Abstract
BACKGROUND: Lateral locked plating remains an important treatment strategy for distal femoral fractures but has been associated with nonunion rates ranging from 6% to 20%. The objective of this study was to identify factors associated with nonunion following lateral locked plating of distal femoral fractures with use of a Bayesian analysis.
METHODS: All consecutive patients ≥18 years of age who were treated with lateral locked plating for a distal femoral fracture at 2 Level-I trauma centers between 2006 and 2024 and who had ≥3 months of follow-up were included. Multivariable Bayesian logistic regression analysis was performed to identify factors associated with nonunion, which was defined as a reoperation to promote healing, and the results are reported as odds ratios (ORs) with 95% credible intervals (CrIs). Probabilities of >95% were considered very strong evidence of an association with nonunion, and probabilities of 90% to 95% were considered strong evidence.
RESULTS: A total of 560 patients (median age, 68 years; 29% male; 90% White; 97% non-Hispanic; 41% with distal periprosthetic fractures) were included. Fifty-four patients (9.6%) underwent reoperation to promote healing. There was very strong evidence that multifragmentary comminution of the metaphysis (versus simple fracture: OR, 2.60; 95% CrI, 0.91 to 8.06), medial cortical comminution of >0 to 25 mm (versus 0 mm: OR, 3.11; 95% CrI, 1.35 to 7.48), and varus (lateral distal femoral angle [LDFA] of ≥84°: OR, 3.04; 95% CrI, 1.46 to 6.51) or valgus (LDFA of ≤78°: OR, 2.42; 95% CrI, 0.96 to 5.99) malalignment increased the odds of nonunion. A screw density of ≤0.60 proximal to the working length reduced the odds of nonunion (versus ≥0.81: OR, 0.40; 95% CrI, 0.16 to 0.95), although the size and certainty of this effect varied in the sensitivity analysis that utilized alternative thresholds. There was strong evidence that obesity increased the odds of nonunion (OR, 1.64; 95% CrI, 0.86 to 3.13) and that intact wedge fractures reduced the odds of nonunion (versus simple fracture: OR, 0.35; 95% CrI, 0.05 to 1.74).
CONCLUSIONS: One in 10 patients developed nonunion and underwent reoperation to promote healing. Surgeons should restore coronal plane alignment and may consider augmenting fixation in the presence of multifragmentary comminution. Constructs in which all screw holes proximal to the working length are filled should be avoided, although the optimal configuration remains unclear and depends on other construct characteristics influencing biomechanics. Overall, the small to moderate effect sizes highlight the multifactorial etiology of nonunion following lateral locked plating of distal femoral fractures.
LEVEL OF EVIDENCE: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.