Antibiotics and Facial Fractures: Evidence-Based Recommendations Compared with Experience-Based Practice

Abstract

Efficacy of prophylactic antibiotics in craniofacial fracture management is controversial. The purpose of this study was to compare evidence-based literature recommendations regarding antibiotic prophylaxis in facial fracture management with expert-based practice. A systematic review of the literature was performed to identify published studies evaluating pre-, peri-, and postoperative efficacy of antibiotics in facial fracture management by facial third. Study level of evidence was assessed according to the American Society of Plastic Surgery criteria, and graded practice recommendations were made based on these assessments. Expert opinions were garnered during the Advanced Orbital Surgery Symposium in the form of surveys evaluating senior surgeon clinical antibiotic prescribing practices by time point and facial third. A total of 44 studies addressing antibiotic prophylaxis and facial fracture management were identified. Overall, studies were of poor quality, precluding formal quantitative analysis. Studies supported the use of perioperative antibiotics in all facial thirds, and preoperative antibiotics in comminuted mandible fractures. Postoperative antibiotics were not supported in any facial third. Survey respondents (n = 17) cumulatively reported their antibiotic prescribing practices over 286 practice years and 24,012 facial fracture cases. Percentages of prescribers administering pre-, intra-, and postoperative antibiotics, respectively, by facial third were as follows: upper face 47.1, 94.1, 70.6; midface 47.1, 100, 70.6%; and mandible 68.8, 94.1, 64.7%. Preoperative but not postoperative antibiotic use is recommended for comminuted mandible fractures. Frequent use of pre- and postoperative antibiotics in upper and midface fractures is not supported by literature recommendations, but with low-level evidence. Higher level studies may better guide clinical antibiotic prescribing practices.

Approximately 3 million individuals suffer craniofacial trauma in the United States on a yearly basis,[1] and approximately 50% of all wounds presenting to emergency rooms involve the head and neck.[2] In 2007, facial fractures accounted for more than 400,000 emergency department admissions.[3] Surgical intervention is often necessary in the management of craniomaxillofacial fractures, and poses a significant public health burden in terms of comorbidity and financial cost.[4,5,6]

The prevention of surgical site infections is a major focus of The Joint Commission Centers for Medicare and Medicaid Services Surgical Care Improvement Project (SCIP),[7] and the efficacy of prophylactic antibiotics has been proven in multiple clinical trials.[8] In the modern era of managed health care, it has become common practice for hospitals to closely monitor and even restrict antibiotic use in surgical patients. Risks of antibiotic treatment, uncertainty regarding the efficacy of antibiotics in specific scenarios, the possibility of antibiotic resistance, prescriber inattention to antibiotic course, and cost containment are all commonsense justification for greater scrutiny of surgical antibiotic prescribing practices.[9,10,11]

Despite increased regulation in health care, physicians maintain great autonomy and individuality in clinical practice. Antibiotics can be administered preoperatively (i.e., from the time of injury or presentation to the time of surgery), perioperatively (i.e., immediately before surgery and continuing through the procedure, but not more than 24 hours postprocedure, often called “prophylactic” antibiotics), or postoperatively (i.e., continuing past the perioperative period). Surgical antibiotic prescribing practices remain largely dependent on surgeon choice, which, although based on personal experience and surgical training, should be supported by objective evidence. However, when looking to the medical literature for guidance with regard

to antibiotic prescribing practices for specific scenarios, there is little available data and controversy abounds for even the most basic questions regarding surgical antibiotic administration (Figure 1). For example, even the efficacy of perioperative antibiotic timing, a major component of SCIP guidelines, has recently been called into question.[7,8,12] Studies generally focus on perioperative antibiotic administration, but do not address the utility of preoperative or postoperative antibiotic use, nor the choice of antibiotic in specific situations.

With regard to antibiotic prescribing practices in craniofacial fracture surgery, there are no current standard recommendations. In extrapolating data from recommendations for orthopedic procedures involving fractures, there is suggested benefit to perioperative antibiotics in open fracture repair, and, in extrapolating from head and neck oncologic procedures, for clean-contaminated procedures that involve an incision through the oral or pharyngeal mucosa.[13] If perioperative antibiotics are used, the first dose should be administered less than 60 minutes before surgical incision, or between 60 and 120 minutes of incision if vancomycin or clindamycin is used. Antibiotic duration should be less than 24 hours without continuation beyond this point. For head and neck procedures not involving mucosal incisions, perioperative antibiotics are not recommended, yet this recommendation conflicts with orthopedic fracture recommendations where antibiotics may be indicated despite skin-only incisions. Preoperative and postoperative antibiotics are not endorsed, but, again, are not specific to craniofacial fracture management.

Unique situations in the management of craniofacial fractures, such as contamination of fracture sites from the sinuses, exposure of fractures to intraoral bacteria from mucosal tears, and delay in fracture management, intuitively suggest that there may be benefit to preoperative and prolonged postoperative antibiotic administration in craniofacial fractures. The antibiotic prescribing practices of craniofacial surgeons are largely unknown, and may conflict with broader surgical antibiotic prescribing recommendations.

The purpose of this study was to compare evidence-based literature recommendations regarding antibiotic administration in operative craniofacial fracture repair with expert-based practice. Further resolution with regard to preoperative, perioperative, and postoperative antibiotic time points, as well as fracture location in the craniofacial skeleton would be useful to guide both clinical practice and identify areas where research efforts would be beneficial.

Methods

A systematic literature review was performed in June 2013 using Medline, Embase, PubMed and Cochrane databases to identify published studies evaluating the use of antibiotics in craniofacial trauma including the upper, middle, and lower thirds of the craniofacial skeleton. Search terms included “frontal sinus,” “nasal bone,” “zygoma,” “orbit,” “mandible,” “fracture,” “antibiotics,” “prophylaxis,” and “facial fracture” alone and in combination. Included studies were limited to the English language, and related articles were used to broaden the search. Identified abstracts and included studies were independently evaluated by three reviewers for inclusion or exclusion based on study design, study population, and indications for antibiotics. Studies were excluded if two out of three reviewers concluded that they did not meet inclusion criteria.

Data from selected studies were tabulated, and grouped according to both fracture area addressed (upper, middle, and/or mandible) and time point of antibiotic administration. Preoperative antibiotics were defined as antibiotics administered from the time of presentation, but before surgical intervention. Perioperative antibiotics were defined as antibiotics administered at the time of surgery, but not continuing for longer than 24 hours postoperatively. Postoperative antibiotics were defined as antibiotics administered beyond the 24 hours postoperative time point. Additional extracted data included first author, year of publication, study population characteristics, study design, number of patients, indications for antibiotics, and choice of antibiotics. Included studies were graded from Levels I–V according to the American Society of Plastic Surgery (ASPS) Evidence Rating Scales.[14] Level V studies were included.[15] Evidence rating was used to make grading recommendations for antibiotic prescribing practices according to ASPS Scale for Grading Recommendation guidelines.[16]

Expert opinions were garnered during the Advanced Orbital Surgery Symposium, held on May 3–5, 2012, in Baltimore, Maryland, in the form of surveys evaluating senior surgeon clinical antibiotic prescribing practices by facial third. Data queried included first, second, and third choices of antibiotic, time points of administration (pre-, peri-, and/or postoperative), and duration of postoperative antibiotic use, if applicable. Statistical evaluation of prescriber practices was performed using twosided Student t-tests. Results were considered significant at a p-value < 0.05.

Results

Systematic literature review identified 44 studies from eight countries addressing antibiotics and facial fracture management (Figure 2).[1,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59] Extracted study data are presented in

Table 1. Summary of included studies grouped by ASPS level of evidence and facial third.

Abbreviations: CCSs, case-control studies; RCTs, randomized controlled trials.

. Overall, studies were of poor quality, precluding formal quantitative analysis; 29.5% (n ¼ 13 were Level I or II. Most studies addressed fractures of the mandible (n ¼ 27, 61.3%), followed by midface (n ¼ 20, 45.5%), and then upper face (n ¼ 7, 15.9%). Four studies (9%) addressed midface and mandible fractures, while three studies (6.8%) addressed all three fracture regions. Level of evidence increased with inferior fracture location, with 14.3% (n ¼ 1), 20.0% (n ¼ 4), and 37.0% (n ¼ 10) of studies reporting Levels I and II evidence for upper, middle, and mandible fractures, respectively. Penicillins, cephalosporins, and clindamycin were the most commonly prescribed antibiotics, but no determination could be made regarding superiority of any antibiotic at any time point.

Studies supported the use of perioperative antibiotics in all facial thirds, especially if mucosal incisions were used (Grade A recommendation).[31,32,47] In contrast, preoperative antibiotics were not recommended for upper and midface fractures (Grade C recommendation),[1,22,42,56] while preoperative antibiotic use was supported for comminuted mandible fractures (Grade A recommendation).[21,22,23,29,50] Postoperative antibiotics were not recommended for upper and midface fractures (Grade C recommendations)[20,27,41] and were associated with increased morbidity in upper face fractures in one Level IV study.[41] The administration of postoperative antibiotics was also not supported in mandible fractures, even if comminuted (Grade A recommendation).[22,23,24,26,31,34,37] There was inconsistent low-level evidence suggesting benefit from prescribing prophylactic antibiotics in patients with premorbid acute or chronic sinusitis with midface fractures (Grade D recommendation).[17,40,43,45,51,53,56,57]

Survey respondents (n ¼ 17) cumulatively reported their antibiotic prescribing practices over 286 practice years, 24,012 facial fracture cases, three countries (United States [n ¼ 15], Canada [n ¼ 1], and Germany [n ¼ 1]), and 13 institutions. Cefazolin and clindamycin were most commonly prescribed in all situations, while vancomycin, metronidazole, and piperacillin/tazobactam were least commonly prescribed (

Table 2. Expert percentage first choice antibiotic use by time point and facial fracture region.

Note: Metronidazole, vancomycin, and piperacillin/tazobactam were not chosen as a first choice antibiotic by any survey respondent.

). In contrast to literature recommendations, percentages of prescribers administering pre-, intra-, and postoperative antibiotics, respectively, by facial third were as follows: upper face 47.1, 94.1, 70.6%; midface 47.1, 100, 70.6%; and mandible 68.8, 94.1, 64.7% (

Table 3. Expert frequency of antibiotic use by facial fracture region and time point.

). For those prescribing postoperative antibiotics, average duration for upper face, midface, and mandible fractures was 3.7, 4.0, and 4.6 days, respectively (range 1–7 days in each facial third). There were no significant differences between prescribing practices by facial third (all time point comparison p-values > 0.22).

Discussion

Overall, prescriber practice differed markedly with literature recommendations with the exception of perioperative antibiotic administration. These differences and their implications for both clinical practice and study design can be conceptualized by evaluating whether study quality and clinical knowledge are either good or poor for a given clinical scenario (Figure 1). The importance of perioperative antibiotic administration, especially in clean-contaminated procedures,[8,13,60] such as procedures involving oral incisions, is well supported in the literature (Grade A recommendation),[31,32,47] and survey respondents indicated their practices reflect this in all facial thirds. For mandible fractures, preoperative antibiotics are supported for compound mandible fractures (Grade A recommendation).[21,22,23,29,50] This may be the only other scenario where surgeon practice was congruous with literature recommendations as reflected in higher reported rates of preoperative antibiotic use in mandible fractures as compared with midand upper face fractures, though these differences were not significant. Unfortunately, our survey did not query more specific attributes (i.e., open/ closed, displaced/nondisplaced, etc.) of fractures in each region, and whether clinicians alter their prescribing practices in response to these fracture features cannot be addressed by our study design.

The incongruity of literature recommendations and prescriber practice identified in most antibiotic prescribing scenarios is due to both poor prescriber adherence to good literature, and clinicians practicing in complex clinical scenarios where little appropriate literature exists (Figure 1). Overall, studies addressing antibiotic use were of poor quality, especially for upper and midface fractures, and their results were often not presented in clinically useful ways.[21] In these situations, practitioners likely revert to the default of prescribing antibiotics, as indicated by high overall rates of preantibiotic and postantibiotic administration in upper and midface fractures despite literature recommendations to the contrary (Grade C recommendations).[1,20,22,27,41,42,56] We attempted to make these data more clinically approachable by clarifying and organizing data presentation for all reviewed studies.

In contrast, in scenarios where high-quality studies have essentially answered clinical questions, prescriber incongruity is likely due to disagreement with specifics of study design/study population, or ignorance of existing literature. This may be the case for mandible fractures, where the literature is overall of higher quality. The identified literature does not support continued postoperative antibiotics (Grade A recommendation),[22,23,24,26,31,34,37] yet 64.7% of practitioners say they administer postoperative antibiotics for an average of 4.6 days postoperatively. Similarly, practitioners may not prescribe preoperative antibiotics in compound mandible fractures despite strong literature evidence to the contrary (Grade A recommendation).[21,22,23,29,50]

Our results highlight several areas where further research is clearly warranted. The issue of premorbid acute or chronic sinusitis and risk for orbital cellulitis in patients with untreated midface fracture was addressed in multiple Level V studies.[17,40,43,45,51,53,56,57] Most authors felt that patients with pre-existing sinusitis and midface fractures should be treated with preoperative antibiotics in the hope of reducing the risk of orbital cellulitis, but this is unproven, and the number needed to treat to prevent orbital cellulitis is unknown. This question could be answered in prospective fashion. Moreover, studies addressing antibiotic use and midface fractures are few in number and of poor quality. Practice recommendations would benefit from better quality studies addressing midface fractures. The same can be said for upper face fractures, where studies were least frequent and of overall lowest quality, with no study reporting Level I evidence.

The question arises, what makes for a “good” study addressing antibiotics and facial fractures? In addition to a prospective, randomized design that will specifically answer a clinical question and a study population with enough power to generate significance, it is our opinion that “good” studies in this field need to possess additional qualities specific to craniofacial trauma. Given the complexity of the craniofacial skeleton and the complexity of managing craniofacial fractures in patients with other traumatic injuries, studies need to have greater specificity with regard to fracture type, number of areas fractured, displacement, comminution, and exposure.[61,62] No study has explored antibiotic efficacy between simple (i.e., one fractured region) or complex (i.e., multiregion or panfacial) craniofacial fractures, and differences could be found between these populations. In addition, craniofacial fractures are frequently treated in polytrauma patients, who can receive antibiotics for several reasons unrelated to their craniofacial injuries.[27] Studies must carefully address this issue, and should control for several confounding variables, such as time from injury to operative intervention,[27,33] the presence of associated cerebrospinal fluid leak,[41] and the presence of basilar skull fractures.63 Overall injury severity should be considered in regression modeling with control of predictor outcome variables by widely accepted global injury scoring systems, such as the injury severity score, or Glasgow coma scale.[63,64] Finally, the timing, choice, and dose of evaluated antibiotics, which were poorly reported in upper and midface studies, should be clearly stated and their administration rigorously controlled. This is best achieved through prospective study design.

Although literature reviews and practitioner surveys have inherent drawbacks, we sought to minimize issues associated with these study designs. The literature was systematically and thoroughly evaluated, and results from studies were tabulated using our anatomical and time point schemes to facilitate comparisons and maximize extraction of study data to clinical practice. Similarly, survey respondents were asked to provide data relevant to clinical practice for the same locations and time points, and were all experts in craniofacial trauma. The study was not designed as a meta-analysis of specific antibiotic superiority for any time point/fracture location combination, and, indeed, we found that data of this resolution would be too sparse to perform meta-analyses in most situations, with the possible exception of preoperative and postoperative antibiotic administration in mandible fractures.[21]

Conclusion

Frequent use of preand postoperative antibiotics in upper and midface fractures is not supported by literature recommendations, but with low-level evidence. Prophylactic antibiotic use was evaluated by higher level of evidence studies for mandible fractures: preoperative antibiotic use in comminuted mandible fractures is supported, but postoperative antibiosis in mandible fractures is not. Well designed and higher level of evidence studies, especially for upper and midface fractures, may better guide clinical antibiotic prescribing practices.