Predictive Model for Reintervention in Total Ankle Arthroplasty

Abstract

Category: Ankle Arthritis Introduction/Purpose: Osteoarthritis (OA) of the tibiotalar joint is associated with a high degree of morbidity, being comparable with final stages of renal and heart failure. The incidence of OA of the tibiotalar joint is as high as 47.7 per 100,000 habitants, of whom most of them will need a surgical intervention to improve their quality of life. Up to date evidence supports the use of the total ankle arthroplasty for its management. A great effort has been made to elucidate the factors associated with failure of the total ankle prothesis, until now no attempt has been made to look for factors associated with the need of surgical reintervention, beyond revision of the prothesis. Methods: Prospective cohort study in patients with tibiotalar OA operated between 2015 and 2022, with a minimum follow-up of 12 months, excluding revision prothesis and lost to follow-up. Reintervention was considered as any complication, related to the prothesis, that needed a surgical intervention that was not planned. We performed a multivariate logistic regression model with 13 variables (diabetes, hypertension, tabaquism, age, etiology, hind foot alignment, subtalar OA, talonavicular OA, calcaneocuboid OA, pes planus, cavo varus, previous subtalar arthrodesis and surgical time). Confounding effect was assessed with the change of the odds ratio of the principal independent variable, by adding or subtracting potentially confounder variables into the model. The variables that were statistically non-significant and those variables that do not cause confusion effect, were discarded. Goodness of fit was evaluated using the Hosmer-Lemeshow test and the discrimination of the model with the ROC curve. The Kaplan-Meier curve was used for survival analysis. Results: 105 total ankle arthoplasties were included, the statistically significant variables were: take-down etiology OR = 171.73 (CI: 6.87 - 4294.95, p < 0.01), post-traumatic etiology OR = 14.56 (CI: 1.98 - 107.04, p < 0.01) and surgical time OR = 0.96 (CI: 0.92 - 0.99, p = 0.04), while age OR = 1.02 (CI: 0.96 - 1.09, p = 0.57) and cavo varus OR = 4.75 (CI: 0.71 - 31.93, p = 0.11), wereincluded in the model because of its confounding effect. Goodness of fit demonstrates that the data fit the model adequately (p = 0.84). The ROC curve demonstrates 89.03% discriminatory power. The survival rate at 8 years was 80.39%, median of survival rate was not determined, because longer follow-up was needed. Conclusion: The take-down and post-traumatic etiology can be considered as important risk factors for requiring a surgical reintervention. Patients with a take-down etiology that have a cavo varus deformity and higher age, have a higher risk to require a surgical reintervention. Longer surgical time can be considered as a protective factor for needing a surgical reintervention. At 8 years 20.61% of patients will need a surgical reintervention, related to their Total Ankle Arthroplasty.