RESUMO
OBJECTIVE: To determine factors associated with nonunion of adult tibial fractures. DESIGN: Retrospective review with data collection for logistic regression and survival analysis. SETTING: Scottish Level I trauma center, 1985-2007. PATIENTS: During this period, 1590 adult tibial fractures were treated by reamed nailing and 1003 fractures met all inclusion criteria for the chosen analysis. INTERVENTION: Reamed intramedullary nailing. MAIN OUTCOME MEASURES: Record of nonunion diagnosis and final union time with characteristics, including age, gender, closed or open injury, OTA/AO classification, Gustilo classification, fasciotomy, infection, polytrauma, smoking, and injury severity score. RESULTS: The overall nonunion rate was 12%, and median time to healing was 18 weeks. Age significantly influenced nonunion, with middle-aged patients at highest risk. Both fracture type (closed/open) and morphology (OTA/AO classification) significantly influenced nonunion risk and time to union. Among closed injuries, the highest nonunion rate was for OTA/AO type B fractures (15%). Among open injuries, the highest nonunion rate was for OTA/AO type C (61%). Both compartment syndrome and smoking did not significantly influence nonunion risk but did significantly extend time to union. CONCLUSIONS: Injury characteristics including fracture morphology and severity of soft tissue injury were strong predictors of compromised fracture healing. Age also influenced nonunion risk in an unexpected way, with highest rates in the middle decades of adulthood. Future studies should consider the possibility of similar age-related effects and clinical studies should seek to identify explanations for why this may arise, including both physiological and socio-behavioral factors. LEVEL OF EVIDENCE: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.