Bagratee et al. have published a well done randomised trial examining the effect of antibiotic prophylaxis on fever and infectious outcomes in elective caesarean delivery. This study received further attention when abstracted in the widely distributed ACOG Clinical Review1. The authors found no statistically significant differences in the outcomes of post-operative fever (8.3% antibiotics vs 7.9% placebo, RR 1.05, 95% CI [0.58–1.92]), endometritis (0.8% vs 1.7%, 0.5 [0.09–2.70]), and wound infection (12.5% vs 13.3%, 0.94 [0.59–1.49]). They concluded that prophylactic antibiotics do not prevent infection in low risk caesarean delivery.
We think it is important to interpret this study in light of other previously published similar papers. We have previously published a meta-analysis of randomised, placebo-controlled studies of antibiotic prophylaxis for elective caesarean deliveries2. Combining the seven studies available at the time, we noted large, statistically significant decreases in the risk of fever (relative risk, 0.25; 95% CI [0.14–0.44]) and endometritis (0.05 [0.01–0.38]) in the treatment arms. The pooled risk of endometritis in the placebo groups of these studies was 11%, much higher than noted in Dr Bagratee's study.
We repeated our literature search and updated the meta-analysis using the Bagratee study and one other recent paper3 that met our original criteria. We found statistically significant decreases in fever (0.47 [0.32–0.68]) and endometritis (0.18 [0.07–0.45], Fig. 1) with prophylaxis. Inclusion of these new studies does not alter the original conclusions and is consistent with the effectiveness of antibiotic prophylaxis.
Bagratee et al. noted a 50% decrease in endometritis in their population, an effect consistent with all other prior works. It is likely that they were observing the prophylactic effect of the antibiotics. However, this will not be a statistically significant difference given the lack of power of a study of this size to detect an outcome as rare as the authors have managed to achieve in their population in the absence of prophylaxis. Their reported decrease is actually consistent with other works showing that antibiotic prophylaxis decreases endometritis risk in women undergoing elective caesarean delivery, as well as our meta-analysis. Decisions on whether to administer prophylactic antibiotics should be made on the baseline risk of endometritis in the individual population, and antibiotic prophylaxis should be administered in populations with significant risk of endometritis.