To determine factors associated with developing nonunion or delayed healing after open fracture.
Prospective cohort between 2001 and 2009.
Three level 1 Canadian trauma centers.
Seven hundred thirty-six (791 fractures) subjects were enrolled. Six hundred eighty-nine (94%) subjects (739 fractures) provided adequate outcome data.
Subjects were followed until fracture(s) healed; phone interviews and chart reviews were conducted 1 year after fracture. Patient, fracture, and injury information, and time to surgery and antibiotics were recorded during hospitalization.
Nonunion defined as unplanned surgical intervention after definitive wound closure or incomplete radiographic healing at 1 year and delayed healing defined as 2 consecutive clinical assessments showing no radiographic progression or incomplete radiographic healing between 6 months and 1 year.
There were 413 (52%) tibia/fibular, 285 (36%) upper extremity, and 93 (13%) femoral fractures. Nonunion developed in 124 (17%) and delayed healing in 63 (8%) fractures. The median time to surgery was not different for fractures that developed nonunion compared with those who did not (P = 0.36). Deep infection [Odd ratio (OR) 12.75; 95% confidence interval (CI) 6.07–26.8], grade 3A fractures (OR 2.49; 95% CI, 1.30–4.78), and smoking (OR 1.73; 95% CI, 1.09–2.76) were significantly associated with developing a nonunion. Delayed healing was also significantly associated with deep infection (OR 4.34; 95% CI, 1.22–15.48) and grade 3B/C fractures (OR 3.69; 95% CI, 1.44–9.44). Multivariate regression found no association between nonunion and time to surgery (P = 0.15) or antibiotics (P = 0.70).
Deep infection and higher Gustilo grade fractures were associated with nonunion and delayed healing.
Prognostic Level I. See Instructions for Authors for a complete description of levels of evidence.
*Department of Surgery, University of Alberta, Edmonton, Alberta, Canada;
†Department of Surgery, University of Calgary, Calgary, Alberta, Canada; and
‡Departments of Physical Therapy, University of Alberta, Edmonton, Alberta, Canada.
Reprints: Lauren A. Beaupre, PT, PhD, 2-50 Corbett Hall, University of Alberta, Edmonton, AB, Canada T6G 2G4 (e-mail: email@example.com).
This study was funded through unrestricted research grants from the Edmonton Orthopedic Research Committee and Zimmer Canada, INC.
Presented in Part at the Canadian Orthopedic Association Annual Meeting, June 2015, Vancouver, BC, Canada, and at the Annual Meeting of the Orthopedic Trauma Association, October 2015, San Diego, CA.
The authors report no conflict of interest.
This study was approved by the regional health ethics board at the University of Alberta (Pro00000936).
Accepted October 28, 2015