Change in Gustilo-Anderson classification at time of surgery does not increase risk for surgical site infection in patients with open fractures: A secondary analysis of a multicenter, prospective randomized controlled trial : OTA International

Secondary Logo

Journal Logo

Clinical/Basic Science Research Article

Change in Gustilo-Anderson classification at time of surgery does not increase risk for surgical site infection in patients with open fractures: A secondary analysis of a multicenter, prospective randomized controlled trial

Axelrod, Daniel MD, MSca,*; Comeau-Gauthier, Marianne MD, MSca; Prada, Carlos MD, MHSca; Bzovsky, Sofia MSca; Heels-Ansdell, Diane MScb; Petrisor, Brad MD, MSca; Jeray, Kyle MDc; Bhandari, Mohit MD, PhD, FRCSCa,b; Schemitsch, Emil MD, FRCSCd; Sprague, Sheila PhDa,b;  on behalf of the FLOW Investigators

Author Information
OTA International: The Open Access Journal of Orthopaedic Trauma 6(1):p e231, March 2023. | DOI: 10.1097/OI9.0000000000000231
  • Open


1. Introduction

Open fractures represent a major source of mortality and morbidity across the world.1,2 Moreover, complications after open fractures are common and are often associated with surgical site infections (SSIs).3,4 Patients who have suffered from a SSI after an open fracture report a quality of life similar to those who have suffered from a major cardiac event.5,6 Administration of appropriate antibiotics within 60 minutes of injury, based on fracture severity defined by the Gustilo-Anderson classification (GAC), has reliably been shown to reduce the incidence of postoperative complications due to infection.7,8

The GAC, first developed in 19699 (and modified to its current version in 198410), remains the standard for evaluation of the skin and soft-tissue injuries associated with open fractures. Any open fracture of the extremity can be graded from I to III based on the length of the open wound, the amount of energy that produces the trauma, and the degree of soft-tissue injury and contamination. Although the GAC was developed as a tool to be applied after irrigation and debridement of an open fracture was completed and a definitive intraoperative evaluation of skin, soft-tissue, and periosteal injury can be performed, a preliminary GAC is also given to patients when they first present after their traumatic injury.11 This initial GAC is often the definitive injury characteristic that guides antibiotic prophylaxis,12 although not originally designed for this purpose.

Furthermore, given that the preoperative GAC drives initial antibiotic use, an incorrect preoperative GAC (when compared with the definitive GAC) may lead to inappropriate antibiotic administration in the immediate hours after the open injury, the most crucial period of open fracture care. With inappropriate antibiotic administration, there is a theoretical increase in the risk of both deep and superficial SSIs. This potential relationship between incorrect GAC and the risk of SSI has not yet been studied.

Moreover, no formal statistical analysis has evaluated how reliably a preoperative GAC predicts the definitive classification once formal debridement is completed, or what may predict a change in score. An understanding of how often the preoperative GAC correctly predicts the definitive GAC, along with what factors predict changes in preliminary to definitive GAC, will help evaluate the utility of assigning a preoperative GAC altogether.

Thus, the objectives of this study were (1) to determine whether a change in GAC was associated with an increased risk of SSI, with the hypothesis that a GAC that was upgraded may not have received appropriate antibiotic prophylaxis initially and, thus, increased the patient's risk of SSI; (2) to determine the injury characteristics associated with either upgrading or downgrading the preoperative GAC to the definitive classification; and (3) to determine which wound characteristics are associated with increases and decreases of the GAC.

2. Methods

2.1. Data Source—The FLOW Trial

This study presents the results of secondary data analysis performed on the Fluid Lavage of Open Wounds (FLOW) study, of which the full objectives, methods, and primary results are provided in a prior publication.13 In brief, the FLOW study was an international, multicenter, blinded, 2 × 3 factorial randomized controlled trial assessing the effect of two different irrigation solutions (normal saline vs. soap) and three different irrigation pressures (high vs. low vs. gravity flow) on 1-year reoperation rates among patients with open fractures.

The FLOW study was approved by the ethics committees at the coordinating center, McMaster University (REB: 08‐268).

From June 2009 to September 2013, patients 18 years and older presenting with an open fracture of an extremity requiring operative fixation were randomized to one of the six study groups, after stratification by center and GAC. For a patient with multiple fractures, the highest GAC that did not meet any exclusion criteria was considered as the fracture of consideration for the FLOW study.

Patients returned for follow-up at 1, 2, and 6 weeks and 3, 6, and 12 months. A total of 2447 patients were included in the final primary analysis, which showed that the rates of reoperation were similar, regardless of irrigation pressure, and higher in the group treated with a soap solution compared with those treated with saline.

2.2. Assessment of SSI

A blinded, independent central adjudication committee determined the occurrence of the end point of SSI based on a review of clinic notes, procedure logs, and antibiotic prescriptions, using a modified version of the Centre for Disease Control criteria.14 SSIs included superficial incisional SSI, deep incisional SSI, and organ/space SSI. The final analysis in this study evaluated all SSIs together as a composite outcome of infection.

2.3. Accuracy of GAC Grading

As part of the FLOW trial, participants were assigned a GAC at the time of randomization, as it was a stratification variable, and then a definitive GAC on the first irrigation and debridement. The GAC was assigned by a treating surgical team member (either orthopaedic fellow or orthopaedic surgeon). As part of our analysis, the GAC change was subclassified into downgraded, upgraded, or no change to score.

2.4. Selection of Baseline Factors

We selected baseline factors associated with either upgrading or downgrading the preoperative GAC to the definitive classification a priori based on biologic rationale and previous reports in the literature. For each potential factor, we proposed a hypothesized effect for the logistic regression outcome. To avoid an overfitted or unstable model, we used a rule of thumb that there should be at least 10 times the number of observations as there are factors in a regression model.15 Using primary outcomes data from the FLOW trial, 325 patients developed a SSI; thus, with less than 30 covariates, we were not at high risk of overfitting the model.

Baseline factors were classified into two main groups: injury characteristics and wound characteristics. Injury characteristics included preoperative GAC, location of injury (classified as tibia or elsewhere, as used in the primary study), and mechanism of injury (high or low energy). Wound characteristics included preoperative assessment of bone loss, skin loss, and muscle loss as well as amount of muscle, skin, bone, and fascia debrided intraoperatively.

2.5. Definition of Baseline Factors

2.5.1. Injury Characteristics

We analyzed preoperative GAC as an ordinal variable, creating four levels of analysis with a separate grade for GA I/II, IIIA, and IIIB. GA I/II fractures were grouped together because they often represent injuries with lower energy mechanisms, compared with those in GA class III (A + B). GA IIIC fractures were excluded from the primary study and subsequent analysis. Location of injury and mechanism of injury were analyzed as categorical variables.

2.5.2. Wound Characteristics

The preoperative assessment of bone, skin, and muscle loss were all recorded as categorical variables (presence or absence). By contrast, the amount of muscle, skin, bone, and fascia debrided were classified in an ordinal fashion (0—no debridement, 1—small amount debrided [<1 cm2], 2—moderate amount debrided [1–5 cm2], and 3—large amount debrided [>5 cm2]).

2.6. Statistical Analysis

Our statistical analysis plan was determined a priori. We included FLOW trial participants with complete data in the logistic regression model. We used descriptive statistics to summarize all factors (frequencies and percentages for categorical and ordinal variables and means, medians, and ranges for continuous variables).

For our definitive analysis, we developed a multinomial regression model, with three potential outcomes: an upgraded GAC, a downgraded GAC, and no change in GAC. This produced two regression outputs, one that identified factors associated with upgrading preoperative to intraoperative GAC and another to identify factors associated with downgrading scores. We included all factors specified above as independent variables in fixed effects.

A logistic regression model was then built with SSI as the dependent variable of interest. Change in GAC (no change, upgrade, downgrade) was the independent variable while the above-listed covariates, including initial GAC, were also incorporated as adjustment variables. Patients were compared with those in their initial GAC (eg, the risk of SSI for patients who were upgraded from GAC II to III was compared with the risk for those with initial GAC II). Regression results were reported as odds ratios (ORs) with corresponding 95% confidence intervals (CIs) and P values. We used R (v3.6.1 open access online) for statistical analyses.

3. Results

3.1. Participant Characteristics

Of the 2447 participants who enrolled and contributed data for the FLOW primary outcome analysis, 2420 participants were ultimately included in this secondary analysis, with 27 participants (1.1%) excluded because of incomplete data. The demographic data for this cohort are similar to the cohort reported in the initial trial (Table 1).

TABLE 1 - Participant Characteristics
Variables FLOW Cohort (N = 2420)
GAC (preoperative), n (%)
 1/2 1682 (69.5)
 3a 529 (21.9)
 3b 209 (8.6)
GAC (intraoperative), n (%)
 1/2 1538 (63.6)
 3a 650 (26.8)
 3b 232 (9.6)
Location, n (%)
 Tibia 1511 (62.4)
 Elsewhere 909 (37.6)
Mechanism of injury, n (%) N = 2419
 High energy 2132 (88.1)
Presence of bone loss (preoperative), n (%) N = 2419
534 (22.1)
Skin loss (preoperative), n (%) N = 2419
426 (17.6)
Muscle loss (preoperative), n (%) N = 2419
345 (14.3)
Width of wound, cm (SD) 2.77 (3.77)
Length of wound, cm (SD) 4.50 (5.81)
Muscle debridement, cm2, n (%) N = 2415
 None 1210 (50.1)
 <1 913 (37.8)
 1–5 237 (9.8)
 >5 55 (2.3)
Skin debridement, cm2, n (%) N = 2416
 None 751 (31.1)
 <1 1327 (54.9)
 1–5 282 (11.7)
 >5 56 (2.3)
Fascia debridement, cm2, n (%) N = 2417
 None 1034 (42.8)
 <1 1123 (46.4)
 1–5 228 (9.4)
 >5 32 (1.3)
Bone debridement, cm2, n (%) N = 2416
 None 1358 (56.2)
 <1 771 (31.9)
 1–5 246 (10.2)
 >5 41 (1.7)
Note: A total of 2447 participants were enrolled in the trial, 2432 received a preoperative GAC and 2420 received a definitive perioperative GAC. These 2420 participants were included in the analysis. Not all patients had complete data on every factor listed.

3.2. Accuracy of GAC Grading

Of the 2420 participants included, 305 participants had their preoperative GAC changed (12.6%). In total, 229 participants (9.5%) were upgraded and 76 participants (3.1%) were downgraded after surgery. Of the 1682 participants classified as GAC I/II preoperatively, 190 participants (11.3%) were upgraded intraoperatively. Of those 529 participants classified as GAC IIIA, 39 (7.4%) were upgraded and 39 (7.4%) were downgraded. Finally, of the 209 participants classified as GAC IIIB, 37 (17.7%) were ultimately downgraded.

3.3. GAC and SSI

Forty-six participants who had their GAC upgraded suffered an SSI (20%), compared with 13 patients who had their GAC downgraded and suffered an SSI (17.1%), and 266 patients who had no change in GAC went onto having an SSI (12.5%). An upgraded GAC was not significantly associated with an increased risk of SSI when considered in the multivariate model (OR: 1.03; 95% CI: 0.99–1.07; P = 0.14). These results were mirrored in patients who had a downgraded GAC, which did not change their risk of SSI (Table 2).

TABLE 2 - Change in GAC and SSI Association
Characteristic OR 95% CI P
Change in GA classification
 Same (Reference)
 Downgrade 0.98 0.92–1.04 0.5
 Upgrade 1.03 0.99–1.07 0.14
Note: These results are adjusted for injury and participant characteristics.

3.4. Factors Associated with Upgrading GAC

The factors associated with upgrading the GAC (from the preoperative score to the definitive assessment) were lower initial GAC (grade I/II compared to grade IIIA OR: 7.91, 95% CI: 4.68–13.40; P < 0.001), fracture site other than the tibia (OR: 1.97, 95% CI: 1.45–2.67; P < 0.001), bone loss at presentation (OR: 1.67, 95% CI: 1.14–2.45; P = 0.009), width of wound (OR: 1.06 per additional cm of wound width, 95% CI: 1.01–1.12; P = 0.02), length of wound (OR: 1.08 per additional cm of wound length, 95% CI: 1.05–1.12; P < 0.001), and skin loss at presentation (OR: 1.75, 95% CI: 1.15–2.66; P = 0.009). Moreover, any amount of muscle debridement, 1–5 cm of skin debridement (95% CI 1.61–4.84, P < 0.001), and bone debridement over 5 cm2 were associated with upgrading GAC (95% CI: 1.32–11.0, P = 0.01; Table 3).

TABLE 3 - Factors Associated With Upgrading GAC
Characteristic OR 95% CI P
GA classification
 Reference (GA IIIA)
 I/II 7.91 4.68–13.4 <0.001
 IIIB/C 0.00 0.00–0.00 <0.001
Location of injury
 Reference (tibia)
 Injuries elsewhere 1.97 1.45–2.67 <0.001
Mechanism of injury 1.68 0.96–2.97 0.07
Presence of bone loss 1.67 1.14–2.45 0.009
Width of wound (per additional cm) 1.06 1.01–1.12 0.02
Length of wound (per additional cm) 1.08 1.05–1.12 <0.001
Presence of skin loss 1.75 1.15–2.66 0.009
Presence of muscle loss 1.18 0.72–1.92 0.52
Amount of skin debrided, cm2
 Reference (no debridement)
 <1 1.44 0.95–2.17 0.08
 1-5 2.79 1.61–4.84 <0.001
 >5 2.19 0.61–7.95 0.23
Amount of muscle debrided, cm2
 Reference (no debridement)
 <1 1.52 1.03–2.23 0.03
 1-5 2.58 1.38–4.81 0.003
 >5 15.1 4.08–55.6 <0.001
Amount of fascia debrided, cm2
 Reference (no debridement)
 <1 1.25 0.85–1.84 0.25
 1-5 0.87 0.45–1.69 0.69
 >5 0.03 0.00–0.44 0.01
Amount of bone debrided, cm2
 Reference (no debridement)
 <1 1.33 0.93–1.91 0.12
 1-5 1.28 0.75–2.17 0.36
 >5 3.81 1.32–11.0 0.01

3.5. Factors Associated with Downgrading GAC

The only covariate associated with downgrading the initial GAC was the actual grade assigned, with higher initial grade more likely to be downgraded (grade IIIB compared with grade IIIA, OR: 9.84, 95% CI: 5.23–18.50; P < 0.001; Table 4).

TABLE 4 - Factors Associated With Downgrading GAC
Characteristic OR 95% CI P
GA classification
 Reference (GA IIIA)
 I/II 0.00 0.00 to 4.19E+13 0.59
 IIIB/C 9.84 5.23 to 18.50 <0.001
Location of injury
 Reference (tibia)
 Injuries elsewhere 1.07 0.62 to 1.84 0.81
Mechanism of injury 0.44 0.15 to 1.27 0.13
Presence of bone loss 0.86 0.42 to 1.76 0.69
Width of wound (per additional cm) 0.92 0.85 to 1.00 0.04
Length of wound (per additional cm) 0.93 0.87 to 0.98 0.01
Presence of skin loss 0.26 0.10 to 0.64 0.004
Presence of muscle loss 0.78 0.30 to 2.01 0.61
Amount of skin debrided, cm2
 Reference (no debridement)
 <1 0.90 0.47 to 1.73 0.76
 1-5 0.53 0.18 to 1.57 0.25
 >5 1.19 0.17 to 8.05 0.86
Amount of muscle debrided, cm2
 Reference (no debridement)
 <1 0.82 0.43 to 1.55 0.54
 1-5 0.31 0.09 to 1.10 0.71
 >5 0.88 0.11 to 7.03 0.91
Amount of fascia debrided, cm2
 Reference (no debridement)
 <1 0.88 0.46 to 1.70 0.71
 1-5 0.68 0.22 to 2.15 0.51
 >5 1.24 0.09 to 17.2 0.88
Amount of bone debrided, cm2
 Reference (no debridement)
 <1 1.19 0.63 to 2.25 0.58
 1-5 0.94 0.35 to 2.57 0.91
 >5 0.63 0.07 to 5.79 0.68

4. Discussion

This FLOW trial secondary analysis identified that 12% of patients with open fractures, in a large prospective randomized controlled trial, had a change in GAC from initial preoperative assessment to final postoperative evaluation. Moreover, this study demonstrated that an upgrade in GAC was not associated with an increased risk of SSI. Finally, patients with open fractures in a location other than the tibia and with bone loss at presentation were more likely to have an upgraded GAC.

Rather than time to debridement, appropriate early antibiotic administration has been shown to be the best predictor of SSI after open fractures.12,16 The 2011 Eastern Association for Surgery of Trauma (EAST) recommends antibiotic choice based on preoperative GAC, with first-generation cephalosporins for type I and II fractures, with the addition of an aminoglycoside for type III fractures, and penicillin for highly contaminated injuries or those occurring in farming environments.12 This antibiotic protocol was followed in the FLOW trial.17 In addition, antibiotic coverage postoperatively should be altered if the GAC changes after definitive operative debridement, although the use of any postoperative prophylaxis has recently come into question.18 Moreover, the time to definitive operative debridement is often delayed by 12 or more hours as the standard of care for allowable time before operative debridement continues to increase.19–21 Thus, with inadequate initial antibiotic administration and a longer time to definitive GAC, patients may be at a substantially higher risk of SSIs.

We found that while injuries where GAC was upgraded had a modestly increased proportion of SSI (20% vs. 17%), regression models, adjusting for confounding factors, did not show a significant association between GAC upgrade and SSI (95% CI 0.99–1.07, P = 0.14). Thus, treating more severe injuries with antibiotic monotherapy did not contribute to infection. However, our findings were quite fragile with a fragility index of 2—if two more patients in the upgraded GAC had suffered a SSI, the results would have been statistically significant.22

It has been suggested that broadened antibiotic coverage for more severe injuries—the inclusion of an aminoglycoside antibiotic—may have little role in the prevention of SSI in open fractures and that time to delivery of antibiotics, along with injury and patient characteristics, play a larger role. Lack (2015), in an evaluation of SSI risk in severe (GAC III) open tibia fractures, reported that only time to antibiotic administration was predictive of infection. In this observational study, cefazolin was the only agent given in 93.4% of cases and the overall infection rate was 17.5%. Patients who received antibiotics within 1 hour of injury had a 6.8% infection rate compared with 27.9% in those receiving antibiotics after 90 minutes. Similarly, Roddy et al23 found that patients who did not receive antibiotics within 120 minutes were at 2.4 times greater risk of SSI, with patients who had GAC 3 injuries twice as likely to develop a SSI. Moreover, a recent publication, which included 1234 patients enrolled in an open fracture randomized trial, reported that only 61.1% of patients with GAC I/II fractures were given appropriate antibiotics (ie, a first-generation cephalosporin).24 Thus, in addition to delayed time to antibiotics, there remains low adherence to antibiotic prophylaxis guidelines for open fractures, which may also contribute to a higher rate of SSI.

In addition, our study shows that certain factors, including bone loss on presentation, may predict an eventual change to a higher GAC after final debridement. These injuries may be classified as low GAC based on wound characteristics alone but have worrisome features of higher energy events with a higher risk of eventual soft-tissue infection. For example, the finding that fractures located in a site other than the tibia were associated with a greater chance of upgrading may be related to underestimating injuries other than those in the tibia and in particular underestimating injuries in bones that are not subcutaneous, such as the femur or humerus. Although this is an interesting finding, it remains to be seen whether there is any clinical relevance associated with a change in GAC.

There are a few limitations to this study. First, the GAC may have been assigned by different physicians at different levels of training, with the possibility of a resident, fellow, or staff physician assigned the GAC. Moreover, the preoperative and intraoperative GAC may have been assigned by different physicians, although we do not have discrete data on which physician assigned which grade. These physicians may have had varying experiences managing orthopaedic trauma, and it remains unclear whether there are underlying surgeon factors that prompt a surgeon to classify an open fracture as they do, beyond the basic assessment of wound size. In addition, while all centers used standardized antibiotic protocols with the goal of administration of antibiotics within 1 hour, there were deviations within the protocol where patients either received inappropriate or delayed antibiotics (approximately 1% each). Unfortunately, owing to limitations in data collection, it was not possible to exclude or adjust for this as a potential confounding variable affecting the risk of SSI.

Moreover, in determining the relationship between GAC change and SSI, all patients with a change in GAC were pooled together because of data limitations. This meant that patients with changes from GAC I to GAC II were treated equivalently to patients who changed from GAC II to GAC III. This prevented us from analyzing each subgroup and could have understated or overstated the importance of misclassification in the development of a SSI. In addition, for the same reason, we are unable to compare patients on a more granular level. Moreover, quantifying the amount of bone, muscle, fascia, and skin debrided can be inexact, and given this limitation, the original study chose to follow an ordinal scale for these measurements. Because an upgraded score was nearly three times more likely than a downgraded one, it is possible that the model constructed to predict downgraded GA did not include sufficient events to generate meaningful statistical findings.

Finally, the GAC classification has many potential limitations because it is based primarily on the length of the skin wound and not the extent of muscle, arterial, or other soft-tissue injury. Alternative classifications, such as the Orthopaedic Trauma Association open fracture classification, which are designed to have more inclusive criteria for rating open fracture, are becoming more prevalent and popular.25 These alternate classifications may reduce the likelihood for change in classification of an open fracture wound between the emergency department and after definitive operative debridement.

5. Conclusion

When treating patients with open fracture wounds, surgeons should consider that when grading injuries according to the GAC and choosing the initial prophylactic antibiotic scheme or triaging care urgency, 12% may initially be misclassified. However, a misclassified GAC may not have a substantial impact on the overall risk of SSI. Fractures with any bone loss at presentation and those located in sites different than the tibia are more likely to be inappropriately underclassified. A focus on early antibiotic delivery, rather than the choice of antibiotics associated with a specific GAC classification, seems to be more important in preventing SSI. Future randomized controlled trials evaluating either limited or broad antibiotic coverage for open fractures in prevention of SSIs would build on the results from this study.

Appendix 1. The FLOW Investigators

Steering Committee: Mohit Bhandari (Chair, McMaster University), Gordon H. Guyatt (Co-Chair, McMaster University), Kyle J. Jeray (Co-Chair, Greenville Health System), Stephen D. Walter (McMaster University), Brad Petrisor (McMaster University), Emil H. Schemitsch (St. Michael's Hospital), Paul Tornetta III (Boston University Medical Center), Jeff Anglen (Eskenazi Health Services, Indiana University), Michael Bosse (Carolinas Health Care System), Susan Liew (The Alfred), and Parag Sancheti (Sancheti Institute for Orthopaedics and Rehabilitation)

Global Methods Center: Mohit Bhandari (principal investigator); Sheila Sprague (research methodologist); Paula McKay, Kim Madden, and Kerry Tai (project management); Diane Heels-Ansdell (statistical analysis); and Lisa Buckingham and Aravin Duraikannan (data management) (McMaster University)

United States Methods Center: Kyle J. Jeray (principal investigator), Stephanie L. Tanner, and Rebecca G. Snider (project management) (Greenville Health System)

Data Monitoring Committee: Douglas Altman (University of Oxford), Rajiv Gandhi (Toronto Western Hospital), and Markus Bischoff (McMaster University)

Adjudication Committee: Mohit Bhandari (McMaster University), Gregory J. Della Rocca (University of Missouri Health Care), Brad Petrisor (McMaster University), Kyle J. Jeray (Greenville Health System), and Emil H. Schemitsch (St. Michael's Hospital)

Participating Clinical Sites:

Canada:Hamilton Health Sciences—Brad Petrisor, Bill Ristevski, Krishan Rajaratnam, Dale Williams, Matthew Denkers, Drew Bednar, John Sadler, Desmond Kwok, Mohit Bhandari, Brian Drew, Ivan Wong, Paula McKay, Kim Madden, and Kerry Tai. St. Michael's Hospital—Jeremy A. Hall, Michael D. McKee, Emil H. Schemitsch, Henry Ahn, Daniel Whelan, James Waddell, Timothy Daniels, Earl Bogoch, Aaron Nauth, Milena R. Vicente, and Jennifer T. Hidy. London Health Sciences Centre—David Sanders, Abdel-Rahman Lawendy, Kevin Gurr, Timothy Carey, Chris Bailey, Mark Macleod, Debra Bartley, and Christina Tieszer. Queen Elizabeth II Health Sciences Centre—Chad Coles, Ross Leighton, C. Glen Richardson, Michael Biddulph, Michael Gross, Michael Dunbar, J. David Amirault, David Alexander, Catherine Coady, Mark Glazebrook, David Johnston, William Oxner, J. Andrew Trenholm, Gerald Reardon, Kelly Trask, and Shelley MacDonald. The Ottawa Hospital Civic Campus—Steven Papp, Wade Gofton, Allan Liew, Stephen Kingwell, Joseph O'Neill, Garth Johnson, Eugene Wai, and Julia Foxall. Vancouver General Hospital—Henry M. Broekhuyse, Peter J. O'Brien, Piotr A. Blachut, Kelly A. Lefaivre, and Raman Johal. Hôpital du Sacré-Coeur de Montréal—Stéphane Leduc, G. Yves Laflamme, Pierre Beaumont, Michel Malo, Benoit Benoit, Dominique Rouleau, Pierre Ranger, Julie Fournier, and Karine Tardif. McGill University Health Centre—Rudy Reindl, Greg Berry, Edward Harvey, William Fisher, Mark Burman, Paul Martineau, Eric Lenczner, Robert Marien, Robert Turcotte, Michael Tanzer, Max Talbot, Peter Jarzem, Mike Weber, and Fiona Houghton. University of British Columbia/Fraser Health Authority—Robert McCormack, Kelly Apostle, Dory Boyer, Farhad Moola, Bertrand Perey, Trevor Stone, Darius Viskontas, H. Michael Lemke, Mauri Zomar, Karyn Moon, and Raely Moon. Sunnybrook Health Sciences Centre — Hans Kreder, Richard Jenkinson, David Stephen, Markku Nousiainen, Terry Axelrod, Veronica Wadey, Michael Ford, Joel Finkelstein, Richard Holtby, Robin Richards, Sebastian Rodriguez-Elizalde, Diane Nam, Albert Yee, Patrick Henry, John Murnaghan, Harsha Malempati, Julian Sernik, Tim Dwyer, Katrine Milner, Monica Kunz, Melanie MacNevin, Wesley Ghent, Fathima Adamsahib, Ria De Gorter, and Michelle Arakgi. Winnipeg Health Sciences Centre—Ted V. Tufescu, Brad Pilkey, Chris Graham, Laurie Barron, Allan Hammond, Nigar Sultana. Queen's University—Ryan T. Bicknell, David Pichora, Aaron Campbell, and Fiona Howells. Centre de Recherche du Centre Hospitalier Universitaire de Sherbrooke (CRCHUS)—Annie Deshaies, Frédéric Balg, François Cabana, Rejean Dumais, Jean-François Joncas, Marc-André Magalhaes-Grave, Nicolas Patenaude, Bernard LaRue, Stéphane Ricard, Chantal Théorêt, François Vézina, Amy Svotelis, and Jennifer Downey. Hôpital de l'Enfant-Jésus—Stéphane Pelet, Jean Lamontagne, Luc Bédard, Alexandre Denault, Pierre Lavallée, Luc Petitclerc, Bernard Laliberté, Martin Bédard, Marie-Eve Roger, Luc Lemire, Hélène Côté, Linda Lépine, and Pascale Lévesque-Bernier. United States:Greenville Health System—Kyle J. Jeray, J. Scott Broderick, David R. Goetz, Thomas M. Schaller, Scott E. Porter, Michael L. Beckish, John D. Adams Jr, Benjamin B. Barden, Grant W. Bennett, David M. Conner, Aaron T. Creek, Melissa M. Earles, Stephen H. Finley, Jonathan L. Foret, Garland K. Gudger Jr, Richard W. Gurich Jr, Austin D. Hill, S. Matthew Hollenbeck, Lyle T. Jackson, Benjamin S. Koch, Kevin K. Kruse, Wesley G. Lackey, Justin W. Langan, Julia Lee, Lauren C. Leffler, Michael J. Maughon, Jr, S. Brennan McClure, Timothy J Miller, R. Lee Murphy Jr, Lawrence K. O'Malley, Dustin M. Price, Lorra M. Sharp, J. Adam Smitherman, John A. Tanksley Jr, Erick G. Torres, Dylan J. Watson, Scott T. Watson, Stephanie L. Tanner, Rebecca G. Snider, Shea A. Bielby, Lauren A. Nastoff, and Robert J. Teasdall. United States Army Institute of Surgical Research and Brooke Army Medical Center—Joseph Hsu, Katherine M. Bedigrew, Tod Gerlinger, Dave Brown, Joseph Alderete, Kevin Kirk, Mickey Cho, Anthony Johnson, Raymond Topp, Damian Rispoli, James Ficke, Eric Ritchie, Anthony Beardmore, Siraj Sayeed, Michael Charlton, Kristen Walick, Dmitry Tuder, Greg Maytok, Travis Burns, Donald Gajewski, Warren Kactmas, Ramnov Andreson, Patrick Osborn, Michael Connally, Donna Lopez, Mary Fan, Dennis Mann, Andrea Garza, and Rina L. Harman. Duke University Medical Center—Steven Olson, Robert Zura, Rachel Reilly, Prerana Patel, Claude T Moorman, Fraser Leversedge, Chard Harbour, Brian Brigman, David Ruch, Nikoletta Leontaritis, Michael Bolognesi, Shalini Ramasunder, Alison Toth, Allen Diane, Grant Garrigues, Dean Taylor, Richard C. Mather III, Kristoff Reid, Robert Lark, Samuel Adams, and Maria Manson. San Francisco General Hospital—Utku Kandemir, Saam Morshed, Murat Pekmezci, Richard Coughlin, Trigg McClellan, Meir Marmor, Eric Meinberg, Tigist Belaye, and Jonathan Kwong. Orthopaedic Associates of Michigan—Clifford B. Jones, James R. Ringler, Terrence J. Endres, David J. Bielema, Michael R. Jabara, Samuel G. Agnew, Debra L. Sietsema, and Jane E. Walker. University of Missouri Health Care —Gregory J. Della Rocca, Brett D. Crist, Yvonne M. Murtha, David A. Volgas, James P. Stannard, Linda K. Anderson, Kelly M. Sullivan, Lori Kramer Clark, Kathleen Markley, and Stacee Clawson. Hennepin County Medical Center—Andrew Schmidt, Patrick Yoon, Thomas Varecka, Matthew Karam, and Jerald R. Westberg. St. Louis University—Lisa K. Cannada, Jason Stoneback, Kevin Kuhn, Erik Nott, and Leslie Dillender. Eskenazi Health Services, Indiana University—Karl Shively, Brian Mullis, Janos Ertl, Ripley Worman, Jeffrey Anglen, Valda Frizzell, and Molly Moore. Miami Valley Hospital—Michael J. Prayson, David Nelles, Jason Vourazeris, Matthew Ross, Richard T. Laughlin, Joseph Cox, Roman Trimba, and Joy M. Bradford-Johnson. Lahey Clinic—Andrew J. Marcantonio, Michael Kain, Richard Wilk, Mark Lemos, Joshua Baumfeld, John Tilzey, Brian Jolley, and John Garfi. University of Pittsburgh Medical Center—Ivan Tarkin, Andrew Evans, Peter Siska, Lisa Blackrick, and Dana J. Farrell. University of Alabama at Birmingham—Emily Keener, Jason Lowe, William Min, Jeffrey Leary, Rena Stewart, David Volgas, Leslie Barnes, Nurit Shadmi, Matthew Robinson, Taylor Vlack, Kathryn Hornbuckle, Melanese Leonard, Nikia Hawkins Malone, Tanya Nix, and Jessica Goldstein. University of California Irvine—David Zamorano, Martin Tynan, Samuel Bederman, Nitin Bhatia, Arthur Kreitenberg, Bang Hoang, Deeba Pourmand, and Deanna Lawson. Scottsdale Healthcare—Anthony Rhorer, Brian Miller, Gilbert Ortega, Lori Wood, and Veronica Place

International:The Alfred (Australia)—Susan Liew, Harvinder Bedi, Ashley Carr, Andrew Chia, Hamish Curry, Steve Csongvay, Craig Donohue, Stephen Doig, Elton Edwards, Eugene Ek, Max Esser, Greg Etherington, Richard Freeman, Andrew Gong, Doug Li, Matthan Mammen, Russell Miller, Ash Moaveni, Mathias Russ, Lu Ton, Tom Treseder, Otis Wang, Zoe Murdoch, Claire Sage, and Adam Dowrick. Oslo University Hospital (Norway)—John Clarke-Jenssen, Frede Frihagen, Lars Nordsletten, Tor Nicolaysen, Hilde Apold, Petter Iversen, Are Stodle, Mette Andersen, Vera Halvorsen, Geir Hjorthaug, Anders Lippert, Ida Sletten, Ellen Langslet, Marius Molund, Asgeir Amundsen, Oliver Muller, Cathrine Aga, Torben Ianssen, Gunnar Flugsrud, Jonas Rydinge, Kim Hemlock, Jan Egil Brattgjerd, John Magne Hoseth, Bernhard Flatoy, Havard Furunes, Peder Bogsti, Guri Ekås, Gilbert Moatshe, Ali Al-Ashtari, Tore Fjalestad, Fredrik Nilsen, Morten Smedsrud, Anne Christine Brekke, Elise Berg Vesterhus, and Sissel Knuts. Sancheti Institute for Orthopaedics and Rehabilitation (India)—Parag Sancheti, Steve Rocha, Chetan Puram, Atul Patil, and Neelam Jhangiani. Highway Hospital (India)—Anil K. Rai and Kamal Narayan Rai. Jabade Hospital (India)—Vivek V. Jabade and Deepali Nassikars. Karne Hospital (India)—Narayan J Karne and Chetan Metha. Apollo Hospital Tondierpet (India)—A Navaladi Shankar. Apollo Hospital Greams Lane (India)—A Navaladi Shankar and R Saravana. Nirmal Hospital (India)—Ajay Gupta and Neeraj Jain. RLB Hospital and Research Centre (India)—Mahesh Bhatia, Vinod Arora, Vivek Tyagi, and Anoop Dubey. Popular Hospital (India)—Vinit Yadav and Rani Rai. Kolhapur Institute of Orthopaedics and Trauma (India)—Kiran M. Doshi and Arjun Patil


1. Schade AT, Khatri C, Nwankwo H, et al. The economic burden of open tibia fractures: a systematic review. Injury. 2021;52:1251–1259.
2. Ifesanya AO, Omololu AB, Ogunlade SO, Alonge TO. The burden of open fractures of the tibia in a developing economy. Nig J Plast Surg. 2010;6:32–39.
3. Patzakis MJ, Wilkins J. Factors influencing infection rate in open fracture wounds. Clin Orthop Relat Res. 1989;243:36–40.
4. Dellinger EP, Miller SD, Wertz MJ, et al. Risk of infection after open fracture of the arm or leg. Arch Surg. 1988;123:1320–1327.
5. Parker B, Petrou S, Masters JPM, et al. Economic outcomes associated with deep surgical site infection in patients with an open fracture of the lower limb. Bone Joint J. 2018;100:1506–1510.
6. Giannoudis PV, Harwood PJ, Kontakis G, et al. Long-term quality of life in trauma patients following the full spectrum of tibial injury (fasciotomy, closed fracture, grade IIIB/IIIC open fracture and amputation). Injury. 2009;40:213–219.
7. Gosselin RA, Roberts I, Gillespie WJ. Antibiotics for preventing infection in open limb fractures. Cochrane Database Syst Rev. 2004;1:CD003764.
8. Kumar G, Narayan B. Prevention of infection in the treatment of one thousand and twenty-five open fractures of long bones. Retrospective and prospective analyses. Class Pap Orthop. 2014;58:527–530.
9. Gustilo RB, Simpson L, Nixon R, et al. Analysis of 511 open fractures. Clin Orthop Relat Res. 1969;66:148–154.
10. Gustilo RB, Mendoza RM, Williams DN. Problems in the management of type III (severe) open fractures: a new classification of type III open fractures. J Trauma. 1984;24:742–746.
11. Yim GH, Hardwicke JT. The evolution and interpretation of the Gustilo and anderson classification. JBJS. 2018;100:e152.
12. Hoff WS, Bonadies JA, Cachecho R, Dorlac WC. East Practice Management Guidelines Work Group: update to practice management guidelines for prophylactic antibiotic use in open fractures. J Trauma Acute Care Surg. 2011;70:751–754.
13. Investigators F. A trial of wound irrigation in the initial management of open fracture wounds. N Engl J Med. 2015;373:2629–2641.
14. Borchardt RA, Tzizik D. Update on surgical site infections: the new CDC guidelines. JAAPA. 2018;31:52–54.
15. Peduzzi P, Concato J, Kemper E, et al. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49:1373–1379.
16. Lack WD, Karunakar MA, Angerame MR, et al. Type III open tibia fractures: immediate antibiotic prophylaxis minimizes infection. J Orthop Trauma. 2015;29:1–6.
17. Bhandari M, Jeray KJ, Petrisor BA, et al. A trial of wound irrigation in the initial management of open fracture wounds. N Engl J Med. 2015;373:2629–2641.
18. Chang Y, Bhandari M, Zhu KL, et al. Antibiotic prophylaxis in the management of open fractures: a systematic survey of current practice and recommendations. JBJS Rev. 2019;7:e1.
19. Obremskey W, Molina C, Collinge C, et al. Current practice in the management of open fractures among orthopaedic trauma surgeons. Part A: initial management. A survey of orthopaedic trauma surgeons. J Orthop Trauma. 2014;28:e198–e202.
20. Skaggs DL, Friend L, Alman B, et al. The effect of surgical delay on acute infection following 554 open fractures in children. JBJS. 2005;87:8–12.
21. Srour M, Inaba K, Okoye O, et al. Prospective evaluation of treatment of open fractures: effect of time to irrigation and debridement. JAMA Surg. 2015;150:332–336.
22. Ridgeon EE, Young PJ, Bellomo R, et al. The fragility index in multicenter randomized controlled critical care trials. Crit Care Med. 2016;44:1278–1284.
23. Roddy E, Patterson JT, Kandemir U. Delay of antibiotic administration greater than 2 hours predicts surgical site infection in open fractures. Injury. 2020;51:1999–2003.
24. Lin CA, O'Hara NN, Sprague S, et al. Low adherence to recommended guidelines for open fracture antibiotic prophylaxis. JBJS. 2021;103:609–617.
25. Roberts CS, Adams EL. The classification of open fractures: are we there yet? Injury. 2013;44:403–405.

open fracture; trauma; randomized trial; secondary analysis; surgical site infection; antibiotics

Copyright © 2022 The Authors. Published by Wolters Kluwer Health, Inc. on behalf of the Orthopaedic Trauma Association.