Alimentary Pharmacology and Therapeutics
Systematic review with meta-analysis: the effects of rifaximin
in hepatic encephalopathy
N. Kimer*,†, A. Krag‡, S. Møller†, F. Bendtsen* & L. L. Gluud*
*Gastrounit, Medical Division,
Copenhagen University Hospital
Hvidovre, Hvidovre, Denmark.
†
Department of Clinical Physiology
and Nuclear Medicine, Centre of
Functional Imaging and Research,
Copenhagen University Hospital
Hvidovre, Hvidovre, Denmark.
‡
Department of Gastroenterology,
Odense University Hospital, University
of Southern Denmark, Odense,
Denmark.
Correspondence to:
Dr N. Kimer, Gastrounit, Medical
Division 360, Copenhagen University
Hospital Hvidovre, Kettegaard Alle
30, 2650 Hvidovre, Denmark.
E-mail: nina.kimer@regionh.dk
Publication data
Submitted 6 March 2014
First decision 27 March 2014
Resubmitted 29 April 2014
Accepted 30 April 2014
EV Pub Online 21 May 2014
As part of AP&T’s peer-review process, a
technical check of this meta-analysis was
performed by Dr Y. Yuan.
ª 2014 John Wiley & Sons Ltd
doi:10.1111/apt.12803
SUMMARY
Background
Rifaximin is recommended for prevention of hepatic encephalopathy (HE).
The effects of rifaximin on overt and minimal HE are debated.
Aim
To perform a systematic review and meta-analysis of randomised controlled
trials (RCTs) on rifaximin for HE.
Methods
We performed electronic and manual searches, gathered information from
the U.S. Food and Drug Administration Home Page, and obtained unpub-
lished information on trial design and outcome measures from authors and
pharmaceutical companies. Meta-analyses were performed and results pre-
sented as risk ratios (RR) with 95% confidence intervals (CI) and the num-
ber needed to treat. Subgroup, sensitivity, regression and sequential
analyses were performed to evaluate the risk of bias and sources of hetero-
geneity.
Results
We included 19 RCTs with 1370 patients. Outcomes were recalculated
based on unpublished information of 11 trials. Overall, rifaximin had a
beneficial effect on secondary prevention of HE (RR: 1.32; 95% CI 1.06–
1.65), but not in a sensitivity analysis on rifaximin after TIPSS (RR: 1.27;
95% CI 1.00–1.53). Rifaximin increased the proportion of patients who
recovered from HE (RR: 0.59; 95% CI: 0.46–0.76) and reduced mortality
(RR: 0.68, 95% CI 0.48–0.97). The results were robust to adjustments for
bias control. No small study effects were identified. The sequential analyses
only confirmed the results of the analysis on HE recovery.
Conclusions
Rifaximin has a beneficial effect on hepatic encephalopathy and may reduce
mortality. The combined evidence suggests that rifaximin may be consid-
ered in the evidence-based management of hepatic encephalopathy.
Aliment Pharmacol Ther 2014; 40: 123–132
123
N. Kimer et al.
INTRODUCTION
Hepatic encephalopathy (HE) is a neuropsychiatric syn-
drome seen in patients with advanced liver disease. The
symptoms include neurological impairment, ranging
from minor defaults in orientation and coordination to
deep coma.1 Hospitalisations are common and the
impact on the quality of life among out-patients is signif-
icant and 1-year survival rates are below 50%.2 The type
of HE can be divided into clinically overt and minimal,
which requires neurological and functional testing.3, 4
The development of HE involves ammonia, inflamma-
tion and nitrosative stress.1, 5 Current treatment regi-
mens aim to reduce the production or increase the
elimination of ammonia.1, 6–11 Several trials have evalu-
ated the effects of branched-chain amino acids and rifax-
imin.12, 13 A meta-analysis of randomised controlled
trials (RCTs) on branched-chain amino acids found a
beneficial effect on manifestations of HE, but no effect
on mortality.14 A large RCT found that rifaximin pre-
vents HE episodes.15 The results of RCTs on rifaximin
for minimal or overt HE are equivocal and none pro-
vides firm evidence to allow for clinical recommenda-
tions. Meta-analyses on rifaximin for patients with HE
have also generated inconsistent results. One meta-analy-
sis including seven RCTs found that rifaximin had a
beneficial effect on HE compared to non-absorbable di-
saccharides.16 A subsequent meta-analysis including eight
controlled trials concluded that the effect of rifaximin
and non-absorbable disaccharides on HE was similar.17
A larger meta-analysis on 12 RCTs also found that the
effect of rifaximin was similar to conventional treat-
ments.12 The combined evidence therefore remains
inconclusive. Accordingly, we performed a systematic
review with meta-analyses of RCTs on rifaximin vs. pla-
cebo, disaccharides or other antibiotics for overt, mini-
mal and recurrent HE.
METHODS
Search strategy and data extraction
This review was carried out and reported based on a reg-
istered (Prospero no CRD42013005776) protocol devel-
oped using the methods described in the Cochrane
Handbook for Systematic Reviews of Interventions and
the PRISMA Statement for Reporting Systematic Reviews
and Meta-analyses.18, 19 The literature searches were per-
formed with help from the Cochrane Hepato-Biliary
Group (Table S1). The electronic searches included The
Cochrane Library, Medline, Embase, The Cochrane Hep-
ato-Biliary Group Controlled Trials Register and Science
124
Citation Index Expanded. Manual searches included
scanning of bibliographies in relevant articles and annual
conference proceedings from the American Association
for the study of Liver Diseases (AASLD), European
Association for the Study of the Liver (EASL), the Asian
Pacific Association for the study of the Liver (APASL),
the International Society for Hepatic Encephalopathies
and Nitrogen Metabolism (ISHEN) and associations
hereunder. Electronic searches were performed in
December 2013 and updated April 1, 2014. Additional
trials were identified through electronic registers includ-
ing the World Health Organization International Clinical
Trials Registry Platform. All corresponding authors as
well as the pharmaceutical companies sponsoring or
assisting in conduction of trials were contacted for addi-
tional data. Information was also sought retrieved
through the Food and Drug Administration (www.fda.
gov) and European Medicines Agency websites (www.
ema.europa.eu).
Randomised clinical trials were included irrespective
of language, publication status, dose, type or duration of
therapy. We included RCTs on prevention of HE and
trials on patients with overt or minimal HE. We
included RCTs on patients with TIPSS in a sensitivity
analysis.
Three authors (NK, AK and LG) extracted data in an
independent manner. The extracted data included year
of publication, study design and population, dose and
duration of treatment, number of patients in each group,
number of dropouts, patient age and sex. The primary
outcomes were HE (prevention, recovery and improved
manifestations), mortality and adverse events. Secondary
outcome measures included surrogate markers of HE
(psychometric tests and ammonia levels).
Assessment of risk of bias
Two authors (NK and LG) performed bias assessment
independently. The control of bias control was assessed
based on the recommendations specified in the Cochra-
ne Handbook for Intervention Reviews. Trials were
classed as having a low, unclear or high risk of bias
based on the individual components: allocation
sequence generation, concealment of allocation, blind-
ing, incomplete outcome data, selective reporting and
‘other bias’.
Statistical analysis
Statistical analyses were performed using Revman version
5 (Nordic Cochrane Centre, Copenhagen, Denmark), STA-
TA version 13 (STATA Corp, College Station, TX, USA)
Aliment Pharmacol Ther 2014; 40: 123-132
ª 2014 John Wiley & Sons Ltd
 Systematic review with meta-analysis: rifaximin and hepatic encephalopathy
and TSA version 9 (Copenhagen Trial Unit,
Copenhagen, Denmark). The meta-analyses were
performed using random effects models with results pre-
sented as risk ratios (RR) or mean differences (MD) with
95% confidence interval (CI), I2 and v2 (presented as Q
and P values) as markers of heterogeneity. We defined I2
values between 30% and 60% as moderate heterogeneity,
60–75% as considerable heterogeneity and values >75%
as substantial heterogeneity. Values below 30% were con-
sidered unimportant.18 For the v2, a P value below 0.1
was considered to indicate significant heterogeneity. To
determine the influence of between trial heterogeneity
and the robustness of the result, the analyses were
repeated using fixed effect models. The results were
reported if the conclusions of the two models differed.
The meta-analyses were performed with outcome mea-
sures that were recalculated based on individual patient
data when available or based on data extracted from
published reports or online information (the FDA, Food
and Drug Administration website as described below).
We calculated the number needed to treat (NNT) for the
primary outcome measures that were statistically signifi-
cant (the 95% CI did not cross 1) in trials with a low
risk of bias and confirmed in sequential analyses. The
NNT was calculated as based on the RR and the average
control group risk (ACR) in trials with a low risk of bias
(based on the formula NNT = 1/(ACR 9 (1 RR) and
in trials with different control groups. We performed
subgroup, sensitivity and regression analyses to evaluate
sources of heterogeneity and bias. Subgroup analyses
evaluated the effect of rifaximin in relation to the type of
HE, control interventions (placebo, disaccharides or anti-
biotics), the type of data (outcomes recalculated based
on individual patient data or extracted from published
trial reports) and bias control. Differences between sub-
groups were expressed as P values (test for subgroup dif-
ferences) with values below 0.05 considered statistically
significant. We also performed a sensitivity analysis that
included RCTs on patients randomised after TIPSS. The
risk of bias and other small study effects were evaluated
through regression analyses (using Harbord’s test for
dichotomous and Egger’s test for continuous outcomes)
and funnel plots for meta-analyses with at least 10 RCTs.
The sequential analyses were based on random effects
models, alpha 5%, power 80% and model-based hetero-
geneity (diversity). The sequential analysis was consid-
ered as confirmatory of the primary meta-analyses when
the z-curve (cumulative results of included trials) crossed
the trial monitoring curve and the required information
size.
Aliment Pharmacol Ther 2014; 40: 123-132
ª 2014 John Wiley & Sons Ltd
RESULTS
In the electronic searches, 468 references including 114
duplicates were identified (Figure 1). Another 17 refer-
ences were identified through manual searches. After
excluding references to articles that clearly did not fulfil
our inclusion criteria, 19 RCTs15, 20–37 (Table 1)
described in 24 references38–42 were included in our
analyses. Two trials were published as abstracts,22, 36 and
the remaining as full paper articles. One additional full
paper RCT on HE after TIPSS was included in sensitivity
analyses.43
We gathered additional information about the trial
design and outcome measures from the FDA website
(three trials),15, 22, 28, 38–41 Alfa Wasserman (10 tri-
als)23–30, 32, 33 and Salix Pharmaceuticals Incorporated
(one trial).15 For the remaining trials, we only had
access to the information provided in the trial publica-
tions.
In total, our primary analyses included 1370 patients
with cirrhosis. The aetiology was alcohol for 35% and
viral hepatitis for 35%. The control groups received
placebo (six trials),15, 20–22, 34, 35 disaccharides (eight
trials)23, 25–29, 31, 36 or other antibiotics (six tri-
als).24, 26, 30, 32, 33, 37 Fifteen trials assessed overt HE,
two trials minimal HE and two RCTs secondary preven-
tion of HE. The HE diagnostics included clinical assess-
ments (Conn and West Haven) and various composite
scores. The dose of rifaximin was 1100–1200 mg/day
and the duration of treatment 5–180 days. One addi-
tional RCT on patients with TIPSS was included in sen-
sitivity analysis.43 The trial included patients who were
randomised to rifaximin 1200 mg/day (n = 25), no treat-
ment (n = 25) or lactitol (n = 25).
Risk of bias
We gathered information on bias from publications and
correspondence with the primary investigators or phar-
maceutical companies (Table S2). The randomisation
methods (allocation sequence generation and allocation
concealment) were classed as adequate in 16 trials
(Table S2). Fourteen RCTs were double blind with
blinding of patients and investigators (including blinded
outcome assessment). Fourteen trials had no missing
outcome data. We were unable to extract data on HE
in one trial33 and classed the trial as having selective
outcome reporting. Three trials were classed as having
potential risk of ‘other biases’ due to type of publica-
tion (abstract only) or because the trials were
terminated before the required sample size was
met.22, 28, 36
125
N. Kimer et al.

Additional records identified through other

Records identified through database searching
sources
(n = 468)
(n = 17)

Records after duplicates removed

(n = 371)

Records screened Records excluded

(n = 371) after reading titles and abstracts
(n = 279)

Full-text articles assessed for Full-text articles excluded:

eligibility observational studies, reviews or
(n = 92) trials that did not assess rifaximin
for HE
(n = 68)

Studies included in qualitative synthesis

(n = 19)
(Described in 24 reports)

Studies included in meta-analysis

(n = 19)
1 additional RCT included in sensitivity
analyses (n = 20)

Figure 1 | Flow chart of trial selection.

Prevention of HE bias based on components of bias control (RR: 1.45: 95%

Two placebo-controlled trials evaluated the effect of rif-
aximin on prevention of overt HE (Figure 2). The
RCTs found that rifaximin prevented overt HE with a
RR of 1.36 (95% CI: 1.06–1.65). When including the
trial on TIPSS, the RR was reduced to 1.24 (95% CI:
1.00–1.53).
Resolution of HE
Full resolution of HE was achieved for 239 of 337
patients randomised to rifaximin vs. 148 of 304 controls
(Figure 3). Random effects meta-analysis showed that
rifaximin had a beneficial effect on this outcome measure
(RR: 1.34, 95% CI: 1.11–1.62). The between trial hetero-
geneity was moderate (I2 = 54%, Q = 21.86, P = 0.02).
The result of the meta-analysis was confirmed in a sensi-
tivity analysis that only included trials with a low risk of
CI: 1.16–1.82). In subgroup analyses, there was no clear
differences between trials stratified by the control group
(P = 0.12) or the type of data (P = 0.08). There was a
clear difference (P < 0.001) between the effect of rifaxi-
min in the subgroup of RCTs on overt HE and the trial
on minimal HE (Figure 3). The between study heteroge-
neity was lower among RCTs on overt HE after exclu-
sion of the trial on minimal HE (I2 = 0%, Q = 7.32,
P = 0.60). In a post hoc sensitivity analysis, we evaluated
the influence of precipitating events. After exclusion of
trials with patients who did not develop HE based on a
precipitating event, rifaximin had a beneficial effect on
HE resolution (RR: 1.34, 95% CI: 1.10–1.63).
The NNT was six patients in the overall analysis.
When the analysis was repeated stratified by the type
of control group, the NNT was four patients (placebo

126 Aliment Pharmacol Ther 2014; 40: 123-132

ª 2014 John Wiley & Sons Ltd
 Systematic review with meta-analysis: rifaximin and hepatic encephalopathy

Table 1 | Trial characteristics of randomised controlled trials on rifaximin for hepatic encephalopathy

No of Treatment
Author, year and origin patients Type of HE Interventions duration, days
Festi et al. 1993, Italy26 35 Overt Rifaximin 1200 mg/Neomycin 3 g 21
Festi et al. 1993, Italy26 21 Overt Rifaximin 1200 mg/Lactulose 40 g 21
Miglio et al. 1997, Italy30 60 Overt Rifaximin 1200 mg/Neomycin 3 g 14
Parini et al. 1992, Italy32 30 Overt Rifaximin 1200 mg/Paromycin 1500 mg 10
Pedretti et al. 1991, Italy33 30 Overt Rifaximin 1200 mg/Neomycin 3 g 21
Testa et al. 1985, Italy37 20 Minimal Rifaximin 1200 mg/Paromycin 1500 mg 5
De Marco et al. 1985, Italy24 32 Overt Rifaximin 1200 mg/Paromycin 1500 mg 6–15
Fera et al. 1993, Italy25 40 Overt Rifaximin 1200 mg/Lactulose 40 g 14
Massa et al. 1993, Italy29 40 Overt Rifaximin 1200 mg/Lactulose 60 g 15
Paik et al. 2005 Korea31 54 Overt Rifaximin 1200 mg/Lactulose 90 mL 7
Song et al. 2000, South Korea36 64 Overt Rifaximin 1200 mg/Lactulose 90 mL 7
Bucci and Palmieri 1993, Italy23 58 Overt Rifaximin 1200 mg/Lactulose 30 g 15
Loguercio et al. 2003, Italy27 40 Overt Rifaximin 1200 mg/Lactitol 60 g 3 9 15
Mas et al. 2003, Spain28, 38, 40, 41 103 Overt Rifaximin 1200 mg/Lactitol 60 g 5–10
Bass et al. 2004, USA22, 38, 40, 41 93 Overt Rifaximin 1200 mg/Placebo 14
Bass et al. 2010, USA/Russia15, 38–41 299 Prevention Rifaximin 1100 mg/Placebo 180
of recurrent HE
Sidhu et al. 2011, India35 94 Minimal Rifaximin 1100 mg/Placebo 56
Bajaj et al. 2011, USA21 42 Minimal Rifaximin 1100 mg/Placebo 56
Sharma et al. 2013, India34 120 Minimal Rifaximin 1200 mg/Placebo 10
Ali et al. 2014, Pakistan20 126 Prevention Rifaximin 1200 mg/Placebo 180
of Recurrent HE

control), seven patients (non-absorbable disaccharides)
and six patients (other antibiotics). No evidence of
publication bias or other small study effects were seen
in regression analysis (P = 0.83) or funnel plots (Fig-
ure S1). The sequential analysis was performed using
the observed control group incidence and with the rel-
ative risk reduction set to 26%. The sequential analysis
confirmed the result of the primary meta-analysis with
the trial monitoring boundary (indicating the
time, when firm evidence was reached) being crossed
in 2013.
Improved manifestations of HE
In total, the primary investigators classed 403 of 486
patients randomised to rifaximin vs. 299 of 429
controls as having improved manifestations of HE.
There was substantial heterogeneity (I2 = 80%,
Q = 75.08, P = 0.001) and a corresponding difference
between the results of the random effects (RR: 1.00,
95% CI: 1.00–1.23) and the fixed effect meta-analysis
(RR: 1.19, 95% CI: 1.11–1.27). Sensitivity analyses lim-
ited to include only RCTs with a low bias risk, found
a beneficial effect of rifaximin in fixed effect
meta-analysis (RR: 1.23; 95% CI: 1.13–1.35), but not
when a random effects model was used (RR: 1.16;
95% CI: 0.99–1.37). In subgroup analyses, there were
no differences between RCTs stratified by the type of
HE (P = 0.51), control group (P = 0.78) or type of
data (P = 0.26). There was evidence of publication bias
or other small study effects in regression analyses

Rifaximin Control Risk Ratio

Events Total Events Total Weight Random effects, 95% CI
Ali 2014 40 63 35 63 29.2%
Figure 2 | Forest plot of Bass 2010 109 140 86 159 46.1%
randomised controlled trials Riggio2005 (TIPS*) 17 25 33 50 24.6%
on rifaximin for prevention of
recurrent HE. The control Total (95% CI) 228 272 100.0%
groups received placebo or Total events 166 154
Heterogeneity: Q = 4.03, P = 0.13; I2 = 50%
disaccharides (Riggio 2005, 0.5 0.7 1 1.5 2
TIPSS patients). Favours Control Favours Rifaximin

Aliment Pharmacol Ther 2014; 40: 123-132 127

ª 2014 John Wiley & Sons Ltd
N. Kimer et al.

Rifaximin Control Risk Ratio

Events Total Events Total Weight Random effects, 95% CI
Overt HE
Bucci 1993 26 30 18 28 11.9%
De Marco 1984 14 18 11 14 10.5%
Fera 1993 17 20 15 20 11.8%
Festi 1993a 16 20 10 15 9.4%
Loguercio 2003 12 27 4 13 3.4%
Mas 2003 26 50 20 53 9.0%
Massa 1993 14 20 11 20 8.0%
MIglio 1997 18 25 15 24 9.9%
Parini 1992 11 15 10 15 8.4%
Sharma 2013 48 63 25 57 11.5%
Subtotal (95% CI) 288 259 93.8%
Total events 202 139
Heterogeneity : Q = 7.32, P = 0.60; I2 = 0%

Minimal HE
Sidhu 2011 37 49 9 45 6.2% Figure 3 | Forest plot of
Subtotal (95% CI) 49 45 6.2% randomised controlled trials
Total events 37 9 on rifaximin for overt, minimal
Heterogeneity: Not applicable
or prevention of recurrent HE.
The outcome measure is full
Total (95% CI) 337 304 100.0% resolution of HE. The control
Total events 239 148 groups received placebo, non-
Heterogeneity: Q = 21.86, P = 0.02; I2 = 54% absorbable disaccharides or
0.1 0.2 0.5 1 2 5 10
Favours control Favours rifaximin other antibiotics.
Test for subgroup differences: P = 0.0005)

(P = 0.09) and funnel plots (Figure S2). In sequential Adverse events

analysis (performed with the observed control group
incidence and a relative risk reduction of 10%), the
monitoring boundary was not crossed. Hence, the
required information size was not reached, suggesting
that additional information is needed to support or
refute the effect of rifaximin on improved HE manifes-
tations.
Mortality
In total, 35 of 533 in the rifaximin group and 52 of 549
controls died (Figure 4). Random effects meta-analysis
showed that rifaximin reduced mortality (RR: 0.64, 95%
CI: 0.43–0.94; between study heterogeneity: I2 = 0%,
Q = 3.76, P = 0.58). The RCTs that contributed to the
analysis had a low risk of bias. No differences were seen
between subgroups stratified by the type of HE
(P = 0.23), control group (P = 0.88) or type of data
(P = 0.29). Inclusion of the trial on patients with TIPSS
did not change the result (RR: 0.66, 95% CI: 0.45–0.97).
In sequential analysis, with the observed control group
incidence and the relative risk reduction set to 40%,
additional trials were necessary to determine if rifaximin
reduces mortality (the trial monitoring boundary was
not crossed and the required information size was not
met).
We were able to gather data on adverse events from 13
trials. None of the trials found a difference between the
rifaximin and control groups regarding serious adverse
events. We were only able to combine data on serious
adverse events from four trials15, 22, 28, 34 (RR 1.04, 95%
CI: 0.77–1.41, I2 = 0%, Q = 1.43, P = 0.70). None of the
included trials found an increased risk of resistance to
rifaximin or other antibiotics or clostridium difficile
enteritis. The nonserious adverse events were mainly
related to gastrointestinal symptoms and included nausea
and diarrhoea. Due to differences in the registration of
nonserious adverse events, combining the data from
individual trials was not possible.
Secondary outcome measures
We were able to extract data on serum ammonia from
12 trials. The trials showed that rifaximin lowered serum
ammonia with a mean difference of 7.10 lg/dL (95%
CI: 12.29 to 1.91). Rifaximin was also associated with
an improved number connection test (mean difference:
5.29 s, 95% CI: 10.05 to 0.53, six trials).
DISCUSSION
This review found clinically relevant effects of rifaximin
in the management of patients with HE. The included

128 Aliment Pharmacol Ther 2014; 40: 123-132

ª 2014 John Wiley & Sons Ltd
 Systematic review with meta-analysis: rifaximin and hepatic encephalopathy

Rifaximin Lactulose/lactitol Risk Ratio

Study or Subgroup Events Total Events Total Weight M- H, Random, 95% CI
Sidhu 2011 0 49 1 45 1.5%
Bass 2010 9 140 11 159 21.1%
57.6%
Sharma 2013 15 63 28 57
Pedretti 1991 0 15 0 15
Fera 1993 0 20 0 20
Loguercio 2003 0 27 0 13
Bass 2004 1 48 0 45 1.5%
Mas 2003 1 50 2 53 2.7%
Riggio 2005 (TIPS) 2 25 3 50 0.0%
Ali 2014 7 63 7 63 15.6%

De Marco 1984 0 18 0 14
Parini 1992 0 15 0 15

Total (95% CI) 508 499 100.0%

Total events 33 49
Heterogeneity: Tau2 = 0.00; Chi2 = 3.76, df = 5 (P = 0.58); I2 = 0%
0.1 0.2 0.5 1 2 5 10
Favours rifaximin Favours lact

Figure 4 | Forest plot of randomised controlled trials on rifaximin for overt, minimal or prevention of recurrent HE.
The outcome measure is mortality. The control groups received placebo, non-absorbable disaccharides or other
antibiotics.

trials compared rifaximin vs. placebo or active interven-
tions including non-absorbable disaccharides or other
antibiotics. Based on our analyses, rifaximin has a benefi-
cial effect on mortality, secondary prevention of HE and
full recovery from HE. No clear differences were seen
between trials with different control interventions or dif-
ferent types of HE. Sensitivity analyses showed no clear
benefit of rifaximin after TIPSS, but the number of
events and patients was too small to make definite con-
clusions regarding this outcome measure. The evidence
on HE recovery did not appear to reflect random or sys-
tematic errors due to cumulative testing or bias control.
Although mortality is an outcome measure that is rela-
tively stable to the potential influence of bias,44 our
analyses indicate that the required information size is
not yet met for this outcome measure. Accordingly,
additional RCTs are needed to determine if rifaximin
reduces mortality. However, the combined evidence sug-
gests that rifaximin should be considered in the evi-
dence-based management of patients with HE.
Hepatic encephalopathy can be classed as an episodic,
minimal or persistent condition,4 that can occur
spontaneously or precipitated by events. The severity of the
condition varies from minor signs that are identified
through psychometric tests to overt coma. Several diagnos-
tic tests and clinical scores are used in the diagnostic assess-
ment.4, 45–47 The evaluation of improved manifestions of
HE varied considerably in included RCTs. The diagnostic
assessment and definition of this outcome measure include
different scores (West Haven, Conn and Portal Systemic
Encephalopathy Index) as well as clinical signs such as aste-
rixis or measurements of ammonia. Accordingly, the
between study heterogeneity was considerable in the assess-
ment of improved manifestations of HE. The heterogeneity
was evident in spite of attempt to overcome differences in
the assessment and reporting through recalculation of
outcomes. We were unable to identify sources of heteroge-
neity in subgroup and regression analyses.
Unlike previous meta-analyses,12, 17, 48 we found that
rifaximin reduces mortality in patients with cirrhosis.
Previous meta-analyses on other interventions for HE
have not been able to find a similar effect. It is possible
that the rifaximin has other effects not directly related
to HE. Potential beneficial effects related to bacterial
translocation may exist.49 Translocation of bacterial
products may deteriorate liver haemodynamics in

Aliment Pharmacol Ther 2014; 40: 123-132 129

ª 2014 John Wiley & Sons Ltd
N. Kimer et al.
patients with cirrhosis.50 A study on long-term rifaxi-
min for patients with alcohol-related decompensated
cirrhosis and ascites showed that patients allocated to
rifaximin had a low risk of variceal bleeding, spontane-
ous bacterial peritonitis and hepatorenal syndrome
compared with matched controls.51 Rifaximin also
improved survival. A case–control study found a similar
beneficial effect of rifaximin on spontaneous bacterial
peritonitis.52 We attempted to analyse the effect of rif-
aximin on infections and hepatorenal syndrome in our
assessment of adverse events, but the data were not suf-
ficiently strong to make any definite conclusions. How-
ever, the combined evidence is promising and suggests
that rifaximin has several clinically relevant effects in
patients with cirrhosis.
Unlike previous meta-analyses, we included data made
available by pharmaceutical companies (Alfa Wasserman
and Salix Pharmaceuticals Incorporated). Based on out-
comes that were recalculated based on the individual
patient data, we were able to include information that
was not described in the published reports. We also
received important information about the trial design
and bias control. The advantage of this approach is the
possibility to perform standardisation of outcomes.
Retrieving unpublished information may also reduce the
risk of publication and reporting bias.
Our meta-analysis includes RCTs published more than
20 years ago. The available concurrent therapies and the
assessment and definitions of outcome measures have
changed considerably over time. The duration of therapy
in the earliest trials was much shorter than the more
recent RCTs. The length of therapy varied between a few
days to several months. The importance of the treatment
duration is likely to depend on the underlying disease.
The prevention of HE is likely to require months of
treatment whereas the treatment of acute HE episodes
may only require weeks. The evidence concerning long--
term rifaximin for prevention of HE in patients who do
not respond to non-absorbable disaccharides is strong.
Whether these two treatments provide synergistic effects
and increase the favourable changes to the intestinal
flora is not clear.
A previous review on patients with overt HE found
that rifaximin has a better safety profile than
non-absorbable disaccharides or other antibiotics.12 We
found that the most common adverse events included
diarrhoea, abdominal pain and gastrointestinal discom-
fort such as nausea, anorexia or weight loss. There were
no clear differences between the rifaximin and control
groups, but the statistical power was low.
130
Previous studies have evaluated the cost-effectiveness
of rifaximin compared to lactulose.53–56 In a decision
analysis, the cost-effectiveness of six strategies in HE
were compared including no treatment, lactulose, lacti-
tol, neomycin, rifaximin and rifaximin salvage (rifaxi-
min administered to lactulose nonresponders). The
analysis showed that the ‘no treatment’, was least effec-
tive and rifaximin salvage was most effective in terms
of quality-adjusted life-year-gained. The authors con-
cluded that rifaximin monotherapy was not cost-effec-
tive based compared with lactulose monotherapy. A
subsequent study on patients with minimal HE assessed
if the detection and treatment reduces costs and mor-
bidity associated with motor vehicle accidents.57 The
study included a cost-effectiveness analysis to assess the
different strategies of diagnosis and treatment. The
analyses showed that rifaximin therapy was
not cost-saving at current prices but would become so
at a monthly cost below $353.
In conclusion, this systematic review found evidence
to support the use of rifaximin in the management of
HE, in terms of prevention as well as the treatment of
acute episodes. Future trials are needed to determine the
effects of combining different HE interventions and to
determine the duration of therapy in different situations.
Additional trials on the potential effects of rifaximin on
bacterial translocation, infections and complications to
cirrhosis are also needed.
AUTHORSHIP
Guarantor of the article: Nina Kimer, MD.
Author contributions: Nina Kimer, Lise L. Gluud and
Aleksander Krag conceived and drafted the review and
extracted data. Nina Kimer and Lise L. Gluud performed
the statistical analyses. All authors participated in the
interpretation of analyses, reviewed and commented on
the article and approved the final version version of the
manuscript.
ACKNOWLEDGEMENT
We are grateful to Sara Louise Klingenberg at the Coch-
rane Hepato-Biliary Group for her invaluable assistance
in trial search, to Preti Pier Alessandro Monici from
Alpha Wasserman and Andy Barrett from Salix Pharma-
ceuticals Inc. for answering our queries about their trials
and their invaluable help with the collection of data.
Declaration of personal interests: Nina Kimer and
Flemming Bendtsen have received rifaximin tablets and
placebo from Norgine Denmark A/S in relation to an
ongoing trial. Lise L. Gluud has participated in clinical
Aliment Pharmacol Ther 2014; 40: 123-132
ª 2014 John Wiley & Sons Ltd
 Systematic review with meta-analysis: rifaximin and hepatic encephalopathy

courses on hepatic encephalopathy sponsored by Norgine.
Søren Møller has received research funding from Novo
Nordisk Foundation and Lundbeck Foundation.
Declaration of funding interests: None.
SUPPORTING INFORMATION
Additional Supporting Information may be found in the
online version of this article:
Figure S1. Funnel plot of publication bias and small
study effects in regard to full resolution of HE.
Figure S2. Funnel plot of publication bias and small
study effects in regard to manifestations of HE.
Table S1. Search strategies for ‘Rifaximin in hepatic
encephalopathy’.
Table S2. Risk of bias assessment.

REFERENCES
1. Prakash R, Mullen KD. Mechanisms,
diagnosis and management of hepatic
encephalopathy. Nat Rev Gastroenterol
Hepatol 2010; 7: 515–25.
2. Jepsen P, Ott P, Andersen PK, Sorensen
HT, Vilstrup H. Clinical course of
alcoholic liver cirrhosis: a Danish
population-based cohort study.
Hepatology 2010; 51: 1675–82.
3. Amodio P, Campagna F, Olianas S,
et al. Detection of minimal hepatic
encephalopathy: normalization and
optimization of the Psychometric
Hepatic Encephalopathy Score. A
neuropsychological and quantified
EEG study. J Hepatol 2008; 49: 346–53.
4. Cordoba J. New assessment of hepatic
encephalopathy. J Hepatol 2011; 54:
1030–40.
5. Schliess F, Gorg B, Haussinger D.
Pathogenetic interplay between osmotic
and oxidative stress: the hepatic
encephalopathy paradigm. Biol Chem
2006; 387: 1363–70.
6. Als-Nielsen B, Gluud LL, Gluud C.
Non-absorbable disaccharides for
hepatic encephalopathy: systematic
review of randomised trials. BMJ 2004;
328: 1046.
7. Bajaj JS. Review article: the modern
management of hepatic encephalopathy.
Aliment Pharmacol Ther 2010; 31:
537–47.
8. Riordan SM, Williams R. Gut flora and
hepatic encephalopathy in patients with
cirrhosis. N Engl J Med 2010; 362:
1140–2.
9. Sharma BC, Sharma P, Agrawal A,
Sarin SK. Secondary prophylaxis of
hepatic encephalopathy: an open-label
randomized controlled trial of lactulose
versus placebo. Gastroenterology 2009;
137: 885–91.e1.
10. Shawcross DL, Davies NA, Williams R,
Jalan R. Systemic inflammatory
response exacerbates the
neuropsychological effects of induced
hyperammonemia in cirrhosis. J
Hepatol 2004; 40: 247–54.
11. Sharma P, Sharma BC, Agrawal A,
Sarin SK. Primary prophylaxis of overt
hepatic encephalopathy in patients with
cirrhosis: an open labeled randomized
controlled trial of lactulose versus no
lactulose. J Gastroenterol Hepatol 2012;
27: 1329–35.
12. Eltawil KM, Laryea M, Peltekian K,
Molinari M. Rifaximin vs. conventional
oral therapy for hepatic
encephalopathy: a meta-analysis. World
J Gastroenterol 2012; 18: 767–77.
13. Mullen KD, Dasarathy S. Protein
restriction in hepatic encephalopathy:
necessary evil or illogical dogma? J
Hepatol 2004; 41: 147–8.
14. Gluud LL, Dam G, Borre M, et al. Oral
branched-chain amino acids have a
beneficial effect on manifestations of
hepatic encephalopathy in a systematic
review with meta-analyses of
randomized controlled trials. J Nutr
2013; 143: 1263–8.
15. Bass NM, Mullen KD, Sanyal A, et al.
Rifaximin treatment in hepatic
encephalopathy. N Engl J Med 2010;
362: 1071–81.
16. Jiang Q, Jiang XH, Zheng MH, Jiang
LM, Chen YP, Wang L. Rifaximin
versus nonabsorbable disaccharides in
the management of hepatic
encephalopathy: a meta-analysis. Eur J
Gastroenterol Hepatol 2008; 20:
1064–70.
17. Wu D, Wu SM, Lu J, Zhou YQ, Xu L,
Guo CY. Rifaximin versus
nonabsorbable disaccharides for the
treatment of hepatic encephalopathy: a
meta-analysis. Gastroenterol Res Pract
2013; 2013: 236963.
18. Higgins JT, Green S. Cochrane
handbook for systematic reviews of
interventions, March 2011. Available at:
http://handbook.cochrane.org. Accessed
February 2014.
19. Liberati A, Altman DG, Tetzlaff J, et al.
The PRISMA statement for reporting
systematic reviews and meta-analyses of
studies that evaluate healthcare
interventions: explanation and
elaboration. BMJ 2009; 339: b2700.
20. Ali B, Zaidi YA, Alam A, Anjum HS.
Efficacy of rifaximin in prevention of
recurrence of hepatic encephalopathy in
patients with cirrhosis of liver. J Coll
Physicians Surg Pak 2014; 24: 269–73.
21. Bajaj JS, Heuman DM, Wade JB, et al.
Rifaximin improves driving simulator
performance in a randomized trial of
patients with minimal hepatic
encephalopathy. Gastroenterology 2011;
140: 478–487.e1.
22. Bass NM, Ahmed A, Johnson L,
Gardner JD. Rifaximin treatment is
beneficial for mild hepatic
encephalopathy. Hepatology 2004; 40:
646A.
23. Bucci L, Palmieri GC. Double-blind,
double-dummy comparison between
treatment with rifaximin and lactulose
in patients with medium to severe
degree hepatic encephalopathy. Curr
Med Res Opin 1993; 13: 109–18.
24. De Marco F, Santamaria Amoto P,
D’arienzo A. Rifaximin in collateral
treatment of portal-systemic
encephalopathy: a preliminary report.
Current Ther Res 1984; 36: 668–74.
25. Fera G, Fransesco A, Nigro M,
Schiraldi O, Ferrieri A. Rifaximin in the
treatment of hepatic encephalopathy.
Eur J Clin Res 1993; 4: 57–66.
26. Festi D, Mazella G, Orsini M, Sottili S,
Sangermano A, Li Bassi S. Rifaximin
in the treatment of chronic hepatic
encephalopathy; results of a multicenter
study of efficacy and safety. Curr Ther
Res 1993; 54: 598–609.
27. Loguercio C, Federico A, De Girolamo
V, Ferrieri A, Del Vecchio Blanco C.
Cyclic treatment of chronic hepatic
encephalopathy with rifaximin. Results
of a double-blind clinical study.
Minerva Gastroenterol Dietol 2003; 49:
53–62.
28. Mas A, Rodes J, Sunyer L, et al.
Comparison of rifaximin and lactitol in
the treatment of acute hepatic

Aliment Pharmacol Ther 2014; 40: 123-132 131

ª 2014 John Wiley & Sons Ltd
N. Kimer et al.
encephalopathy: results of a
randomized, double-blind, double-
dummy, controlled clinical trial. J
Hepatol 2003; 38: 51–8.
29. Massa P, Vallerino F, Dodero M.
Treatment of hepatic encephalopathy
with rifaximin. Eur J Clin Res 1993; 4:
7–18.
30. Miglio F, Valpiani D, Rossellini SR,
Ferrieri A. Rifaximin, a non-absorbable
rifamycin, for the treatment of hepatic
encephalopathy. A double-blind,
randomised trial. Curr Med Res Opin
1997; 13: 593–601.
31. Paik YH, Lee KS, Han KH, et al.
Comparison of rifaximin and lactulose for
the treatment of hepatic encephalopathy:
a prospective randomized study. Yonsei
Med J 2005; 46: 399–407.
32. Parini P, Cipolla A, Ronchi M, Salzetta
A, Mazella G, Roda E. Effect of
rifaximin and paromycin in the
treatment of portal-systemic
encephalopathy. Curr Ther Res 1992;
52: 34–9.
33. Pedretti G, Calzetti C, Missale G,
Fiaccadori F. Rifaximin versus
neomycin on hyperammoniemia in
chronic portal systemic encephalopathy
of cirrhotics. A double-blind,
randomized trial. Ital J Gastroenterol
1991; 23: 175–8.
34. Sharma BC, Sharma P, Lunia MK,
Srivastava S, Goyal R, Sarin SK. A
randomized, double-blind, controlled
trial comparing rifaximin plus
lactulose with lactulose alone in
treatment of overt hepatic
encephalopathy. Am J Gastroenterol
2013; 108: 1458–63.
35. Sidhu SS, Goyal O, Mishra BP, Sood
A, Chhina RS, Soni RK. Rifaximin
improves psychometric performance
and health-related quality of life in
patients with minimal hepatic
encephalopathy (the RIME
Trial). Am J Gastroenterol 2011; 106:
307–16.
36. Song KH, Lee K, Kim MH, et al. The
clinical efficacy of rifaximin in the
treatment of hepatic encephalopaty
(comparison with lactulose). Hepatology
2000; 32: 407A.
37. Testa R, Eftimiadi C, Sukkar GS,
et al. A non-absorbable rifamycin for
treatment of hepatic
encephalopathy. Drugs Exp Clin Res
1985; 11: 387–92.
38. Dimick-Santos L. Meeting of The
Gastrointestinal Drugs Advisory
Committe (GIDAC) to evaluate
Rifaximin 550 mg tablets for the
proposed indication of maintenance of
remission of Hepatic Encephalopathy
(HE) in patients 18 years old and older.
NDA 22-554 Clinical Review: FDA;
132
2010. Available at: http://www.fda.gov/
downloads/AdvisoryCommittees/
CommitteesMeetingMaterials/Drugs/
GastrointestinalDrugsAdvisory
Committee/UCM201079.pdf. Accessed
February 2014.
39. Salix Pharmaceuticals Inc. Xifaxan
(rifaximin) tablets 550 mg for hepatic
encephalopathy. Committee
USFaDAGDA, 2010. Available at:
http://www.fda.gov/downloads/
AdvisoryCommittees/
CommitteesMeetingMaterials/Drugs/
GastrointestinalDrugs
AdvisoryCommittee/UCM203248.pdf.
Accessed February 2014.
40. Salix Pharmaceuticals Inc. Xifaxan
(rifaximin) tablets, 550 mg NDA 22-
554. Briefing document for the
Gastrointestinal Drugs Advisory
Committee Meeting 23 February 2010,
FDA; 2010. Available at: http://www.
fda.gov/downloads/Advisory
Committees/CommitteesMeeting
Materials/Drugs/GastrointestinalDrugs
AdvisoryCommittee/UCM201081.pdf.
Accessed February 2014.
41. Regina Alivisatos EC. 21-361 Xifaxan
Medical Review Part 6 Access Data
FDA. Salix Pharmaceuticals, FDA, 2002.
Available at: http://www.accessdata.fda.
gov/drugsatfda_docs/nda/2004/21-361
_Xifaxan_Medr_P6.pdf. Accessed
February 2014.
42. Sanyal A, Younossi ZM, Bass NM,
et al. Randomised clinical trial:
rifaximin improves health-related
quality of life in cirrhotic patients
with hepatic encephalopathy – a
double-blind placebo-controlled study.
Aliment Pharmacol Ther 2011; 34:
853–61.
43. Riggio O, Masini A, Efrati C, et al.
Pharmacological prophylaxis of hepatic
encephalopathy after transjugular
intrahepatic portosystemic shunt: a
randomized controlled study. J Hepatol
2005; 42: 674–9.
44. Savovic J, Jones H, Altman D, et al.
Influence of reported study design
characteristics on intervention effect
estimates from randomised controlled
trials: combined analysis of meta-
epidemiological studies. Health Technol
Assess 2012; 16: 1–82.
45. Mullen KD, Prakash RK. New
perspectives in hepatic encephalopathy.
Clin Liver Dis 2012; 16: 1–5.
46. Bajaj JS, Cordoba J, Mullen KD, et al.
Review article: the design of clinical
trials in hepatic encephalopathy – an
International Society for Hepatic
Encephalopathy and Nitrogen
Metabolism (ISHEN) consensus
statement. Aliment Pharmacol Ther
2011; 33: 739–47.
47. Randolph C, Hilsabeck R, Kato A,
et al. Neuropsychological assessment of
hepatic encephalopathy: ISHEN
practice guidelines. Liver Int 2009; 29:
629–35.
48. Morgan MY, Morris RW. Rifaximin for
the treatment of hepatic
encephalopathy; a meta-analysis of
randomised studies. Gut 2012; 61
(Suppl. 2): A203–4.
49. Fiorucci S, Distrutti E, Mencarelli A,
Barbanti M, Palazzini E, Morelli A.
Inhibition of intestinal bacterial
translocation with rifaximin
modulates lamina propria monocytic
cells reactivity and protects against
inflammation in a rodent model
of colitis. Digestion 2002; 66: 246–56.
50. Kalambokis GN, Mouzaki A, Rodi M,
et al. Rifaximin improves systemic
hemodynamics and renal function in
patients with alcohol-related cirrhosis
and ascites. Clin Gastroenterol Hepatol
2012; 10: 815–8.
51. Vlachogiannakos J, Viazis N,
Vasianopoulou P, Vafiadis I,
Karamanolis DG, Ladas SD. Long-
term administration of rifaximin
improves the prognosis of patients
with decompensated alcoholic
cirrhosis. J Gastroenterol Hepatol 2013;
28: 450–5.
52. Danulescu RM, Ciobica A, Stanciu C,
Trifan A. The role of rifaximine in the
prevention of the spontaneous bacterial
peritonitis. Rev Med Chir Soc Med Nat
Iasi 2013; 117: 315–20.
53. Huang E, Esrailian E, Spiegel BM. The
cost-effectiveness and budget impact of
competing therapies in hepatic
encephalopathy – a decision analysis.
Aliment Pharmacol Ther 2007; 26:
1147–61.
54. Kimer N, Krag A, Gluud LL. Safety,
efficacy, and patient acceptability of
rifaximin for hepatic
encephalopathy. Patient Prefer
Adherence 2014; 8: 331–8.
55. Leevy CB, Phillips JA. Hospitalizations
during the use of rifaximin versus
lactulose for the treatment of hepatic
encephalopathy. Dig Dis Sci 2007; 52:
737–41.
56. Neff G. Pharmacoeconomics of hepatic
encephalopathy. Pharmacotherapy 2010;
30: 28S–32S.
57. Bajaj JS, Pinkerton SD, Sanyal AJ,
Heuman DM. Diagnosis and treatment
of minimal hepatic encephalopathy to
prevent motor vehicle accidents: a cost-
effectiveness analysis. Hepatology 2012;
55: 1164–71.
Aliment Pharmacol Ther 2014; 40: 123-132
ª 2014 John Wiley & Sons Ltd