Factors predicting failure of patella fixation

Abstract

Surgical fixation of patella fractures is frequently indicated due to disruption of the quadriceps mechanism. Operative technique varies; however, failure rates can be high. The purpose of this study was to compare the effectiveness of various techniques for the fixation of patella fractures and the etiology of fixation failure. We retrospectively reviewed 173 patella fractures treated operatively at two Level I trauma centers. Patients with less than 90 days of follow-up, inadequate radiographic studies, and partial or total excision were excluded. Failure was defined as hardware breakage, nonunion, or displacement of fragments from their initial reduced position. Twelve factors were examined independently for predictive value using both univariate and multivariate analyses. A comparison between groups based on reoperation and hardware removal was also performed. One hundred nine patients met the inclusion criteria, and 13 were found to have failed (12%). Both older patient age (p < 0.02) and use of K-wires, with or without tension-band wires (p < 0.04), were found to be significant predictors of failure. Increasing follow-up time was the only significant predictor of reoperation (p < 0.001) and hardware removal (p < 0.001). As anticipated, increasing age was found to correlate with higher failure rates. Use of K-wires with or without tension-band wires correlated with higher failure rates, compared with the use of screws, both K-wires and screws, or other fixation. Increasing follow-up time predicted both reoperation and hardware removal, with patients having symptomatic hardware and other complications naturally returning to clinic for evaluation and treatment.