Manipulation under anaesthetic following primary knee arthroplasty is associated with a higher rate of subsequent revision surgery

Aim: To determine the association between manipulation under anaesthetic (MUA) following primary knee arthroplasty and subsequent revision surgery. Methods: Patients undergoing primary knee arthroplasty April 2011 to April 2016 with minimum 1-year follow up to April 2017 were identified from the national hospital episode statistics for England. The first arthroplasty per patient, per side, was included; cases with a record of subsequent infection or periprosthetic fracture were excluded. Patients undergoing MUA within 1-year to the same knee were identified, defining the populations for the MUA and non-MUA cohorts. Mortality adjusted Kaplan-Meier survival analysis (revision arthroplasty), was performed to a maximum of 6-years. A Cox proportional hazards model was used to determine the hazard for revision, adjusting for type of primary arthroplasty, gender, age group, year, comorbidity index, obesity, regional deprivation, rurality, and ethnicity. Results: A total of 309,650 primary arthroplasty cases (309,650 patients) were included. Manipulation under anaesthetic within 1-year was recorded in 6882 patients (2.22%; 95% CI 2.17-2.28) defining the MUA cohort; all others were included in the parallel non-MUA cohort. At 6-years, the mortality-adjusted estimated implant survival rate in the MUA cohort was 91.2% (95% CI 90.0-92.2) in comparison to 98.1% (95% CI 98.0-98.2) in the non-MUA cohort. In the fully adjusted model, this corresponded to an adjusted hazard for revision of 5.03 (hazard ratio [HR]; 95% CI 4.55-5.57). Conclusion: Patients who underwent MUA within 1-year of primary arthroplasty were at a five-fold increased risk of subsequent revision even after excluding cases of infection or fracture. Further investigation of the aetiology of ‘stiffness’ following primary knee arthroplasty and the optimal treatment options to improve outcomes is justified.