Open Reduction Is Not Enough: Residual Dominant-Fragment Displacement and Nonunion After Intramedullary Nailing of Comminuted Subtrochanteric Femur Fractures
Description
Retrospective single-centre cohort of 131 patients (AO/OTA 32B2(a), 32B3(a), 32C3(a)) treated with intramedullary nailing between 2014 and 2022. Background: Nonunion is common after intramedullary nailing of comminuted subtrochanteric femur fractures. Whether the dominant comminuted fragment is adequately reduced, and whether open reduction alone achieves this, is unclear. Methods: Retrospective single-centre cohort of 131 patients (AO/OTA 32B2(a), 32B3(a), 32C3(a)) treated with intramedullary nailing between 2014 and 2022, reported per STROBE. Residual displacement of the dominant fragment was measured on immediate post-operative radiographs by two blinded assessors (intraclass correlation 0.998). The primary outcome was nonunion at nine months, examined by reduction technique and adjunctive cerclage. With few events, regression was exploratory (Firth penalised). Results: Nonunion occurred in 13 patients (10%) and delayed union in 21 (16%). Residual displacement was greater in nonunions (17.0 ± 18.2 mm) than unions (6.6 ± 5.2 mm) or delayed unions (8.9 ± 7.0 mm) (P < 0.001). Eleven of 13 nonunions had open reduction. Displacement varied by technique: open reduction without fragment capture left the most (11.9 ± 12.9 mm), open reduction with cerclage the least (4.1 ± 3.4 mm), below closed reduction (9.2 ± 6.4 mm) (P < 0.001). Discrimination was modest (AUC 0.67; sensitivity 46%, specificity 93% at 14.3 mm; penalised odds ratio 11.3). Fragment length, width and location were not associated with healing. Conclusion: Residual displacement of the dominant fragment not its size or location is associated with nonunion. Open reduction alone does not reliably reduce the fragment; without active capture it left the greatest displacement, whereas cerclage achieved the least. Active fragment control is a modifiable target warranting prospective study.