LAUSR

INTRODUCTION Removal of symptomatic implants is a common procedure performed by orthopaedic trauma surgeons. No guidance is available regarding which factors contribute to the likelihood of an implant becoming symptomatic. Our objective was to determine whether radiographic parameters associated with distal interlocks in retrograde femoral nails are associated with the rate of symptomatic screw removal. PATIENTS AND METHODS We conducted a retrospective review at a Level I trauma center. Study patients (n=442) had femoral fractures treated with retrograde intramedullary nails from 2007 to 2014 and at least 1year of follow-up. The main outcome measurement was symptomatic distal screw removal as predicted by radiographic parameters. RESULTS Symptomatic screw removal occurred in 12% of the patients. Increased distance between the most distal screw and the articular surface of the femur significantly reduced likelihood of symptomatic screw removal. A cutoff of 40mm from the articular block was predictive of removal (≥40mm, 0% removal; <40mm, 18% removal, p<0.0001). In patients with distal screws placed within 40mm of the articular surface of the femur, a ratio of screw length to distance between medial and lateral femoral cortices that was ≥1 was a strong predictor of symptomatic screw removal (area under Receiver Operating Characteristic curve, 0.75; p<0.0001). CONCLUSIONS More distal screws and screws that radiographically extend to or beyond the medial cortex are more likely to cause pain and require removal in femoral fractures treated with retrograde intramedullary nails. We identified a specific distance from the joint (<40mm) and a ratio of screw length to bone width (≥1) that significantly increased the likelihood of symptomatic screw removal. Clinicians can use these data to inform patients of the likely risk of implant removal and perhaps to better guide placement and length of screws when the clinical scenario allows some flexibility in location and length of screws.