Circumcision and neonatal tetanus: disclosure of risk and its reduction by topical antibiotics

Background Previous case-control studies have paradoxically suggested that circumcisions protect against neonatal tetanus (NNT), but these observations have not been adjusted for differences in the length of survival of cases and controls. Methods Boy cases (n = 133) and their sex-matched controls (n = 399) were extracted from a population-based study of NNT undertaken in Punjab Province, Pakistan. In the resulting file, circumcisions were censored such that analysis was restricted to only those that occurred before onset in cases or before age of onset in the matched case for controls. The effect of topical antibiotics in circumcision wounds was then evaluated. Results After adjusting for confounders, circumcision before onset posed a significant risk for NNT (matched odds ratio [OR] = 3.1, 95% CI: 1.2-8.0). The risk of NNT in those circumcised before onset and treated with topical antibiotics did not differ significantly from the referent group who had not been circumcised before onset (matched OR = 1.1, 95% CI: 0.2-6.8), whereas the lack of topical use was associated with significant risk (matched OR = 4.2, 95% CI: 1.4-12.6). This suggests that topical antibiotics are likely to be highly effective in preventing NNT from circumcision wounds. We estimated an overall risk of about 16 fatal NNT cases per 1000 live boy births with circumcision wounds that were not protected by topical antibiotics, and that circumcision and umbilical wounds each accounted for about half of this overall risk in these boys. Conclusions Topical antibiotics should be routinely applied to all wounds created by traditional circumcisions, to prevent NNT and sepsis from these frequently unsterile procedures.