Contemporary Comparison of Short-Term Outcomes after Total Ankle Replacement and Ankle Arthrodesis

Abstract

Category: Ankle, Ankle Arthritis Introduction/Purpose: During the last twenty years, studies comparing total ankle replacement (TAR) and ankle arthrodesis (AA) appear to demonstrate lower complication rates with TAR than with AA. However, advances in implant technology and surgical techniques have dramatically reduced complication rates. As a result, studies comparing TAR and AA require more patients to detect differences in rare events. Despite this, few epidemiologic studies have been performed examining short-term outcomes after TAR and AA using a contemporary patient population. The purpose of the current study was to compare perioperative outcomes after TAR and AA using patient data from the American College of Surgeons-National Surgical Quality Improvement Program (ACS-NSQIP) database between 2012 and 2017. Methods: We reviewed patient data from ACS-NSQIP database collected between 2012 and 2017 using Current Procedural Terminology (CPT) codes 27700 (TAR), 27702 (TAR), 29899 (AA) and 27870 (AA). Patients were then excluded if they were treated for fractures, infections, non-foot or ankle-related conditions or had revision procedures. Patients were also excluded if they were older than 90 years as ACS-NSQIP does not report age above 90 years. The study population included those treated in inpatient and outpatient settings. The main outcomes of interest were readmission and reoperation related to initial surgery, surgical site complications and hospital length of stay (LOS). Predictors of adverse outcomes were evaluated through multivariate regression of patient demographics, comorbidities and treatment characteristics. Results: Out of 1214 patients included in the study, 187 (15.4%) patients were treated with AA, and 1027 (84.6%) underwent TAR. Patients with AA were younger, had higher body-mass index, higher white blood cell count, more often had diabetes mellitus (DM) treated with insulin, received more dialysis treatment, had higher anesthesia risk classification and were treated in the outpatient setting more often than patients with TAR. Among outcomes, AA patients had longer hospital LOS, more deep surgical site infections and more reoperations than TAR patients. Post-operative readmissions were not significant but were higher in AA patients (2.7% vs. 0.9%, p=0.101). Combining these adverse outcomes, multivariate regression revealed that higher anesthesia risk category (p=0.0007), DM (p=0.029) and AA (p=0.049) had positive correlations with adverse outcomes. Conclusion: Ankle arthrodesis appears to be independently associated with perioperative complications compared to TAR, consistent with previous reports. Although complications were rare, patients with DM and higher anesthesia risk seem to be important factors to consider. Interestingly, patients with DM had fewer adverse outcomes with TAR than AA (3.8% vs. 7.4%). The difference was even greater in DM patients treated with insulin (4.3% vs. 13.3%) although only 38 patients had DM controlled with insulin in the cohort. Further studies are needed to identify patient populations at risk of complications, specifically those with DM.