A Scoring System for Predicting Nonunion After Intramedullary Nailing of Femoral Shaft Fractures

Abstract

Introduction: Femoral shaft nonunion negatively affects patient quality of life. Although multiple risk factors have been identified for femoral shaft nonunion after intramedullary nail (IMN) fixation, there is no quantitative model for predicting nonunion.

Study description: The study is a retrospective cohort study of patients with femoral shaft fractures treated at two level one trauma centers who were followed to fracture union or nonunion. Patient, injury, and surgical characteristics were analyzed to create a quantitative model for nonunion risk after intramedullary nailing.

Methods: Eight hundred one patients aged 18 years and older with femoral shaft fractures treated with reamed, locked IMNs were identified. Risk factors including demographics, comorbidities, surgical variables, and injury-related characteristics were evaluated. Multivariate analysis was conducted, and several variables were included in a scoring system to predict nonunion risk.

Results: The overall nonunion rate was 7.62% (61/801). Multivariate analysis showed significant association among pulmonary injury (odds ratio [OR] = 2.19, P = 0.022), open fracture (OR=2.36, P = 0.02), current smoking (OR=3.05, P < 0.001), postoperative infection (OR=12.1, P = 0.007), AO/OTA fracture pattern type A or B (OR=0.43, P = 0.014), and percent cortical contact obtained intraoperatively ≥25% (OR=0.41, P = 0.021) and nonunion. The scoring system created to quantitatively stratify nonunion risk showed that a score of 3 or more yielded an OR of 6.38 for nonunion (c-statistic = 0.693, P < 0.0001).

Conclusions: Femoral shaft nonunion risk is quantifiable based on several independent injury, patient, and surgical factors. This scoring system is an additional tool for clinical decision making when caring for patients with femoral shaft fractures treated with IMNs.

Figure 1

Flowchart demonstrating inclusion of patients. 2469 patients with CPT codes 27245, 27506, and 27511 were initially identified. 1137 patients with femoral shaft fracture remained after radiographic screening to confirm AO/OTA type 32 fractures, with 801 meeting eligibility criteria. 61 patients with nonunion were identified and compared with 244 randomly selected patients with union in a 1:4 ratio.

Figure 2

Immediate postoperative AP and lateral fluoroscopic images showing an AO type B femoral shaft fracture after intramedullary nailing. As an example of determining cortical contact, this case was assigned 50% cortical contact, 25% for the anterior cortex and 25% for the lateral cortex. The medial and posterior cortices had residual fracture gaps of >1 mm and were assigned 0% each.