Meta-analysis of randomized and quasi-randomized clinical trials of topical antibiotics after primary closure for the prevention of surgical-site infection

    loading  Checking for direct PDF access through Ovid

Abstract

Background

Surgical-site infections (SSIs) increase patient morbidity and costs. The aim was to identify and synthesize all RCTs evaluating the effect of topical antibiotics on SSI in wounds healing by primary intention.

Methods

The search included Ovid MEDLINE, Ovid Embase, the Cochrane Wounds Specialized Register, Central Register of Controlled Trials and EBSCO CINAHL from inception to May 2016. There was no restriction of language, date or setting. Two authors independently selected studies, extracted data and assessed risk of bias. When sufficient numbers of comparable trials were available, data were pooled in meta-analysis.

Results

Fourteen RCTs with 6466 participants met the inclusion criteria. Pooling of eight trials (5427 participants) showed that topical antibiotics probably reduced the risk of SSI compared with no topical antibiotic (risk ratio (RR) 0·61, 95 per cent c.i. 0·42 to 0·87; moderate-quality evidence), equating to 20 fewer SSIs per 1000 patients treated. Pooling of three trials (3012 participants) for risk of allergic contact dermatitis found no clear difference between antibiotics and no antibiotic (RR 3·94, 0·46 to 34·00; very low-quality evidence). Pooling of five trials (1299 participants) indicated that topical antibiotics probably reduce the risk of SSI compared with topical antiseptics (RR 0·49, 0·30 to 0·80; moderate-quality evidence); 43 fewer SSIs per 1000 patients treated. Pooling of two trials (541 participants) showed no clear difference in the risk of allergic contact dermatitis with antibiotics or antiseptic agents (RR 0·97, 0·52 to 1·82; very low-quality evidence).

Conclusion

Topical antibiotics probably prevent SSI compared with no topical antibiotic or antiseptic. No conclusion can be drawn regarding whether they cause allergic contact dermatitis.

Conclusion

May work, poor evidence

Related Topics

    loading  Loading Related Articles