Risk of early revision in total hip arthroplasty: the relative contribution of the surgeon versus the hospital

Abstract Background Early revision for total hip arthroplasty is a serious adverse outcome. There are multiple contributing risk factors for early revision. Risk factors can exist at the level of the surgeon and the level of the institution. The primary research question of this study was to determine the relative contribution of surgeon‐level and hospital‐level variance to rates of early revision (overall and for infection) after primary total hip arthroplasty. Methods This is a registry‐based study from the Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR). Data for the most commonly used stem (Exeter V40) were used to reduce prosthesis variation from the analysis. A mixed effects Cox Model (also known as a frailty model) with crossed random effects for surgeon and hospital was used. Outcomes were early revision (within 2 years) for all causes and for infection. This model allowed for the risk of early revision to be explained by the variability at the surgeon level or hospital level. Results There were 32 031 procedures performed by 735 surgeons across 250 hospitals between 1 January 2015 and 31 December 2019. Surgeon variability significantly contributed to overall variation in revision for any cause and revision for infection ( P < 0.0001). There was no significant contribution of hospital‐level variation to overall revision or for infection. Conclusions Surgeon‐level factors play a more important role than institution‐level factors in early revision after primary total hip arthroplasty. If surgeons are identified as having a higher risk of revision, there is potential for surgeon‐level practice change to reduce the risk of early revision.