- Email: [email protected]
Are Prophylactic Oral Antibiotics Indicated for the Treatment of Intraoral Wounds?
TRAUMA/BEST AVAILABLE EVIDENCE
Are Prophylactic Oral Antibiotics Indicated for the Treatment of Intraoral Wounds? Dustin G. Mark, MD Eric J. Granquist, DMD
From the Department of Emergency Medicine (Mark) and Department of Oral and Maxillofacial Surgery (Granquist), Hospital of the University of Pennsylvania, Philadelphia, PA.
From the Editor—Emergency physicians must often make decisions about patient management without clear-cut data of sufficient quality to support clinical guidelines or evidence-based reviews. Topics in the Best Available Evidence section must be relevant to emergency physicians, are formally peer reviewed, and must have a sufficient literature base to draw a reasonable conclusion but not such a large literature base that a traditional “evidence-based” review, meta-analysis, or systematic review can be performed. [Ann Emerg Med. 2008;52:368-372.]
ABSTRACT Patients with intraoral and oral-cutaneous wounds commonly present to emergency departments (EDs), usually after sustaining blunt facial trauma.1 Many of these wounds are minor, small (⬍1 cm), hemostatic on presentation, and confined to the oral mucosa, requiring no specific intervention. Some of these injuries, however, do require wound repair, in particular large mucosal wounds (⬎1 to 2 cm) and those that communicate with the overlying facial skin (“through-andthrough” wounds). Most emergency medicine textbooks consider these larger wounds to be at significant risk for infection because of oral bacterial flora and thus recommend that a course of prophylactic antibiotics be prescribed when primary repair of these wounds is performed in the ED, in addition to routine oral and local wound care.2,3 However, studies have shown that other wounds commonly considered to be at significant risk for secondary infection, such as simple lacerations of the hand, do not appear to conclusively benefit from prophylactic antibiotic therapy when careful wound preparation is performed before primary repair.4 We therefore reviewed the available literature to determine the strength of evidence supporting the use of prophylactic oral antibiotics for intraoral wounds treated and repaired in the ED.
SEARCH STRATEGY An Ovid MEDLINE (1950 to present) search was performed in October 2007 with the key words “intraoral,” “oral,” “laceration,” “wound,” “antibiotics,” “anti-bacterial agents,” “prophylaxis,” “orofacial,” and “orocutaneous.” Searches were limited to “human.” Additional searches were performed using the Cochrane Library through 2007, Emergency Medicine Abstracts from 1977 through September 2007, and online 368 Annals of Emergency Medicine
resources including BestBETS, with the search terms “oral,” “intraoral,” “laceration,” and “antibiotics.” In addition, bibliographies of eligible articles and several core clinical textbooks were reviewed for citations of any additional eligible studies. Articles were considered for inclusion if they concerned original clinical research evaluating prophylactic oral antibiotics for the aftercare of intraoral or through-and-through wounds, sustained through either blunt or penetrating mechanisms. Excluded from this summary were articles dealing exclusively with wounds resulting from animal or human (other than the host) bites, those associated with underlying bony or dentoalveolar fractures, wounds resulting from burns, wounds with evidence of established infection on presentation, and tongue wounds, which are considered distinct from oral mucosa wounds. A total of 264 articles were yielded with the search criteria. Of these, 254 were excluded according to their titles (study subjects not involving intraoral lacerations and studies those involving human or animal bites, underlying bone fractures, or maxillofacial surgical procedures). Six of these citations were reviews of facial wound management principles and on examination offered no original data about the clinical question of interest. A total of 4 clinical research articles were left for inclusion in this review. No existing evidentiary reviews addressing the topic of interest were identified.
ARTICLE SUMMARIES Goldberg5 This prospective observational study enrolled a “large number” (exact amount not stated) of adult patients with lower lip lacerations involving mucosa only or mucosa plus skin (ie, through-and-through), presenting to a single ED. Patients were eligible for inclusion if they underwent primary wound repair in the ED. The author excluded patients with extensive facial lacerations, those complicated by mandibular fractures, patients with extensive tissue loss, or those that required flap procedures for closure. No mention is made about whether wounds with evidence of infection on presentation were excluded. The decision to treat with antibiotic prophylaxis was based on the discretion of the treating clinician. No mention is made of which antibiotics were used for prophylaxis. All wounds were irrigated, debrided, and closed in layers. No specific mention was made of instructions given to patients about adjuvant home wound care. Twelve patients with wounds more than 24 hours Volume , . : October
Mark & Granquist
Oral Antibiotics for Intraoral Wounds
Table 1. Study designs. Study Steele et al, 1989
Altieri et al, 1986
Paterson et al, 1970
Goldberg, 1965
Patients
Interventions
Comparisons
Outcomes
Randomized 76 adult ED patients with full-thickness intraoral or through-and-through lacerations, presenting within 24 h. Age cutoff not specified. Excluded tongue lacerations, wounds with associated fractures, patients already receiving antibiotics, patients with penicillin allergies. Randomized 100 pediatric (age 1 mo to 18.5 y) ED patients presenting with intraoral lacerations (not specified). Excluded patients already receiving antibiotics, lacerations more than 24 h old, lacerations with gingivitis or draining dental abscess, associated dental avulsion, or underlying bony fracture.
Penicillin VK 500-mg capsules 4 times a day for 5 days, with saltwater rinses
Placebo 4 times a day for 5 days, with saltwater rinses
Subjective and objective signs and symptoms of wound infection up to a minimum of 5 days of follow-up
Penicillin VK 25-50 mg/kg/day 4 times a day during 3 days, or erythromycin for penicillin-allergic patients (dose not specified), both with saltwater rinses. Nonstandardized antibiotic prophylaxis at discretion of the treating physician
No antibiotics with saltwater rinses.
Subjective and objective signs and symptoms of wound infection, including fever, swelling, erythema, and tenderness, either at followup visit or by telephone interview with parents within 48 to 72 h.
No antibiotic prophylaxis
Subjective clinical determination (not specified) of wound infection on first, third, and fifth days after presentation.
Nonstandardized antibiotic prophylaxis “left to the discretion of the surgeon.”
No antibiotic prophylaxis
Subjective clinical determination of wound infection (severe edema, erythema and pus) between fourth and sixth postoperative days
Prospective randomization of 151 wounds in unstated number of patients (no age criteria specified) presenting to the ED with intraoral or through-and-through lacerations resulting from trauma, in whom corrective care was necessary. No exclusion criteria specified. Included lacerations resulting from bites (human and animal). Prospective observational study of “large number” of ambulatory adults presenting to an ED with lower lip lacerations involving the mucosa only or mucosa and skin (through and through). Excluded patients with extensive facial lacerations, those complicated by mandibular fractures, patients with extensive tissue loss, or those that required flap procedures for closure. Wounds more than 24 h old were followed up but not included in final analysis.
old were also followed but were not included in the final analysis (none of these wounds were repaired at presentation). Patients were reassessed at 4 to 6 days after the initial presentation for evidence of wound infection, according to subjective clinical criteria. Only those patients completing follow-up were included in the final analysis (116 total patients). There was no blinding to treatment group allocation among the clinicians assessing outcomes. Wounds in the antibiotic treatment group were on average more extensive than those in the control group; other baseline data were not reported. In patients treated with antibiotics, wound infection was present in 14.6% versus 23.1% of those receiving no antibiotics (relative risk [RR] 0.63; 95% confidence interval [CI] 0.27 to 1.50; P⫽.33). Subgroup analysis of patients with through-and-through lacerations likewise showed no significant difference in infection rates, though an increased trend towards benefit is suggested (RR 0.5; 95% CI 0.18 to 1.38; P⫽.23). All of the patients with wounds more than 24 hours old at presentation were found to have wound infections at follow-up, despite antibiotic prophylaxis. The authors conclude that prophylactic antibiotics are likely useful in patients with Volume , . : October
through-and-through lacerations. They discourage the routine use of antibiotics for intraoral wound prophylaxis. Paterson et al6 This prospective trial randomized an unreported number of patients (all ages) who presented to a single ED with traumatic intraoral or through-and-through wounds to either oral antibiotic prophylaxis or no treatment. Randomization was based on an alternating treatment assignment block schedule (each block lasted 4 months). The type and duration of antibiotic treatment in the intervention arm were left to the discretion of the treating clinician, and thus clinicians were aware of allocation assignment at enrollment. All wounds were repaired primarily in the ED. Wounds were prepared and cleansed in a standardized fashion, but the method of wound closure was left to the discretion of the treating clinician. No exclusion criteria existed (patients with animal and human bites were included; no mention was made of time from injury to presentation). No baseline data about allocation groups were reported, and no mention is made of postintervention wound care instruction. Annals of Emergency Medicine 369
Oral Antibiotics for Intraoral Wounds
Mark & Granquist
Table 2. Assessment of criteria subject to important bias among reviewed studies. Criteria
Steele et al, 1989
Altieri et al, 1986
Paterson et al, 1970
Randomization
Randomization based on a table of random digits
Randomization based on day of birth (even vs odd days)
None
Concealment
Not stated
None, patients and treating physicians aware of group assignment
Intention-to-treat analysis
Yes, also secondary analysis excluding noncompliant patients No differences between groups in age, sex, time to presentation, number of wounds, wound locations, wound types, wound length, or wound closures. Noted difference in causes of wounds (more from assault in placebo group). Presumably, but not specified
Yes
Randomization based on rotating treatment allocations (4-mo blocks) None, patients and treating physicians aware of group assignment Yes (for those completing follow-up)
No differences between groups in wounds requiring sutures, length or depth of wound, mechanism of injury, or time since injury.
Not reported
Predominately male population (80%). Wounds in antibiotic group longer on average. No specific other baseline data reported.
Clinicians assessing patients denied access to ED charts; no mention of other precautions
None
Follow-up
Fourteen patients (18%) lost to follow-up, equally distributed between groups. No differences between groups in treatment compliance, assessed by pill counts.
Seven patients (7%) lost to follow-up, 63% returned for complete evaluation, 30% assessed by telephone only.
Cointervention (wound care)
Standard wound care with copious irrigation and predefined suture materials. All patients were advised to use saltwater washes during 5 days. Decision to repair was not protocolized; 97% of patients underwent primary wound repair.
Standard wound care with copious irrigation and predefined suture materials. All patients were advised to use saline mouthwash for 48 h. 14% of patients underwent primary wound repair.
Clinicians at follow-up reportedly blinded to treatment group; method of blinding not specified. Only patients completing follow-up included; rate of follow-up not reported. Patients treated at 1, 3, and 5 days after presentation. Standard wound care with irrigation, debridement and pHisoHex cleansing. No predefined suture technique or materials. All patients underwent primary wound repair.
Baseline comparisons
Blinding
Goldberg, 1965
None
Yes
“Approximately” 70% follow-up rate (final analysis included 116 patients).
Standard wound care with copious irrigation, debridement of nonviable tissue. Wounds closed in 2 (mucosa, skin) or 3 layers (muscle, mucosa, skin). All patients underwent primary wound repair
Table 3. Measured outcome (wound infection) in reviewed studies. Studies
Antibiotics, %
Control, %
Relative Risk (95% CI)*
6.7 (n⫽30) 0 (n⫽24) 4.4 (n⫽46) 18.5 (n⫽81) 14.6 (n⫽41) 16.7 (n⫽24)
18.8 (n⫽32) 17.9 (n⫽28) 8.5 (n⫽47) 4.3 (n⫽70) 23.1 (n⫽65) 33.3 (n⫽33)
0.36 (0.08 to 1.63; P⫽.26) 0 (n/a; P⫽.054) 0.51 (0.10 to 2.65; P⫽.68) 4.32 (1.30 to 14.31; P⫽.01) 0.63 (0.27 to 1.50; P⫽.33) 0.5 (0.18 to 1.38; P⫽.23)
Steele et al (intention to treat) Steele et al (compliant only) Altieri et al Paterson et al Goldberg (all oral wounds) Goldberg (through-and-through only) n/a, Not applicable. *P values calculated with 2-tailed Fisher exact probability test.
Patients were examined for subjective signs of wound infection at days 1, 3, and 5 after presentation, though the authors do not state when the determination of wound infection was ultimately made. Those patients who failed to complete follow-up were excluded from the final analysis (151 intraoral 370 Annals of Emergency Medicine
wounds were ultimately included in the analysis from an unknown number of patients). Clinicians assessing wounds were reportedly blinded to treatment allocation, though no specific details about this process were provided. Wound infection was ultimately deemed to be present in 18.5% of wounds in patients Volume , . : October
Mark & Granquist treated with antibiotics versus 4.3% of wounds in patients receiving no treatment (RR 4.32; 95% CI 1.30 to 14.31; P⫽.01). None of the patients with mucosa-only wounds developed wound infections; all infections were observed in patients with through-and-through lacerations. This is also notable because twice as many through-and-through lacerations were randomized to the antibiotic group (48 versus 24 wounds). The authors conclude that the role of prophylactic antibiotics is questionable in preventing local infection. Altieri et al7 This prospective study randomized 100 pediatric patients (aged 1 month to 18.5 years) presenting to a single ED with intraoral wounds to either prophylactic therapy with penicillin VK for 3 days or no treatment. Randomization was based on day of birth (odd versus even). Patients with penicillin allergies were treated with erythromyocin. Treating clinicians were aware of allocation assignment at enrollment. Patients were excluded if they were already taking antibiotics, presented more than 24 hours from the time of injury, or had associated gingivitis, draining dental abscesses, dental avulsion, or underlying bony fracture. Wound repair and preparation were standardized, though the decision to repair wounds was left to the discretion of the emergency physician. Only 14% of the study patients underwent primary wound repair. All patients were instructed to rinse their mouths with saltwater several times a day for the first 48 hours after presentation. No baseline differences were present with regard to wound length, wound depth, mechanism of injury, or time since injury. Wounds were assessed between 48 and 72 hours after presentation by one of 3 clinicians, all of whom were blinded to treatment allocation. Blinding was achieved by preventing the assessing clinicians access to the initial ED charts. No mention is made of how patients and families were prevented from revealing treatment allocation. Patients were assessed either in person (63% of patients) or over the telephone with assistance from parents (30% of patients). A total of 7% of patients were lost to follow-up, leaving a total of 93 patients in the final analysis. Of patients treated with prophylactic antibiotics, 4.4% had evidence of wound infection at follow-up versus 8.5% of patients in the control arm (RR 0.51; 95% CI 0.10 to 2.65; P⫽.68). Subgroup analysis of wounds requiring suture repair showed no significant difference in infection rates between treatment allocation groups. The authors conclude that, given a low background rate of infection, the study was markedly underpowered, severely limiting their ability to make any definite conclusions about the efficacy of prophylactic antibiotics. Steele et al8 This study randomized 76 adults presenting within 24 hours of injury to a single ED with full-thickness intraoral or throughand-through lacerations to either treatment with oral antibiotic prophylaxis (penicillin VK for 5 days) or placebo therapy. Volume , . : October
Oral Antibiotics for Intraoral Wounds Randomization was achieved with a table of random digits. Treating clinicians were reportedly unaware of allocation assignment at enrollment (method of concealment not specified). All repairs were performed with standardized wound preparation and repair materials, but the decision to repair was left to the discretion of the treating physician. Ninety-seven percent of study patients underwent primary wound repair. The authors excluded patients with isolated tongue lacerations, wounds with associated fractures, patients already receiving antibiotics, and patients with penicillin allergies. All patients were instructed to perform oral saltwater rinses for 5 days. Patients were reexamined at a minimum interval of 5 days from the time of presentation. Eighteen percent of patients were lost to follow-up, leaving a total of 62 patients for final analysis. Clinicians assessed wound infection according to subjective criteria (erythema, swelling, pain, induration, pus, and odor), in addition to systemic temperature. The decision to assess compliance with treatment for subgroup analysis was part of the study design and was accomplished at follow-up by pill counts. No differences between groups existed in baseline characteristics about age, sex, time to presentation, number of wounds, wound locations, wound types, wound length, or number of wound closures. There was a slight difference in the cause of wounding between groups, with more patients having wounds from assault in the placebo group (26 of 32 versus 15 of 30). Among patients treated with penicillin prophylaxis, a total of 6.7% developed wound infection versus 18.8% in the placebo group (RR 0.36; 95% CI 0.08 to 1.63; P⫽.26). In a planned subgroup analysis of patients who were perfectly compliant with therapy by pill counts, there were no infections in the penicillin arm versus 17.9% in the placebo arm (P⫽.054 using 2-tailed Fisher exact test). A second subgroup analysis of through-and-though lacerations revealed that 7% (1 of 14) versus 27% (4 of 15) of patients developed wound infection in the treatment and placebo groups, respectively (RR 0.27; 95% CI 0.03 to 2.12; P⫽.33). Overall, the authors conclude that although they cannot conclusively recommend prophylactic penicillin for adults with intraoral lacerations treated within 24 hours of injury, there may be a benefit according to subgroup analysis of compliant patients. The authors also suggest that patients with through-and-through wounds may benefit proportionally more from antibiotic prophylaxis. Summarized study designs, assessment criteria, and outcomes of interest are presented in Tables 1, 2 and 3.
THE BOTTOM LINE There is only 1 placebo-controlled, double-blinded, randomized clinical trial (Steele et al8) evaluating the efficacy of prophylactic oral antibiotics to prevent wound infection in simple intraoral wounds treated within 24 hours of injury in the ED. This study had small enrollment numbers and was ultimately unable to conclusively demonstrate a statistically significant benefit in favor of antibiotics in the overall analysis. However, a planned subgroup analysis of those patients who were adherent to the medication plan did reveal a clinically Annals of Emergency Medicine 371
Oral Antibiotics for Intraoral Wounds
Mark & Granquist
relevant trend (P⫽.054) toward decreased rates of wound infection. Another randomized trial (Paterson et al6) conversely reported harm in the form of significantly increased rates of wound infection in patients taking prophylactic antibiotics. However, this latter study had substantial methodologic flaws, including unequal randomization of those lacerations with the highest risk for infection (through-and-through lacerations), making the results difficult to interpret and of questionable validity. Of the 2 remaining studies, one (Altieri et al7) was substantially underpowered, given a low background rate of infection (8.5%), presumably because of the inclusion of mostly minor wounds that did not require repair. The study results do suggest that minor intraoral wounds have such a low rate of background infection that a substantial number of other patients need to be treated for a single patient to benefit (number needed to treat⫽25 according to the available study results). In such a scenario, the benefit of preventing an adverse cosmetic outcome and wound infection must be weighed against the cost of widespread use of penicillin in the age of antibiotic-resistant bacteria. The final study (Goldberg5) was also unable to demonstrate a significant decrease in wound infection rates among patients treated with prophylactic antibiotics, despite universal wound repair. However, a potential trend toward benefit may have been present, particularly among those patients with throughand-through lacerations. Also of note is the apparent ineffectiveness of antibiotics to prevent infection in patients presenting with wounds more than 24 hours from the time of injury, given a 100% incidence of infection. Unfortunately, the nonrandomized study design renders its findings suggestive at best. In summary, according to the best available evidence, prophylactic oral antibiotics play an inconclusive role in the treatment of intraoral wounds. All randomized studies to date have been inadequately powered to demonstrate a statistically significant difference in wound infection rates when antibiotics are compared with placebo or routine wound care. Although nonsignificant trends toward benefit have been demonstrated in subgroup analyses of high-risk wounds (ie, through-and-though lacerations) and in medication compliance– based analysis, until
a larger clinical trial is performed, treatment decisions must be guided by clinical and scientific rationale rather than evidencebased medicine. doi:10.1016/j.annemergmed.2007.12.028
Supervising editor : Judd E. Hollander, MD Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article, that might create any potential conflict of interest. The authors have stated that no such relationships exist. See the Manuscript Submission Agreement in this issue for examples of specific conflicts covered by this statement. Publication date: Available online March 4, 2008. Address for reprints: Dustin G. Mark, MD, Department of Emergency Medicine, Hospital of the University of Pennsylvania, Ground Ravdih, Philadelphia, PA 19104, 215662-6305, fax 215-349-8038; E-mail [email protected]. REFERENCES 1. Allonby-Neve CL, Okereke CD. Current management of facial wounds in UK accident and emergency departments. Ann R Coll Surg Engl. 2006;88:144-150. 2. Tintinalli JE, Kelen GD, Stapczynski JS, eds. Emergency Medicine: A Comprehensive Study Guide. 6th ed. New York, NY: McGraw-Hill; 2004. 3. Marx JA, Hockberger RS, Walls RM, eds. Marx: Rosen’s Emergency Medicine: Concepts and Clinical Practice. 6th ed. Philadelphia, PA: Mosby Elsevier; 2006. 4. Zehtabchi S. Evidence-based emergency medicine/critically appraised topic. The role of antibiotic prophylaxis for prevention of infection in patients with simple hand lacerations. Ann Emerg Med. 2007;49:682-689, 9 e1. 5. Goldberg MH. Antibiotics and oral and oral-cutaneous lacerations. J Oral Surg. 1965;23:117-122. 6. Paterson JA, Cardo VA Jr, Stratigos GT. An examination of antibiotic prophylaxis in oral and maxillofacial surgery. J Oral Surg. 1970;28:753-759. 7. Altieri M, Brasch L, Getson P. Antibiotic prophylaxis in intraoral wounds. Am J Emerg Med. 1986;4:507-510. 8. Steele MT, Sainsbury CR, Robinson WA, et al. Prophylactic penicillin for intraoral wounds. Ann Emerg Med. 1989;18:847-852.
Did you know? You can track the impact of your article with citation alerts that let you know when your article (or any article you’d like to track) has been cited by another Elsevier-published journal.
Visit www.annemergmed.com today to see what else is new online!
372 Annals of Emergency Medicine
Volume , . : October
Are Prophylactic Oral Antibiotics Indicated for the Treatment of Intraoral Wounds? Dustin G. Mark, MD Eric J. Granquist, DMD
From the Department of Emergency Medicine (Mark) and Department of Oral and Maxillofacial Surgery (Granquist), Hospital of the University of Pennsylvania, Philadelphia, PA.
From the Editor—Emergency physicians must often make decisions about patient management without clear-cut data of sufficient quality to support clinical guidelines or evidence-based reviews. Topics in the Best Available Evidence section must be relevant to emergency physicians, are formally peer reviewed, and must have a sufficient literature base to draw a reasonable conclusion but not such a large literature base that a traditional “evidence-based” review, meta-analysis, or systematic review can be performed. [Ann Emerg Med. 2008;52:368-372.]
ABSTRACT Patients with intraoral and oral-cutaneous wounds commonly present to emergency departments (EDs), usually after sustaining blunt facial trauma.1 Many of these wounds are minor, small (⬍1 cm), hemostatic on presentation, and confined to the oral mucosa, requiring no specific intervention. Some of these injuries, however, do require wound repair, in particular large mucosal wounds (⬎1 to 2 cm) and those that communicate with the overlying facial skin (“through-andthrough” wounds). Most emergency medicine textbooks consider these larger wounds to be at significant risk for infection because of oral bacterial flora and thus recommend that a course of prophylactic antibiotics be prescribed when primary repair of these wounds is performed in the ED, in addition to routine oral and local wound care.2,3 However, studies have shown that other wounds commonly considered to be at significant risk for secondary infection, such as simple lacerations of the hand, do not appear to conclusively benefit from prophylactic antibiotic therapy when careful wound preparation is performed before primary repair.4 We therefore reviewed the available literature to determine the strength of evidence supporting the use of prophylactic oral antibiotics for intraoral wounds treated and repaired in the ED.
SEARCH STRATEGY An Ovid MEDLINE (1950 to present) search was performed in October 2007 with the key words “intraoral,” “oral,” “laceration,” “wound,” “antibiotics,” “anti-bacterial agents,” “prophylaxis,” “orofacial,” and “orocutaneous.” Searches were limited to “human.” Additional searches were performed using the Cochrane Library through 2007, Emergency Medicine Abstracts from 1977 through September 2007, and online 368 Annals of Emergency Medicine
resources including BestBETS, with the search terms “oral,” “intraoral,” “laceration,” and “antibiotics.” In addition, bibliographies of eligible articles and several core clinical textbooks were reviewed for citations of any additional eligible studies. Articles were considered for inclusion if they concerned original clinical research evaluating prophylactic oral antibiotics for the aftercare of intraoral or through-and-through wounds, sustained through either blunt or penetrating mechanisms. Excluded from this summary were articles dealing exclusively with wounds resulting from animal or human (other than the host) bites, those associated with underlying bony or dentoalveolar fractures, wounds resulting from burns, wounds with evidence of established infection on presentation, and tongue wounds, which are considered distinct from oral mucosa wounds. A total of 264 articles were yielded with the search criteria. Of these, 254 were excluded according to their titles (study subjects not involving intraoral lacerations and studies those involving human or animal bites, underlying bone fractures, or maxillofacial surgical procedures). Six of these citations were reviews of facial wound management principles and on examination offered no original data about the clinical question of interest. A total of 4 clinical research articles were left for inclusion in this review. No existing evidentiary reviews addressing the topic of interest were identified.
ARTICLE SUMMARIES Goldberg5 This prospective observational study enrolled a “large number” (exact amount not stated) of adult patients with lower lip lacerations involving mucosa only or mucosa plus skin (ie, through-and-through), presenting to a single ED. Patients were eligible for inclusion if they underwent primary wound repair in the ED. The author excluded patients with extensive facial lacerations, those complicated by mandibular fractures, patients with extensive tissue loss, or those that required flap procedures for closure. No mention is made about whether wounds with evidence of infection on presentation were excluded. The decision to treat with antibiotic prophylaxis was based on the discretion of the treating clinician. No mention is made of which antibiotics were used for prophylaxis. All wounds were irrigated, debrided, and closed in layers. No specific mention was made of instructions given to patients about adjuvant home wound care. Twelve patients with wounds more than 24 hours Volume , . : October
Mark & Granquist
Oral Antibiotics for Intraoral Wounds
Table 1. Study designs. Study Steele et al, 1989
Altieri et al, 1986
Paterson et al, 1970
Goldberg, 1965
Patients
Interventions
Comparisons
Outcomes
Randomized 76 adult ED patients with full-thickness intraoral or through-and-through lacerations, presenting within 24 h. Age cutoff not specified. Excluded tongue lacerations, wounds with associated fractures, patients already receiving antibiotics, patients with penicillin allergies. Randomized 100 pediatric (age 1 mo to 18.5 y) ED patients presenting with intraoral lacerations (not specified). Excluded patients already receiving antibiotics, lacerations more than 24 h old, lacerations with gingivitis or draining dental abscess, associated dental avulsion, or underlying bony fracture.
Penicillin VK 500-mg capsules 4 times a day for 5 days, with saltwater rinses
Placebo 4 times a day for 5 days, with saltwater rinses
Subjective and objective signs and symptoms of wound infection up to a minimum of 5 days of follow-up
Penicillin VK 25-50 mg/kg/day 4 times a day during 3 days, or erythromycin for penicillin-allergic patients (dose not specified), both with saltwater rinses. Nonstandardized antibiotic prophylaxis at discretion of the treating physician
No antibiotics with saltwater rinses.
Subjective and objective signs and symptoms of wound infection, including fever, swelling, erythema, and tenderness, either at followup visit or by telephone interview with parents within 48 to 72 h.
No antibiotic prophylaxis
Subjective clinical determination (not specified) of wound infection on first, third, and fifth days after presentation.
Nonstandardized antibiotic prophylaxis “left to the discretion of the surgeon.”
No antibiotic prophylaxis
Subjective clinical determination of wound infection (severe edema, erythema and pus) between fourth and sixth postoperative days
Prospective randomization of 151 wounds in unstated number of patients (no age criteria specified) presenting to the ED with intraoral or through-and-through lacerations resulting from trauma, in whom corrective care was necessary. No exclusion criteria specified. Included lacerations resulting from bites (human and animal). Prospective observational study of “large number” of ambulatory adults presenting to an ED with lower lip lacerations involving the mucosa only or mucosa and skin (through and through). Excluded patients with extensive facial lacerations, those complicated by mandibular fractures, patients with extensive tissue loss, or those that required flap procedures for closure. Wounds more than 24 h old were followed up but not included in final analysis.
old were also followed but were not included in the final analysis (none of these wounds were repaired at presentation). Patients were reassessed at 4 to 6 days after the initial presentation for evidence of wound infection, according to subjective clinical criteria. Only those patients completing follow-up were included in the final analysis (116 total patients). There was no blinding to treatment group allocation among the clinicians assessing outcomes. Wounds in the antibiotic treatment group were on average more extensive than those in the control group; other baseline data were not reported. In patients treated with antibiotics, wound infection was present in 14.6% versus 23.1% of those receiving no antibiotics (relative risk [RR] 0.63; 95% confidence interval [CI] 0.27 to 1.50; P⫽.33). Subgroup analysis of patients with through-and-through lacerations likewise showed no significant difference in infection rates, though an increased trend towards benefit is suggested (RR 0.5; 95% CI 0.18 to 1.38; P⫽.23). All of the patients with wounds more than 24 hours old at presentation were found to have wound infections at follow-up, despite antibiotic prophylaxis. The authors conclude that prophylactic antibiotics are likely useful in patients with Volume , . : October
through-and-through lacerations. They discourage the routine use of antibiotics for intraoral wound prophylaxis. Paterson et al6 This prospective trial randomized an unreported number of patients (all ages) who presented to a single ED with traumatic intraoral or through-and-through wounds to either oral antibiotic prophylaxis or no treatment. Randomization was based on an alternating treatment assignment block schedule (each block lasted 4 months). The type and duration of antibiotic treatment in the intervention arm were left to the discretion of the treating clinician, and thus clinicians were aware of allocation assignment at enrollment. All wounds were repaired primarily in the ED. Wounds were prepared and cleansed in a standardized fashion, but the method of wound closure was left to the discretion of the treating clinician. No exclusion criteria existed (patients with animal and human bites were included; no mention was made of time from injury to presentation). No baseline data about allocation groups were reported, and no mention is made of postintervention wound care instruction. Annals of Emergency Medicine 369
Oral Antibiotics for Intraoral Wounds
Mark & Granquist
Table 2. Assessment of criteria subject to important bias among reviewed studies. Criteria
Steele et al, 1989
Altieri et al, 1986
Paterson et al, 1970
Randomization
Randomization based on a table of random digits
Randomization based on day of birth (even vs odd days)
None
Concealment
Not stated
None, patients and treating physicians aware of group assignment
Intention-to-treat analysis
Yes, also secondary analysis excluding noncompliant patients No differences between groups in age, sex, time to presentation, number of wounds, wound locations, wound types, wound length, or wound closures. Noted difference in causes of wounds (more from assault in placebo group). Presumably, but not specified
Yes
Randomization based on rotating treatment allocations (4-mo blocks) None, patients and treating physicians aware of group assignment Yes (for those completing follow-up)
No differences between groups in wounds requiring sutures, length or depth of wound, mechanism of injury, or time since injury.
Not reported
Predominately male population (80%). Wounds in antibiotic group longer on average. No specific other baseline data reported.
Clinicians assessing patients denied access to ED charts; no mention of other precautions
None
Follow-up
Fourteen patients (18%) lost to follow-up, equally distributed between groups. No differences between groups in treatment compliance, assessed by pill counts.
Seven patients (7%) lost to follow-up, 63% returned for complete evaluation, 30% assessed by telephone only.
Cointervention (wound care)
Standard wound care with copious irrigation and predefined suture materials. All patients were advised to use saltwater washes during 5 days. Decision to repair was not protocolized; 97% of patients underwent primary wound repair.
Standard wound care with copious irrigation and predefined suture materials. All patients were advised to use saline mouthwash for 48 h. 14% of patients underwent primary wound repair.
Clinicians at follow-up reportedly blinded to treatment group; method of blinding not specified. Only patients completing follow-up included; rate of follow-up not reported. Patients treated at 1, 3, and 5 days after presentation. Standard wound care with irrigation, debridement and pHisoHex cleansing. No predefined suture technique or materials. All patients underwent primary wound repair.
Baseline comparisons
Blinding
Goldberg, 1965
None
Yes
“Approximately” 70% follow-up rate (final analysis included 116 patients).
Standard wound care with copious irrigation, debridement of nonviable tissue. Wounds closed in 2 (mucosa, skin) or 3 layers (muscle, mucosa, skin). All patients underwent primary wound repair
Table 3. Measured outcome (wound infection) in reviewed studies. Studies
Antibiotics, %
Control, %
Relative Risk (95% CI)*
6.7 (n⫽30) 0 (n⫽24) 4.4 (n⫽46) 18.5 (n⫽81) 14.6 (n⫽41) 16.7 (n⫽24)
18.8 (n⫽32) 17.9 (n⫽28) 8.5 (n⫽47) 4.3 (n⫽70) 23.1 (n⫽65) 33.3 (n⫽33)
0.36 (0.08 to 1.63; P⫽.26) 0 (n/a; P⫽.054) 0.51 (0.10 to 2.65; P⫽.68) 4.32 (1.30 to 14.31; P⫽.01) 0.63 (0.27 to 1.50; P⫽.33) 0.5 (0.18 to 1.38; P⫽.23)
Steele et al (intention to treat) Steele et al (compliant only) Altieri et al Paterson et al Goldberg (all oral wounds) Goldberg (through-and-through only) n/a, Not applicable. *P values calculated with 2-tailed Fisher exact probability test.
Patients were examined for subjective signs of wound infection at days 1, 3, and 5 after presentation, though the authors do not state when the determination of wound infection was ultimately made. Those patients who failed to complete follow-up were excluded from the final analysis (151 intraoral 370 Annals of Emergency Medicine
wounds were ultimately included in the analysis from an unknown number of patients). Clinicians assessing wounds were reportedly blinded to treatment allocation, though no specific details about this process were provided. Wound infection was ultimately deemed to be present in 18.5% of wounds in patients Volume , . : October
Mark & Granquist treated with antibiotics versus 4.3% of wounds in patients receiving no treatment (RR 4.32; 95% CI 1.30 to 14.31; P⫽.01). None of the patients with mucosa-only wounds developed wound infections; all infections were observed in patients with through-and-through lacerations. This is also notable because twice as many through-and-through lacerations were randomized to the antibiotic group (48 versus 24 wounds). The authors conclude that the role of prophylactic antibiotics is questionable in preventing local infection. Altieri et al7 This prospective study randomized 100 pediatric patients (aged 1 month to 18.5 years) presenting to a single ED with intraoral wounds to either prophylactic therapy with penicillin VK for 3 days or no treatment. Randomization was based on day of birth (odd versus even). Patients with penicillin allergies were treated with erythromyocin. Treating clinicians were aware of allocation assignment at enrollment. Patients were excluded if they were already taking antibiotics, presented more than 24 hours from the time of injury, or had associated gingivitis, draining dental abscesses, dental avulsion, or underlying bony fracture. Wound repair and preparation were standardized, though the decision to repair wounds was left to the discretion of the emergency physician. Only 14% of the study patients underwent primary wound repair. All patients were instructed to rinse their mouths with saltwater several times a day for the first 48 hours after presentation. No baseline differences were present with regard to wound length, wound depth, mechanism of injury, or time since injury. Wounds were assessed between 48 and 72 hours after presentation by one of 3 clinicians, all of whom were blinded to treatment allocation. Blinding was achieved by preventing the assessing clinicians access to the initial ED charts. No mention is made of how patients and families were prevented from revealing treatment allocation. Patients were assessed either in person (63% of patients) or over the telephone with assistance from parents (30% of patients). A total of 7% of patients were lost to follow-up, leaving a total of 93 patients in the final analysis. Of patients treated with prophylactic antibiotics, 4.4% had evidence of wound infection at follow-up versus 8.5% of patients in the control arm (RR 0.51; 95% CI 0.10 to 2.65; P⫽.68). Subgroup analysis of wounds requiring suture repair showed no significant difference in infection rates between treatment allocation groups. The authors conclude that, given a low background rate of infection, the study was markedly underpowered, severely limiting their ability to make any definite conclusions about the efficacy of prophylactic antibiotics. Steele et al8 This study randomized 76 adults presenting within 24 hours of injury to a single ED with full-thickness intraoral or throughand-through lacerations to either treatment with oral antibiotic prophylaxis (penicillin VK for 5 days) or placebo therapy. Volume , . : October
Oral Antibiotics for Intraoral Wounds Randomization was achieved with a table of random digits. Treating clinicians were reportedly unaware of allocation assignment at enrollment (method of concealment not specified). All repairs were performed with standardized wound preparation and repair materials, but the decision to repair was left to the discretion of the treating physician. Ninety-seven percent of study patients underwent primary wound repair. The authors excluded patients with isolated tongue lacerations, wounds with associated fractures, patients already receiving antibiotics, and patients with penicillin allergies. All patients were instructed to perform oral saltwater rinses for 5 days. Patients were reexamined at a minimum interval of 5 days from the time of presentation. Eighteen percent of patients were lost to follow-up, leaving a total of 62 patients for final analysis. Clinicians assessed wound infection according to subjective criteria (erythema, swelling, pain, induration, pus, and odor), in addition to systemic temperature. The decision to assess compliance with treatment for subgroup analysis was part of the study design and was accomplished at follow-up by pill counts. No differences between groups existed in baseline characteristics about age, sex, time to presentation, number of wounds, wound locations, wound types, wound length, or number of wound closures. There was a slight difference in the cause of wounding between groups, with more patients having wounds from assault in the placebo group (26 of 32 versus 15 of 30). Among patients treated with penicillin prophylaxis, a total of 6.7% developed wound infection versus 18.8% in the placebo group (RR 0.36; 95% CI 0.08 to 1.63; P⫽.26). In a planned subgroup analysis of patients who were perfectly compliant with therapy by pill counts, there were no infections in the penicillin arm versus 17.9% in the placebo arm (P⫽.054 using 2-tailed Fisher exact test). A second subgroup analysis of through-and-though lacerations revealed that 7% (1 of 14) versus 27% (4 of 15) of patients developed wound infection in the treatment and placebo groups, respectively (RR 0.27; 95% CI 0.03 to 2.12; P⫽.33). Overall, the authors conclude that although they cannot conclusively recommend prophylactic penicillin for adults with intraoral lacerations treated within 24 hours of injury, there may be a benefit according to subgroup analysis of compliant patients. The authors also suggest that patients with through-and-through wounds may benefit proportionally more from antibiotic prophylaxis. Summarized study designs, assessment criteria, and outcomes of interest are presented in Tables 1, 2 and 3.
THE BOTTOM LINE There is only 1 placebo-controlled, double-blinded, randomized clinical trial (Steele et al8) evaluating the efficacy of prophylactic oral antibiotics to prevent wound infection in simple intraoral wounds treated within 24 hours of injury in the ED. This study had small enrollment numbers and was ultimately unable to conclusively demonstrate a statistically significant benefit in favor of antibiotics in the overall analysis. However, a planned subgroup analysis of those patients who were adherent to the medication plan did reveal a clinically Annals of Emergency Medicine 371
Oral Antibiotics for Intraoral Wounds
Mark & Granquist
relevant trend (P⫽.054) toward decreased rates of wound infection. Another randomized trial (Paterson et al6) conversely reported harm in the form of significantly increased rates of wound infection in patients taking prophylactic antibiotics. However, this latter study had substantial methodologic flaws, including unequal randomization of those lacerations with the highest risk for infection (through-and-through lacerations), making the results difficult to interpret and of questionable validity. Of the 2 remaining studies, one (Altieri et al7) was substantially underpowered, given a low background rate of infection (8.5%), presumably because of the inclusion of mostly minor wounds that did not require repair. The study results do suggest that minor intraoral wounds have such a low rate of background infection that a substantial number of other patients need to be treated for a single patient to benefit (number needed to treat⫽25 according to the available study results). In such a scenario, the benefit of preventing an adverse cosmetic outcome and wound infection must be weighed against the cost of widespread use of penicillin in the age of antibiotic-resistant bacteria. The final study (Goldberg5) was also unable to demonstrate a significant decrease in wound infection rates among patients treated with prophylactic antibiotics, despite universal wound repair. However, a potential trend toward benefit may have been present, particularly among those patients with throughand-through lacerations. Also of note is the apparent ineffectiveness of antibiotics to prevent infection in patients presenting with wounds more than 24 hours from the time of injury, given a 100% incidence of infection. Unfortunately, the nonrandomized study design renders its findings suggestive at best. In summary, according to the best available evidence, prophylactic oral antibiotics play an inconclusive role in the treatment of intraoral wounds. All randomized studies to date have been inadequately powered to demonstrate a statistically significant difference in wound infection rates when antibiotics are compared with placebo or routine wound care. Although nonsignificant trends toward benefit have been demonstrated in subgroup analyses of high-risk wounds (ie, through-and-though lacerations) and in medication compliance– based analysis, until
a larger clinical trial is performed, treatment decisions must be guided by clinical and scientific rationale rather than evidencebased medicine. doi:10.1016/j.annemergmed.2007.12.028
Supervising editor : Judd E. Hollander, MD Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article, that might create any potential conflict of interest. The authors have stated that no such relationships exist. See the Manuscript Submission Agreement in this issue for examples of specific conflicts covered by this statement. Publication date: Available online March 4, 2008. Address for reprints: Dustin G. Mark, MD, Department of Emergency Medicine, Hospital of the University of Pennsylvania, Ground Ravdih, Philadelphia, PA 19104, 215662-6305, fax 215-349-8038; E-mail [email protected]. REFERENCES 1. Allonby-Neve CL, Okereke CD. Current management of facial wounds in UK accident and emergency departments. Ann R Coll Surg Engl. 2006;88:144-150. 2. Tintinalli JE, Kelen GD, Stapczynski JS, eds. Emergency Medicine: A Comprehensive Study Guide. 6th ed. New York, NY: McGraw-Hill; 2004. 3. Marx JA, Hockberger RS, Walls RM, eds. Marx: Rosen’s Emergency Medicine: Concepts and Clinical Practice. 6th ed. Philadelphia, PA: Mosby Elsevier; 2006. 4. Zehtabchi S. Evidence-based emergency medicine/critically appraised topic. The role of antibiotic prophylaxis for prevention of infection in patients with simple hand lacerations. Ann Emerg Med. 2007;49:682-689, 9 e1. 5. Goldberg MH. Antibiotics and oral and oral-cutaneous lacerations. J Oral Surg. 1965;23:117-122. 6. Paterson JA, Cardo VA Jr, Stratigos GT. An examination of antibiotic prophylaxis in oral and maxillofacial surgery. J Oral Surg. 1970;28:753-759. 7. Altieri M, Brasch L, Getson P. Antibiotic prophylaxis in intraoral wounds. Am J Emerg Med. 1986;4:507-510. 8. Steele MT, Sainsbury CR, Robinson WA, et al. Prophylactic penicillin for intraoral wounds. Ann Emerg Med. 1989;18:847-852.
Did you know? You can track the impact of your article with citation alerts that let you know when your article (or any article you’d like to track) has been cited by another Elsevier-published journal.
Visit www.annemergmed.com today to see what else is new online!
372 Annals of Emergency Medicine
Volume , . : October