Antibiotic prophylaxis for percutaneous endoscopic gastrostomy insertion in patients with non-malignant disease: authors' reply


  • AP&T correspondence columns are restricted to letters discussing papers that have been published in the journal. A letter must have a maximum of 300 words, may contain one table or figure, and should have no more than 10 references. It should be submitted electronically to the Editors via

Sirs, We are grateful for the interest in our paper shown by Hull and Robins.

We agree that the incidence of peristomal sepsis in our study may be a little higher than that reported in some other series, but this does not invalidate our conclusions that antibiotic prophylaxis reduces both peristomal and systemic sepsis in patients with non-malignant disease undergoing PEG insertion.

We agree that the use of scoring systems to diagnosis peristomal sepsis is prone to observer bias. However, the presence of pus as a diagnostic criterion for peristomal sepsis was not prone to observer bias in our study. There were 12 patients who had sepsis diagnosed by this criterion alone, i.e. with a total ‘peristomal sepsis score’ otherwise below the threshold level of 8 for diagnosing sepsis. Of these 12 patients (nine from placebo and three from the antibiotic group), 11 had positive microbiological cultures from their PEG site swabs, so this was clearly genuine pus rather than refluxed PEG feed.

The correspondents’ second paragraph is difficult to follow, but it is nevertheless a little difficult to accept their implication that nasal MRSA colonization, being the only identified risk factor for surgical infections in a cohort of orthopaedic patients in another study, is of direct relevance to the conclusions of our study. As shown in our Table 4, the subjects who developed MRSA peristomal sepsis in our study actually constituted the minority of those with positive cultures from PEG site infections.

We agree with the correspondents that policies for antibiotic prophylaxis might be best guided by local microbial prevalence patterns, and screening for nasal MRSA colonization prior to PEG insertion would certainly constitute good practice.

However, their final comment is ambiguous in failing to define a ‘significant’ PEG site infection and in failing to cite the incidence of MRSA infections in their series. It could be argued that in our own study there were no ‘significant’ PEG site infections even in the placebo group, as no patient required PEG removal for this reason. However, one of the goals in improving outcome after PEG insertion must be to minimize the incidence of systemic sepsis, and our study has shown that this is achieved with intravenous antibiotic prophylaxis. It seems unlikely that this would be achieved with topical prophylaxis only, although this point may deserve further study.