Background: Traditional recommendations hold that open fractures in both children and adults require urgent surgical débridement for a number of reasons, including the preservation of soft-tissue viability and vascular status as well as the prevention of infection. Following the widespread use of early administration of antibiotics, a number of single-institution studies challenged the belief that urgent surgical débridement decreases the risk of acute infection.
Methods: We performed a retrospective, multicenter study of open fractures that had been treated at six tertiary pediatric medical centers between 1989 and 2000. The standard protocol at each medical center was for all children to be given intravenous antibiotics upon arrival in the emergency department. The medical records of all children with open fractures were reviewed to identify the location of the fracture, the interval between the injury and the time of surgery, the Gustilo and Anderson classification, and the occurrence of acute infection.
Results: The analysis included 554 open fractures in 536 consecutive patients who were eighteen years of age or younger. The overall infection rate was 3% (sixteen of 554). The infection rate was 3% (twelve of 344) for fractures that had been treated within six hours after the injury, compared with 2% (four of 210) for those that had been treated at least seven hours after the injury; this difference was not significant (p = 0.43). When the fractures were separated according to the Gustilo and Anderson classification system, there were no significant differences in the infection rate between those that had been treated within six hours after the injury and those that had been treated at least seven hours after the injury. Specifically, these infection rates were 2% (three of 173) and 2% (two of 129), respectively, for type-I fractures, 3% (three of 110) and 0% (zero of forty-four), respectively, for type-II fractures, and 10% (six of sixty-one) and 2% (two of thirty-seven), respectively, for type-III fractures (p > 0.05 for all three comparisons).
Conclusions: In the present retrospective, multicenter study of children with Gustilo and Anderson type-I, II, and III open fractures, the rates of acute infection were similar regardless of whether surgery was performed within six hours after the injury or at least seven hours after the injury. The findings of the present study suggest that, in children who receive early antibiotic therapy following an open fracture, surgical débridement within six hours after the injury offers little benefit over débridement within twenty-four hours after the injury with regard to the prevention of acute infection.
Level of Evidence: Therapeutic Level III. See Instructions to Authors for a complete description of levels of evidence.
1 Division of Orthopedic Surgery, Childrens Hospital Los Angeles, MS# 69, 4650 Sunset Boulevard, Los Angeles, CA 90027
2 Hospital for Sick Children, 555 University Avenue, Toronto, ON M5G 1X8, Canada
3 Pediatric Orthopedic and Scoliosis Center, 3030 Children's Way, Suite 410, San Diego, CA 92123-4208
4 Children's Orthopaedics of Atlanta, 5545 Meridian Mark Road, Suite 250, Atlanta, GA 30342
5 The Children's Hospital of Philadelphia, 34th and Civic Center Boulevard, Wood Building, 2nd Floor, Philadelphia, PA 19104
In the introduction of their seminal study on the prevention of infection in patients with open fractures, Gustilo and Anderson stated that “There is universal agreement that open fractures require emergency treatment.”1
The use of débridement to help to prevent wound infection dates back to the time of Hippocrates (460-377 B.C.). Over the centuries, the belief that urgent surgical débridement helps to decrease the risk of infection has been a cornerstone of the treatment of open fractures1. In 1881, Carl Reyher found that débridement of open fractures decreased the risk of infection and mortality during the Franco-Prussian War2. To the best of our knowledge, only one scientific study has demonstrated that surgical débridement within six hours after the injury lowers the risk of acute infection. That study, performed by Friedrich in the pre-antibiotic era (in 1898), involved tissue-contamination experiments at the site of soft-tissue wounds in guinea pigs3.
Following the widespread use of early administration of antibiotics, a number of studies demonstrated no difference between the infection rates associated with fractures treated with early as opposed to late débridement1,2,4-10. Patzakis and Wilkins5, for example, reported the results of a large study in which the rate of acute infection for fractures that had been débrided less than twelve hours after the injury (7%; twenty-seven of 396) was the same as that for fractures that had been débrided more than twelve hours after the injury (7%; fifty of 708). In a retrospective study of 104 pediatric patients who had been managed at one institution, Kreder and Armstrong reported that the infection rate was 3% (one of forty) for fractures that had been treated within six hours after the injury and 2% (one of sixty-four) for those that had been treated at least seven hours after the injury10. This difference was not significant. However, those studies appear to have had limited influence on the orthopaedic community.
The purpose of the present study was to determine whether a delay of seven hours or more between the time of the injury and the time of surgical débridement influences the rate of acute infection at the site of an open fracture in a child. Tertiary referral centers often admit patients after substantial delays related to transfer, thus providing a population of children with open fractures for whom surgery is delayed for seven hours or more.
Materials and Methods
A retrospective chart review was performed to identify patients in whom an open fracture had been treated between 1989 and 2000 at one of six tertiary pediatric referral centers in North America. Five hundred and eighty-seven consecutive patients with 605 open fractures were identified. Patients with injuries that resulted in death were not included in the study. At some institutions, gunshot wounds are not treated with operative irrigation and débridement; therefore, this potentially confounding group was not included in the study. Patients for whom data regarding the interval between the injury and surgery were unavailable (forty-five), patients with incomplete follow-up (three), and patients with a delay of more than seventy-two hours before surgery (three) were excluded from further study. Patients were followed until there was radiographic and clinical evidence of union with absence or resolution of infection. Thus, 536 patients with 554 open fractures were included in this study.
The standard protocol at all participating institutions was for intravenous antibiotics to be administered to all children with open fractures upon arrival in the emergency department and then for intravenous antibiotics to be continued for at least twenty-four hours9. Antibiotic selection, wound closure, and fixation method were determined by the preference of the surgeon. The time between the injury and the surgical incision for débridement was calculated from transfer and hospital records. All times were rounded to the nearest hour.
Fracture groups were stratified according to the severity of soft-tissue damage as determined with use of the Gustilo and Anderson system for the classification of open fractures (Table I) as well as according to the time from the injury to definitive surgical treatment1,10. Subtypes of type-III fractures were not studied separately because of the limited number of patients with such fractures. The location of the fracture was recorded. A fracture was considered to be infected if any one of three conditions was met: (1) the patient had positive intra-operative cultures and was subsequently managed with antibiotics or additional surgical débridement, (2) the patient did not have positive cultures but was managed with subsequent surgery or antibiotics for the treatment of infection, or (3) the patient had a clinical diagnosis of infection. These criteria were intentionally stringent to make certain that we did not miss any cases of infection. Patients who had a single positive culture but had no clinical evidence of infection and no additional treatment were not considered to have an infection. Both deep and superficial infections were considered collectively. There were no known cases of late infection. Pin-site infections resulting from external fixators were not considered to be wound infections.
The demographic variables were evaluated with univariate analysis to determine the distribution of the data. The data were then stratified into three groups on the basis of the delay in treatment: six hours or less, seven to twenty-four hours, and twenty-five hours or more. Because only eight open fractures had been treated after a delay of twenty-five hours or more, the latter two groups (seven to twenty-four hours and twenty-five hours or more) were combined for statistical purposes. To determine the association between the rate of acute infection and the delay in treatment, we first stratified the fractures according to the Gustilo and Anderson classification system as type I, II, or III. Fisher's exact test was then performed to compare the rate of infection and the delay in treatment for type-I, II, and III fractures individually.
In addition, the Mantel-Haenszel chi-square test was performed to determine if the site of infection (upper extremity as opposed to lower extremity) was a confounding factor. Finally, a power analysis was performed to determine the number of patients needed to detect a 20% difference in the rate of infection with a p value of <0.05 and a power of 80%. All analyses were performed with use of the Statistical Analysis System (SAS version 8.2; SAS Institute, Cary, North Carolina). The level of significance was set at p < 0.05.
Five hundred and fifty-four open fractures in 536 children met the criteria for inclusion in the study. The mean age at the time of the injury was 8.8 ± 4.0 years (range, 0.2 to eighteen years). The distribution of fractures according to location is listed in Table II.
According to the system of Gustilo and Anderson, 302 fractures (55%) were classified as type I, 154 fractures (28%) were classified as type II, and ninety-eight fractures (18%) were classified as type III. The interval between the injury and definitive surgical treatment was six hours or less for 344 fractures, seven to twenty-four hours for 202 fractures, and twenty-five hours or more for eight fractures (Table III).
The overall infection rate was 3% (sixteen of 554) (Table IV). The infection rate was 3% (twelve of 344) for fractures that had been treated after a delay of six hours or less, compared with 2% (four of 210) for those that had been treated after a delay of seven hours or more; this difference was not significant (p = 0.43). When the fractures were stratified according to the classification system of Gustilo and Anderson, there was no significant difference in the rate of acute infection between type-I fractures that had been treated after a delay of six hours or less and those that had been treated after a delay of seven hours or more (p = 1.00). Similarly, for type-II and III fractures, there was no significant difference in the rate of infection between fractures that had been treated after a delay of six hours or less and those that had been treated after a delay of seven hours or more (p = 0.56 and 0.71, respectively). A power analysis revealed that 11,390 cases would be needed in order to detect a 20% difference in the infection rate.
An infection developed at the sites of four tibial or fibular fractures, three femoral fractures, five radial or ulnar fractures, one humeral fracture, one hand or metacarpal fracture, one foot or phalanx fracture, and one multiple lower extremity fracture that was not further characterized. The site of injury (the upper extremity as opposed to the lower extremity) was determined not to be a confounding factor affecting the rate of infection (p = 0.28).
A premise of modern orthopaedic care of open fractures has been that urgent surgical irrigation and débridement minimizes the risk of acute infection. The report on open fractures by Gustilo and Anderson is frequently cited in support of this belief1. That article concludes, “Open fractures require emergency treatment, including adequate débridement and copious irrigation.” This conclusion was not supported by the data in that study as the relationship between the surgical delay and the infection rate was not independently addressed. Gustilo and Anderson reported that the acute infection rate decreased from 10% (fourteen of 135) to 5% (twenty-four of 458) with the advent of four changes in their approach: (1) type-III wounds were treated with delayed primary closure, (2) internal fixation was no longer used, (3) open fractures were treated as emergencies, and (4) antibiotics were administered before surgery. There is no way to determine from that retrospective study which of the four changes were responsible for the observed improvement in the acute infection rate.
Subsequent clinical studies that have been performed during the antibiotic era have indicated that the timing of surgical débridement of open fractures may not play as critical a role in the prevention of acute infection4,7-14. Merritt, in a study of seventy patients with open fractures, concluded that “the time between injury and treatment in the emergency room was not correlated with infection rate, nor was the time between treatment in the emergency room and débridement in the operating room correlated with infection rate”.14 An interesting finding of that study was that there was little correlation between the bacterial counts in the first piece of tissue taken at the time of débridement and the development of infection, whereas there was a significant correlation between the bacterial count in the last piece of tissue taken at the time of débridement and the development of infection. Bednar and Parikh examined eighty-two open fractures of the lower extremity that had been caused by blunt trauma in adults, of which 76% (sixty-two) were débrided seven to twenty-four hours after the injury11. The overall rate of infection was 5% (four of eighty-two), with no increase in the rate of infection being observed in association with fractures that had been treated with delayed surgery.
We are aware of two previous reports on this subject that have been limited to children. Kreder and Armstrong, in a review of fifty open tibial fractures in children, reported that a delay in surgical treatment of seven hours or more was associated with an infection rate of 25% (two of eight) whereas a delay of six hours or less was associated with an infection rate of 12% (five of forty-two)10. Clearly, these numbers are too small to be meaningful as one fewer infection in the group that was treated after a delay of seven hours or more would have made the infection rates nearly equal. Wilkins and Patzakis, in a retrospective study of 104 children with open fractures that had been treated at a single institution, reported that the infection rate was 3% (one of forty) for fractures that had been treated within six hours after the injury and 2% (one of sixty-four) for those that had been treated at least seven hours after the injury8. This difference was not significant (p = 0.77).
Clinical studies have demonstrated that the timing of antibiotic administration influences the risk of infection. Patzakis and Wilkins, in a review of 1025 open fractures, reported that the infection rate was 4.7% (seventeen of 364) when antibiotics had been started within three hours after the injury and 7.4% (forty-nine of 661) when antibiotics had been started four hours or more after the injury5. They concluded that “the single most important factor in reducing the infection rate was the early administration of antibiotics.”
The limitations of the present study are inherent to the retrospective study design. Surgical delay is common at tertiary referral centers because of the transfer of patients from other medical facilities. A prospective, randomized trial involving surgical delay in the care of children with open fractures would be unethical and impractical. The present study is not of sufficient size for us to conclude that there is no difference in the infection rates between fractures that are treated within six hours after the injury and those that are treated seven hours or more after the injury. Similarly, because of the limited numbers of patients in the present study, no factors other than time, infection, and Gustilo and Anderson classification were evaluated. Numerous other factors of interest, such as the type of fixation, the method of wound closure, and the type of antibiotics, to name a few, could not be evaluated in the present study because of its limited size.
We would like to state clearly that the present study only evaluated the effect of surgical delay on the rate of infection following open fractures, which is one of many factors that may be influenced by the timing of surgery. Open fractures may require emergent surgical treatment for reasons other than the prevention of infection, such as the preservation of soft-tissue viability or vascular status. The findings of the present study suggest that, in children who receive early antibiotic therapy following an open fracture, surgical débridement within six hours after the injury offers little benefit over débridement within twenty-four hours after the injury with regard to the prevention of acute infection. ▪
The authors did not receive grants or outside funding in support of their research or preparation of this manuscript. They did not receive payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.
Investigation performed at Childrens Hospital Los Angeles, Los Angeles, California
1. , Anderson JT. Prevention of infection in the treatment of one thousand and twenty-five open fractures of long bones: retrospective and prospective analyses. J Bone Joint Surg Am. 1976;58: 453-8.
2. , Wangensteen SD. Carl Reyher (1846-1890), great Russian military surgeon: his demonstration of the role of débridement in gunshot wounds and fractures. Surgery. 1973;74: 641-9.
3. . Die aseptische Versorgung frischer Wunden. Arch Klin Chir. 1898;57: 288-310.
4. , Steyers C, Stoddard M. Infections in open fractures of the hand. J Hand Surg [Am]. 1991;16: 108-12.
5. , Wilkins J. Factors influencing infection rate in open fracture wounds. Clin Orthop. 1989;243: 36-40.
6. , Fry DE, Polk HC Jr. Infectious morbidity in extremity fractures. J Trauma. 1986;26: 757-61.
7. , Kautz SM, Kay RM, Tolo VT. Effect of delay of surgical treatment on rate of infection in open fractures in children. J Pediatr Orthop. 2000;20: 19-22.
8. , Patzakis M. Choice and duration of antibiotics in open fractures. Orthop Clin North Am. 1991;22: 433-7.
9. , Simpson L, Nixon R, Ruiz A, Indeck W. Analysis of 511 open fractures. Clin Orthop. 1969;66: 148-54.
10. , Armstrong P. A review of open tibia fractures in children. J Pediatr Orthop. 1995;15: 482-8.
11. , Parikh J. Effect of time delay from injury to primary management on the incidence of deep infection after open fractures of the lower extremities caused by blunt trauma in adults. J Orthop Trauma. 1993;7: 532-5.
12. , Riggins RS, Lawrence RM, Hoeprich PD, Huston AC, Harrison JA. Treatment of open fractures: a prospective study. J Trauma. 1983;23: 25-30.
13. , Beaupre LA, Jones CA, Dulai SK, Weber DW. The effect of time to definitive treatment on the rate of nonunion and infection in open fractures. J Orthop Trauma. 2002;16: 484-90.
14. . Factors increasing the risk of infection in patients with open fractures. J Trauma. 1988;28: 823-7.
15. . Management of the multiply injured child. In: Beaty JH, Kasser JR, editors. Rockwood and Wilkin's fractures in children. 5th ed. Philadelphia: Lippincott Williams and Wilkins; 2001. p 83.