I read with interest the paper in which Bagratee et al.1 concluded that antibiotic prophylaxis with cefoxitin did not reduce maternal infectious morbidity in elective caesarean section. These results conflict with the Zurich University Hospital data obtained in a similar number of women2 and prompt the following comments:
- 1Elective vs non-elective section is only one of several possible criteria of pre-operative risk in caesarean section. However, Bagratee et al. restrict themselves to the single criterion of elective section (excluding rupture of the membranes) in the apparent belief that this will leave them with an exclusively low risk population. Yet their infection rates—especially for wound infection and fever—are too high to justify such an assumption, compared not only to our own figures but also to those in the literature. The absence of difference between the placebo and cefoxitin groups may have been due to both groups containing both high and low risk patients, and not necessarily in identical proportions.
- 2Cefoxitin may not have been used at its optimal dosage schedule. At the dose given (2 g, in a single iv injection), cefoxitin remains above its minimal inhibitory concentration in tissue and urine for a maximum of only 6–8 hours, due to its short half-life (0.8 hours). Lower infection rates can be achieved using higher and, above all, more frequent doses (e.g. 3 × 1 g2 or 3 × 2 g3) or using a long acting antibiotic as ceftriaxone given as a single dose of 1 g2.
- 3Wound infection—13.3% (placebo) and 12.5% (cefoxitin)—was three times more prevalent than in our study (including our high risk patients). These data, which suggest an influence of pre-operative risk, are surprising: in low risk patients, the rates should have been lower in both study groups.
- 4Endometritis is a typical post-operative risk reduced by peri-operative antibiotic prophylaxis, as the data of Bagratee et al. suggest (cefoxitin: 0.8% vs placebo: 1.7%). The difference may not be statistically significant—endometritis is the least frequent infection in all studies—but a 50% decrease is an appreciable clinical bonus.
- 5Urinary tract infection rates were very low. They would have been higher, although still probably similar in both groups, had they included asymptomatic bacteriuria, a risk which is particularly important to avoid after a pregnancy complicated by such indications for caesarean section as prematurity, diabetes and pre-eclampsia. The Zurich University Hospital data showed that a long acting antibiotic such as ceftriaxone significantly reduces the rates of asymptomatic bacteriuria vs cefoxitin2.
- 6The anaesthesia data are inconsistent. The text states that ‘over 90% of women in both groups… had spinal anaesthesia’, whereas Table 3 shows that only 2.1% and 5.1% had spinal anaesthesia, while 93.8% in the placebo group had an epidural and 95% of the cefoxitin group had general anaesthesia. I suspect that the text is correct and that the tabular data are mistakes. However, as spinal and epidural anaesthesia differ from general anaesthesia in their effects on post-operative mortality4, readers need reassurance that infection rates were compared only between groups receiving similar anaesthesia.