Skip navigation

Antibiotics do not prevent infection in simple wounds

The team that brought us the analysis of antibiotics in dog bites ( Bandolier 16 ) has also produced a meta-analysis of the use of antibiotics in simple non-bite wounds [1].

Inclusion criteria

The search was for randomised trials of prophylactic systemic antibiotics for non-bite wounds managed in emergency departments. The date of the last search was December 1993. Studies were included if the authors stated they had randomly assigned patients with uninfected wounds to an antibiotic treatment or a control group, and then followed patients to see if the wound became infected.
Nine studies were found; two were excluded. The seven included studies had information on 939 patients treated with antibiotics and 762 controls and were published between 1975 and 1983. Five studies were done in the UK and two in the United States. All studies used either cephalosporins or penicillins, some of which were resistant to penicillinase. Five studies were limited to hand wounds, and five to sutured wounds. The nature of the controls in these studies was not clearly described in the meta-analysis.


The rates of follow up in the seven studies were high, between 90 and 95%. The cumulative rate of infection in controls was 1.1% to 12% with a mean of 6%. Patients on antibiotics had a higher rate of infection, with a combined odds ratio of 1.2 (95%CI 0.8 to 1.8). Subgroup analysis did not reveal any groups which had any significant difference from controls, which was hardly surprising since no single study showed any difference between patients treated with systemic antibiotics and control subjects. So in 1,204 patients treated with penicillinase resistant antibiotics the odds ratio was 1.0 (0.6 to 1.7).


These results are quite negative. There is no evidence that antibiotics prevent infection in simple wounds. After dog bites, antibiotics were effective, with a pooled NNT of 16 ( Bandolier 16 ).


  1. P Cummings, MA Del Beccaro. Antibiotics to prevent infection of simple wounds: a meta-analysis of randomized studies. American Journal of Emergency Medicine 1995 13: 396-400.

previous or next story in this issue