For a long time when encountering a patient who had suffered a dog bite I reflexively placed the patient on antibiotics. Generally, either amoxicillin/clavulanate or clindamycin. Along the way I began to question whether or not there is evidence to support this practice, since I also haven’t seen many infected dog bites. As you may already be aware, most dog bites are from an animal known to the victim. One study from Australia noted that >4/5 were known. In children under 5 the bites are more likely to be in the head/neck (60-70%). See Patronek at al. for more. In older children and adults the extremities – more commonly the dominant hand – are involved.

The exact rate of infections is hard to estimate since many patients with dog bites likely do not seek care. But, in those that do seek care >24 hours after the bite the rate of infections seems to be higher. Overall though, infections from dog bites are more infrequent than almost any other animal – see Capellan from Emerg Med Clin North Amer for more. If a wound needs to be closed copious irrigation is a must. Debriding dead/compromised tissue is key, as is avoiding deep sutures and any tissue adhesive. It is OK to allow the wound to drain. Ultimately, despite these practices I still wonder whether or not antibiotics truly make a difference. This is, unfortunately, where the literature is a bit thin.

Dog-bite lacerations: a controlled trial of primary wound closure
C Maimaris and D N Quinton
Arch Emerg Med. 1988 Sep; 5(3): 156–161

This was randomized trial of 96 patients of all ages (though most were <30 years). Together they had 169 dog bites. 92/169 were closed and 77/169 were not. No patient got antibiotics. Overall 13/169 wounds developed infection – with no difference between sutured or not (7.6 versus 7.8 percent).

Antibiotics to prevent infection in patients with dog bite wounds: a meta-analysis of randomized trials
Cummings, P
Ann Emerg Med. 1994 Mar;23(3):535-40.

Eight randomized trials were reviewed with the cumulative incidence of infection in controls 16%. The relative risk for infection in antibiotics vs not was RR=0.56 (95% CI, 0.38 to 0.82). The number needed to treat was approximately 14.

Relative risk estimates and 95% CI for individual studies and the summary relative risk estimate for all studies. A relative risk estimate less than 1.0 indicates that antibiotic treatment reduces the risk of infection. The reference number is given for individual studies. *Study has a relative risk estimate of infinity.

Relative risk estimates and 95% CI for individual studies and the summary relative risk estimate for all studies. A relative risk estimate less than 1.0 indicates that antibiotic treatment reduces the risk of infection. The reference number is given for individual studies. *Study has a relative risk estimate of infinity.

Primary closure of mammalian bites
Chen et al.
Acad Emerg Med. 2000 Feb;7(2):157-61.

This observational study evaluated 145 bites that underwent primary closure. Dogs made up more than half 88/145, but cat and human bites weer also included muddying the waters a bit with regard to the question at hand. majority of the wounds (57 percent) were on the head and neck. Only 8 patients (5.5%) developed infection despite having received antibiotics. The study was purely descriptive.

Antibiotic prophylaxis for mammalian bites
Medeiros et al.
Cochrane Database Syst Rev. 2001

This Cochrane review of 8 RCTs trials saw no statistically significant reduction of local wound infection in antibiotic prophylaxis (4%, 10 of 225 patients) versus controls (5.5%, 13 of 238 patients). The odd ratio was 0.74 (95% CI 0.30 to 1.85). Subanalysis did show that antiobiotic prophylaxis lowered the rate of infection in dog bites of the hand – 28% to 2%, with an OR = 0.10 (95% CI 0.01 to 0.86). The number needed to treat for hand infections was 4.

Primary repair of facial dog bite injuries in children
Wu et al.
Pediatr Emerg Care. 2011 Sep;27(9):801-3.

This descriptive report of 87 consecutive patients under age 18 with dog bite facial injuries, all of which received an antibiotic. None developed wound infection.

Why did I choose this topic for a Why We Do What We Do? Well, I do feel that this is an area where we practice reflexively. There really is limited evidence for what seems to be a routine practice. I do suspect that there is a bias towards patients who seek care in an ED/office. There are probably thousands of dog bites that happen at home and do not result in an ED visit. Do those get infected? Probably not in all honesty. Nevertheless, I am left in a bit of a quandry, as I feel that we should always be practicing evidence based medicine. So, based on the available literature it makes sense to put patients on prophylactic antibiotics for a dog bite wound if;

  • The wound has been primarily closed
  • Moderate to severe wounds (crush injury, devitalized or missing tissue, deep bites to tendons or bone)
  • Puncture wounds
  • Bites to the hands, feet, genitals or face
  • Asplenic or immunocompromised patients

If you are going to prescribe antibiotics (though not the exact focus of this post), I would use Amoxicillin/clavulanate, a broad spectrum cephalosporin if penicillin allergic, or clindamycin. If you aren’t going to prescribe prophylactic antibiotics have a good discussion with the patient and their family.