Randomized controlled trial of selective bowel decontamination for prevention of infections following liver transplantation.

Nonabsorbable antibiotics for selective bowel decontamination (SBD) sometimes are administered to liver transplant patients to prevent postoperative infections, but the efficacy of SBD is not known. Accordingly, we prospectively studied 69 patients randomly assigned to receive conventional prophylaxis with systemic antibiotics (control patients) or conventional prophylaxis plus oral nonabsorbable antibiotics for SBD (SBD patients). Overall rates of bacterial and/or yeast infections were nearly equal among control patients (42%) and SBD patients (39%). However, the infection rate at SBD key sites (abdomen, bloodstream, surgical wound, and lungs) was lower among patients who received the SBD regimen > or = 3 days before transplantation (23%) than among control patients (36%). Administration of the SBD regimen was complicated by gastrointestinal intolerance and noncompliance but not by increased stool colonization with antibiotic-resistant gram-negative bacilli. Practical problems associated with administering an SBD regimen to patients awaiting cadaver liver transplants limit the regimen's usefulness, but we found a trend toward reduced key site infection when the regimen was given > or = 3 days before transplantation.