REVIEW

CME EDUCATIONAL OBJECTIVE: Readers will evaluate and manage acute upper gastrointestinal bleeding effectively
CREDIT
Mazen Albeldawi, MD Mohammed A. Qadeer, MD, MPH John J. Vargo, MD, MPH*
Department of Internal Medicine, Department of Gastroenterology and Hepatology, Department of Gastroenterology and
Cleveland Clinic Digestive Disease Institute, Cleveland Clinic Hepatology, Digestive Disease Institute,
Cleveland Clinic

Managing acute upper GI bleeding,

preventing recurrences
■ ■Abstract
U pper gastrointestinal (GI) bleeding
is common, costly, and potentially life-
threatening. It must be managed promptly and
Acute upper gastrointestinal (GI) bleeding is common
and potentially life-threatening and needs a prompt appropriately to prevent adverse outcomes.
assessment and aggressive medical management. All More people are admitted to the hospital
patients need to undergo endoscopy to diagnose, assess, for upper GI bleeding than for congestive heart
failure or deep vein thrombosis. In the United
and possibly treat any underlying lesion. In addition,
States, the annual rate of hospitalization for
patients found to have bleeding ulcers should receive a upper GI bleeding is estimated to be 165 per
proton pump inhibitor, the dosage and duration of treat- 100,000—more than 300,000 hospitalizations
ment depending on the endoscopic findings and clinical per year, at a cost of $2.5 billion.1,2
factors. Furthermore, despite advances in therapy,
the case-fatality rate has remained unchanged
■ ■Key Points at 7% to 10%.3 This may be because today’s
The first priority is to ensure that the patient is hemody- patients are older and have more comorbidi-
ties than those in the past.4
namically stable, which often requires admission to the
intensive care unit for monitoring and fluid resuscitation.
■■ cAUSES OF UPPER GI BLEEDING
Peptic ulcers account for most cases of upper GI bleed- Peptic ulcers account for about 60% of se-
ing, but bleeding from varices has a much higher case- vere cases of upper GI bleeding,5 and they are
fatality rate and always demands aggressive treatment. the focus of this paper. Fortunately, up to 80%
of bleeding ulcers stop bleeding spontaneously
Patients with ulcer disease should be tested and treated without any intervention.6
for Helicobacter pylori infection. Gastroduodenal erosions account for
about 12%.3
Varices due to cirrhosis are less common
Patients with a history of bleeding ulcers who need
but more dangerous. Variceal bleeding ac-
long-term treatment with aspirin or a nonsteroidal anti- counts for a relatively small percentage (6%)
inflammatory drug should also be prescribed a proton of upper GI bleeding, but the mortality rate
pump inhibitor. from a single episode of variceal bleeding is
30%, and 60% to 70% of patients die within 1
year, mostly of underlying liver disease.
Less frequent causes include Mallory-
Weiss tears, erosive duodenitis, Dieulafoy ul-
cer (a type of vascular malformation), other
*
Dr. Vargo has disclosed that he has received consulting fees from Ethicon Endo- vascular lesions, neoplasms, aortoenteric fistu-
Surgery and honoraria for teaching and speaking from Olympus America, Inc. la, gastric antral vascular ectasia, and prolapse
doi:10.3949/ccjm.77a.09035 gastropathy.5
CL EVEL AND CL I NI C J O URNAL O F M E DI CI NE    V O L UM E 77  •   NUM BE R 2   F E BRUARY   2010  131
 Acute Upper Gastrointestinal Bleeding

■■ Hematemesis and melena • Endoscopic stigmata or recent hemor-

rhage—0 points for a clean-based ulcer or
The most common presenting signs of acute flat pigmented spot; 2 points for blood in
upper GI bleeding are hematemesis (vomiting the upper GI tract, active bleeding, a non-
of blood), “coffee grounds” emesis, and mele- bleeding visible vessel, or adherent clot.
na (tarry black stools). About 30% of patients The Rockall score can thus range from 0
with bleeding ulcers present with hematem- to 11 points, with an overall score of 0, 1, or 2
esis, 20% with melena, and 50% with both.7 associated with an excellent prognosis.10
Hematochezia (red blood in the stool) usu- The Blatchford scoring system uses only
ally suggests a lower GI source of bleeding, since clinical and laboratory factors and has no endo-
blood from an upper source turns black and tarry scopic component (TABLE 1). In contrast to the
as it passes through the gut, producing melena. Rockall score, the main outcome it predicts is
However, up to 5% of patients with bleeding ul- the need for clinical intervention (endoscopy,
cers have hematochezia,7 and it indicates heavy surgery, or blood transfusion). The Blatchford
bleeding: bleeding of approximately 1,000 mL score ranges from 0 to 23; most patients with a
into the upper GI tract is needed to cause hema- score of 6 or higher need intervention.11
tochezia, whereas only 50 to 100 mL is needed Other systems that are used less often in-
to cause melena.8,9 Hematochezia with signs and clude the Baylor severity scale and the Acute
symptoms of hemodynamic compromise such as Physiology and Chronic Health Evaluation
syncope, postural hypotension, tachycardia, and (APACHE) II score.
shock should therefore direct one’s attention to
an upper GI source of bleeding. Does the patient have varices?
Nonspecific features include nausea, vom- All variceal bleeding should be considered se-
iting, epigastric pain, vasovagal phenomena, vere, since the 1-year death rate is so high (up
and syncope. to 70%). Clues pointing to variceal bleeding
include previous variceal bleeding, thrombo-
■■ WHAT IS the PATIENT’S RISK? cytopenia, history of liver disease, and signs of
More people liver disease on clinical examination.
are admitted An assessment of clinical severity is the first All patients suspected of having bleeding
critical task, as it helps in planning treatment. varices should be admitted to the intensive
for upper GI Advanced age, multiple comorbidities, and care unit for close monitoring and should be
bleeding than hemodynamic instability call for aggressive given the highest priority, even if they are he-
treatment. Apart from this simple clinical modynamically stable.
for congestive rule, scoring systems have been developed.
heart failure The Rockall scoring system, the most Is the patient hemodynamically stable?
widely used, gives estimates of the risks of re- Appropriate hemodynamic assessment in-
current bleeding and death. It is based on the cludes monitoring of heart rate, blood pres-
three clinical factors mentioned above and on sure, and mental status. Tachycardia at rest,
two endoscopic ones, awarding points for: hypotension, and orthostatic changes in vi-
• Age—0 points if less than 60; 1 point if 60 tal signs indicate a considerable loss of blood
to 79; or 2 points if 80 years or older volume. Low urine output, dry mucous mem-
• Shock—1 point if the pulse is more than branes, and sunken neck veins are also useful
100; 2 points if the systolic blood pressure signs. (Tachycardia may be blunted if the pa-
is less than 100 mm Hg tient is taking a beta-blocker.)
• Comorbid illness—2 points for ischemic If these signs of hypovolemia are present,
heart disease, congestive heart failure, or the initial management focuses on treating
other major comorbidity; 3 points for renal shock and on improving oxygen delivery to
failure, hepatic failure, or metastatic disease the vital organs. This involves repletion of
• Endoscopic diagnosis—0 points if no le- the intravascular volume with intravenous
sion found or a Mallory-Weiss tear; 1 point infusions or blood transfusions. Supplemental
for peptic ulcer, esophagitis, or erosive dis- oxygen also is useful, especially in elderly pa-
ease; 2 points for GI malignancy tients with heart disease.12
132  CLEV ELA N D C LI N I C JOURNAL OF MEDICINE   VOL UME 77  •  N UM BE R 2   F E BRUARY   2010
 Albeldawi and Colleagues

Inspection of nasogastric aspirate

TABLE 1
In the initial assessment, it is useful to insert a
nasogastric tube and inspect the aspirate. If it The Blatchford scoring system
contains bright red blood, the patient needs an
Variables Points
urgent endoscopic evaluation and an intensive
level of care13,14; if it contains coffee-grounds Systolic blood pressure (mm Hg)
material, the patient needs to be admitted to 100–109 1
the hospital and to undergo endoscopic evalu- 90–99 2
ation within 24 hours. < 90 3
However, a normal aspirate does not rule Blood urea nitrogen (mmol/L)
out upper GI bleeding. Aljebreen et al15 found 6.5–7.9 2
that 15% of patients with upper GI bleeding 8.0–9.9 3
and normal nasogastric aspirate still had high- 10.0–24.9 4
risk lesions (ie, visible bleeding or nonbleed- > 25 6
ing visible vessels) on endoscopy. Hemoglobin (men; g/dL)
12.0–12.9 1
■■ acid-suppression helps ulcers heal 10.0–11.9 3
< 10.0 6
Acid and pepsin interfere with the healing of
Hemoglobin (women; g/dL)
ulcers and other nonvariceal upper GI lesions.
10.0–11.9 1
Further, an acidic environment promotes < 10.0 2
platelet disaggregation and fibrinolysis and
impairs clot formation.16 This suggests that in- Other variables
hibiting gastric acid secretion and raising the Pulse > 100 1
gastric pH to 6 or higher may stabilize clots. Presentation with melena 1
Hepatic disease 2
Moreover, pepsinogen in the stomach is con-
Cardiac failure 2
verted to its active form (pepsin) if the pH is
less than 4. Therefore, keeping the pH above Total ___ Old age,
4 keeps pepsinogen in an inactive form. comorbidities,
Most patients need intervention if their score is 6 or higher.

Histamine-2 receptor antagonists

Conversely, few patients need intervention if their systolic
blood pressure is 110 mm Hg or less, their blood urea
and
Histamine-2 receptor antagonists were the nitrogen is less than 6.5 mmol/L, their hemoglobin level is 13 hemodynamic
first drugs to inhibit acid secretion, reversibly g/dL or higher (in men) or 12 g/dL or higher (in women), and
blocking histamine-2 receptors on the baso- their pulse is less than 100. instability
lateral membrane of parietal cells. However, Adapted from Blatchford O, Murray WR, Blatchford M. A
risk score to predict need for treatment for upper-gastro-
should prompt
these drugs did not prove very useful in man- intestinal haemorrhage. Lancet 2000; 356:1318–1321.
aggressive
with permission from Elsevier,
aging upper GI bleeding in clinical trials.17,18 In
treatment
www.sciencedirect.com/science/journal/01406736.
their intravenous form, they often fail to keep
the gastric pH at 6 or higher, due to tachy­
phylaxis.19 The use of this class of drugs has Started after endoscopy. Randomized
declined in favor of proton pump inhibitors. controlled trials have found proton pump in-
hibitors to be effective when given in high
Proton pump inhibitors doses intravenously for 72 hours after success-
Proton pump inhibitors reduce both basal and ful endoscopic treatment of bleeding ulcers
stimulated acid secretion by inhibiting hydro- with high-risk endoscopic signs, such as active
gen-potassium adenosine triphosphatase, the bleeding or nonbleeding visible vessels.22,23
proton pump of the parietal cell. A meta-analysis indicated that these drugs
Multiple studies have shown that proton decrease the incidence of recurrent peptic ul-
pump inhibitors raise the gastric pH and keep cer bleeding, the need for blood transfusions,
it high. For example, an infusion of omeprazole the need for surgery, and the duration of hos-
(Prilosec) can keep the gastric pH above 6 for pitalization, but not the mortality rate.24,25
72 hours without inducing tachyphylaxis.20,21 These studies also illustrate the benefit of fol-
CL EVEL AND CL I NI C J O URNAL O F M E DI CI NE    V O L UM E 77  •   NUM BE R 2   F E BRUARY   2010  133
 Acute Upper Gastrointestinal Bleeding

lowing up endoscopic treatment to stop the tients with uncontrolled nonvariceal bleeding
bleeding with an intravenous infusion of a who are awaiting endoscopy, since it is rela-
proton pump inhibitor. tively safe to use.
The recommended dose of omeprazole for
patients with high-risk findings on endoscopy ■■ ALL PATIENTS NEED ENDOSCOPY
is an 80-mg bolus followed by an 8-mg/hour
infusion for 72 hours. After the patient’s con- All patients with upper GI bleeding need an
dition stabilizes, oral therapy can be substi- upper endoscopic examination to diagnose
tuted for intravenous therapy. In patients with and assess the risk posed by the bleeding le-
low-risk endoscopic findings (a clean-based sion and to treat the lesion, reducing the risk
ulcer or flat spot), oral proton pump inhibitors of recurrent bleeding.
in high doses are recommended.
In either case, after the initial bleeding How urgently does endoscopy
is treated endoscopically and hemostasis is need to be done?
achieved, a proton pump inhibitor is recom- Endoscopy within the first 24 hours of upper
mended for 6 to 8 weeks, or longer if the pa- GI bleeding is considered the standard of care.
tient is also positive for Helicobacter pylori or Patients with uncontrolled or recurrent bleed-
is on daily treatment with aspirin or a non- ing should undergo endoscopy on an urgent
steroidal anti-inflammatory drug (NSAID) basis to control the bleeding and reduce the
that is not selective for cyclo-oxygenase 2 (see risk of death.
below). However, how urgently endoscopy needs
Started before endoscopy, these drugs re- to be done is often debated. A multicenter
duced the frequency of actively bleeding ul- randomized controlled trial compared out-
cers, the duration of hospitalization, and the comes in patients who underwent endoscopy
need for endoscopic therapy in a randomized within 6 hours of coming to the emergency
controlled trial.26 A meta-analysis found that department vs within 24 hours after the initial
significantly fewer patients had signs of recent evaluation. The study found no significant dif-
All variceal bleeding on endoscopy if they received a pro- ference in outcomes between the two groups;
bleeding ton pump inhibitor 24 to 48 hours before the however, the group that underwent endoscopy
procedure, but it did not find any significant sooner needed fewer transfusions.29
should be difference in important clinical outcomes
considered such as death, recurrent bleeding, or surgery.27 For a better view of the stomach
Nevertheless, we believe that intravenous pro- Gastric lavage improves the view of the
severe ton pump inhibitor therapy should be started gastric fundus but has not been proven to im-
before endoscopy in patients with upper GI prove outcome.30
bleeding. Promotility agents such as erythromycin
and metoclopramide (Reglan) are also used to
Somatostatin analogues empty the stomach for better visualization.31–35
Octreotide (Sandostatin), an analogue of the Erythromycin has been shown to improve
hormone somatostatin, decreases splanchnic visualization, shorten the procedure time,
blood flow, decreases secretion of gastric acid and prevent the need for additional endos-
and pepsin, and stimulates mucus production. copy attempts in two randomized controlled
Although it is beneficial in treating upper GI studies.33,34 Furthermore, a cost-effectiveness
bleeding due to varices, its benefit has not study confirmed that giving intravenous eryth-
been confirmed in patients with nonvariceal romycin before endoscopy for acute upper GI
upper GI bleeding. bleeding saved money and resulted in an in-
A meta-analysis revealed that outcomes crease in quality-adjusted life-years.35
were better with high-dose intravenous proton
pump inhibitor therapy than with octreotide Endoscopy to diagnose bleeding
when these drugs were started after endoscop- and assess risk
ic treatment of acute peptic ulcer bleeding.28 Upper endoscopy is 90% to 95% diagnostic
Nevertheless, octreotide may be useful in pa- for acute upper GI bleeding.36
134  CLEV ELA N D C LI N I C JOURNAL OF MEDICINE   VOL UME 77  •  N UM BE R 2   F E BRUARY   2010
 Albeldawi and Colleagues

Nonbleeding visible vessel (high risk) Dieulafoy ulcer (high risk)

Clean-based ulcer (low risk)

Red spot in a gastric ulcer (low risk)

A normal
aspirate does
not rule out
upper GI
FIGURE 1. Endoscopic stigmata of bleeding peptic ulcer (arrows) and risk of recurrent bleeding
bleeding and death.

Furthermore, some of the clinical scor- on the lesser curvature of the stomach, ulcers
ing systems are based on endoscopic findings in the posterior or superior duodenal bulb, ul-
along with clinical factors on admission. These cers larger than 2 cm in diameter, and ulcers
scoring systems are valuable for assessing pa- that are actively bleeding at the time of endos-
tients with nonvariceal upper GI bleeding, as copy.37 Other endoscopic findings that predict
they predict the risk of death, longer hospi- clinical outcome are summarized in Table 2.
tal stay, surgical intervention, and recurrent Patients at high risk (ie, older than 60 years,
bleeding.37,38 Patients with endoscopic find- with severe comorbidity, or hemodynamically
ings associated with higher rates of recurrent compromised) who have active bleeding (ie,
bleeding and death (FIGURE 1) need aggressive witnessed hematemesis, red blood per naso-
management. gastric tube, or fresh blood per rectum) or a
Certain factors, primarily clinical and en- nonbleeding visible vessel should be admitted
doscopic, predict that endoscopic treatment to a monitored bed or intensive care unit. Ob-
will fail to stop ulcer bleeding. Clinical factors servation in a regular medical ward is appro-
include a history of peptic ulcer bleeding and priate for high-risk patients found to have an
hemodynamic compromise at presentation. adherent clot. Patients with low-risk findings
Endoscopic factors include ulcers located high (eg, a clean ulcer base) are at low risk of recur-
CL EVEL AND CL I NI C J O URNAL O F M E DI CI NE    V O L UM E 77  •   NUM BE R 2   F E BRUARY   2010  135
 Acute Upper Gastrointestinal Bleeding

TABLE 2
Endoscopic findings as predictors of clinical outcome
Endoscopic finding Prevalence % recurrent bleeding (%) Surgery (%) Mortality (%)

Active bleeding 18 55 35 11
Visible vessel 17 43 34 11
Adherent clot 17 22 10 7
Flat spot 20 10 6 3
Clean-base ulcer 42 5 0.5 2
Adapted from Laine L, Peterson WL. Bleeding peptic ulcer. N Engl J Med 1994; 331:717–727.
Copyright 1994 Massachusetts medical society. All rights reserved.

rent bleeding and may be considered for early ated with higher rates of recurrent bleeding
hospital discharge with appropriate outpatient following endoscopic therapy for nonvariceal
follow-up. upper GI bleeding.47
Patients with refractory bleeding are can-
Endoscopy to treat bleeding didates for angiography or surgery. However,
About 25% of endoscopic procedures per- even when endoscopic hemostasis fails, en-
formed for upper GI bleeding include some doscopy is important before angiography or
type of treatment,39 such as injections of epi- surgery to pinpoint the site of bleeding and
nephrine, normal saline, or sclerosants; ther- diagnose the cause.
mal cautery; argon plasma coagulation; electro- A second endoscopic procedure is gener-
cautery; or application of clips or bands. They ally not recommended within 24 hours after
are all equally effective, and combinations of the initial procedure.48 However, it is appro-
these therapies are more effective than when priate in cases in which clinical signs indicate
they are used individually. A recent meta-anal- recurrent bleeding or if hemostasis during the
In cases of ysis found dual therapy to be superior to epi- initial procedure is questionable. A meta-anal-
active bleeding nephrine monotherapy in preventing 40recurrent ysis found that routinely repeating endoscopy
bleeding, need for surgery, and death. reduces the rate of recurrent bleeding but not
or nonbleeding Endoscopic therapy is recommended for the need for surgery or the risk of death.49
visible vessels, patients found to have active bleeding or non-
bleeding visible blood vessels, as outcomes are ■■ All patients Should be admitted
continue better with endoscopic hemostatic treatment
IV omeprazole than with drug therapy alone (Table 3).41–44 All patients with upper GI bleeding should be
for 72 hours How to manage adherent clots is contro- admitted to the hospital, with the level of care
versial, but recent studies have revealed a dictated by the severity of their clinical condi-
significant benefit from removing them and tion (FIGURE 2).
treating the underlying lesions compared with
drug therapy alone.43,45 ■■ Variceal bleeding
Flat, pigmented spots and nonbleeding
ulcers with a clean base do not require endo- Variceal bleeding, a severe outcome of portal
scopic treatment because the risk of recurrent hypertension secondary to cirrhosis, carries
bleeding is low. a 6-week mortality rate of 10% to 20%.50 In
Endoscopic therapy stops the bleeding in view of the risk, primary prevention is indi-
more than 90% of patients, but bleeding recurs cated in patients with high-risk varices.
after endoscopic therapy in 10% to 25%.46 Re- The mainstays of primary and secondary
versal of any severe coagulopathy with trans- prevention are the nonselective beta-blockers
fusions of platelets or fresh frozen plasma is such as nadolol (Corgard) and propranolol
essential for endoscopic hemostasis. However, (Inderal). Several randomized controlled tri-
coagulopathy at the time of initial bleeding als have shown lower rates of recurrent bleed-
and endoscopy does not appear to be associ- ing and death with propranolol or nadolol
136  CLEV ELA N D C LI N I C JOURNAL OF MEDICINE   VOL UME 77  •  N UM BE R 2   F E BRUARY   2010
 Albeldawi and Colleagues

TABLE 3
Signs of ulcer hemorrhage and risk of recurrent bleeding
with endoscopic hemostasis vs medical therapy
SIGNS Risk of recurrent bleeding Risk of recurrent bleeding
with Medical therapy alone with Endoscopic hemostasis

Active arterial bleeding (spurting) 85%–95% 10%–20%

Nonbleeding visible vessel 50% 5%–-10%
Nonbleeding adherent clot 35% < 5%
Ulcer oozing 10%–25% < 5%
Flat spots 7% Not indicated
Clean-based ulcer 3% Not indicated
Data are from Kovacs and Jensen41,42 and Jensen and Machicado.44

than with placebo.51 In doses that decrease doscopic therapy. It is not the primary thera-
the heart rate by 25%, beta-blockers have py because it doubles the risk of encephalopa-
been shown to delay and decrease variceal thy and has a high stent occlusion rate (up to
hemorrhage. However, most patients require 60%, lower with covered stents).
prophylactic endoscopic variceal ligation be-
cause they cannot tolerate beta-blocker ther- ■■ GI Bleeding can cause
apy. acute Myocardial infarction
In suspected acute variceal bleeding, a
somatostatin analogue should be started to The simultaneous presentation of acute myo-
decrease the portal pressure, and antibiotics cardial infarction (MI) and GI hemorrhage is
should be started to reduce the risks of infec- very serious and unfortunately common. Bleeding recurs
tion and death. Vasoactive drugs, ie, soma- An acute MI occurring simultaneously after
tostatin analogues, should be started before with or after GI bleeding is usually precipitat-
endoscopy and continued for 5 days to reduce ed by massive bleeding causing hypovolemia, endoscopic
the chances of recurrent bleeding.52,53 hemodynamic compromise, and hypoperfu- treatment in
Terlipressin is the only drug proven to im- sion. Conversely, the anticoagulant, anti-
prove the odds of survival in acute variceal platelet, or thrombolytic drugs given to treat
10% to 25%
bleeding. Although widely used in Europe, it MI can precipitate GI bleeding (see below). of cases
has not been approved for use in the United This distinction is important because the
States. two scenarios have different clinical courses
Octreotide, another option, improves he- and prognoses. GI bleeding that precipitates
mostasis to the same extent, although it does an acute MI tends to be massive, whereas GI
not increase the survival rate.54,55 The recom- bleeding after treatment of acute MI tends to
mended dose of octreotide for patients with be self-limited and often resolves with rever-
variceal bleeding is a 50-μg intravenous bo- sal of underlying coagulopathy.57
lus, followed by a 50-μg/hour infusion for 5 Endoscopy carries a higher than average
days. risk in patients with recent acute MI, with
Combining endoscopic and drug therapy all-cause mortality rates as high as 1%.58 (The
improves the chances of stopping the bleed- usual rate is 0.0004%.59) Nevertheless, endos-
ing and reduces the risk of recurrent bleeding copy can be safely performed early on in pa-
compared with endoscopic therapy alone.56 tients with acute MI if it is done under strict

Transjugular intrahepatic portosystemic monitoring in a coronary care unit.
shunting is indicated in recurrent variceal Several studies have shown that MI pa-
hemorrhage or in those with initial bleeding tients who present with upper GI bleeding as
that is refractory to standard medical and en- the inciting event or patients with acute MI
CL EVEL AND CL I NI C J O URNAL O F M E DI CI NE    V O L UM E 77  •   NUM BE R 2   F E BRUARY   2010  137
 Acute Upper Gastrointestinal Bleeding

Acute upper GI bleeding

Ulcer Varices

Active bleeding or Adherent clot Flat, pigmented Clean base

visible vessel spot

Intravenous proton Intravenous proton Oral proton pump Oral proton pump Intravenous octreotide
pump inhibitor plus pump inhibitor plus inhibitor; no endo- inhibitor; no endo- (Sandostatin) plus
endoscopic endoscopic scopic treatment scopic treatment endoscopic treatment
treatment treatment required required

Intensive care unit Ward 3 days Ward 1–3 days Discharge Intensive care unit
1 day, then ward 2 1–2 days; ward 2–3
days days

FIGURE 2. Algorithm for patients with acute upper gastrointestinal bleeding.

who are vomiting blood or who are hemody- doscopy can be safely performed in patients
namically unstable due to GI bleeding are sig- with acute GI bleeding whose INR is between
nificantly more likely to have a high-risk lesion 2.0 and 3.0.62,63 Some suggest that both the
and so have the greatest need for endoscopic length of warfarin therapy and the INR affect
therapy. Therefore, endoscopic intervention the risk of bleeding.64,65
When starting may be offered to MI patients at high risk who Managing patients with an INR higher
warfarin, have been started on antiplatelet agents. than 3.0 who have an episode of GI bleed-
ing is always a challenge. It is not uncom-
evaluate ■■ warfarin can precipitate bleeding mon to find pathologic lesions causing GI
patients for bleeding in patients who are on warfarin
Acute upper GI bleeding can be a severe com- with a supratherapeutic INR, and thus, en-
risk factors plication of long-term oral anticoagulation, doscopy is indicated. However, before en-
for upper GI not because the drugs cause ulcers, but rather doscopy, reversal of anticoagulation should
bleeding because they exacerbate ulcers that are already be considered.
present.60 Therefore, when starting warfarin
(Coumadin), patients should be evaluated to ■■ Bleeding in patients
determine if they have other risk factors for GI on antiplatelet drugs
bleeding, such as ulcers.
The number of people presenting with up- Aspirin
per GI bleeding while on warfarin therapy is Aspirin decreases production of prostaglandins
increasing because of the expanding indica- in the GI tract, thereby decreasing the protec-
tions for long-term anticoagulation therapy, tive and restorative properties of the gastric and
such as atrial fibrillation and deep venous duodenal mucosa and predisposing to ulcers and
thrombosis. bleeding.
The risk of GI bleeding in patients who The higher the aspirin dose, the higher
use oral anticoagulants is estimated to be 2.3 the risk. Aspirin doubles the risk of upper GI
to 4.9 times higher than in nonusers.61 bleeding at daily doses of 75 mg and quadruples

The goal international normalized ratio it at doses of 300 mg.66 Even doses as low as 10
(INR) for patients on warfarin therapy is usu- mg can decrease gastric mucosal prostaglandin
ally 2.0 to 3.0. Recent studies found that en- production.67 Thus, it appears that there is no
138  CLEV ELA N D C LI N I C JOURNAL OF MEDICINE   VOL UME 77  •  N UM BE R 2   F E BRUARY   2010
 Albeldawi and Colleagues

risk-free dose of aspirin, and enteric-coated or clopidogrel group (P = .0019).75 These studies
buffered formulations do not appear to reduce suggest that a once-daily proton pump inhibi-
the risk.68–70 tor combined with aspirin is a safer alternative
The most important risk factor for upper GI than clopidogrel alone.
bleeding in patients taking aspirin is a history
of peptic ulcer bleeding. Approximately 15% Clopidogrel plus a proton pump inhibitor
of aspirin users who have bleeding from ulcers Interestingly, recent studies have shown that
have recurrent bleeding within 1 year.71 omeprazole decreases the antiplatelet ef-
As aspirin-induced GI bleeding becomes fect of clopidogrel, possibly by inhibiting the
more common, health care providers often feel CYP2C19 enzyme.76 However, concomitant
caught between the GI risk and the cardiovas- use of pantoprazole (Protonix), lansoprazole
cular benefit. When considering whether to (Prevacid), and esomeprazole did not have
discontinue antiplatelet therapy, a cardiologist this effect, suggesting that although all proton
should be consulted along with a gastroenter- pump inhibitors are metabolized to a varying
ologist to weigh the risks of GI bleeding vs degree by CYP2C19, the interaction between
thrombosis. To date, there have been no clini- proton pump inhibitors and clopidogrel is not
cal trials published to suggest when antiplate- a class effect.77–79 Therefore, pantoprazole,
let therapy should be stopped to optimize GI lansoprazole, and esomeprazole may be the ap-
and cardiovascular outcomes. An alternative propriate proton pump inhibitors to use with
is to replace aspirin with another antiplatelet clopidogrel in patients who have a clear in-
drug that does not induce ulcers. dication for the medication, consistent with
current guideline recommendations.
Clopidogrel
Clopidogrel (Plavix) is recommended for hos- Helicobacter pylori infection
pitalized patients with acute coronary syn- in antiplatelet drug users
drome who cannot tolerate the GI side effects Before starting any long-term antiplatelet
of aspirin, according to the joint guidelines of therapy, patients with a history of ulcers
the American College of Cardiology and the should be tested and treated for H pylori Health care
American Heart Association, with the high- (TABLE 4).80 Confirmation of eradication is providers
est level of evidence.72 This recommendation required after H pylori treatment in patients
was largely based on the safety data from the with upper GI bleeding. Some suggest that often feel
CAPRIE (Clopidogrel Versus Aspirin in Pa- for patients with a history of bleeding ulcer caught between
tients at Risk of Ischemic Events) trial, in who need aspirin, eradication of H pylori sub-
which the incidence of major GI bleeding stantially reduces the risk of recurrent ulcer
the GI risk
was lower in the clopidogrel group (0.52%) bleeding.81 and the
than in the aspirin group (0.72%; P < .05).73 cardiovascular
■■ Treatment and prevention
Aspirin plus a proton pump inhibitor of NSAID-related GI injury benefit of
Patients who have had an episode of upper GI aspirin
bleeding and who need long-term aspirin ther- About 1 in 20 users of NSAIDs develop GI
apy should also receive a proton pump inhibi- complications and ulcers of varying degrees of
tor indefinitely to prevent ulcer recurrence. severity, as do one in seven NSAID users over
In a recent double-blind randomized con- the age of 65. In fact, NSAID use accounts
trolled trial in patients with a history of as- for 30% of hospitalizations for upper GI bleed-
pirin-induced bleeding, the combination of ing and deaths from this cause.82–85 In addition,
low-dose aspirin plus esomeprazole (Nexium) approximately 15% to 30% of NSAID users
twice a day was superior to clopidogrel by have clinically silent but endoscopically evi-
itself in terms of the rate of recurrent bleed- dent peptic ulcers.86
ing (0.7% vs 8.6%; P < .05).74 A similar trial NSAIDs contribute to ulcer development
showed nearly identical results: 0% upper GI by depleting prostaglandins. Thus, misopros-
bleeding in the group receiving aspirin plus tol (Cytotec), a synthetic prostaglandin, has
esomeprazole 20 mg daily, vs 13.6% in the been used to reduce this side effect.
CL EVEL AND CL I NI C J O URNAL O F M E DI CI NE    V O L UM E 77  •   NUM BE R 2   F E BRUARY   2010  139
 Acute Upper Gastrointestinal Bleeding

TABLE 4
Preferred therapies for Helicobacter pylori infection
Regimen Duration Eradication rate Comments

Triple therapy
Proton pump inhibitor twice a day 10–14 days 70%–85% Consider in non-penicillin-allergic
Clarithromycin (Biaxin) 500 mg twice a day patients who have not previously
Amoxicillin 1,000 mg twice a day received a macrolide
or
Proton pump inhibitor twice a day 10–14 days 75%–85% Consider in penicillin-allergic patients
Clarithromycin 500 mg twice a day who have not previously received a
Metronidazole (Flagyl) 500 mg twice a day macrolide or who cannot tolerate
bismuth quadruple therapy
Quadruple therapy
Proton pump inhibitor twice a day 10–14 days 75%–90% Consider in penicillin-allergic patients
Bismuth subsalicylate 525 mg twice a day
Metronidazole 250 mg four times a day
Tetracycline 500 mg four times a day

In a clinical trial, misoprostol reduced the Numerous clinical trials using endoscopic
incidence of NSAID-associated GI compli- end points showed that proton pump inhibi-
cations by 40%.87 Furthermore, it has been tors in standard doses significantly reduce the
shown to be better than placebo in preventing incidence of ulcers associated with the use of
recurrent gastric ulcers in patients with a his- NSAIDs.89 Proton pump inhibitor therapy has
tory of gastric ulcer who were receiving low- achieved a significant reduction in relative risk
dose aspirin.88 of upper GI bleeding in patients who received
However, misoprostol is rarely used because low-dose aspirin therapy, as confirmed by epi-
it can cause diarrhea and abdominal cramp- demiologic studies.90,91 The number of NSAID-
ing. Rather, the preferred drugs for preventing related ulcers found on endoscopy could be
and treating NSAID- and aspirin-related GI reduced by an estimated 90% simply by using
lesions are proton pump inhibitors. proton pump inhibitors.92 ■

■■ REFERENCES for emergency endoscopy. Scand J Gastroenterol 1985; 20:72–78.

8. Jensen DM, Machicado GA. Diagnosis and treatment of severe he-
1. Lewis JD, Bilker WB, Brensinger C, Farrar JT, Strom BL. Hospitaliza-
matochezia. The role of urgent colonoscopy after purge. Gastroen-
tion and mortality rates from peptic ulcer disease and GI bleeding in
terology 1988; 95:1569-1574.
the 1990s: relationship to sales of non-steroidal anti-inflammatory
9. Daniel WA Jr, Egan S. The quantity of blood required to produce a
drugs and acid suppression medications. Am J Gastroenterol 2002;
tarry stool. J Am Med Assoc 1939; 113:2232.
97:2540–2549.
10. Rockall TA, Logan RF, Devlin HB, Northfield TC. Risk assessment after
2. Viviane A, Alan BN. Estimates of costs of hospital stays for variceal
acute upper gastrointestinal hemorrhage. Gut 1996; 38:316–321.
and nonvariceal upper gastrointestinal bleeding in the United
11. Blatchford O, Murray WR, Blatchford M. A risk score to predict need
States. Value Health 2008; 11:1–3.
3. Yavorski RT, Wong RK, Maydonovitch C, Battin LS, Furnia A, for treatment for upper-gastrointestinal hemorrhage. Lancet 2000;
Amundson DE. Analysis of 3,294 cases of upper gastrointestinal 356:1318–1321.
bleeding in military medical facilities. Am J Gastroenterol 1995; 12. Gralnek IM, Barkun AN, Bardou M. Management of acute bleeding
90:568–573. from a peptic ulcer. N Engl J Med 2008; 359:928–937.
4. Kaplan RC, Heckbert SR, Koepsell TD, et al. Risk factors for hospital- 13. Silverstein FE, Gilbert DA, Tedesco FJ, Buenger NK, Persing J. The
ized gastrointestinal bleeding among older persons. Cardiovascular national ASGE survey on upper gastrointestinal bleeding II. Clinical
Health Study Investigators. J Am Geriatr Soc 2001; 49:126–133. prognostic factors. Gastrointest Endosc 1981; 27:80–93.
5. Longstreth GF. Epidemiology of hospitalization for acute upper 14. Corley DA, Stefan AM, Wolf M, Cook EF, Lee TH. Early indicators of
gastrointestinal hemorrhage: a population-based study. Am J Gas- prognosis in upper gastrointestinal hemorrhage. Am J Gastroenterol
troenterol 1995; 90:206–210. 1998; 93:336–340.
6. Laine L, Peterson WL. Bleeding peptic ulcer. N Engl J Med 1994; 15. Aljebreen AM, Fallone CA, Barkun AN. Nasogastric aspirate predicts
331:717–727. high-risk endoscopic lesions in patients with acute upper-GI bleed-
7. Wara P, Stodkilde H. Bleeding pattern before admission as guideline ing. Gastrointest Endosc 2004; 59:172–178.

140  CLEV ELA N D C LI N I C JOURNAL OF MEDICINE   VOL UME 77  •  N UM BE R 2   F E BRUARY   2010
 Albeldawi and Colleagues

16. Barkun AN, Cockeram AW, Plourde V, Fedorak RN. Review article: fectiveness of endoscopy in patients admitted to the intensive care
acid suppression in non-variceal acute upper gastrointestinal bleed- unit with upper GI hemorrhage. Gastrointest Endosc 2001; 53:6–13.
ing. Aliment Pharmacol Ther 1999; 13:1565–1584. 37. Lau JY, Chung SC, Leung JW, Lo KK, Yung MY, Li AK. The evolution
17. Levine JE, Leontiadis JI, Sharma VK, Howden CW. Meta-analysis: the of stigmata of hemorrhage in bleeding peptic ulcers: a sequential
efficacy of intravenous H2-receptor antagonists in bleeding peptic endoscopic study. Endoscopy 1998; 30:513–518.
ulcer. Aliment Pharmacol Ther 2002; 16:1137–1142. 38. Chung IK, Kim EJ, Lee MS, et al. Endoscopic factors predisposing
18. Walt RP, Cottrell J, Mann SG, Freemantle NP, Langman MJ. Continu- to rebleeding following endoscopic hemostasis in bleeding peptic
ous intravenous famotidine for hemorrhage from peptic ulcer. ulcers. Endoscopy 2001; 33:969–975.
Lancet 1992; 340:1058–1062. 39. Elta GH. Acute nonvariceal upper gastrointestinal hemorrhage. Curr
19. Labenz J, Peitz U, Leusing C, Tillenburg B, Blum AL, Börsch G. Ef- Treat Options Gastroenterol 2002; 5:147–152.
ficacy of primed infusion with high dose ranitidine and omeprazole 40. Marmo R, Rotondano G, Piscopo R, Bianco MA, D’Angella R,
to maintain high intragastric pH in patients with peptic ulcer bleed- Cipolletta L. Dual therapy versus monotherapy in the endoscopic
ing: a prospective randomized controlled study. Gut 1997; 40:36–41. treatment of high-risk bleeding ulcers: a meta-analysis of controlled
20. Merki HS, Wilder-Smith CH. Do continuous infusions of omeprazole trials. Am J Gastroenterol 2007; 102:279–289.
and ranitidine retain their effect with prolonged dosing? Gastroen- 41. Kovacs TO, Jensen DM. Recent advances in the endoscopic diagnosis
terology 1994; 106:60–64. and therapy of upper gastrointestinal, small intestinal, and colonic
21. Netzer P, Gaia C, Sandoz M, et al. Effect of repeated injection and bleeding. Med Clin North Am 2002; 86:1319–1356.
continuous infusion of omeprazole and ranitidine on intragastric pH 42. Kovacs TO, Jensen DM. Endoscopic treatment of ulcer bleeding.
over 72 hours. Am J Gastroenterol 1999; 94:351–357. Curr Treat Options Gastroenterol 2007; 10:143–148.
22. Lin HJ, Lo WC, Cheng YC, Perng CL. Role of intravenous omeprazole 43. Jensen DM, Kovacs TO, Jutabha R, et al. Randomized trial of medi-
in patients with high-risk peptic ulcer bleeding after successful en- cal or endoscopic therapy to prevent recurrent ulcer hemorrhage in
doscopic epinephrine injection: a prospective randomized compara- patients with adherent clots. Gastroenterology 2002; 123:407–413.
tive trial. Am J Gastroenterol 2006; 101:500–505. 44. Jensen DM, Machicado GA. Endoscopic hemostasis of ulcer hemor-
23. Lau JY, Sung JJ, Lee KK, et al. Effects of intravenous omeprazole on rhage with injection, thermal, and combination methods. Tech-
recurrent bleeding after endoscopic treatment of bleeding peptic niques Gastrointest Endosc 2005; 7:124–131.
ulcers. N Engl J Med 2000; 343:310–316. 45. Bleau BL, Gostout CJ, Sherman KE, et al. Recurrent bleeding from
24. Leontiadis GI, Sharma VK, Howden CW. Proton pump inhibitor peptic ulcer associated with adherent clot: a randomized study
treatment for acute peptic ulcer bleeding. Cochrane Database Syst comparing endoscopic treatment with medical therapy. Gastrointest
Rev 2006; CD002094. Endosc 2002; 56:1–6.
25. Andriulli A, Annese V, Caruso N, et al. Proton-pump inhibitors and 46. Lau JY, Sung JJ, Lam YH, et al. Endoscopic retreatment compared
outcome of endoscopic hemostasis in bleeding peptic ulcers: a series with surgery in patients with recurrent bleeding after initial endo-
of meta-analyses. Am J Gastroenterol 2005; 100:207–219. scopic control of bleeding ulcers. N Engl J Med 1999; 340:751–756.
26. Lau JY, Leung WK, Wu JC, et al. Omeprazole before endoscopy 47. Wolf AT, Wasan SK, Saltzman JR. Impact of anticoagulation on
in patients with gastrointestinal bleeding. N Engl J Med 2007; rebleeding following endoscopic therapy for nonvariceal upper gas-
356:1631–1640. trointestinal hemorrhage. Am J Gastroenterol 2007; 102:290–296.
27. Dorward S, Sreedharan A, Leontiadis GI, Howden CW, Moayyedi 48. Barkun A, Bardou M, Marshall JK; Nonvariceal Upper GI Bleeding
P, Forman D. Proton pump inhibitor treatment initiated prior to Consensus Conference Group. Consensus recommendations for
endoscopic diagnosis in upper gastrointestinal bleeding. Cochrane managing patients with nonvariceal upper gastrointestinal bleed-
Database Syst Rev 2006; CD005415. ing. Ann Intern Med 2003; 139:843–857.
28. Bardou M, Toubouti Y, Benhaberou-Brun D, Rahme E, Barkun 49. Marmo R, Rotondano G, Bianco MA, Piscopo R, Prisco A, Cipolletta
AN. Meta-analysis: proton-pump inhibition in high-risk patients L. Outcome of endoscopic treatment for peptic ulcer bleeding: is a
with acute peptic ulcer bleeding. Aliment Pharmacol Ther 2005; second look necessary? A meta-analysis. Gastrointest Endosc 2003;
21:677–686. 57:62–67.
29. Bjorkman DJ, Zaman A, Fennerty MB, Lieberman D, Disario JA, 50. Dell’Era A, deFrancis R, Iannuzzi F. Acute variceal bleeding: pharma-
Guest-Warnick G. Urgent vs elective endoscopy for acute non-va- cological treatment and primary/secondary prophylaxis. Best Pract
riceal upper-GI bleeding: an effectiveness study. Gastointest Endosc Res Clin Gastroenterol 2008; 22:279–294.
2004; 60:1–8. 51. Jalan R, Hayes PC. UK guidelines on the management of variceal
30. Lee SD, Kearney DJ. A randomized controlled trial of gastric lavage hemorrhage in cirrhotic patients. British Society of Gastroenterol-
prior to endoscopy for acute upper gastrointestinal bleeding. J Clin ogy. Gut 2000; 46(suppl 3–4):III1–III15.
Gastroenterol 2004; 38:861–865. 52. Bernard B, Lebrec D, Mathurin P, Opolon P, Poynard T. Beta-adren-
31. Tack J, Janssens J, Vantrappen G, et al. Effect of erythromycin on ergic antagonists in the prevention of gastrointestinal rebleeding in
gastric motility in controls and in diabetic gastroparesis. Gastroen- patients with cirrhosis: a meta-analysis. Hepatology 1997; 25:63–70.
terology 1992; 103:72–79. 53. De Franchis R. Evolving consensus in portal hypertension. Report of
32. Xynos E, Mantides A, Papageorgiou A, Fountos A, Pechlivanides the Baveno IV consensus workshop on methodology of diagnosis
G, Vassilakis JS. Erythromycin accelerates delayed gastric emptying and therapy in portal hypertension. J Hepatol 2005; 43:167–176.
of solids in patients after truncal vagotomy and pyloroplasty. Eur J 54. Levacher S, Letoumelin P, Pateron D, Blaise M, Lapandry C, Pourriat
Surg 1992; 158:407–411. JL. Early administration of terlipressin plus glyceryl trinitrate to
33. Coffin B, Pocard M, Panis Y, et al; Groupe des endoscopistes de control active upper gastrointestinal bleeding in cirrhotic patients.
garde á l’AP-HP. Erythromycin improves the quality of EGD in pa- Lancet 1995; 346:865–868.
tients with acute upper GI bleeding: a randomized controlled study. 55. Abraldes JG, Bosch J. Somatostatin and analogues in portal hyper-
Gastrointest Endosc 2002; 56:174–179. tension. Hepatology 2002; 35:1305–1312.
34. Frossard JL, Spahr L, Queneau PE, et al. Erythromycin intravenous 56. Banares R, Albillos A, Rincon D, et al. Endoscopic treatment versus
bolus infusion in acute upper gastrointestinal bleeding: a ran- endoscopic plus pharmacological treatment for acute variceal bleed-
domized, controlled, double-blind trial. Gastroenterology 2002; ing: a meta analysis. Hepatology 2002; 35:609–615.
123:17–23. 57. Cappell M. Gastrointenstinal bleeding associated with myocardial
35. Winstead NS, Wilcox CM. Erythromycin prior to endoscopy for acute infarction. Gastroenterol Clin North Am 2000; 29:423–444.
upper gastrointestinal hemorrhage: a cost-effectiveness analysis. Ali- 58. Lin S, Konstance R, Jollis J, Fisher DA. The utility of upper endoscopy
ment Pharmacol Ther 2007; 26:1371–1377. in patients with concomitant upper gastrointestinal bleeding and
36. Chak A, Cooper GS, Lloyd LE, Kolz CS, Barnhart BA, Wong RC. Ef- acute myocardial infarction. Dig Dis Sci 2006; 51:2377–2383.

CL EVEL AND CL I NI C J O URNAL O F M E DI CI NE    V O L UM E 77  •   NUM BE R 2   F E BRUARY   2010  141

 Acute Upper Gastrointestinal Bleeding

59. Silvis SE, Nebel O, Rogers G, Sugawa C, Mandelstam P. Endoscopic Effects of pantoprazole and esomeprazole on platelet inhibition by
complications. Results of the 1974 American Society for Gastrointesti- clopidogrel. Am Heart J 2009; 157:148.e1-e5.
nal Endoscopy Survey. JAMA 1976; 235:928–930. 78. Small DS, Farid NA, Payne CD, et al. Effects of proton pump inhibitor
60. Lanas A, Garcia-Rodriguez LA, Arroyo MT, et al; Investigators of the lansoprazole on the pharmacokinetics and pharmacodynamics of
Asociación Española de Gastroenterología (AEG). Effect of antisecre- prasugel and clopidogrel. J Clin Pharmacol 2008; 48:475–484.
tory drugs and nitrates on the risk of ulcer bleeding associated with 79. Ishizaki T, Horai Y. Review article: cytochrome P450 and the metabo-
nonsteroidal anti-inflammatory drugs, antiplatelet agents, and lism of proton pump inhibitors—emphasis on rabeprazole. Aliment
anticoagulants. Am J Gastroenterol 2007; 102:507–515. Pharacol Ther 1999; 13(suppl 3):27–36.
61. Shorr RI, Ray WA, Daugherty JR, Griffin MR. Concurrent use of 80. Lanas A, Fuentes J, Benito R, Serrano P, Bajador E, Sainz R. Helico-
nonsteroidal anti-inflammatory drugs and oral anticoagulants places bacter pylori increases the risk of upper gastrointestinal bleeding
elderly persons at high risk for hemorrhagic peptic ulcer disease. in patients taking low-dose aspirin. Aliment Pharmacol Ther 2002;
Arch Intern Med 1993; 153:1665–1670. 16:779–786.
62. Tabibian N. Acute gastrointestinal bleeding in anticoagulated pa- 81. Chan FK. NSAID-Induced peptic ulcers and Helicobacter pylori
tients: a prospective evaluation. Am J Gastroenterol 1989; 84:10–12. infection: implications for patient management. Drug Saf 2005;
63. Choudari CP, Rajgopal C, Palmer KR. Acute gastrointestinal hemor- 28:287–300.
rhage in anticoagulated patients: diagnoses and response to endo- 82. Bombardier V, Laine L, Reicin A, et al; VIGOR Study Group. Compari-
scopic treatment. Gut 1994; 35:464–466. son of upper gastrointestinal toxicity of rofecoxib and naproxen in
64. Petty GW, Brown RD Jr, Whisnant JP, Sicks JD, O’Fallon WM, Wiebers patients with rheumatoid arthritis: VIGOR Study Group. N Eng J Med
DO. Frequency of major complications of aspirin, warfarin, and intra- 2000; 343:1520–1528.
venous heparin for secondary stroke prevention: a population-based 83. Griffin MR, Ray WA, Schaffner W. Nonsteroidal anti-inflammatory
study. Ann Intern Med 1999; 130:14–22. drug use and death from peptic ulcer in elderly persons. Ann Intern
65. Landefeld CS, Rosenblatt MW, Goldman L. Bleeding in outpatients Med 1988; 109:359–363.
treated with warfarin: relation to the prothrombin time and impor- 84. Griffin MR, Piper JM, Daugherty JR, Snowden M, Ray WA. Nonsteroi-
tant remediable lesions. Am J Med 1989; 87:153–159. dal anti-inflammatory drug use and increased risk for peptic ulcer
66. Weil J, Colin-Jones D, Langman M, et al. Prophylactic aspirin and risk disease in elderly persons. Ann Intern Med 1991; 114:257–263.
of peptic ulcer bleeding. BMJ 1995; 310:827–830. 85. Smalley WE, Ray WA, Daugherty JR, Griffin MR. Nonsteroidal anti-
67. Cryer B, Feldman M. Effects of very low dose daily, long-term aspirin inflammatory drugs and the incidence of hospitalizations for peptic
therapy on gastric, duodenal, and rectal prostaglandin levels and on ulcer disease in elderly persons. Am J Epidemiol 1995; 141:539–545.
mucosal injury in healthy humans. Gastroenterology 1999; 117:17–25. 86. Laine L. Approaches to nonsteroidal anti-inflammatory drug use in
68. De Abajo FJ, Garcia Rodriguez LA. Risk of upper gastrointestinal the high-risk patient. Gastroenterology 2001; 120:594–606.
bleeding and perforation associated with low-dose aspirin as plain 87. Silverstein FE, Graham DY, Senior JR, et al. Misoprostol reduces serious
and enteric-coated formulations. BMC Clin Pharmacol 2001; 1:1. gastrointestinal complications in patients with rheumatoid arthritis re-
69. Kelly JP, Kaufman DW, Jurgelon JM, Sheehan J, Koff RS, Shapiro S. ceiving nonsteroidal anti-inflammatory drugs. A randomized, double
Risk of aspirin-associated major upper gastrointestinal bleeding with blind, placebo controlled trial. Ann Intern Med 1995; 123:241–249.
enteric coated or buffered product. Lancet 1996; 348:1413–1416. 88. Goldstein JL, Huang B, Amer F, Christopoulos NG. Ulcer recurrence
70. Garcia Rodriguez LA, Hernandez-Diaz S, De Abajo FJ. Association in high-risk patients receiving nonsteroidal anti-inflammatory drugs
between aspirin and upper gastrointestinal complications: system- plus low dose aspirin: results of a post hoc subanalysis. Clin Ther
atic review of epidemiological studies. Br J Clin Pharmacol 2001; 2004; 26:1637–1643.
52:563–571. 89. Berger JS, Stebbins A, Granger CB, et al. Initial aspirin dose and
71. Wilcox CM, Ladabaum U. A patient with high risk of gastrointesti- outcome among ST-elevation myocardial infarction patients treated
nal bleeding requiring nonsteroidal anti-inflammatory drugs. Clin with fibrinolytic therapy. Circulation 2008; 117:192–199.
Gastroenterol Hepatol 2006; 4:1090–1093. 90. Lanas A, Garcia-Rodriguez LA, Arroyo MT, et al; Investigators of the
72. Anderson JL, Adams CD, Antman EM, et al. ACC/AHA 2007 guidelines Asociación Española de Gastroenterología (AEG). Effect of antisecre-
for the management of patients with unstable angina/non-ST-eleva- tory drugs and nitrates on the risk of ulcer bleeding associated with
tion myocardial infarction. J Am Coll Cardiol 2007; 50:e1–e157. nonsteroidal anti-inflammatory drugs, antiplatelet agents, and
73. CAPRIE Steering Committee. A randomised, blinded, trial of clopi- anticoagulants. Am J Gastroenterol 2007; 102:507–515.
dogrel versus aspirin in patients at risk of ischaemic events (CAPRIE). 91. Chin MW, Yong G, Bulsara MK, Rankin J, Forbes GM. Predictive and
Lancet 1996; 348:1329–1339. protective factors associated with upper gastrointestinal bleeding
74. Chan FK, Ching JY, Hung LC, et al. Clopidogrel versus aspirin and es- after percutaneous coronary intervention: a case-control study. Am J
omeprazole to prevent recurrent ulcer bleeding. N Engl J Med 2005; Gastroenterol 2007; 102:2411–2416.
352:238–244. 92. Hunt RH, Bazzoli F. Should NSAID/low dose aspirin takers be tested
75. Lai KC, Chu KM, Hui WM, et al. Esomeprazole with aspirin versus routinely for H. Pylori infection and treated if positive? Implications
clopidogrel for prevention of recurrent gastrointestinal ulcer compli- for primary risk of ulcer and ulcer relapse after initial healing. Ali-
cations. Clin Gastroenterol Hepatol 2006; 4:860–865. ment Pharmacol Ther 2004; 19(suppl 1):9–16.
76. Ho MP, Maddox TM, Wang L, et al. Risk of adverse outcomes associat-
ed with concomitant use of clopidogrel and proton pump inhibitors ADDRESS: John J. Vargo, MD, MPH, Department of Gastroenterology and
following acute coronary syndrome. JAMA 2009; 301:937–944. Hepatology, A30, Cleveland Clinic, 9500 Euclid Avenue, Cleveland, OH
77. Siller-Matula JM, Spiel AO, Lang IM, Kreiner G, Christ G, Jilma B. 44195; e-mail: vargoj@ccf.org.

Sign up for
Cleveland Clinic Journal of Medicine eTOCs
(electronic Table of Contents)
a free service providing the Table of Contents or a notification of availability
when new issues of Cleveland Clinic Journal of Medicine are published online.
To sign up, go to www.ccjm.org and click on “E-MAIL ALERTS (FREE)”

142  CLEV ELA N D C LI N I C JOURNAL OF MEDICINE   VOL UME 77  •  N UM BE R 2   F E BRUARY   2010