0041-1337/02/7312-1904/0 TRANSPLANTATION Vol. 73, 1904–1909, No. 12, June 27, 2002 Copyright © 2002 by Lippincott Williams & Wilkins, Inc. Printed in U.S.A. A RANDOMIZED, PROSPECTIVE, DOUBLE-BLINDED EVALUATION OF SELECTIVE BOWEL DECONTAMINATION IN LIVER TRANSPLANTATION WALTER C. HELLINGER, 1,9 JOSEPH D. YAO, 1 SALVADOR ALVAREZ, 1 JANIS E. BLAIR, 2 JOHN J. CAWLEY, 3 CARLOS V. PAYA, 4 PETER C. O’BRIEN, 5 JAMES R. SPIVEY, 6 ROLLAND C. DICKSON, 6 DENISE M. HARNOIS, 6 DAVID D. DOUGLAS, 7 CHRISTOPHER B. HUGHES, 8 JUSTIN H. NGUYEN, 8 DAVID C. MULLIGAN, 7 AND JEFFREY L. STEERS 8 Division of Infectious Diseases, Mayo Clinic, Jacksonville, FL; Division of Infectious Diseases, Mayo Clinic, Scottsdale, AZ; Microbiology Laboratory, Mayo Clinic, Jacksonville, FL; Division of Infectious Diseases and Internal Medicine, Mayo Clinic, Rochester, MN; Section of Biostatistics, Mayo Clinic, Rochester, MN; Division of Gastroenterology and Hepatology, Mayo Clinic, Jacksonville, FL; Division of Transplantation Medicine, Mayo Clinic, Scottsdale, AZ; and Department of Surgery, Mayo Clinic, Jacksonville, FL Background. Bacterial infection is a frequent, mor- bid, and mortal complication of liver transplantation. Selective bowel decontamination (SBD) has been re- ported to reduce the rate of bacterial infection after liver transplantation in uncontrolled trials, but bene- fits of this intervention have been less clear in con- trolled studies. Methods. Eighty candidates for liver transplantation were randomly assigned in a double-blinded fashion to an SBD regimen consisting of gentamicin 80 mgpolymyxin E 100 mgnystatin 2 million units (37 patients) or to nystatin alone (43 patients). Both treat- ments were administered orally in 10 ml (increasing to 20 ml, according to predefined criteria), four times daily, through day 21 after transplantation. Anal fecal swab cultures were performed on days 0, 4, 7, and 21. Rates of infection, death, and charges for medical care were assessed from day 0 through day 60. Results. More than 85% of patients in both treatment groups began study treatment more than 3 days before transplantation. Rates of infection (32.4 vs. 27.9%), death (5.4 vs. 4.7%), or charges for medical care (medi- an $194,000 vs. $163,000) were not reduced in patients assigned to SBD. On days 0, 4, 7, and 21, growth of aerobic gram-negative flora in fecal cultures of pa- tients assigned to SBD was significantly less than that of patients taking nystatin alone; growth of aerobic gram-positive flora, anaerobes, and yeast was not sig- nificantly different. Conclusion. Routine use of SBD in patients undergo- ing liver transplantation is not associated with signif- icant benefit. INTRODUCTION Selective bowel decontamination refers to interventions that reduce the aerobic gram-negative bacterial and yeast populations of the gastrointestinal tract, without elimination of the anaerobic microbial flora. Treatment typically consists of the administration of unabsorbed oral antibiotics that have selective antimicrobial activity, with or without a brief period of systemic antibiotic therapy. The goal of selective bowel decontamination is the preven- tion of infection in patients at high risk for hospital-acquired infection. More than 30 studies of selective bowel decontam- ination have since been conducted in critically ill, hospital- ized patients. Recent metaanalyses (1, 2) of these studies confirmed a reduction of risk of pneumonia conferred by the use of selective bowel decontamination, a finding of earlier analyses (3–5), and observed a small survival benefit con- fined to certain subpopulations of critically ill patients (1) or to those treated with systemic as well as enteral antimicro- bial therapy (2). However, the use of selective bowel decon- tamination has not been widely adopted in intensive care units in the United States because of uncertainty regarding its net benefit to patients and because of concern that it may promote the spread of antibiotic resistance. Nonetheless, bacterial infections are frequent, morbid, and mortal complications of liver transplantation (6–8), and some form of selective bowel decontamination is often used in centers performing liver transplantation (9 –13). This prac- tice is related partly to early reports of effectiveness in un- controlled studies (14 –18) and partly to the theoretical ap- peal of applying selective bowel decontamination to liver transplantation, where early postoperative infections are of- ten of gut origin and related to circumscribed interventions of the transplant procedure (10). However, of the four random- ized, prospective studies of selective bowel decontamination in liver transplantation (19 –22), only two enrolled sufficient numbers of patients for analysis. Of these, only one demon- strated a decrease in bacterial infections with selective bowel decontamination (19), although benefit in the second was confined to a subset of all treated patients (20). A randomized, prospective, double-blinded study of selec- tive bowel decontamination in liver transplantation was therefore performed to carefully assess the impact of this 1 Division of Infectious Diseases, Mayo Clinic, Jacksonville, FL. 2 Division of Infectious Diseases, Mayo Clinic, Scottsdale, AZ. 3 Microbiology Laboratory, Mayo Clinic, Jacksonville, FL. 4 Division of Infectious Diseases and Internal Medicine, Mayo Clinic, Rochester, MN. 5 Section of Biostatistics, Mayo Clinic, Rochester, MN. 6 Division of Gastroenterology and Hepatology, Mayo Clinic, Jack- sonville, FL. 7 Division of Transplantation Medicine, Mayo Clinic, Scottsdale, AZ. 8 Department of Surgery, Mayo Clinic, Jacksonville, FL. 9 Address correspondence to: W.C. Hellinger, MD, Division of In- fectious Diseases, Mayo Clinic, 4500 San Pablo Road, Jacksonville, FL 32224. 1904