Abstract
The Revision Hip Complexity Classification (RHCC) was developed by modified Delphi system in 2022 to provide a comprehensive, reproducible framework for the multidisciplinary discussion of complex revision hip surgery. The aim of this study was to assess the validity, intra-relater and inter-relater reliability of the RHCC.
Radiographs and clinical vignettes of 20 consecutive patients who had undergone revision of Total Hip Arthroplasty (THA) at our unit during the previous 12-month period were provided to observers. Five observers, comprising 3 revision hip consultants, 1 hip fellow and 1 ST3-8 registrar were familiarised with the RHCC. Each revision THA case was classified on two separate occasions by each observer, with a mean time between assessments of 42.6 days (24–57). Inter-observer reliability was assessed using the Fleiss™ Kappa statistic and percentage agreement. Intra-observer reliability was assessed using the Cohen Kappa statistic. Validity was assessed using percentage agreement and Cohen Kappa comparing observers to the RHCC web-based application result.
All observers were blinded to patient notes, operation notes and post-operative radiographs throughout the process.
Inter-observer reliability showed fair agreement in both rounds 1 and 2 of the survey (0.296 and 0.353 respectively), with a percentage agreement of 69% and 75%.
Inter-observer reliability was highest in H3-type revisions with kappa values of 0.577 and 0.441.
Mean intra-observer reliability showed moderate agreement with a kappa value of 0.446 (0.369 to 0.773).
Validity percentage agreement was 44% and 39% respectively, with mean kappa values of 0.125 and 0.046 representing only slight agreement.
This study demonstrates that classification using the RHCC without utilisation of the web-based application is unsatisfactory, showing low validity and reliability. Reliability was higher for more complex H3-type cases. The use of the RHCC web app is recommended to ensure the accurate and reliable classification of revision THA cases.
