hip replacement? A theoretical model of the potential impact of evidence-based referral thresholds using data from a retrospective review of clinic records from an English musculoskeletal referral hub

Objectives: To estimate the relationship between patient characteristics and referral decisions made by musculoskeletal hubs, and to assess the possible impact of an evidence-based referral tool. Design: Retrospective analysis of medical records and decision tree model evaluating policy changes using local and national data. Setting: One musculoskeletal interface clinic (hub) in England. Participants: 922 adults aged ≥50 years referred by general practitioners with symptoms of knee or hip osteoarthritis. Interventions: We assessed the current frequency and determinants of referrals from one hub and the change in referrals that would occur at this centre and nationally if evidence-based thresholds for referral (Oxford Knee and Hip Scores, OKS/OHS) were introduced. Main outcome measure: OKS/OHS, referrals for surgical assessment, referrals for arthroplasty, costs and quality-adjusted life years. Results: Of 110 patients with knee symptoms attending face-to-face hub consultations, 49 (45%) were referred for surgical assessment; the mean OKS for these 49 patients was 18 (range: 1-41). Of 101 hip patients, 36 (36%) were referred for surgical assessment (mean OHS: 21, range: 5-44). No patients referred for surgical assessment were above previously reported economic thresholds for OKS (43) or OHS (45). Setting thresholds of OKS ≤31 and OHS ≤35 might have resulted in an additional 22 knee referrals and 26 hip referrals in our cohort. Extrapolating hub results across England suggests a possible increase in referrals nationally, of around 13 000 additional knee replacements and 4500 additional hip replacements each year. Conclusions: Musculoskeletal hubs currently consider OKS/OHS and other factors when making decisions about referral to secondary care for joint replacement. Those referred typically have low OHS/OKS, and introducing evidence-based OKS/OHS thresholds would prevent few inappropriate (high-functioning, low-pain) referrals. However, our findings suggest that some patients not currently referred could benefit from arthroplasty based on OKS/OHS. More research is required to explore other important patient characteristics currently influencing hub decisions.