Lactulose Is Associated With Decreased Risk of Clostridium difficile Infection in Decompensated Cirrhosis Anant Agarwalla, * Andrew Weber, ‡ Sonya Davey, § Keith Hamilton, k David Goldberg, ¶,# Andrew D. Rhim, ** ,b and Yu-Xiao Yang ¶,#,b *Department of Medicine, § Perelman School of Medicine, k Division of Infectious Diseases, ¶ Division of Gastroenterology, and # Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania, Philadelphia, Pennsylvania; ‡ Department of Internal Medicine, University of California, Los Angeles, California; and **Department of Gastroenterology, Hepatology, and Nutrition, M.D. Anderson Cancer Center, Houston, Texas C lostridium difficile infection (CDI) is the leading cause of nosocomial infectious diarrhea in devel- oped countries. 1 The composition of human gut micro- biome is associated with the development of CDI. This theory is bolstered by the increase in CDI risk after antibiotic exposure and the success of fecal microbiota transplant in treating refractory CDI. 2 Use of non-digestible oligosaccharides (eg, lactulose, oligofructose) as prebiotics may be a safe and effective approach to restoring or strengthening colonization resistance against CDI by selectively promoting the growth of a host’s indigenous microbiota. 3,4 Lactulose has been shown to alter the colonic microenvironment in ways that inhibit C difficile growth. 3 We sought to determine whether lactulose use was associated with a decreased risk of CDI. Methods We conducted a matched nested case-control study among end-stage liver disease (ESLD) patients admitted to an academic health system between 2005 and 2013. Patients were identified by the presence of ESLD on discharge diagnosis codes. 5 Admissions were excluded if the patient was hospitalized for less than 24 hours, had incomplete medical records, had undergone prior liver transplantation, or had a history of CDI. Cases were identified by incident diagnosis of CDI during admission on the basis of stool toxin or poly- merase chain reaction testing for toxigenic genes. For each case, controls were selected by using incidence density sampling matching on age, sex, admission date, and duration of hospital stay. The exposure of interest was at least 10 g/day of lactulose for more than 48 hours. Conditional logistic regression was used to calculate the odds ratio (OR) for CDI risk associated with lactulose use. We examined the following potential confounders: Model for End Stage Liver Disease score, intensive care unit stay, race, type of hepatic decompensation, use of rifaximin, metronidazole, or other antibiotics grouped by class. Variables were included in the multivariate model if their inclusion changed the crude OR for CDI risk by more than 5%. Results The study included 112 eligible cases matched with 928 controls. The 2 groups differed in history of hepatic encephalopathy (HE), spontaneous bacterial perito- nitis (SBP), and use of rifaximin, metronidazole, and cephalosporins (Table 1). The prevalence of lactulose use was significantly lower among cases than controls (crude OR, 0.48; 95% confidence interval [CI], 0.31–0.74; P < .001). A diag- nosis of HE as well as exposure to rifaximin, metroni- dazole, and cephalosporins significantly changed the crude OR by >5%. HE was automatically excluded from the regression model because it was highly correlated with rifaximin use. The adjusted OR accounting for these variables was 0.60 (95% CI, 0.37–0.97; P ¼ .04). In secondary analyses, the effect estimates were un- changed when excluding patients who received rifaximin (adjusted OR, 0.52; 95% CI, 0.31–0.89; P ¼ .02) or had SBP (OR, 0.60; 95% CI, 0.36–1.01; P ¼ .053). There was no difference in the frequency of CDI testing between the lactulose users (23%) vs non-users (19%), P ¼ .3. Discussion Lactulose use was associated with a reduced risk of CDI among hospitalized ESLD patients. This association was independent of comorbidities or concurrent b Andrew Rhim and Yu-Xiao Yang share co-senior authorship. Abbreviations used in this paper: CDI, Clostridium difficile infection; CI, confidence interval; ESLD, end-stage liver disease; HE, hepatic enceph- alopathy; OR, odds ratio; SBP, spontaneous bacterial peritonitis. Most current article © 2017 by the AGA Institute 1542-3565/$36.00 http://dx.doi.org/10.1016/j.cgh.2017.01.012 Clinical Gastroenterology and Hepatology 2017;15:953–954