ABSTRACT
OBJECTIVE: Ankle arthrodesis (AA) is often fixed using cannulated screws. The irritation from metalwork is a relatively common complication, but there is no consensus regarding the need to remove the screws on a systematic basis. The aim of this study was to determine (1) the proportion of screws removed after AA and (2) whether predictors of screw removal could be identified. METHODS: This PRISMA-compliant systematic review was part of a larger previous protocol registered on the PROSPERO platform. Multiple databases were searched including studies in which patients undergone AA using screws as exclusive fixation method were followed. Data were harvested regarding the cohort, the study design, the surgical technique, the nonunion and complication rate at the longest follow-up. Risk of bias was assessed using the modified Coleman Methodology Score (mCMS). RESULTS: Forty-four series of patients from thirty-eight studies (1990 ankles, 1934 patients) were selected. The average follow-up was 40.8 months (range 12-110). In all studies, hardware was removed due to symptoms reported by patients and related to the screws. The pooled proportion of removal of metalwork was 3% (95% CI 2-4). The pooled proportion of fusion was 96% (95%CI 95-98), while the pooled proportion of complications and reoperations (excluding the removal of metalwork) stood at 15% (95% CI 11-18) and 3% (95% CI 2-4), respectively. The mean mCMS (50.8 ± 8.1, range 35-66) revealed only an overall fair quality of studies. The univariate analysis and the multivariate model showed that the year of publication (R = - 0.004; p = 0.01) and the number of screws (R = 0.08; p = 0.01) were associated with the screw removal rate. Specifically, we found that over time the removal rate decreased by 0.4% per year and that the use of three screws instead of two reduced the risk of removal of metalwork by 8%. CONCLUSIONS: In this review, removal of metalwork after ankle arthrodesis using cannulated screws was needed in 3% of cases at an average follow-up of 40.8 months. It was indicated only in case of symptoms related to soft tissue irritation from screws. The use of three screws was paradoxically related to a reduced risk of removal of screws as compared to two-screw constructs. LEVEL OF EVIDENCE: Level IV, systematic review of Level IV.
