Influence of screw configuration on reduction and stabilization of simulated complete lateral condylar fracture in equine limbs.

Abstract

To determine the influence of screw configuration on the reduction and stabilization of simulated complete lateral condylar fracture. Randomized experimental crossover study. A lateral condylar fracture was simulated in 18 cadaver limbs from nine horses. Each limb underwent repair with 4.5 mm diameter cortex screws, tightened to 4 Nm, in a linear and triangular configuration. Computed tomography (CT) of each repair was performed with the limbs in unloaded and loaded conditions. Fracture gaps were measured at the dorsal, palmar, and middle locations of the third metacarpal condyle. Fracture gap measurements were graded 0-4, based on voxels. Following descriptive analysis, a Bayesian network (BN) model was fitted to the data. The median fracture grade was 0 (range: 0-4) for unloaded linear repairs and 2 (0-4) for loaded linear repairs. The median fracture grade was 0 (0-3) for unloaded triangular repairs and 1 (0-3) for loaded triangular repairs. Bayesian network sensitivity analysis showed that the construct configuration reduced the uncertainty in the measured fracture outcome by 0.8%. Triangular screw configuration resulted in better fracture reduction and stability in comparison with linear screw configuration. However, the BN sensitivity analysis results showed that the effect of construct configuration on fracture outcome was weak. These findings indicate a low probability that triangular repair of lateral condylar fracture will result in improved outcomes, when compared with linear repair.