ORIGINAL ARTICLE: Clinical Endoscopy

Effect of erythromycin before endoscopy in patients presenting

with variceal bleeding: a prospective, randomized, double-blind,
placebo-controlled trial
Ibrahim Altraif, MD, FRCP, Fayaz A. Handoo, MD, MRCP, Abdulrahman Aljumah, FRCP(I),
Abduljaleel Alalwan, MD, Mutasim Dafalla, MD, MRCP, Abdullah Mohamed Saeed, MD, MRCP,
Abdulrahman Alkhormi, MD, Abdul Karim Albekairy, PhD, Hani Tamim, PhD
Riyadh, Saudi Arabia

Background: Blood in the stomach and esophagus in patients with variceal bleeding often obscures the
endoscopic view and makes endoscopic intervention difficult to perform. Erythromycin, a motilin agonist,
induces gastric emptying.
Objective: To assess the effect of erythromycin on endoscopic visibility and its outcome in patients with variceal
bleeding.
Design: Randomized, double-blind, placebo-controlled trial.
Setting: Tertiary care hospital.
Patients: Adult patients with liver cirrhosis presenting with hematemesis within the previous 12 hours.
Intervention: Either 125 mg erythromycin or placebo administered intravenously 30 minutes before endoscopy.
Main Outcome Measurements: Endoscopic visibility during index endoscopy and mean duration of
procedure.
Secondary Outcome Measurements: Need for repeat endoscopy and blood transfusions within 24 hours,
endoscopy-related complications, and length of hospital stay.
Results: A total of 102 patients received either erythromycin or placebo (53 erythromycin and 49 placebo).
Forty-seven patients in the erythromycin group and 43 in the placebo group had variceal bleeding and were
considered for final analysis. A completely empty stomach was seen in 48.9% of the erythromycin group versus
23.3% of the placebo group (P ⬍ .01). Mean endoscopy duration was significantly shorter in the erythromycin
group than in the placebo group (19.0 minutes vs 26.0 minutes, respectively; P ⬍ .005). Length of hospital stay
was significantly shorter in the erythromycin group than in the placebo group (3.4 days vs 5.1 days, respectively;
P ⬍ .002). The need for repeat endoscopy and the mean number of units of blood transfused did not differ
significantly in the 2 groups. No adverse events were observed with erythromycin.
Limitations: Sample size not sufficient to measure the need for repeat endoscopy and survival benefit.
Conclusions: Erythromycin infusion before endoscopy in patients with variceal bleeding significantly improves
endoscopic visibility and shortens the duration of the index endoscopy. (Clinical trial registration number: NCT
01060267.) (Gastrointest Endosc 2011;73:245-50.)

DISCLOSURE: All authors disclosed no financial relationships relevant to
this publication.
Copyright © 2011 by the American Society for Gastrointestinal Endoscopy
0016-5107/$36.00
doi:10.1016/j.gie.2010.09.043
Received June 13, 2010. Accepted September 27, 2010.
Current affiliations: Division of Hepatology (I.A., F.A.H., A. Aljumah, A. Alalwan,
M.D., A.M.S., A. Alkhormi), Hepatobiliary Sciences, and Liver Transplantation,
Department of Clinical Pharmacy (A.K.A.), Department of Medical Research
Centre (H.T.), King Abdul-Aziz Medical City, Riyadh, Saudi Arabia.
Reprint requests: Ibrahim H. Altraif, FRCP, Division of Hepatology,
Hepatobiliary Sciences and Liver Transplantation King Abdulaziz Medical
City, National Guard Health Affairs PO Box 22490, Riyadh 11426, KSA.
If you would like to chat with an author of this article, you may contact Dr
Altraif at ibrahimtraif@yahoo.com.

www.giejournal.org Volume 73, No. 2 : 2011 GASTROINTESTINAL ENDOSCOPY 245

Effect of erythromycin before endoscopy in patients with variceal bleeding
Variceal bleeding is a major complication of liver cir-
rhosis with a high mortality rate of 15% to 20%1; endo-
scopic band ligation or sclerotherapy can control bleeding
in approximately 80% of patients.2,3 However, the success
rate of these interventions largely depends on the quality
of the endoscopic view, which is often obscured by the
residual blood in the stomach, which makes intervention
difficult to perform, exposes the patient to the potential
risk of bronchopulmonary aspiration, and necessitates a
second-look endoscopy in as many as 15% of cases.4
Erythromycin, a macrolide antibiotic, is a potent motilin
agonist that induces rapid gastric emptying when given in-
travenously in doses ranging from 1 to 3 mg/kg in healthy
individuals.5,6 This effect is secondary to an increase in gastric
tone and phasic motor activity observed within minutes of
starting an infusion and persists for at least 2 hours.7-9 Previ-
ous studies showed that erythromycin helps in stomach
cleansing and improves the quality of endoscopic examina-
tion in acute upper GI bleeding but with conflicting reports
on its clinical benefits.10-13 In addition, these studies were
conducted with patients with predominantly nonvariceal up-
per GI bleeding. The aim of our study was to determine
whether intravenous erythromycin would improve the qual-
ity of the endoscopic view and thereby the effectiveness of
endoscopic therapeutic measures in patients with variceal
bleeding. The impact of this therapy on the rate of
endoscopy-related complications such as bronchopulmo-
nary aspiration, the need for second-look endoscopy within
24 hours, the need for blood transfusions, and the duration of
hospital stay were evaluated.
METHODS
Between January 2007 and May 2009, all adult patients
with liver cirrhosis or signs of chronic liver disease seen at
King Abdul Aziz Medical City Riyadh with hematemesis
during the previous 12 hours were considered for study if
they met the following criteria: older than 18 years, in-
formed consent provided, and hemodynamically stable at
the time of endoscopy after resuscitation. Patients meeting
any of the following criteria were excluded: known allergy
to erythromycin, previous use of other prokinetic agents
or gastric lavage on admission, previous gastric surgery,
history of cardiac arrhythmias, current use of antiarrhyth-
mic drugs or antihistaminic drugs with possible drug in-
teractions with erythromycin, or pregnant or lactating.
Study intervention
All eligible patients were assessed for severity of liver
cirrhosis by Child-Pugh-Turcotte scoring and were randomly
assigned to receive either erythromycin or placebo. All pa-
tients were continuously monitored in an intensive care set-
ting, resuscitated with intravenous fluids (isotonic saline so-
lution) and blood transfusions aiming at a hemoglobin level
of 9 to 10 g/dL. All patients received an intravenous bolus of
50 ␮g octreotide on admission before endoscopy followed
246 GASTROINTESTINAL ENDOSCOPY Volume 73, No. 2 : 2011
Altraif et al
Take-home Message
● Intravenous erythromycin administered before
endoscopy in patients with variceal bleeding improves
the quality of endoscopic visualization, makes the
endoscopic procedure easier and quicker, and results in
better outcome in terms of therapeutic intervention. It
also decreases the length of hospital stay, which could
possibly decrease the cost of management.
by continuous infusion at 50 ␮g/h for 48 to 72 hours, intra-
venous bolus of 40 mg esmoprazole before endoscopy, and
ceftrioxone 2 g intravenously on admission and then once
daily for 3 to 5 days. Platelets were transfused when the
platelet count was less than 50,000, aiming at a correction to
50,000, and fresh frozen plasma was transfused when the
international normalized ratio was more than 1.5. The pa-
tients in the erythromycin group received an intravenous
bolus infusion of 125 mg erythromycin lactobionate in 50 mL
of normal saline solution, and those in the placebo group
received 50 mL of normal saline solution over 10 minutes.
Endoscopy was performed 30 minutes after completion of
the erythromycin or placebo infusion.
Randomization generation and
allocation concealment
The 2 treatments were delivered in identical packages
labeled as the study protocol drug from the central phar-
macy of our hospital in a double-blind manner after
computer-generated randomization and assignment list.
Both the endoscopist and the patient were blinded to the
study protocol drug used. The treatment identity and al-
locations were released by the pharmacy only after com-
pletion of the study for final analysis. Informed consent
was obtained from each patient or from a family member
if the patient was encephalopathic before inclusion in the
study. The study protocol was approved by the hospital
ethics and research committee. Using an Olympus video
endoscope GIF XQ-260 (Olympus, Tokyo, Japan), 3 senior
hepatologists performed endoscopies. The hepatologists
discussed the study design and grading scale before initi-
ating the study and testing the interobserver variation in a
pilot study of 10 patients with variceal bleeding who
agreed to receive intravenous erythromycin before endos-
copy at our center to ensure satisfactory interobserver
reproducibility. Patients were continuously monitored for
pulse, O2 saturation, blood pressure, and electrocardiog-
raphy during protocol drug infusion and for 2 hours after
the endoscopic procedure. Procedure duration was timed
from insertion to the withdrawal of the endoscope. A
systematic examination of the esophagus, fundus, corpus,
antrum, duodenal bulb, and second part of the duodenum
was performed. Bleeding esophageal varices were treated
by band ligation or polidocanol injection or a combination
www.giejournal.org
Altraif et al
TABLE 1. Objective scoring system for endoscopic
visibility
Score Surface visible Vision
0 ⬍25% Worst
1 25%-50% Poor
2 ⬎50%-75% Satisfactory
3 ⬎75%-95% Good
4 ⬎95% Best
of them, determined by feasibility of performing each
intervention and achieving satisfactory bleeding control
and fundal varices by histacryl or glubran injection. A
standardized endoscopy report was generated immedi-
ately after completion of the procedure.
Outcome measurements
The primary study objective was to determine the effect
of erythromycin on gastric cleansing and thereby the qual-
ity of endoscopic visualization. Whether cleansing oc-
curred was the main study outcome. We used a modified
and expanded objective scoring system (previously used
by Frossard et al11) to score each area from 0 to 4 (Table 1).
The esophagus, fundus, corpus, and antrum were scored
separately. The maximum total score obtainable was 16. A
score of 15 or more was considered as a clear esophagus
and stomach. The cutoff of 15 was determined based on
results of previous experience in the pilot study. The mean
duration of the endoscopic procedure was used as an
alternative measure of the primary study objective because
it reflected the ease with which the procedure was per-
formed in a clear stomach. The secondary outcomes were
(1) the need for a second-look endoscopy within 24 hours
of the initial procedure because of either failure at the
initial endoscopy to identify the bleeding site and unsat-
isfactory bleeding control or clinical suspicion of rebleed-
ing in the form of hematemesis with a decrease in hemo-
globin of more than 2 g from initial levels, systolic blood
pressure less than 90 mm Hg, or the need for a transfusion
of more than 2 units of blood in first 24 hours after
endoscopy; (2) the number of units of blood transfused
within 24 hours after the initial endoscopy; (3) the inci-
dence of endoscopy-related complications such as bron-
chopulmonary aspiration; (4) drug-related complications
such as cardiac arrhythmias and GI symptoms; and (5) the
length of hospital stay.
Statistical analysis
The sample size was calculated to provide sufficient
power to test the main study outcome. In our experience
with variceal bleeding, at least two thirds (66%) of the pa-
tients have a stomach filled with blood clots at the time of
www.giejournal.org
Effect of erythromycin before endoscopy in patients with variceal bleeding
endoscopy. In our pilot study with erythromycin, we ob-
served an empty stomach in approximately two thirds of
patients. We proposed improving the endoscopic yield
from 33% to 66% (␣ error ⫽ .05, ␤ error ⫽ .10). We
calculated that 50 patients had to be included in each
group. Data suitable for analysis by parametric statistics
were reported as summary statistics and expressed as
mean (standard deviation). Statistical analysis was per-
formed by using Pearson’s ␹2 test, Yates correction for
continuity where appropriate for categorical variables, and
a t test for continuous variables. Finally, we used multi-
variate logistic regression analysis to determine which fac-
tors, including erythromycin, age, sex, hematemesis dura-
tion before endoscopy, systolic blood pressure on
admission, pre-endoscopy hemoglobin level, and Child-
Pugh-Turcotte score, were associated with an empty stom-
ach. A P value ⬍.05 was considered significant. With
regard to the issue of multiple testing of outcome data
arising from individual patients, it should be noted that the
P values for the univariate statistical tests were not cor-
rected for multiple testing because those tests were ex-
ploratory and should be taken as descriptive only. The
subsequent multivariate logistic regression analysis was
considered the main definitive result because it deter-
mined those variables independently associated with an
empty stomach and, hence, good endoscopic visibility in
patients with variceal bleeding, after adjusting for the con-
tribution of the other variables. Analyses were performed
with SPSS software version 17.0 (SPSS Inc, Chicago, Ill).
RESULTS
A total of 134 patients with chronic liver disease who
presented with upper GI bleeding were screened for inclu-
sion in the study; of them, 102 eligible patients who met the
inclusion criteria were randomized: 53 patients received
erythromycin and 49 received placebo. Forty-seven patients
in the erythromycin group and 43 in the placebo group were
found to have variceal bleeding and were considered for
final analysis. The patients with nonvariceal bleeding (peptic
ulcer, esophagitis) were not included in the final analysis.
The patient characteristics, including age, sex, systolic blood
pressure, basal hemoglobin, Child-Pugh-Turcotte score, the
duration of bleeding before endoscopy, the time from hos-
pital admission to endoscopy, and the presence of diabetes
mellitus, did not differ significantly between the 2 groups
(Table 2). Resultant summary statistics for the main and
secondary outcome measures are shown in Table 3. The
number of patients with an empty stomach as defined by
objective criteria as a total score of 15 or more was signifi-
cantly higher in the erythromycin group than in the placebo
group (48.9% vs 23.3%; P ⬍ .01). The mean endoscopic score
was significantly higher in the erythromycin group than in
the placebo group (12.5 vs 9.8, respectively; P ⬍ .05). The
mean duration of endoscopy was significantly shorter in the
erythromycin group than in the placebo group (19.0 minutes
Volume 73, No. 2 : 2011 GASTROINTESTINAL ENDOSCOPY 247
Effect of erythromycin before endoscopy in patients with variceal bleeding Altraif et al

TABLE 2. Characteristics of patients at admission

Variables Erythromycin Placebo P value

No. of patients with variceal bleeding 47 43

Sex, no. (%)

Male 32 (68.1) 31 (72.1) .68

Female 15 (31.9) 12 (27.9) .68

Age, y, mean (SD) 62.3 (9.8) 62.7 (14.7) .87

Systolic BP (mm Hg) 107.7 (18.3) 101.0 (16.5) .07

CPT score (SD) 7.9 (2.5) 8.1 (2.3) .66

Hemoglobin, g/L (SD) 95.7 (24.4) 92.0 (23.5) .46

Hematemesis duration before endoscopy (h) 10.4 (2.8) 9.8 (3.4) .41

Time from admission to endoscopy (h) 3.6 (1.3) 3.3 (1.1) .61

Diabetes mellitus, no. (%) 22 (46.8) 14 (32.5) .17

Source of bleeding

Esophageal varices, no. (%) 43 (91.4) 38 (88.3) .72

Fundal varices, no. (%) 3 (6.4) 3 (7.0) .91

Portal gastropathy, no. (%) 1 (2.1) 2 (4.6) .33

Intervention, no. (%)

Variceal banding 33 (70.2) 21 (48.8) ⬍.04

Sclerotherapy 5 (10.6) 7 (16.3) .43

Sclerotherapy ⫹ variceal banding 5 (10.6) 7 (16.3) .43

Fundal varices injection 3 (6.4) 3 (7.0) .91

No intervention 1 (2.1) 5 (11.6) .07

BP, Blood pressure; CPT, Child-Pugh-Turcotte; SD, standard deviation.

vs 26.0 minutes, respectively; P ⬍ .005). Bleeding control by
band ligation was possible in a significantly higher number of
patients in the erythromycin group (n ⫽ 33, 70.2%) than in
the placebo group (n ⫽ 21, 48.8%) (P ⬍ .04). The need for
sclerotherapy or a combination of banding and sclerotherapy
did not differ in the 2 groups. Five patients in the placebo
group and 1 in the erythromycin group had no intervention
at the first endoscopy. Among these 3 patients in the placebo
group had a stomach full of blood and no identification of the
bleeding source at the first endoscopy and 2 were bleeding
from severe portal gastropathy. In the erythromycin group, 1
patient was bleeding from severe portal gastropathy. The
length of hospital stay was significantly shorter in the eryth-
romycin group than in the placebo group (3.4 days vs 5.1
days, respectively; P ⬍ .002). The need for a second-look
endoscopy within 24 hours was less in the erythromycin
group than in the placebo group (4.3% vs 9.3%, respectively;
not statistically significant). A second-look endoscopy was
performed in 2 patients in the erythromycin group because
of rebleeding and 4 patients in the placebo group (3 for
unsatisfactory bleeding control because of blood in the stom-
ach at the first endoscopy and 1 for rebleeding). The number
of patients who needed blood transfusions and the mean
number of units of blood transfused during the 24 hours after
the endoscopy did not differ significantly in the 2 groups. The
frequency of aspiration pneumonia was higher in the pla-
cebo group than in the erythromycin group (4.7% vs 0.0%,
respectively; not statistically significant). The in-hospital mor-
tality rate was similar in the 2 groups. No adverse reactions
related to erythromycin were observed. Finally, in a multi-
variate logistic regression analysis, we found that erythromy-
cin administration was the only independent predictor of an
empty stomach (Table 4).
DISCUSSION
This study was performed to determine the effects of
intravenous erythromycin on endoscopic visibility in patients
with liver cirrhosis with variceal bleeding. Erythromycin,
when given intravenously, is known to have a powerful

248 GASTROINTESTINAL ENDOSCOPY Volume 73, No. 2 : 2011 www.giejournal.org

Altraif et al Effect of erythromycin before endoscopy in patients with variceal bleeding

TABLE 3. Effects of erythromycin and placebo in the study population

Variables Erythromycin Placebo P value

No. of patients 47 43

Empty stomach, no. (%) 23 (48.9) 10 (23.3) ⬍.012

Endoscopic score, mean (SD) 12.5 (3.6) 9.8 (4.2) ⬍.05

Endoscopy duration, min, mean (SD) 19.0 (9.8) 26.0 (13.4) ⬍.005

Hospital stay, d, mean (SD) 3.4 (2.4) 5.1 (2.9) ⬍.002

No. of units of blood transfused in 24 h, mean (SD) 3.0 (1.8) 3.6 (2.8) .64

Transfusion of PRBCs, no. (%) of patients 12 (25.5) 11 (25.6) .99

Second-look endoscopy, no. (%) 2 (4.3) 4 (9.3) .33

Aspiration pneumonia, no. (%) 0.0 2 (4.7) .23

In-hospital mortality, no. (%) 4 (8.5) 6 (13.9) .41

PRBCs, Packed red blood cells; SD, standard deviation.

ing the ease of performance of the procedure and therapeutic

TABLE 4. Multivariate logistic regression analysis of intervention and less time spent washing, cleaning, and vi-
variables associated with an empty stomach sualizing the rest of the stomach. Bleeding control was pos-
sible by band ligation in a significantly higher number of
Variable Odds ratio 95% CI P value
patients in the erythromycin group than in the placebo group
Erythromycin 2.64 1.02-6.82 ⬍.05 (70.2% vs 48.8%; P ⬍ .04). An empty stomach without reflux
Age 0.99 0.96-1.03 .8 of blood into the esophagus during endoscopy gave a better
view of esophageal varices, making band ligation possible in
Sex 1.39 0.50-3.88 .5
more patients in the erythromycin group. In addition, there
Systolic blood pressure 1.04 1.00-1.07 .7 was a significant decrease in the mean duration of hospital
Hemoglobin 0.99 0.98-1.02 .8 stay in the erythromycin group. Although there was a trend
toward a decrease in the need for repeat endoscopy and
Hematemesis duration 1.08 0.91-1.27 .4
endoscopy-related pulmonary complications, these did not
CPT score 0.98 0.79-1.21 .9 reach statistical significance, perhaps because of the small
CI, Confidence interval; CPT, Child-Pugh-Turcotte. sample size. A previous study by Frossard et al11 of 105
patients with hematemesis randomized to receive either
erythromycin (250 mg intravenously over 5 minutes) or pla-
effect on gastric clearance by inducing antral contractions cebo showed that erythromycin shortened endoscopy dura-
similar to phase 3 of the interdigestive migrating motor com- tion and improved the endoscopic yield, resulting in a de-
plex.14,15 These motor effects appear rapidly after intrave- creased need for second-look endoscopy. Coffin et al12
nous injection, both in fasting and postprandial states.8,9 The reported similar results in 41 patients with acute upper GI
effects persist for at least 2 hours and are seen in severely ill bleeding comparing erythromycin (3 mg/kg intravenously)
patients as well.9,16 The optimal dose that increases gastric with no treatment. Although this study found a significant
emptying appears to be 2 to 3 mg/kg given intravenously improvement in the quality of endoscopy, the need for a
over 10 to 30 minutes.5 The results of 2 previous randomized second-look endoscopy showed only a trend (P ⫽ .089)
studies suggested the usefulness of erythromycin in acute toward a decrease in the erythromycin group. In our study,
upper GI bleeding.11,12 However, the studies were con- we found a clear stomach in 48.9% of patients who received
ducted with a mixed population of mostly nonvariceal bleed- erythromycin compared with 82% in the study by Frossard et
ing and a small number of variceal bleeding patients. In the al and 90% in the study by Coffin et al. This could be partly
current study, which is the first randomized, double-blind, because of differences among the studies in defining primary
placebo-controlled study of patients with variceal bleeding, outcome. With the objective scoring system used by Frossard
we demonstrated that intravenous erythromycin improved et al, more than 75% visible surface and a total score of more
the quality of the endoscopic view because it significantly than 6 out of maximum total of 8 defined a clear stomach,
increased the number of patients with a clear stomach. This whereas in our study, more than 95% visible surface and a
resulted in a shortened mean duration of endoscopy, reflect- total score of 15 or more out of maximum of 16 defined a

www.giejournal.org Volume 73, No. 2 : 2011 GASTROINTESTINAL ENDOSCOPY 249

Effect of erythromycin before endoscopy in patients with variceal bleeding
clear stomach. Similarly, Coffin et al used a scale of 0 to 3
with a score of 2 or more judged to be a good quality
esophagogastroduodenal examination. Accordingly, some
patients classified as having a clear stomach in these 2 trials
would not have been classified as having a clear stomach in
our trial. Also in the study by Coffin et al,12 all patients had
gastric lavage before undergoing endoscopy, whereas no
patient in our study had nasogastric tube or gastric lavage
before undergoing endoscopy. In the studies of both
Frossard et al and Coffin et al, erythromycin was adminis-
tered at a dose of 250 mg (3 mg/kg), which has been shown
to have the most powerful effects on gastric clearance,17
whereas in our study, we used a standard dose of 125 mg,
which might have been suboptimal in some of our patients,
although erythromycin at a dose of 70 mg is as effective for
gastric emptying as 200 mg in critically ill patients.18 Finally,
this could be partly a result of differences in the study pop-
ulation. Our study was exclusively of cirrhotic patients with
variceal bleeding, whereas the other 2 studies were predom-
inantly of patients with nonvariceal bleeding. Although re-
sidual blood in the stomach does not interfere with the
diagnosis of active esophageal variceal bleeding, it does
obscure visualization and injection of bleeding gastric varices
and does prolong the endoscopy procedure time spent
cleaning and visualizing the rest of the stomach. In addition,
the other 2 studies assessed the need for repeat endoscopy
from 1 to 8 days after the initial endoscopy, whereas our
study assessed the need for repeat endoscopy during first 24
hours. This could explain the nonsignificant difference in the
need for a second-look endoscopy between the erythromy-
cin and placebo groups in our study. Our study showed a
significant decrease in the mean duration of hospital stay in
the erythromycin group, a secondary outcome, and an indi-
rect indicator of the efficacy of endoscopic intervention. The
patients in the erythromycin group were discharged earlier
than patients in the placebo group. The blood transfusion
requirement and all-cause hospital mortality were similar in
both groups. The study was not designed to find outcome
differences in terms of survival benefit. The 2 groups were
well matched in patient characteristics such as age, sex, mean
duration of bleeding before endoscopy, baseline hemoglo-
bin, systolic blood pressure, severity of liver disease, and
source of bleeding, and none of these factors independently
influenced the study outcome, as shown by multivariate
logistic regression analysis. Finally, no drug-related adverse
effects were seen.
CONCLUSIONS
Our study showed that erythromycin administered be-
fore endoscopy in cirrhotic patients with variceal bleeding
improves the quality of endoscopic visualization, shortens
the procedure time, and possibly decreases the hospital
stay. It may decrease the need for a second-look endos-
copy. With an excellent safety profile and low cost, intra-
250 GASTROINTESTINAL ENDOSCOPY Volume 73, No. 2 : 2011
Altraif et al
venous erythromycin may be considered to clear the stom-
ach before endoscopy in patients with variceal bleeding.
The results of this study confirm the results of previously
reported similar clinical trials.
ACKNOWLEDGMENTS
We thank the endoscopy staff for their cooperation and
our department secretary, Lourdes Maristela, for her sec-
retarial assistance.
REFERENCES
1. Carbonell N, Pauwels A, Serfaty L, et al. Improved survival after variceal
bleeding in patients with cirrhosis over the past two decades. Hepatol-
ogy 2004;40:652-9.
2. Besson I, Person B, Boutroux D, et al. Sclerotherapy with or without oc-
treotide for acute variceal bleeding. N Engl J Med 1995;333:555-60.
3. Bohnacker S, Sriram PV, Soehendra N. The role of endoscopic therapy in
the treatment of bleeding varices. Baillieres Best Pract Res Clin Gastro-
enterol 2000;14:477-94.
4. Stollman NH, Putcha RV, Neustater BR, et al. The uncleared fundal pool
in acute upper gastrointestinal bleeding; implications and outcomes.
Gastrointest Endosc 1997;46:324-7.
5. Peeters TL. Erythromycin and other macrolides as prokinetic agents.
Gastroenterology 1993;105:1886-99.
6. Mantides A, Xynos E, Chrysos F, et al. The effect of erythromycin in gas-
tric emptying of solids and hypertonic liquids in healthy subjects. Am J
Gastroenterol 1993;88:198-202.
7. Bruley des Varannes S, Parys V, Ropert A, et al. Erythromycin enhances
fasting and postprandial proximal gastric tone in humans. Gastroenter-
ology 1995;109:32-9.
8. Sarna SK, Soergel KH, Koch TR, et al. Gastrointestinal motor effects of
erythromycin in humans. Gastroenterology 1991;101:1488-96.
9. Annese V, Janssens J, Vantrappen G, et al. Erythromycin accelerates gas-
tric emptying by inducing antral contractions and improved gastrodu-
odenal coordination. Gastroenterology 1992;102:823-8.
10. Nion I, Andant C, Jouet P, et al. Role of intravenous erythromycin in the
preparation for endoscopy in case of upper digestive hemorrhage. Gas-
troenterol Clin Biol 1998;22:554-5.
11. Frossard JL, Spahr L, Queneau PE, et al. Erythromycin intravenous bolus
infusion in acute upper gastrointestinal bleeding: a randomized con-
trolled double blind trial. Gastroenterology 2002;123:17-23.
12. Coffin B, Pocard M, Panis Y, et al. Erythromycin improves the quality of
EGD in patients with acute upper GI bleeding: a randomized controlled
study. Gastrointest Endosc 2002;56:174-9.
13. Carbonell N, Pauwels A, Serfaty L, et al. Erythromycin infusion prior to
endoscopy for acute upper gastrointestinal bleeding; a randomized,
controlled, double– blind trial. Am J Gastroenterol 2006;101:1211-5.
14. Weber FH Jr, Richards RD, McCallum RW. Erythromycin: a motillin ago-
nist and gastrointestinal prokinetic agent. Am J Gastroenterol 1993;88:
485-90.
15. Peeters T, Matthijs G, Depoortere I, et al. Erythromycin is a motilin recep-
tor agonist. Am J Physiol 1989;257:470-4.
16. Chapman MJ, Fraser RJ, Kluger MT, et al. Erythromycin improves gastric
emptying in critically ill patients intolerant of nasogastric feeding. Crit
Care Med 2000;28:2334-7.
17. Michel A, Boivin MD, Michael C, et al. Erythromycin accelerates gastric
emptying in a dose–response manner in healthy subjects. Pharmaco-
therapy 2003;23:5-8.
18. Ritz MA, Chapman MJ, Fraser RJ, et al. Erythromycin dose of 70 mg ac-
celerates gastric emptying as effectively as 200 mg in the critically ill.
Intensive Care Med 2005;31:949-54.
www.giejournal.org