Supplemental on-line material:

Author / Year / Type / n / Duration of study / Study Findings / Conclusion
Prospective randomized studies
Patzakis et al. (20) / 1974 / Prospective randomized trial / 310 patients with 333 open fractures, 255 caused by direct trauma vs. gunshot / 12 months / Cephalothin (10 days of first-generation cephalosporin) more effective vs. penicillin-streptomycin and vs. placebo (2.3% vs. 9.7% and 13.9%; p<0.05) in preventing infection of open fractures. 62% and 20% of pathogenic organisms resistant to penicillin-streptomycin and cephalothin, respectively. / Cephalothin is an effective antibiotic for prophylactic therapy of open fractures due to direct trauma. Only broad-spectrum antibiotics that can be expected to be effective against coagulase-positive Staphylococcus aureus should be selected and guided by susceptibility tests to detect resistant organisms.
Patzakis et al. (13) / 2000 / Prospective, double-blind randomized trial / 171 patients with 203 open fractures / 12 months / No significant difference in infection rates with ciprofloxacin vs. combination cefamandole + gentamicin for Types I & II fractures (5.8% vs. 6%, p=1.000); but ciprofloxacin alone associated with higher rates of infection in Type III fractures (31% vs. 7.7%, p=0.079). / Single-agent antibiotic therapy with ciprofloxacin is effective in treatment of Type I & II open fracture wounds. However, cannot recommend ciprofloxacin alone for Type III wounds. Can possibly use fluoroquinolones in combination therapy, specifically as an alternate to an aminoglycoside.
Dellinger et al. (21) / 1988 / Prospective, double-blind randomized trial / 248 patients with open long bone fractures / 2 years / No difference in infection rates with 1 or 5 days of cefonicid vs. 5 days cefamandole (13%, 12%, 13%). / A brief course of antibiotic administration is not inferior to a prolonged course of antibiotics for prevention of postoperative fracture-site infections.
Braun et al. (22) / 1987 / Prospective, double-blind randomized trial / 100 patients with open fractures / n/a / Reduction in infection rate with 10 days cloxacillin vs. placebo (4% vs. 24%, p<0.05). / In the treatment of open fractures with operative fixation, the use of a narrow-spectrum antibiotic active against staphylococci can reduce the frequency of infections.
Bergman (23) / 1982 / Prospective randomized trial / 90 patients with various open fractures (55 Grade I, 35 Grade II-III) / 30 months / No significant difference in superficial infections between 2 days of dicloxacillin vs. benzyl penicillin vs. placebo. MRSA predominantly cultured from patients receiving penicillin and dicloxacillin. Gram negative bacteria were infrequent. No significant difference in deep infections between antibiotic treatment groups but significant difference compared to placebo (p<0.01). / There is not a clear-cut superiority for either of the two antibiotics used. It might be stated that penicillin seems to be better than no antibiotic at all and that the use of dicloxacillin is a further improvement. There seems to be no need for broad spectrum antibiotics, which might cause an ecological disturbance.
Saveli et al. (24) / 2013 / Prospective randomized trial / 130 patients with open fractures / 28 months / No significant difference in SSI rates for cefazolin vs. vancomycin + cefazolin given until 24 hrs after surgical intervention (15% vs. 19%, p=0.62). Nasal colonization of methicillin sensitive S. aureus and MRSA among the sample were 20% and 3%, respectively. S. aureus caused 55% of the deep incisional/organ space SSI (18% MRSA). A significantly higher rate of MRSA SSI was observed among MRSA carriers compared with non-carriers (33% vs.. 1%, respectively, p=0.003). / S. aureus nasal colonization in trauma patients with open fractures is similar to that of the general community. The addition of vancomycin to standard antibiotic prophylaxis was found safe, but its efficacy should be evaluated in a larger multi-institutional trial.
Benson et al. (25) / 1983 / Prospective, double-blind randomized trial / 82 patients with open long bone fractures / n/a / 5 days of clindamycin (2/34) as effective as cefazolin (3/42) in open fracture infection rate. / Infection rates not different with time of wound closure, cefazolin vs. clindamycin, and internal fixation.
Gustilo, Anderson (8) / 1976 / Prospective, non-randomized (and retro-spective) / 352 patients with open long bone fractures (81% Types I/II and 19% Type III) / 5 years / Oxacillin + ampicillin IV for 3 days after surgery or secondary wound closure +/- gentamicin if wound “severely contaminated”. 70.3% with contaminated wound by positive culture either on admission or at wound closure. Eight (2.4%) became infected. No antibiotic comparison performed. / Cephalosporins (cephalothin and cefazolin) appeared to be the most effective antibiotics for prophylaxis after open fractures – effective against all gram-positive organisms including Staphylococcus coagulase positive and negative, and the majority of E. coli, Klebsiella, and Proteus, excluding Pseudomonas. We did not find that anaerobes were particularly significant, presumably due to unsatisfactory culture technique.
Carsenti-Estesse et al. (10) / 1999 / Double-blind, placebo-controlled, randomized trial / 616 patients from 43 centers with grade I/II open leg fractures / 5 years / No difference in infection between pefloxacin (IV, single dose) vs. combination 2 days IV cefazolin + 3 days oral oxacillin for Grade I/II fractures at 3 months (6.6% vs. 8%, p=0.51). Protocol included anaerobic prophylaxis with penicillin G (2 days) or metro/ornidazole. Gram-positives more prevalent in infection isolates from pefloxacin vs. gram-negatives in combination group. Most cases were nosocomial bacteria. / No difference between pefloxacin vs. cefazolin + oxacillin. Infecting species were those escaping the spectrum of prescribed antimicrobrial prophylaxis, and most infections isolates were nosocomial bacteria.
Longitudinal cohort, retrospective and in vitro studies
Patzakis, Wilkins (14) / 1989 / Longitudinal cohort / 1104 open fractures / 2 years / Cephalothin alone (5.6%) more effective than penicillin + streptomycin (10%), and equally effective as cefamandole + tobramycin (4.5%) in preventing open fracture infection. / Recommend starting broad-spectrum antibiotics as soon as possible after injury and continue for 3 days.
Dunkel et al. (26) / 2013 / Retrospective case-control / 1492 patients with open fractures / 14 years / No difference with 1 day of antibiotics vs. 2-3 days (odds ratio (OR) 0.6 (95% CI 0.2 to 2.0)), 4-5 days (OR 1.2 (95% CI 0.3 to 4.9)), or >5 days (OR 1.4 (95% CI 0.4 to 4.4)) in infection risk for open fractures. Similar results with analysis of Grade III fractures only. / Infection in open fractures is related to the extent of tissue damage but not to the
duration of prophylactic antibiotic therapy. Even for grade III fractures, a one-day course of prophylactic antibiotics might be as effective as prolonged prophylaxis.
Holtom et al. (27) / 2000 / In vitro / Fluoroquinolones inhibit osteoblast activity and fracture healing in vitro; yet to be demonstrated in clinical trials. / In vitro and animal (Huddleston) studies have shown fluoroquinolones may inhibit fracture healing. Further in vivo studies are advocated to evaluate the relevance of the in vitro cytotoxicity reported here to bone healing in orthopaedic patients.

1