Secondary Logo

Journal Logo

Myths in Emergency Medicine: More Than You Can Chew Antibiotics for Bite Wounds?

May, Natalie MBCHB, MPHE

doi: 10.1097/01.EEM.0000512777.86151.c6
Myths in Emergency Medicine

Dr. Mayis a specialist in emergency medicine with paediatric emergency medicine in the UK and a registrar in prehospital emergency medicine in Sydney, Australia. She writes regularly for Follow her on Twitter @_NMay, and read her past articles at



Wound management, while perhaps not as glamorous as resuscitation, is a hallmark skill of the emergency physician. As a subset of wounds presenting to the ED, bites are particularly interesting. Potential inoculation of wound spaces with oral flora creates, at least in theory, increased likelihood of infection, and standard care has traditionally been to provide antibiotic prophylaxis.

But is that really necessary, especially in the era of evidence-based medicine and antibiotic stewardship?

What do we know about bites? Observational data suggest about 12.9 dog bite-related presentations to the ED per 10,000 population, or 914 new dog bite injuries presenting to the ED per day across the United States. (JAMA 1998;279[1]:51.) A more recent study from Australia suggests upper limbs are the most common bite injury location and that 46 percent of patients receive intravenous antibiotics. (Med J Aust 2016;204[3]:114.) Almost all bites sustained by children under 15 years in their study were dog bites: Overall, the proportions by animal were dog, 71 percent; cat, 17 percent; human, five percent; and other, seven percent.

Bite wound management is inconsistent. A U.K.-based survey described considerable variation in cleaning wounds, from irrigation and debridement to wiping with saline- or Betadine-soaked swabs (Emerg Med J 2003;20[3]:253), although hopefully by now you are all using good old tap water. (EMN 2016;38[7]:21.)

A prospective study of primary closure of 169 dog bite wounds without antibiotic prophylaxis found an infection rate of just 7.7 percent with significantly greater infection rates for hand wounds (Arch Emerg Med 1988;5[3]:156), comparable with a more recent study of 145 mammalian bites (in which 81% received antibiotics but infection rates were not significantly different between antibiotic and no-antibiotic groups), which found an overall infection rate of six percent with primary closure. (Acad Emerg Med 2000;7[2]:157.)

Do antibiotics reduce infection risk? It has been more than 20 years since Cummings' meta-analysis of eight trials of antibiotic therapy found a cumulative incidence of infection of 16 percent, a relative risk of infection of 0.56 in those treated with antibiotics and an NNT of 14. The subsequent Cochrane review identified six papers about dog bites and found no statistically significant difference in infection rates whether antibiotics were used or not (4% vs 5.5%). (Cochrane Database Syst Rev 2001[2]:CD001738.) A significant reduction in infection rates for hand wounds was seen where these had been studied separately (2% in the antibiotic group vs 28% in the control group: NNT=4).

Each of the studies in the Cochrane review had methodology issues. (Pediatr Emerg Care 1992;8[4]:194; Infection 1986;14[3]:134; Am J Emerg Med 1985;3[1]:19; Postgrad Med J 1985;61[717]:593; Am J Emerg Med 1983;1[1]:17); Ann Emerg Med 1982;11[5]:248.) All had small sample sizes (those that did included a power calculation failed to meet the required sample size), randomization was often poorly described, and blinding was absent from many. The overall low infection rates in these predetermined low risk patients, however, are encouraging, notably in three of these papers. (Am J Emerg Med 1985;3[1]:19; Am J Emerg Med 1983;1[1]:17; Ann Emerg Med 1982;11[5]:248.) Infections occurred only in patients with hand wounds.

Quinn, et al., published a double-blind, randomized, controlled trial of antibiotic therapy for low-risk dog bites in the ED in 2010. (West J Emerg Med 2010;11[5]:435.) Although also underpowered, infection rates were very low: two percent overall, four percent in the placebo arm. The authors had set out to determine an infection rate on the premise that it would not be cost-effective to treat below an incidence of five percent, and although their sample size meant wide confidence intervals for their infection rate (95% confidence interval 4-14%) which exceeded this value, the study provides some reassurance when considered in the context of the Cochrane review.

It seems unlikely then that prophylactic antibiotics for low-risk bite wounds in healthy patients presenting in a timely fashion to the ED will significantly reduce the risk of infection. A safer and more cost-effective treatment strategy is to provide good wound care and clear advice to return should signs of infection occur. Luckily, this is in line with good antibiotic stewardship, a responsibility for all clinicians. (Mayo Clin Proc 2011;86[11]:1113.)

Which patients should receive antibiotic prophylaxis? Most of the studies looking at this topic exclude patients with increased risk of infection. Patients with immunosuppression should be assessed for antibiotic prophylaxis for bite wounds case by case. Remember that this may include patients with diabetes and those who are malnourished, such as alcohol-dependent patients.

Also carefully consider these bite-related factors:

  • Evidence of infection at presentation
  • Delayed presentation (usually given in studies as more than 24 hours)
  • Puncture wounds
  • Wounds involving underlying structures: tendon, periosteum or bone, or associated neurovascular injury
  • Bites involving the hand. In line with the Cochrane review, significant reduction has been demonstrated in infection rates for hand bites treated with antibiotic prophylaxis; an NNT of 4 is impressive!
  • Human bites. None of the evidence given here relates to human bites, which have also been associated with higher infection rates, significantly reduced by antibiotic prophylaxis (Plast Reconstr Surg 1991;88[1]:111), although there is some suggestion that “low risk” human bites that only involve the epidermis can also be managed without antibiotics. (Am J Emerg Med 2004;22[1]:10.)

How should we treat bite wounds in the ED? Irrigation and debridement were important in Callaham's 1980 study of infection of dog bite wounds. (Ann Emerg Med 1980;9[8]:410.) We should remember to inquire about immunization status and provide tetanus prophylaxis if indicated. For human bites, consider management of possible exposure to bloodborne viruses.

Consider the need for antibiotic prophylaxis in high-risk bite wounds; others can be managed with wound care advice and clear guidance on the need to consult a health care professional in the event of subsequent signs of infection.

Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.