This study explores the relationship between
delay to surgical debridement and deep infection in a series of
364 consecutive patients with 459 open fractures treated at an academic
level one trauma hospital in North America. The mean delay to debridement for all fractures was 10.6 hours
(0.6 to 111.5). There were 46 deep infections (10%). There were
no infections among the 55 Gustilo-Anderson grade I open fractures.
Among the grade II and III injuries, a statistically significant
increase in the rate of deep infection was found for each hour of
delay (OR = 1.033: 95% CI 1.01 to 1.057). This relationship shows
a linear increase of 3% per hour of delay. No distinct time cut-off
points were identified. Deep infection was also associated with
tibial fractures (OR = 2.44: 95% CI 1.26 to 4.73), a higher Gustilo-Anderson
grade (OR = 1.99: 95% CI 1.004 to 3.954), and contamination of the
fracture (OR = 3.12: 95% CI 1.36 to 7.36). These individual effects
are additive, which suggests that delayed debridement will have
a clinically significant detrimental effect on more severe open
fractures. Delayed treatment appeared safe for grade 1 open fractures. However,
when the negative prognostic factors of tibial site, high grade
of fracture and/or contamination are present we recommend more urgent
operative debridement. Cite this article:
Traditional recommendations suggest that open fractures require urgent surgical debridement to reduce infection. Although many papers comparing early vs late debridement have found no difference in infection rates, these papers have not taken into account important confounding factors. We attempt to answer whether delay between injury and surgical debridement in open fractures is associated with a higher infection rate after accounting for these important confounders. Five hundred and twenty three open extremity fractures in 417 patients were identified using the Sunnybrook trauma and orthopaedic department registries. Thirty patients (36 fractures) did not have complete follow-up. Seven patients were excluded due to incomplete data (complete follow-up rate=91%). A further 14 patients died during their hospitalization. A total of 459 fractures in 364 patients were reviewed. Data was collected on demographics, ISS score, ASA, time to initial operative debridement, timing of antibiotic administration, mechanism of injury, presence of significant contamination, and Gustillo-Anderson fracture grade. Deep infection was defined as an unplanned return to the operating room for treatment of infection. The influence of time to initial debridement was examined in an unadjusted analysis as a continuous variable and at thresholds of 6 and 12 hours of delay. A multivariable logistic regression was used to analyze the effect of delay while controlling for important confounding variables.Purpose
Method
