Higher hospital volume reduces early failure rates in single-stage revision TKR for infection: An analysis of the United Kingdom National Joint Registry and National Administrative Databases

Revision knee replacement (RevKR) for infection is rare but increasing. It is hypothesised that higher hospital volume reduces adverse outcomes. The aim was to estimate the association of surgical unit volume with outcomes following first, single-stage RevKR for infection. This population-based cohort study merged data from the United Kingdom National Joint Registry, Hospital Episode Statistics, National Patient Reported Outcome Measures and the Civil Registrations of Death. Patients undergoing procedures between 1 January 2009 and 30 June 2019 were included. Early outcomes were chosen to reflect the quality of the surgical provision and included re-revision at 2 years, mortality, serious medical complications, length of stay and patient-reported outcome measures (PROMs). Adjusted fixed effect multivariable regression models were used to examine the association between surgical unit mean annual caseload and the risk of adverse outcomes. A total of 1477 patients underwent first-time single-stage RevKRs for infection across 267 surgical units and 716 surgeons. Following adjustment for age, gender, American Society of Anaesthesiologists grade, surgeon volume, year of surgery and operation funder and modelling surgical unit volume with restricted cubic spline, a greater mean annual volume was associated with a lower risk of re-revision at 2 years. The odds of re-revision in hospitals performing fewer than or equal to 12 cases per year was 2.53 (95% confidence interval = 1.50-4.31) times more likely than hospitals performing three to four cases per month. Annual variation in surgical unit volume was not associated with mortality and serious medical complications within 90 days. Only 99 out of 1477 (7%) of patients had linked PROMs which precluded subsequent analysis. Overall, higher volume surgical units had lower rates of early re-revision following the first RevKR for infection. We were unable to provide recommended specific volume thresholds for units; however, the probability of re-revision appears to be lowest in the highest volume units. Level III, retrospective cohort study of prospectively collected data.