What is the Fate of Undisplaced Femoral Neck Fractures Treated With Cannulated Screws?

Background We aimed to report implant survival in Garden type I and II femoral neck fractures treated with cannulated screws in elderly patients. Methods We retrospectively studied 232 consecutive unilateral Garden I and II patients (232 fractures) treated with cannulated screws. Mean age was 81 years (range, 65 to 100), and a body mass index of 25 (range, 15.8 to 38.3). No between-group differences were found in demographic variables and/or baseline measurements (P > .05). Mean follow-up was 36 months (range, 1 to 171). Two observers measured baseline radiographic variables with good-to-excellent interobserver reliability. The posterior tilt angle, measured on a cross-table lateral x-ray, was used to classify the cohort into <20° (n = 183) and ≥20° (n = 49). The cumulative incidence with competing risk analysis was used to predict association between posterior tilt and subsequent conversion to arthroplasty. Patient survival was calculated with the Kaplan-Meier estimate. Results Implant survival was 86.3% (95% confidence interval (CI) 80 to 90) at 12 months and 77.3% (95% CI 64 to 86) at 70 months. The 12-month cumulative incidence failure was 12.6% (95% CI 8 to 17). After controlling for confounders, posterior tilt ≥20° had higher risk of subsequent arthroplasty when compared to posterior tilt <20° (38.8 [95% CI 25 to 52] versus 5% [95% CI 2.8 to 9], subhazard ratio 8.3, 95% CI 3.8 to 18), without any other radiologic or demographic factor being associated with failure. Patient survival was 88.2% (95% CI 83 to 91.7) at 12 months, 79.5% (95% CI 73 to 84) at 24 months, and 57% (95% CI 48 to 65) at 70 months. Conclusion Cannulated screws were a reliable treatment for Garden I and II fractures, except when there was posterior tilt ≥20°, where arthroplasty should be considered.