Rifaximin Treatment of Pathogen‐Negative Travelers’ Diarrhea

Herbert L. DuPont MD, Robert Haake PhD, David N. Taylor MD, Charles D. Ericsson MD, Zhi‐Dong Jiang PhD, Pablo C Okhuysen MD, Robert Steffen MD
DOI: http://dx.doi.org/10.1111/j.1708-8305.2006.00084.x 16-19 First published online: 1 January 2007


Background Antibacterial drugs appear to be effective in shortening the illness of a majority of cases of travelers’ diarrhea.

Methods This was a subanalysis from two randomized, double‐blind, placebo‐controlled trials in adult travelers with acute diarrhea treated with rifaximin 200 mg three times a day or placebo for 3 days. Efficacy was assessed by the interval beginning with the first dose of medication and ending with the last unformed stool passed after becoming well [time to last unformed stool (TLUS)]; number of unformed stools passed; percent with clinical improvement; and incidence of wellness achieved.

Results Stool pathogens were not identified in pretreatment samples in 122 of 322 (38%) patients and 106 of 230 (46%) randomized to rifaximin and placebo, respectively. Among pathogen‐negative patients, rifaximin was more effective than placebo for median TLUS (33 vs 68 h, p < 0.005), mean number of unformed stools passed (6.5 vs 8.6, p < 0.0001), and clinical wellness (77% vs 61%, p = 0.01). The adverse‐event profiles between rifaximin and placebo were similar.

Conclusions More than one third of patients with travelers’ diarrhea had pathogen‐negative illness. Rifaximin was effective in treating the illness without associated side effects. These results are consistent with the hypothesis that undetected bacterial pathogens are the most likely cause of travelers’ diarrhea without definable cause.

Bacterial pathogens are the most common cause of travelers’ diarrhea1–3 with antibacterial therapy representing the optimal empiric management.4–6 Enteropathogens cannot be identified from pretreatment stool samples in 24% to 50% of patients in clinical studies of travelers’ diarrhea.2,7,8 It is assumed but unproven that unidentified bacterial pathogens are the likely cause of illness in a majority of travelers’ diarrhea cases having negative pretreatment stool samples despite extensive microbiological workup. There is evidence to suggest that known pathogens such as diarrheagenic Escherichia coli could have been missed as well as unknown bacterial pathogens in this illness category.9

Rifaximin is a poorly nonabsorbed (<0.4%), oral antibiotic introduced in the United States in 2004 for the treatment of travelers’ diarrhea caused by noninvasive strains of diarrheagenic E coli.10 Rifaximin has a tolerability profile similar to placebo and has been employed in multiple studies at various doses and treatment durations. Rifaximin has not been associated with systemic drug interactions or clinically relevant bacterial resistance10–12 and is currently available in 18 countries.

While travelers’ diarrhea of uncertain etiology has been shown to respond to antibiotics,7,8,13 the current study systematically evaluated data from two placebo‐controlled trials to determine the efficacy and safety of rifaximin in the treatment of pathogen‐negative illness.


Post hoc analysis evaluated data from two randomized, double‐blind, placebo‐controlled clinical trials of rifaximin 200 mg three times a day for 3 days for the treatment of travelers’ diarrhea occurring during short‐term stay in Mexico, Guatemala, or Kenya;14 or Mexico, Guatemala, and India.15 A breakdown of etiologic agents found in the cases of diarrhea is available in the second publication. Patients with stool samples in which pathogens could not be identified prior to treatment were further evaluated for efficacy and safety.

Diarrhea was defined as passage of greater than or equal to three unformed stools in 24 hours with more than or equal to one sign per symptom of enteric infection including abdominal cramps and pain, nausea, vomiting, fever, tenesmus, and moderate to severe intestinal gas‐related symptoms, lasting £72 hours.16 Clinical efficacy parameters in the trial included: median time from taking the first dose of study drug until passage of last unformed stool after which wellness was achieved or time to last unformed stool (TLUS); number of unformed stools passed during the subjects’ illness; percent of subjects not achieving wellness during the 5 days of therapy; and adverse experiences.

In the trials, conventional bacterial and parasitic pathogens were sought in addition to enterotoxigenic E coli (ETEC) and enteroaggregative E coli (EAEC) by published methods.1,2

Statistical methods

For TLUS, Kaplan–Meier plots were created over the time period of the study and compared with Wald statistic on Cox proportional hazards model; for number of unformed stools passed, repeated measures analysis of variance model was employed; and for incidence of wellness, Cochran–Mantel–Haenszel test was applied. For demographic comparisons, Fisher’s exact test for categorical variables and t‐test for continuous variables were employed.


Data from the two trials indicated that pathogens could not be identified in pretreatment stool samples in 122 of 322 (37.9%) patients randomized to received rifaximin and 106 of 230 (46.1%) of patients randomized to receive placebo, respectively. There were no relevant differences between rifaximin and placebo in demographic characteristics for patients with pathogen‐negative illness: 1. Mean age in years (SD)—28.7 (11.5) versus 28.7 (11.3), p = 0.9822; 2. Mean weight in kg (SD)—71.2 (15.4) versus 70.0 (14.6), p = 0.5674; 3. Male, n (%)—64 (52.5) versus 52 (49.1), p = 0.6905; and 4. Race, n (%)—white 95 (77.9) versus 84 (79.2) and Hispanic 19 (15.6) versus 13 (12.3), p = 0.8721.

Rifaximin significantly shortened the duration of post initiation of therapy diarrhea in patients with pathogen‐free illness, demonstrated by faster median TLUS (p < 0.005; Table 1). In addition, treatment with rifaximin significantly decreased the mean number of unformed stools passed during the 5 days after enrollment in patients with pathogen‐negative illness compared with the placebo (p < 0.0001). Furthermore, rifaximin more effectively led to clinical wellness among patients with pathogen‐negative illness compared with placebo (p = 0.01; Table 1).

View this table:
Table 1

Clinical response of travelers’ diarrhea without definable cause to rifaximin treatment versus placebo14,15

ParameterRifaximin (n =122)Placebo (n =106)p Value
Median TLUS* (h)33680.0015
Mean number of unformed stools during illness6.58.60.0001
Patients achieving wellness during study (%)77610.01
  • TLUS = time to last unformed stool.

  • * Time from initiation of treatment until passage of the last unformed stool after which subjects are declared well.

The adverse‐event profile of rifaximin did not differ from that of placebo among patients with pathogen‐negative illness.


In this combined analysis of two studies, more than 40% of patients in the two placebo‐controlled clinical trials of rifaximin treatment of travelers’ diarrhea had pathogen‐negative illness.14,15 In these studies, testing for important bacterial and protozoal parasite causes of diarrhea of travelers were determined by validated, published methods.2 Note, however, that in these studies, viral enteropathogens were not sought.

In the current subanalysis of pathogen‐negative diarrheal illness in travelers, rifaximin was more effective than placebo in shortening the duration of illness and decreasing the passage of unformed stools. These data confirm earlier observations that antibacterial drugs clinically reduce the duration and intensity of pathogen‐negative travelers’ diarrhea. While pathogen‐negative travelers’ diarrhea is of uncertain etiology, it may be caused by a number of enteropathogens, including ETEC and EAEC. In the studies included herein, we used DNA probe methods to detect ETEC.17,18 However, if we had employed the more sensitive polymerase chain reaction (PCR) method for ETEC detection in studies of travelers’ diarrhea etiology, we could have possibly increased the identification of this common pathogen in our studies.19 Although more sensitive methods help to identify cases of ETEC, the pathogen for a large percentage of patients with travelers’ diarrhea remains unidentified. In a recent study of US students in Mexico, noroviruses were responsible for 17% of cases of diarrhea, making it the third most common pathogen after ETEC and EAEC. In the case of noroviruses, there are no data to support a likely clinical response to antibacterial drugs.

Our results are consistent with the probability that undetected bacterial pathogens cause a majority of cases of illness in pathogen‐negative travelers’ diarrhea. Current methods of detecting diarrheagenic E coli (ETEC and EAEC) include the study of five E coli‐like organisms from each diarrheal stool by DNA probe, enzyme linked immunosorbent assay (ELISA) or PCR, and the HEp‐2 cell assay.20,21 Additional studies are needed in travelers’ diarrhea to determine if diarrheagenic E coli would be better identified by studying a larger number of coliform bacteria from diarrheal stools or by direct testing of stool with molecular methods aimed at specific pathogens or virulence products. In addition, pathogens other than strains of diarrheagenic E coli may be the cause of travelers’ diarrhea and should also be considered, including enterotoxigenic Bacteroides fragilis and Arcobacter.22,23

Antibacterial drugs remain the mainstay of treatment of acute travelers’ diarrhea. Rifaximin is effective against bacterial pathogens and in the pathogen‐negative cases as seen in the analysis of these two randomized, double‐blind, placebo‐controlled clinical trials.

There is no ideal drug for all cases of travelers’ diarrhea. Rifaximin has a safety profile similar to placebo and its nonabsorbed nature makes it unlikely to interfere with other drugs or to induce clinically important antimicrobial resistance. Also rifaximin exhibits a low potential for inducing clinical resistance or producing major alterations of intestinal flora.24,25 While fluoroquinolones may transiently eliminate coliform flora26 and currently represent a major predisposing class of drugs leading to the development of antibiotic‐associated Clostridium difficile diarrhea and colitis27,28 , this class of drugs has an advantage of greater activity against invasive bacterial pathogens in patients with travelers’ diarrhea, particularly that acquired in Southern Asia.

Future studies should be directed in detecting the cause of these cases of otherwise pathogen‐unknown illness. While most cases will undoubtedly be caused by as yet undiscovered bacterial pathogens, viral pathogens and in a small percentage parasitic pathogens are likely to be identified.

Declaration of interests

H. L. D. and C. D. E. are primary or coinvestigators of Salix Pharmaceuticals and Pfizer funded research. They are also consultants for and have received honoraria for speaking for Salix Pharmaceuticals. R. H. was a Salix Pharmaceuticals employee at time of study (June 2002 to January 2006). D. N. T. is currently a full‐time employee for Salix Pharmaceuticals. C. D. E. is primary or coinvestigator of Salix Pharmaceuticals and Pfizer funded research. He is also a consultant for and has received honoraria for speaking for Salix Pharmaceuticals. P. C. O. has accepted fees for speaking and consulting from Salix Pharmaceuticals. R. S. has accepted fees for speaking, organizing and chairing education, consulting and/or serving on advisory boards, also reimbursement for attending meetings and funds for research from Astral, Berna Biotech/Crucell, Baxter, Chiron Behring (now Novartis Vaccine), GlaxoSmithKline, Novartis, Optimer, Salix Pharmaceuticals, and/or Sanofi Pasteur MSD. He owns shares in Novartis.


This work was supported by Salix Pharmaceuticals, Morrisville, NC, and in part by grants from Public Health Service (Grant DK 56338), which funds the Texas Gulf Coast Digestive Diseases Center; NIH NCRR to the University of Texas GCRC: General Clinical Research Center, Grant M01‐RR 02558 and NIH NIAID R01 AI54948.


  • This article has been presented at the National Meeting of the American Society of Tropical Medicine and Hygiene, Washington, DC, December 14, 2005.

    Guest Editor: David R. Shlim, MD


View Abstract