Predicting Functional Outcomes Following Fracture Nonunion Repair-Development and Validation of a Risk Profiling Tool.

Abstract

To develop a tool that can be used preoperatively to identify patients at risk of poor functional outcome following operative repair of fracture nonunion. Retrospective analysis of prospectively collected data. Academic medical center. Three hundred twenty-eight patients who underwent operative repair of a fracture nonunion were prospectively followed for a minimum of 12 months post-operatively. After randomization, 223 (68%) patients comprised an experimental cohort and 105 (32%) patients comprised a separate validation cohort. Within the experimental cohort, forward stepwise multivariate logistic regression was applied to 17 independent variables to generate a predictive model identifying patients at risk of having a poor functional outcome [Predicting Risk of Function in Trauma-Nonunion (PRoFiT-NU) Score]. Functional outcomes were assessed using the Short Musculoskeletal Function Assessment (SMFA). Poor outcome was defined as an SMFA function index greater than 10 points above the mean at 12 months post-operatively. Significant predictors of poor outcome were lower extremity nonunion [odds ratio (OR) = 3.082; P = 0.021], tobacco use (OR = 2.994; P = 0.009), worker's compensation insurance (OR = 3.986; P = 0.005), radiographic bone loss (OR = 2.397; P = 0.040), and preoperative SMFA function index (OR = 1.027; P = 0.001). The PRoFiT-NU model was significant and a good predictor of poor functional outcome (χ(5) = 51.98, P < 0.0005; area under the receiver operating curve = 0.79). Within the separate validation cohort, 16% of patients had a poor outcome at a PRoFiT-NU score below 25% (low risk), 39% of patients had a poor outcome at a PRoFiT-NU score between 25% and 50% (intermediate risk), and 63% of patients had a poor outcome at a PRoFiT-NU score above 50% (high risk). The PRoFiT-NU score is an accurate predictor of poor functional outcome following fracture nonunion repair. Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence description of levels of evidence.