Objective: To determine the incidence of Clostridium difficile (CD) diarrhoea in feed-intolerant, critically ill patients who received erythromycin-based prokinetic therapy.
Design and setting: Prospective observational study in a mixed intensive care unit.
Methods: The development of diarrhoea (> 3 loose, liquid stool per day with an estimated total volume > or = 250ml/day) was assessed in 180 consecutive critically ill patients who received prokinetic therapy (erythromycin only, n = 53; metoclopramide, n 37; combination erythromycin/metoclopramide, n = 90) for feed intolerance. Stool microscopy, culture and CD toxin assay were performed in all patients who developed diarrhoea during and after prokinetic therapy. Diarrhoea was deemed to be related to CD infection if CD toxin was detected.
Results: Demographics, antibiotic use and admission diagnosis were similar amongst the three patients groups. Diarrhoea developed in 72 (40%) patients, 9.9 +/- 0.8 days after commencement of therapy, none of whom was positive for CD toxin or bacterial infection. Parasitic infections were found in four aboriginal men from an area endemic for these infections. Diarrhoea was most prevalent in patients who received combination therapy (49%) and was more common than in those who received erythromycin alone (30%) and metoclopramide alone (32%). Diarrhoea was short-lasting with a mean duration of 3.6 +/- 1.2 days.
Conclusions: In critical illness, diarrhoea following the administration of erythromycin at prokinetic doses is not associated with CD but may be related to pro-motility effects of the agent. Prokinetic therapy should be stopped at the onset of diarrhoea and prophylactic use should be strictly avoided.