1. health and fitness
2. disease

rEVIEWS
Diagnosis and management of lower
gastrointestinal bleeding
Jürgen Barnert and Helmut Messmann
abstract | Lower gastrointestinal bleeding (LGIB) can present as an acute and life‑threatening event or
as chronic bleeding, which might manifest as iron‑deficiency anemia, fecal occult blood or intermittent
scant hematochezia. Bleeding from the small bowel has been shown to be a distinct entity, and LGIB is
defined as bleeding from a colonic source. Acute bleeding from the colon is usually less dramatic than
upper gastrointestinal hemorrhage and is self‑limiting in most cases. Several factors might contribute to
increased mortality, a severe course of bleeding and recurrent bleeding, including advanced age, comorbidity,
intestinal ischemia, bleeding as a result of a separate process, and hemodynamic instability. Diverticula,
angiodysplasias, neoplasms, colitis, ischemia, anorectal disorders and postpolypectomy bleeding are the
most common causes of LGIB. Volume resuscitation should take place concurrently upon initial patient
assessment. Colonoscopy is the diagnostic and therapeutic procedure of choice, for acute and chronic
bleeding. Angiography is used if colonoscopy fails or cannot be performed. The use of radioisotope scans is
reserved for cases of unexplained intermittent bleeding, when other methods have failed to detect the source.
Embolization or modern endoscopy techniques, such as injection therapy, thermocoagulation and mechanical
devices, effectively promote hemostasis. Surgery is the final approach for severe bleeding.
Barnert, J. & Messmann, H. Nat. Rev. Gastroenterol. Hepatol. 6, 637–646 (2009); doi:10.1038/nrgastro.2009.167
Continuing Medical Education online
This activity has been planned and implemented in accordance
with the Essential Areas and policies of the Accreditation Council
for Continuing Medical Education through the joint sponsorship of
MedscapeCME and Nature Publishing Group.
MedscapeCME is accredited by the Accreditation Council for
Continuing Medical Education (ACCME) to provide continuing
medical education for physicians.
MedscapeCME designates this educational activity for a maximum
of 1.0 aMa pra Category 1 CreditstM. Physicians should only claim
credit commensurate with the extent of their participation in the
activity. All other clinicians completing this activity will be issued
a certificate of participation. To participate in this journal CME
activity: (1) review the learning objectives and author disclosures;
(2) study the education content; (3) take the post‑test and/or
complete the evaluation at http://cme.medscape.com/public/
naturereviews; and (4) view/print certificate.
learning objectives
Upon completion of this activity, participants should be able to:
1 Describe clinical features of lower gastrointestinal bleeding
(LGIB).
2 Distinguish among different causes of acute LGIB and
hematochezia.
3 Describe the diagnostic approach for angiodysplasia in
patients with LGIB.
4 Identify the most common causes of anemia associated
with chronic LGIB.
5 Describe effective treatment strategies for different types of
LGIB.
Competing interests
The authors, the Journal Editor N. Wood and the CME questions
author D. Lie declare no competing interests.
Introduction
Hemorrhage from the lower gastrointestinal tract accounts
for about 20% of all cases of acute gastrointestinal bleeding.1 lower gastrointestinal bleeding (lGiB) has traditionally been defined as an abnormal loss of blood beyond
the ligament of treitz. However, results from capsule
and double-balloon endoscopy have revolutionized the
management algorithm of small bowel bleeding, and it
has been shown that bleeding from the small bowel represents a distinct entity.2 it therefore seems reasonable to
divide gastrointestinal bleeding into three categories:
upper, middle, and lower bleeding. in this article, we
define lGiB as the acute or chronic loss of blood from a
source in the colon or anorectum. acute lGiB is defined
as being of recent duration (arbitrarily designated as less
than 3 days) and might result in instability of vital signs,
anemia and/or the need for blood transfusion. Chronic
lGiB is the passage of blood from the rectum over a
period of several days or longer and usually implies that
blood loss is intermittent or slow. a patient with chronic
bleeding can show occult fecal blood, iron-deficiency
anemia, occasional episodes of melena, hematochezia or
maroon stools, or small quantities of visible blood per
rectum.3 alternatively, however, lGiB can be subdivided
into two categories: clinically overt gastrointestinal bleeding (melena, hematochezia) or occult bleeding, identified
by an unexplained iron deficiency and/or positive fecal
occult blood testing result.
the incidence of lGiB in the us ranges from 20.5
to 27 cases per 100,000 adults.4 in the netherlands,
an overall incidence of 8.9/100,000 per year has been
nature reviews | gastroenterology & hepatology
© 2009 Macmillan Publishers Limited. All rights reserved
Department of Internal
Medicine III, Klinikum
Augsburg, Augsburg,
Germany (J. Barnert,
h. Messmann).
Correspondence:
H. Messmann, Klinikum
Augsburg,
Stenglinstrasse 2,
D‑86156 Augsburg,
Germany
helmut.messmann@
klinikum‑augsburg.de
volume 6 | novemBer 2009 | 637
rEVIEWS
Key points
■ The severity of acute lower gastrointestinal bleeding (LGIB) is variable, but
overall mortality is low
■ In most cases, bleeding will stop spontaneously
■ Mortality is higher in older adults, and in those with intestinal ischemia and
other comorbidities
■ Colonoscopy is the diagnostic and therapeutic mainstay in the management of
acute and chronic LGIB
■ Visceral angiography is the preferred diagnostic and therapeutic method in
patients with hemodynamic instability
■ The existing scoring systems are time consuming and offer little help in clinical
decision‑making
reported.5 Differences in incidence might be explained
by the different populations, use of medicine, endoscopic and pharmaceutical management, and different
selection criteria. the age range of patients with lGiB
is 63–77 years;4 the incidence increases with age, with a
>200-fold increase between 20 and 80 years. lGiB occurs
more often in men than in women.6
Compared with acute upper gastrointestinal bleeding, patients with acute lGiB are significantly less likely
to experience shock (19% versus 35%, respectively),
require fewer blood transfusions (36% versus 64%) and
have a significantly higher hemoglobin level (84% versus
61%).7,8 Colonic bleeding necessitates fewer blood transfusions compared with bleeding from the small intestine.8
acute bleeding in the lower gastrointestinal tract stops
spontaneously in the majority (80–85%) of patients. the
overall mortality rate ranges from 2% to 4%.4
in this review, we aim to provide a critical overview
of the existing published data concerning the evaluation
and management of lGiB.
Initial evaluation and resuscitation
on presentation with lGiB, a number of factors relating to the patient’s history should be considered: intake
of aspirin or nsaiDs, occurrence of vascular disease,
past bleeding episodes, radiation therapy for prostatic
or pelvic cancer, occurrence within the previous 2 weeks
of colonoscopy or polypectomy, Hiv infection, liver cirrhosis, iBD, coagulopathy (including anticoagulation
therapy), and indication of colorectal cancer (family
history, weight loss, changes in bowel habits). the duration of bleeding, frequency of bleeding episodes and
stool color are of further importance. Hematochezia is
indicated by bright red or maroon blood per rectum and
must be differentiated from melena, the passage of tarry
stool, the presence of which is suggestive of an upper
gastrointestinal bleeding source (although bleeding from
the cecum and right-sided colon occasionally present in
this manner). up to 11% of patients with hematochezia
have massive upper gastrointestinal bleeding.9 Passage
of bright red blood resulting from upper gastrointestinal
bleeding is associated with hemodynamic instability
(shock or orthostatic hypotension). the presence of
blood in nasogastric tube aspirates is highly predictive
of bleeding proximal to the ligament of treitz, but
this source cannot be excluded if blood is absent from
the aspirate.
Careful digital rectal examination should be performed to exclude anorectal pathology and to confirm
the patient’s description of stool color. Physical examination also helps to assess the severity of bleeding and to
stratify patients according to the severity of hemorrhage.
Blood loss of <200 ml has no effect on heart rate or blood
pressure; however, loss of >800 ml causes a drop in
blood pressure of 10 mmHg and an increase of 10 heartbeats/min. extensive blood loss (>1500 ml) might induce
shock. other signs of hemodynamic compromise include
postural changes, fatigue, pallor, palpitations, chest pain,
dyspnea and tachypnea. the initial laboratory evaluation
should contain a complete blood count, a coagulation
profile, serum chemistry and a sample for blood type
and crossmatch.
Patients with clinical evidence of ongoing or aggressive
bleeding, those with a transfusion requirement of greater
than two units of packed red blood cells, and those with a
significant morbidity should be monitored in an intensive
care unit (iCu) setting. Conversely, a young, nonanemic
patient with scant bleeding, who is otherwise healthy, can
be managed as an outpatient. Coagulopathy (defined as
an international normalized ratio of prothrombin time
>1.5) or thrombocytopenia (<50,000 platelets/μl) should
be treated using fresh frozen plasma or platelets, respectively. in patients receiving warfarin, anticoagulation
should be reversed with vitamin K, although the onset
of action is delayed compared with the use of fresh frozen
plasma or prothrombin complex. the ideal plasma
substitute for volume replacement remains a matter of
debate. in general, crystalloids (for example, lactated
ringer’s solution) are recommended for volume substitution in the us,10 whereas colloids, such as low and
medium molecular weight hydroxyethylstarch preparations with low molar substitution, are considered to
be more appropriate for treating intravascular volume
deficits in europe.11 if there is ongoing hemorrhage or
severe anemia, red blood cells should be used. the ideal
hemoglobin concentration/hematocrit depends on the
patient’s age, rate of bleeding, and any comorbidities. a
young and otherwise healthy person will tolerate a hemoglobin concentration of less than 7–8 g/dl (hematocrit
<20–25%), whereas older patients develop symptoms
at this level. maintaining the hemoglobin concentration around 10 g/dl (hemato crit 30%) in high-risk
patients (for instance, an elderly patient with coronary
heart disease) would be reasonable. However, it must be
emphasized that all these recommendations have been
given on an empirical basis.
Etiologies
the sources of acute lGiB and their frequency as
reported in the literature are listed in table 1. Colonic
diverticula (Figure 1) seem to be the most frequent
source of hematochezia, followed by angiodysplasias,
638 | NOVEMBER 2009 | VOluME 6
www.nature.com/nrgastro
© 2009 Macmillan Publishers Limited. All rights reserved
rEVIEWS
iBD and postpolypectomy bleeding. the causes of
chronic lGiB and their frequency are more difficult
to determine, although they cause 18–30% of irondeficiency anemia, the most common manifestation
of chronic gastrointestinal bleeding (table 2). Chronic
intermittent passage of small amounts of visible red
blood most often originates from anal lesions, such
as hemorrhoids or anal fissures. otherwise, bleeding lesions are usually located in the rectum or distal
colon. in middle-aged and older indivi duals, further
examination might be needed, even if an anal or rectal
lesion seems to be responsible for bleeding, to exclude a
coexisting colon neoplasm.
Diverticular disease
Bleeding from colonic diverticula (Figure 1), which
presents as acute, painless hematochezia, is arterial
and can occur either at the dome or the neck of the
diverticulum. the prevalence of colonic diverticula
increases with age, affecting about two thirds of people
over 80 years. although these structures are reportedly
the most frequent source of acute lGiB (table 1), they
are often cited as the source of bleeding in the colon
for lack of evidence of another source. Colon diverticula were identified as the source in 22% of cases of
acute lGiB, based on active bleeding or the presence
of stigmata, such as a visible vessel or an adherent clot.12
the majority (>75%) of diverticula are found in the left
colon. using colonoscopy, 60% of cases of diverticular
bleeding can be observed in the left colon.6 However,
the bleeding diverticulum is localized more often in the
right colon when using angiography for diagnosis.3,4
Bleeding ceases spontaneously in about 80% of cases.
the cumulative risk of rebleeding for these patients is
25% after 4 years.6 Data on the recurrence rate of bleeding after endoscopic therapy are inconsistent: Jensen and
colleagues12 observed no recurrence, whereas Bloomfeld
and colleagues13 reported an early recurrence in 38%
of patients.
angiodysplasia
angiodysplasias (also known as vascular ectasias or
angioectasias) (Figure 2) are cited in up to 30% of cases
as the source of acute lGiB (table 1); a rate of 3–12%
is probably more realistic. 14 angiodysplasias are also
said to be an important cause of chronic lGiB (table 2).
the majority of colonic angiodysplasias are located in the
right hemicolon, often as multiple lesions; the number of
lesions increases with age. angiodysplasias are seldom
found during routine colonoscopy (0.83%15 and 1.4%16).
most angiodysplasias do not bleed, 15,16 and so most
patients are asymptomatic. overt bleeding often occurs
as a result of coagulopathy or platelet dysfunction, and
nsaiDs, aspirin or anticoagulants might have a role in
triggering bleeding.17
angiodysplasias appear endoscopically as red, circumscribed mucosal lesions measuring from one millimeter
to a few centimeters (Figure 2). as most do not bleed,18,19
Table 1 | Sources of hematochezia reported in the literature (based on14)
source/finding
Frequency (%)
Diverticulum
17–40
Angiodysplasia
9–21
Colitis (ischemic, infectious, chronic IBD, radiation injury)
2–30
Neoplasia, postpolypectomy bleeding
11–14
Anorectal disease (including rectal varices)
4–10
Upper gastrointestinal bleeding
0–11
Small bowel bleeding
2–9
Figure 1 | Bleeding from a small diverticulum, recognizable
by a streak of red blood.
therapy is not always advocated. Furthermore, angiodysplasias seen during emergency colonoscopy should not
automatically be identified as the bleeding source unless
they are acutely bleeding or show stigmata (for example,
a visible vessel, an adherent clot or submucosal bleeding). it is important to avoid the use of opiates and coldwater lavage of the mucosa during colonoscopy, because
these substances reduce mucosal blood, decreasing
diagnostic yield.20–22
Chronic radiation proctopathy is a common complication of radiation therapy in the pelvic region that might
cause rectal bleeding. the term ‘proctitis’ is a misnomer,
as there is no inflammation on biopsy; instead, radiationinduced endarteritis obliterans induces neovascularization. multiple telangiectasias among a pale mucosa can
be seen by endoscopy, which can lead to considerable and
recurrent blood loss. Following radiation therapy for prostatic carcinoma, 4–13% of patients report rectal bleeding.23,24 in hereditary hemorrhagic telangiectasia (HHt;
also known as osler–weber–rendu syndrome), bleeding occurs in one-third of patients.25 the most common
localization of HHt is the stomach and the small bowel;
the colon is less often affected.26 Patients over 60 years are
particularly at risk, but—unlike angiodysplasias—HHt
affects younger people as well.
nature reviews | gastroenterology & hepatology
© 2009 Macmillan Publishers Limited. All rights reserved
volume 6 | novemBer 2009 | 639
rEVIEWS
Table 2 | Colonic causes of iron‑deficiency anemia based on literature reports90–94
Finding
Frequency (%)
Colon carcinoma
4.5–11
Angiodysplasia
0.9–8.5
Colon polyps
2.8–7.2
Colitis
1.4–2
Colonic causes in total
18–30
Figure 2 | Angiodysplasia in the ascending colon, slightly
elevated above the mucosa.
Ischemic colitis
the occurrence of hematochezia caused by colonic
ischemia seems to increase in frequency in our clinical
practice, most likely owing to an increase in frequency
of aged patients with cardiovascular diseases. However,
there are no reliable data available on the incidence of
nonocclusive ischemia of the colon. the cause of ischemic colitis is a sudden, often temporary, reduction in
mesenteric blood flow resulting from episodes of low
blood pressure or vasospasm. typically, the watershed
areas of the colon—the splenic flexure and rectosigmoid
junction—are affected. Colonic ischemia mostly affects
older persons with advanced athero sclerosis and/or
cardiac disease, although vasculitis might be a rare
cause. a specific precipitating event can almost never
be identified. Patients experience a sudden onset of
mild abdominal pain, usually followed by hematochezia or bloody diarrhea. Bleeding is self-limiting
in almost all cases; but colonic ischemia is associated
with an increased risk of mortality. 27 By endoscopy,
submucosal hemorrhage and mucosal nodularity are
present during the early stages or, alternatively, a livid
or pale, nonedematous mucosa can be seen. in the later
stages, when blood and edema have been reabsorbed, the
endoscopic appearance is less characteristic. the mucosa
might be red, the vessel pattern might be obscured, and
ulcerations can appear.
Mucosal inflammation
massive hemorrhaging leads to hospitalization in 0.1%
of patients with ulcerative colitis and 1.2% of patients
with Crohn’s disease.28 in half of the patients, bleeding
stops spontaneously, but rebleeding occurs in 35% of
these patients.29 in infectious colitis (including pseudomembranous colitis), life-threatening hemorrhage is
rare. in patients with Hiv, the most common causes of
lGiB differ from those in other patients, and include
cytomegalovirus colitis (25%), lymphoma (12%), and
idiopathic (unidentifiable) colitis (12%).30 Colonic histoplasmosis, Kaposi’s sarcoma in the colon, and bacterial
colitis are further sources of bleeding. mortality related
to bleeding in patients with Hiv is around 14%.30
the use of nsaiDs is associated with an increased risk
of gastrointestinal bleeding. lanas et al.31 found a strikingly high prevalence (86%) of nsaiD use (especially
aspirin) in patients with lGiB. nsaiDs are also commonly used in patients with iron-deficiency anemia.
Pre-existing lesions, such as diverticula, angiodysplasias,
polyps, cancer and colitis, are prone to bleeding with
nsaiD use. Furthermore, nsaiDs can cause mucosal
damage, colonic inflammation, erosions and ulcerations.
nsaiD use might also exacerbate underlying colitis. the
mucosal aspect of nsaiD-induced colopathy can resemble infectious colitis or chronic iBD, but the lesions often
present themselves as flat and usually irregularly shaped
erosions and ulcerations in a normal appearing mucosa.
neoplasia
Carcinomas account for 2–9% of cases of hematochezia,32
and are by far the most frequent cause of iron-deficiency
anemia and source of chronic lGiB (table 2). erosions
and ulceration on the surface of the tumor might bleed, and
bleeding is often exacerbated by the use of nsaiDs.
Carcinoma in the left colon, especially in the sigmoid,
often leads to rectal bleeding early in the course of the
disease. in the right colon, by contrast, carcinoma often
presents as iron-deficiency anemia.
Colonic polyps are cited in 5–11% of patients as the
source of acute lGiB,14 and in 3–7% as the cause of irondeficiency anemia (table 2). usually, only polyps larger
than 1 cm bleed.
postpolypectomy bleeding
Postpolypectomy bleeding is the most frequent complication of colonoscopy and is seen in 0.2–1.8% of cases
of colonoscopic polypectomy.33–41 this type of bleeding accounts for approximately 2–8% of cases of acute
lGiB.4 Delayed bleeding might occur up to 14 days after
polypectomy.35,36 massive bleeding arising concomitant
with the polypectomy procedure is typically arterial and
results from inadequate hemostasis of the blood vessel in
the polyp stalk. Kim and colleagues33 used multivariate
analysis to assess polypectomy-induced immediate bleeding and identified several risk factors: size of the polyp,
cutting mode of the electrosurgical current, incidence of
inadvertent cutting of a polyp before current application,
640 | NOVEMBER 2009 | VOluME 6
www.nature.com/nrgastro
© 2009 Macmillan Publishers Limited. All rights reserved
rEVIEWS
gross morphology of the polyp (pedunculated or laterally spreading), comorbidity (cardiovascular and renal
disease), bowel preparation, age (>65 years) and experience of the endoscopist. they also found out that even a
mild degree of immediate bleeding during polypectomy
was significantly associated with the development of
delayed bleeding. three groups35,36,41 looked for factors
increasing the risk of delayed postpolypectomy bleeding. all identified polyp size as a risk factor. the munich
study group41 noted that polyp location in the right colon
was a further risk factor. watabe et al.36 identified hypertension as an additional risk factor, but this condition was
not confirmed by sawhney and colleagues.35 the latter
group reported that the resumption of anticoagulation
therapy within 1 week of polypectomy increased the
risk of severe, delayed, postpolypectomy bleeding. By
contrast, the continued use of aspirin or nsaiDs was
not associated with an increase in postpolypectomy
bleeding.42 only a few patients in this study used newer
antiplatelet agents like clopidogrel and ticlopidine or
CoX ii inhibitors. none of these patients developed postpolypectomy bleeding, but the small number of patients
does not enable a clear conclusion to be drawn.
anorectal diseases
Hemorrhoids are the source of acute lGiB in 2–9% of
patients (for an overview, see reference14). acute lGiB
from anal fissures is rare. Fissures are relatively easily
diagnosed by inspecting the anus. the patient typically
has severe pain upon spreading the anus, but the lesion
can be carefully and painlessly inspected after injecting a local anesthetic. Bleeding from fissures usually
ceases spontaneously.
local ischemia seems to be important in the pathogenesis of solitary rectal ulcer. internal rectal prolapse
or lack of inhibition of the puborectalis muscle during
straining have been implicated as causal factors. Heavy
bleeding is rare. Bleeding from rectal varices can be seen
in patients with portal hypertension. rectal varices have
a gray–blue color and might be confounded by mucosal
folds. Bleeding is usually severe. the incidence of
bleeding from rectal varices is reported to be 18%.43
Chronic intermittent passage of small amounts of
visible red blood indicates in most patients a bleeding source that is in the anus, rectum or sigmoid.
Hemorrhoids and chronic anal fissures are the most
common bleeding source in young (<40 years) patients
with chronic lGiB. in patients over 50 years of age colorectal neoplasia must be excluded by colonoscopy or Ct
colonography. this approach is valid even if a benign
(potential) anorectal source of bleeding is present, such as
hemorrhoids or an anal fissure. in patients with radiation
history, chronic radiation proctopathy must be considered
as a potential cause of chronic scant bleeding.
Dieulafoy’s lesion
Dieulafoy’s lesion is a rare source of acute colonic bleeding. it is caused by an exposed artery that arises within
Hematochezia
No
Hemodynamically stable?
Stabilization of the patient
Not possible
Colonoscopy
Possible
EGD
Yes
Yes
Bleeding site
localized
No
Endoscopic
therapy
Bleeding site
localized
Negative
Endoscopic
therapy
EGD without
localizing
bleeding site
Unsuccessful
Angiography +/–
embolization
Surgery
Severe
hemorrhage—
no potential
to localize
bleeding site
Capsule
endoscopy
and/or
double-balloon
enteroscopy
Figure 3 | Algorithm showing recommended evaluation of acute lower
gastrointestinal bleeding from initial presentation of the patient with
hematochezia. Stabilization of the patient should be carried out before
colonoscopy is performed. If this is not possible, angiography should be performed
first. In other cases, colonoscopy is the mainstay in the diagnostic algorithm.
Abbreviation: EGD, esophagogastroduodenoscopy.
a minute mucosal defect and, unless bleeding is visible,
might be difficult to detect.
Detection of LGIB
our recommended approach to patients with acute lGiB
is summarized in Figure 3. For chronic lGiB, colonoscopy and anoscopy are the cornerstones for investigation. if repeat colonoscopies and upper endoscopies
are negative, the small bowel must be investigated as a
potential source of the bleeding.
endoscopy
Flexible endoscopy is now considered the mainstay
for evaluation of acute and chronic colonic bleeding.
the incidence of serious complications is low (about
1 in 1,000 procedures). Patients should be continuously monitored during urgent endoscopy using eCG
and noninvasive measurement of oxygen saturation. in
case of hemodynamic instability, patients must undergo
volume resuscitation before endoscopy.
in patients with hematochezia and concurrent hemodynamic instability, esophagogastroduodenoscopy
(eGD) should be performed first to exclude an upper
gastrointestinal bleeding source. otherwise, colonoscopy is recommended as the first step in the evaluation
of acute lGiB.44 the timing of colonoscopy after initial
presentation varies among studies from 12 h to 48 h.44
Colonoscopy can determine the source and type of
bleeding, and can help to identify patients with ongoing
hemorrhage or those who are at high risk of rebleeding.
Furthermore, endoscopic hemostasis can be performed,
if necessary.
nature reviews | gastroenterology & hepatology
© 2009 Macmillan Publishers Limited. All rights reserved
volume 6 | novemBer 2009 | 641
rEVIEWS
the diagnostic yield for urgent colonoscopy in acute
lGiB is 89–97%.45,46 Current recommendations advise
thorough cleansing of the colon in acute lGiB,3,47 as this
procedure facilitates endoscopic visualization, improves
diagnostic yield, and improves (presumably) the safety
of the procedure by decreasing the risk of perforation. For optimal purge of the colon, the patient must
consume 3–6 l of a polyethylene glycol-based solution.
Patients generally tolerate consumption of 1–2 l per
hour. it might be helpful to administer a prokinetic antiemetic such as metoclopramide (10 mg intravenously)
or to administer the solution through a nasogastric
tube. Colonoscopy can be started with the appearance
of liquid discharge because diluted blood and clots can
be aspirated or washed away. if the discharge becomes
free of blood during the preparation, the endoscopic
evaluation can be done on an elective basis the following day. the cecum should be reached if at all possible,
because a substantial proportion of bleeding sites are
located in the right hemicolon. Flowing blood from
above in the terminal ileum is a clear sign of a proximal
bleeding site. in patients with severe and ongoing bleeding, urgent colonoscopy must sometimes be carried out
without purge. in patients with hemodynamic instability, an urgent angiography is recommended. For chronic
lGiB, colonoscopy and anoscopy should be performed
first in patients with scant intermittent hematochezia or
iron-deficiency anemia. eGD should be carried out in
patients with iron-deficiency anemia if colonoscopy fails
to detect the source of bleeding. if both colonoscopy and
eGD fail to localize the source in acute and chronic Gi
bleeding, additional endoscopic methods can be performed to examine the small intestine. Push enteroscopy
enables visualization of about 50–120 cm of the proximal jejunum. Double-balloon enteroscopy can make the
whole small intestine visible, especially if bidirectional
enteroscopy is performed—that is, if the scope is introduced successively by mouth and anus. using wireless video capsule endoscopy, the small bowel can be
completely visualized in about 80% of cases.48
nonendoscopic methods
Nuclear scintigraphy
nuclear scintigraphy is a sensitive method for detecting gastrointestinal bleeding at a rate of 0.1 ml/min.
the method is more sensitive, but less specific, than
angiography.49 the role of nuclear scans and, in particular, of technetium-99m-labeled erythrocytes is limited
for obscure gastrointestinal bleeding. the major disadvantages are that nuclear imaging localizes bleeding
only to an area of the abdomen and that the intraluminal
blood is moved away by intestinal motility. when scans
are positive within 2 h after injection of the labeled
erythrocytes, localization is correct in 95–100% of cases.
However, when scans are positive after more than 2 h
after injection, the accuracy decreases to 57–67%. 8
scintigraphy might be useful, especially for recurrent
bleeding, when other methods have failed.
Radiology
visceral angiography is estimated to only detect active
bleeding when the rate is at least 0.5–1 ml/min.50,51 the
specificity of this procedure is 100%, but sensitivity varies
from 47% with acute lGiB to 30% with recurrent bleeding.52 unfortunately, bleeding is frequently intermittent
and might be slow, thereby limiting detection of the causative lesion. angiography should be reserved for patients
who have massive bleeding that precludes colonoscopy,
or for whom endoscopies were negative. visceral angiography has a complication rate of 9.3%. 53 studies have
shown that Ct angiography is highly sensitive and specific
for diagnosing colonic angiodysplasia.54,55 this approach
seems to be equivalent to visceral angiography in the
detection of acute gastrointestinal hemorrhage,56 showing
accuracy rates of 54–79% for colonic bleeding.57,58
there is no role for barium studies in the detection
of acute lGiB. Plain abdominal radiography and/or
Ct might be carried out, depending on the clinical
presentation and suspected etiology (such as ischemic or
inflammatory colitis, or in cases where bowel obstruction
or perforation are suspected).
in cases of chronic lGiB, Ct colonography can be used
to examine the colon. However, it must be considered that
small or flat lesions such as angiodysplasias are usually
not recognized by this method.
Therapy
endoscopic hemostasis
endoscopic treatment modalities for lGiB include injection, contact and noncontact thermal coagulation, and
mechanical devices such as metallic clips and band ligation. the use of these techniques depends on the site and
the features of the bleeding lesion, the clinician’s personal
experience with the devices, and access to the bleeding
site. randomized studies on the management of specific
lesions are rare; most reports are case histories (table 3).
Thermal coagulation
thermal coagulation is one of several hemostatic techniques. in bipolar (BiCaP) and monopolar electrocoagulation, an electrical current from a probe heats up
the tissue, whereas a heater probe directly delivers heat. all
three probes induce coagulation upon contact. Perforation
in the thin-walled right hemicolon occurs in up to 2.5%
of patients undergoing thermal coagulation.19 the BiCaP
probe shows an increase in energy deposition when greater
pressure is applied, or when increased angulation causes
an increase in tissue contact. For the heater probe, which
has a non-stick teflon coating, these associations have not
been observed.59 BiCaP and heater probe enable water jet
irrigation through an opening on the probe tip.
argon plasma coagulation (aPC) transmits energy to the
tissue without contact by means of ionized argon gas.
the penetration depth is limited by the extent of tissue
desiccation, whereas the coagulation depth depends
on the power setting, duration of application, and distance from the probe tip to the target tissue.60 the risk
642 | NOVEMBER 2009 | VOluME 6
www.nature.com/nrgastro
© 2009 Macmillan Publishers Limited. All rights reserved
rEVIEWS
Table 3 | Endoscopic management of specific lesions: a summary of case studies
Method
Diverticulum
angiodysplasia
Dieulafoy
neoplasia
postpolypectomy
Colitis
nsaID
hemorrhoids
rectal varices
BICAP
++12,13
++86
++
+
+
+28
+
++91
–
Heater
+
++
++
+
+
+
+
++
–
Laser
–
+32
+
++
–
–
–
–
–
Argon plasma
coagulation
++
++; P
++; P
++; P
+
+
+
–
–
Epinephrine
injection
++12,13
+
+91
+
++14
++28
++
+
–
Sclerosant
injection
–
+
+
++88
+
+89
–
++92
+
Acryl glue
injection
–
–
–
–
–
–
–
–
++; P
Clip
++; P
+
++91
+
++14,87; P
++90; P
++; P
++
+
rubber band
+85
–
+
–
–
–
–
++92
+
Cryotherapy
–
–
–
–
–
–
–
93
+
–
Infrared
coagulation
–
–
–
–
–
–
–
+92,94
–
Low voltage
current
–
–
–
–
–
–
–
+84,93,94
–
86
87
91,94
Abbreviations: +, can be done; ++, recommended; –, not recommended; P, personal preference of the authors.
of perforation in the colon seems to be almost nonexistent.17 Probes that direct the plasma stream parallel or
perpendicular to the axis of the catheter are available.
in laser-mediated coagulation, the high-energy laser
light causes vaporization of the tissue. the deep penetration of a laser pulse (especially by nd:YaG laser) increases
the perforation risk.
Injection therapy
injection therapy is an inexpensive and straightforward
method. epinephrine (at a 1:10,000 dilution) is usually
used, causing vasoconstriction and physical compression of the vessel. addition of a sclerosant (for example,
ethanolamine) seems to confer no additional advantage.
in bleeding rectal varices, a cyanoacrylate glue is injected
to repair the lesion.
Mechanical devices
metallic clips can be used to definitively and securely close
bleeding lesions. Both reusable and single-use clipping
devices are available. the reusable device, manufactured
by olympus (olympus america inc., Center valley, Pa),
is less costly and available in different sizes, but requires a
skilled assistant for loading and handling; after several uses,
the rotation capability of the device becomes erratic. singleuse pre-loaded clipping devices, such as the QuickClip2
(Boston scientific, natick, ma) are easy and quick to use.
ligation with rubber bands is used for bleeding hemorrhoids and bleeding rectal varices and, in certain circumstances, for treatment of focal lesions that are <2 cm in
diameter in nonfibrotic tissue, the technology and equipment used for the treatment of esophageal varices must be
used in this setting. the amount of tissue suctioned into
the cap before application of the rubber band must be
carefully monitored owing to the danger of full-thickness
tissue entrapment and the subsequently increased risk
of perforation.
Interventional angiography
transcatheter embolization is an effective method of controlling gastrointestinal hemorrhage. embolization proximal to the mesenteric border of the colon was initially
carried out via large catheters (5 French), which led to a
significant rate of bowel infarction (13–33%).4 However, the
availability of microcatheters (2.7 French) and embolization methods such as microcoils, gelfoam, and polyvinyl
alcohol particles renewed enthusiasm for this technique.
success rates range from 70% to 90% without major ischemic complications;4 recurrent hemorrhage occurs in less
than 15% of cases.61 the location and etiology of bleeding
have important therapeutic implications for angiotherapy.
Bleeding from the right colon and cecum might be less
amenable to embolization than bleeding in the left colon.
angiodysplasias are more difficult to treat using embolization compared to diverticular bleeding, and have a greater
tendency (7–40%) to rebleed. 62,63 angiography and
embolization should be reserved for those patients who
have massive bleeding with hemodynamic instability that
precludes urgent colonoscopy, or for whom colonoscopy
has failed to identify a bleeding source.
intra-arterial infusion of vasopressin or its longer acting
analog, terlipressin, is successful in controlling gastrointestinal hemorrhage, but rebleeding occurs in up to 50%
of patients after cessation of the infusion.64 vasopressin
and terlipressin also cause significant adverse effects,
including abdominal pain, and are contraindicated in
patients with coronary heart disease. this technique is
now used only in exceptional cases.
nature reviews | gastroenterology & hepatology
© 2009 Macmillan Publishers Limited. All rights reserved
volume 6 | novemBer 2009 | 643
rEVIEWS
surgery
most patients with lGiB will not require surgery,
although this is the therapy of choice in patients with
bleeding related to neoplasia. surgery should also be considered in patients for whom a bleeding source has clearly
been identified and conservative therapies have failed. in
fulminant bleeding or in patients with recurrent bleeding
without localization of the bleeding source, surgery is the
last resort. Blind segmental colectomy is associated with
unacceptably high rates of morbidity (rebleeding rate as
high as 75%) and mortality (up to 50% of patients who
have undergone surgery) and should therefore be avoided
at all costs. whenever possible, intraoperative endoscopy
should be carried out in such patients to clearly localize the bleeding source. Directed segmental resection
is the treatment of choice because of its low morbidity,
mortality (about 4%) and rebleeding rate (about 6%).65
pharmacotherapy
Hormonal therapy of angiodysplasia using estrogen and
progestagen has not been successful. somatostatin and its
analog, octreotide, have been reported to reduce blood
loss from intestinal angiodysplasias.66,67 octreotide was
also successful in a patient with bleeding from portal
colopathy.68 in angiodysplasias, antagonists of vascular
endothelial growth factor, such as thalidomide, seem
to show reasonable efficacy. one study showed that
thalidomide prevented rebleeding in patients with angiodysplasias and in patients with severe bleeding related
to Crohn’s disease.69 However, the use of thalidomide is
associated with severe adverse effects and teratogenicity.
Clinical outcome and recurrence
Patients with chronic lGiB presenting as chronic iron
deficiency anemia or positive occult blood in stool
require colonoscopy because examination of the entire
colon has been demonstrated to significantly reduce the
rate of mortality from colon cancer. For acute lGiB, two
studies demonstrated that early colonoscopy is significantly associated with a shorter hospital stay.70,71 By
contrast, another study found that early colonoscopy is
not superior to a standard care algorithm.72 one benefit
of urgent colonoscopy is endoscopic risk stratification.
Jensen et al.12 and Grisolano et al.73 have shown that
evidence of active bleeding, visible vessels and adherent
clots are associated with a severe course of diverticular
bleeding and a high rate of rebleeding, but this result was
contradicted by another study.13 in general, diverticular
bleeding is characterized by a high rate of spontaneous
cessation of bleeding and a considerable risk of rebleeding
in the long term.6
a reliably predictive model that can accurately forecast
the outcome of an episode of acute lGiB in terms of risk of
recurrent bleeding, need for therapeutic intervention, and
more importantly, the risk of mortality, would be extremely
useful for triage of patients in order to assure appropriate
and cost-effective levels of care. this approach would consequently improve medical resource utilization by reducing
unnecessary iCu admissions and the frequency of urgent
interventions. Kollef and colleagues74,75 have identified
several risk factors for predicting adverse outcomes of both
acute upper Gi bleeding and lGiB: ongoing bleeding, low
systolic blood pressure, elevated prothrombin time, erratic
mental status, and the presence of an unstable comorbid
disease. although high-risk and low-risk patients were
defined using these criteria, this BleeD classification has
not been widely adopted clinically.
strate et al.76 identified several predictors for severe
course and bleeding recurrence in acute lGiB. these
predictors include a heart rate of ≥100/min, systolic
pressure ≤115 mmHg, syncope, nontender abdominal
examination, bleeding per rectum during the first 4 h of
evaluation, a history of aspirin use and the presence
of more than two active comorbidities.
using these criteria, patients were stratified into three
risk groups: low (no risk factor), moderate (1–3 risk
factors), and high (>3 risk factors). the magnitude of the
risk score was significantly correlated with major clinical
outcomes including surgery, death, blood transfusions,
and length of hospitalization.77
the same group27 has published another study using
multivariate analysis to identify variables in acute lGiB
that are associated with mortality. these variables include
age (patients above 70 years versus patients aged below
50 years), the occurrence of intestinal ischemia, two or
more comorbidities, bleeding while hospitalized for a
separate process, a coagulation defect (or chronic anticoagulant usage), hypovolemia, transfusion of packed red
cells and sex (male). among these variables, advanced
age, intestinal ischemia and comorbid illness were the
strongest predictors of mortality, whereas the presence
of colorectal polyps and hemorrhoids was associated with
a lower risk of mortality. the all-cause in-hospital mortality rate of lGiB reported in this study was 3.9%. Das and
colleagues78 used a model based on an artificial neural
network to predict clinical outcome in lGiB and compared it to a multiple logistic regression model and the
BleeD classification. 75 using a commercial neural
network computer program and 19 clinical measures as
input variables, they could predict mortality, recurrent
bleeding and need for intervention with a high degree
of accuracy. Patients at low risk of adverse outcomes, in
particular, could be accurately identified by this method.
moreover, the positive predictive value for death was low.
the predictive accuracy of the artificial neural network
was similar to the multiple logistic regression model
and was superior to the BleeD classification.75
Conclusion
although not as common as, and less fatal than, upper
gastrointestinal bleeding, lGiB occurs increasingly with
advancing age and is associated with substantial morbidity and mortality. advances especially in endoscopic and
angiographic intervention approaches have improved
therapy for lGiB and, subsequently, patient outcome,
although the benefits are limited by the increasing age of
644 | NOVEMBER 2009 | VOluME 6
www.nature.com/nrgastro
© 2009 Macmillan Publishers Limited. All rights reserved
rEVIEWS
the patients and their frequent comorbidities. there are
several risk scores to stratify patients with acute lGiB to
appropriate levels of care and intervention. these scores
can be helpful in clinical decision-making, but none comes
close to being an ideal risk score. Good prospective clinical
studies are still rare and often include only a few patients.
treatment decisions are therefore usually based on personal expertise and preference. some publications indicate that urgent endoscopy will have a similar role in lGiB
as it currently does for upper gastrointestinal bleeding,
but further studies are needed to clarify this issue.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
Zuccaro, G. Jr. Management of the adult patient
with acute lower gastrointestinal bleeding.
American College of Gastroenterology. Practice
Parameters Committee. Am. J. Gastroenterol. 93,
1202–1208 (1998).
Prakash, C. & Zuckerman, G. r. Acute small
bowel bleeding: a distinct entity with significantly
different economic implications compared with GI
bleeding from other locations. Gastrointest.
Endosc. 58, 330–335 (2003).
Davila, r. E. et al. ASGE Guideline: the role of
endoscopy in the patient with lower‑GI bleeding.
Gastrointest. Endosc. 62, 656–660 (2005).
Farrell, J. J. & Friedman, L. S. review article: the
management of lower gastrointestinal bleeding.
Aliment. Pharmacol. Ther. 21, 1281–1298 (2005).
Leerdam, M. E., ramsoekh, D., rauws, E. A. &
Tytgat, G. N. Epidemiology of acute lower
intestinal bleeding [abstract]. Gastrointest.
Endosc. 57, 93 (2003).
Longstreth, G. F. Epidemiology and outcome of
patients hospitalized with acute lower
gastrointestinal hemorrhage: a population‑based
study. Am. J. Gastroenterol. 92, 419–424 (1997).
Peura, D. A., Lanza, F. L., Gostout, C. J. &
Foutch, P. G. The American College of
Gastroenterology Bleeding registry:
preliminary findings. Am. J. Gastroenterol. 92,
924–928 (1997).
Zuckerman, G. r. & Prakash, C. Acute lower
intestinal bleeding: part I: clinical presentation
and diagnosis. Gastrointest. Endosc. 48,
606–617 (1998).
Jensen, D. M. & Machicado, G. A. Diagnosis and
treatment of severe hematochezia. The role of
urgent colonoscopy after purge. Gastroenterology
95, 1569–1574 (1988).
Vermeulen, L. C. Jr, ratko, T. A., Erstad, B. L.,
Brecher, M. E. & Matuszewski, K. A. A paradigm
for consensus. The University Hospital
Consortium guidelines for the use of albumin,
nonprotein colloid, and crystalloid solutions. Arch.
Intern. Med. 155, 373–379 (1995).
Boldt, J. Volume replacement in the surgical
patient‑‑does the type of solution make a
difference? Br. J. Anaesth. 84, 783–793 (2000).
Jensen, D. M., Machicado, G. A., Jutabha, r. &
Kovacs, T. O. Urgent colonoscopy for the
diagnosis and treatment of severe diverticular
hemorrhage. N. Engl. J. Med. 342, 78–82 (2000).
Bloomfeld, r. S., rockey, D. C. & Shetzline, M. A.
Endoscopic therapy of acute diverticular
hemorrhage. Am. J. Gastroenterol. 96,
2367–2372 (2001).
Zuckerman, G. r. & Prakash, C. Acute lower
intestinal bleeding. Part II: etiology, therapy, and
outcomes. Gastrointest. Endosc. 49, 228–238
(1999).
Review criteria
A MEDLINE search was performed for English‑language
papers published between 1966 and June 2009
using the terms “gastrointestinal hemorrhage”,
“gastrointestinal bleeding”, “occult gastrointestinal
bleeding”, and “colonoscopy”. reference lists from
relevant publications were also analyzed to identify
additional applicable publications not found by the
database searches. If deemed appropriate,
abstracts from the Digestive Disease Week (DDW)
conferences from the past 4 years were also used.
15. Foutch, P. G., rex, D. K. & Lieberman, D. A.
Prevalence and natural history of colonic
angiodysplasia among healthy asymptomatic
people. Am. J. Gastroenterol. 90, 564–567
(1995).
16. richter, J. M., Hedberg, S. E., Athanasoulis, C. A.
& Schapiro, r. H. Angiodysplasia. Clinical
presentation and colonoscopic diagnosis. Dig.
Dis. Sci. 29, 481–485 (1984).
17. Kwan, V. et al. Argon plasma coagulation in the
management of symptomatic gastrointestinal
vascular lesions: experience in 100 consecutive
patients with long‑term follow‑up. Am. J.
Gastroenterol. 101, 58–63 (2006).
18. richter, J. M., Christensen, M. r., Colditz, G. A. &
Nishioka, N. S. Angiodysplasia. Natural history
and efficacy of therapeutic interventions. Dig. Dis.
Sci. 34, 1542–1546 (1989).
19. Foutch, P. G. Angiodysplasia of the
gastrointestinal tract. Am. J. Gastroenterol. 88,
807–818 (1993).
20. Brandt, L. J. & Spinnell, M. K. Ability of naloxone
to enhance the colonoscopic appearance of
normal colon vasculature and colon vascular
ectasias. Gastrointest. Endosc. 49, 79–83
(1999).
21. Brandt, L. J. & Mukhopadhyay, D. Masking of
colon vascular ectasias by cold water lavage.
Gastrointest. Endosc. 49, 141–142 (1999).
22. Deal, S. E., Zfass, A. M., Duckworth, P. F. &
McHenry, L. Arteriovenous malformations (AVMs):
are they concealed by meperidine [abstract]? Am.
J. Gastroenterol. 86, 1552 (1991).
23. Teshima, T., Hanks, G. E., Hanlon, A. L.,
Peter, r. S. & Schultheiss, T. E. rectal bleeding
after conformal 3D treatment of prostate cancer:
time to occurrence, response to treatment and
duration of morbidity. Int. J. Radiat. Oncol. Biol.
Phys. 39, 77–83 (1997).
24. Crook, J., Esche, B. & Futter, N. Effect of pelvic
radiotherapy for prostate cancer on bowel,
bladder, and sexual function: the patient’s
perspective. Urology 47, 387–394 (1996).
25. Kjeldsen, A. D. & Kjeldsen, J. Gastrointestinal
bleeding in patients with hereditary hemorrhagic
telangiectasia. Am. J. Gastroenterol. 95, 415–418
(2000).
26. Longacre, A. V. et al. Diagnosis and management
of gastrointestinal bleeding in patients with
hereditary hemorrhagic telangiectasia. Am. J.
Gastroenterol. 98, 59–65 (2003).
27. Strate, L. L., Ayanian, J. Z., Kotler, G. & Syngal, S.
risk factors for mortality in lower intestinal
bleeding. Clin. Gastroenterol. Hepatol. 6,
1004–1010 (2008).
28. Pardi, D. S. et al. Acute major gastrointestinal
hemorrhage in inflammatory bowel disease.
Gastrointest. Endosc. 49, 153–157 (1999).
nature reviews | gastroenterology & hepatology
© 2009 Macmillan Publishers Limited. All rights reserved
29. robert, J. r., Sachar, D. B. & Greenstein, A. J.
Severe gastrointestinal hemorrhage in Crohn’s
disease. Ann. Surg. 213, 207–211 (1991).
30. Bini, E. J., Weinshel, E. H. & Falkenstein, D. B.
risk factors for recurrent bleeding and mortality
in human immunodeficiency virus infected
patients with acute lower GI hemorrhage.
Gastrointest. Endosc. 49, 748–753 (1999).
31. Lanas, A., Sekar, M. C. & Hirschowitz, B. I.
Objective evidence of aspirin use in both ulcer
and nonulcer upper and lower gastrointestinal
bleeding. Gastroenterology 103, 862–869
(1992).
32. Barnert, J. Acute and chronic lower
gastrointestinal bleeding. in Atlas of Colonscopy
(ed. Messmann, H.) 118–142 (Thieme, Stuttgart
New York, 2006).
33. Kim, H. S. et al. risk factors for immediate
postpolypectomy bleeding of the colon:
a multicenter study. Am. J. Gastroenterol. 101,
1333–1341 (2006).
34. Consolo, P. et al. Efficacy, risk factors and
complications of endoscopic polypectomy:
ten year experience at a single center. World J.
Gastroenterol. 14, 2364–2369 (2008).
35. Sawhney, M. S., Salfiti, N., Nelson, D. B.,
Lederle, F. A. & Bond, J. H. risk factors for severe
delayed postpolypectomy bleeding. Endoscopy
40, 115–119 (2008).
36. Watabe, H. et al. risk assessment for delayed
hemorrhagic complication of colonic
polypectomy: polyp‑related factors and patient‑
related factors. Gastrointest. Endosc. 64, 73–78
(2006).
37. Gibbs, D. H. et al. Postpolypectomy colonic
hemorrhage. Dis. Colon Rectum 39, 806–810
(1996).
38. rosen, L., Bub, D. S., reed, J. F. 3rd &
Nastasee, S. A. Hemorrhage following
colonoscopic polypectomy. Dis. Colon Rectum 36,
1126–1131 (1993).
39. Smith, L. E. Fiberoptic colonoscopy:
complications of colonoscopy and polypectomy.
Dis. Colon Rectum. 19, 407–412 (1976).
40. Dafnis, G., Ekbom, A., Pahlman, L. & Blomqvist, P.
Complications of diagnostic and therapeutic
colonoscopy within a defined population in
Sweden. Gastrointest. Endosc. 54, 302–309
(2001).
41. Heldwein, W. et al. The Munich Polypectomy
Study (MUPS): prospective analysis of
complications and risk factors in 4000 colonic
snare polypectomies. Endoscopy 37,
1116–1122 (2005).
42. Hui, A. J. et al. risk of colonoscopic polypectomy
bleeding with anticoagulants and antiplatelet
agents: analysis of 1657 cases. Gastrointest.
Endosc. 59, 44–48 (2004).
volume 6 | novemBer 2009 | 645
rEVIEWS
43. Ganguly, S., Sarin, S. K., Bhatia, V. & Lahoti, D.
The prevalence and spectrum of colonic lesions
in patients with cirrhotic and noncirrhotic portal
hypertension. Hepatology 21, 1226–1231
(1995).
44. Eisen, G. M. et al. An annotated algorithmic
approach to acute lower gastrointestinal
bleeding. Gastrointest. Endosc. 53, 859–863
(2001).
45. Chaudhry, V., Hyser, M. J., Gracias, V. H. &
Gau, F. C. Colonoscopy: the initial test for acute
lower gastrointestinal bleeding. Am. Surg. 64,
723–728 (1998).
46. Ohyama, T. et al. Analysis of urgent colonoscopy
for lower gastrointestinal tract bleeding. Digestion
61, 189–192 (2000).
47. Elta, G. H. Urgent colonoscopy for acute lower‑GI
bleeding. Gastrointest. Endosc. 59, 402–408
(2004).
48. Eliakim, r. Video capsule endoscopy of the small
bowel. Curr. Opin. Gastroenterol. 24, 159–163
(2008).
49. Dusold, r., Burke, K., Carpentier, W. & Dyck, W. P.
The accuracy of technetium‑99m‑labeled red cell
scintigraphy in localizing gastrointestinal
bleeding. Am. J. Gastroenterol. 89, 345–348
(1994).
50. Zuckerman, D. A., Bocchini, T. P. & Birnbaum, E. H.
Massive hemorrhage in the lower gastrointestinal
tract in adults: diagnostic imaging and
intervention. Am. J. Roentgenol. 161, 703–711
(1993).
51. Nusbaum, M. & Baum, S. radiographic
demonstration of unknown sites of
gastrointestinal bleeding. Surg. Forum 14,
374–375 (1963).
52. Fiorito, J. J., Brandt, L. J., Kozicky, O.,
Grosman, I. M. & Sprayragen, S. The diagnostic
yield of superior mesenteric angiography:
correlation with the pattern of gastrointestinal
bleeding. Am. J. Gastroenterol. 84, 878–881
(1989).
53. Egglin, T. K., O’Moore, P. V., Feinstein, A. r. &
Waltman, A. C. Complications of peripheral
arteriography: a new system to identify patients
at increased risk. J. Vasc. Surg. 22, 787–794
(1995).
54. Junquera, F. et al. Accuracy of helical computed
tomographic angiography for the diagnosis of
colonic angiodysplasia. Gastroenterology 119,
293–299 (2000).
55. Mindelzun, r. E. & Beaulieu, C. F. Using biphasic
CT to reveal gastrointestinal arteriovenous
malformations. Am. J. Roentgenol. 168, 437–438
(1997).
56. Yoon, W. et al. Acute massive gastrointestinal
bleeding: detection and localization with arterial
phase multi‑detector row helical CT. Radiology
239, 160–167 (2006).
57. Ernst, O. et al. Helical CT in acute lower
gastrointestinal bleeding. Eur. Radiol. 13,
114–117 (2003).
58. Tew, K., Davies, r. P., Jadun, C. K. & Kew, J. MDCT
of acute lower gastrointestinal bleeding. Am. J.
Roentgenol. 182, 427–430 (2004).
59. Swain, C. P. et al. Which electrode? A comparison
of four endoscopic methods of electrocoagulation
in experimental bleeding ulcers. Gut 25,
1424–1431 (1984).
60. Watson, J. P., Bennett, M. K., Griffin, S. M. &
Matthewson, K. The tissue effect of argon plasma
coagulation on esophageal and gastric mucosa.
Gastrointest. Endosc. 52, 342–345 (2000).
61. Busch, O. r., van Delden, O. M. & Gouma, D. J.
Therapeutic options for endoscopic haemostatic
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
failures: the place of the surgeon and radiologist
in gastrointestinal tract bleeding. Best Pract. Res.
Clin. Gastroenterol. 22, 341–354 (2008).
Funaki, B. et al. Superselective microcoil
embolization of colonic hemorrhage. Am. J.
Roentgenol. 177, 829–836 (2001).
Peck, D. J., McLoughlin, r. F., Hughson, M. N. &
rankin, r. N. Percutaneous embolotherapy of
lower gastrointestinal hemorrhage. J. Vasc. Interv.
Radiol. 9, 747–751 (1998).
Browder, W., Cerise, E. J. & Litwin, M. S. Impact of
emergency angiography in massive lower
gastrointestinal bleeding. Ann. Surg. 204,
530–536 (1986).
Vernava, A. M. 3rd, Moore, B. A., Longo, W. E. &
Johnson, F. E. Lower gastrointestinal bleeding.
Dis. Colon Rectum 40, 846–858 (1997).
rossini, F. P., Arrigoni, A. & Pennazio, M.
Octreotide in the treatment of bleeding due to
angiodysplasia of the small intestine. Am. J.
Gastroenterol. 88, 1424–1427 (1993).
Torsoli, A., Annibale, B., Viscardi, A. &
Dellafave, G. Treatment of bleeding due to diffuse
angiodyplasia of the small intetsine with
somatostaine analogue. Eur. J. Gastroenterol.
Hepatol. 3, 785–787 (1991).
Yoshie, K. et al. Octreotide for severe acute
bleeding from portal hypertensive colopathy:
a case report. Eur. J. Gastroenterol. Hepatol. 13,
1111–1113 (2001).
Bauditz, J., Schachschal, G., Wedel, S. &
Lochs, H. Thalidomide for treatment of severe
intestinal bleeding. Gut 53, 609–612 (2004).
Schmulewitz, N., Fisher, D. A. & rockey, D. C.
Early colonoscopy for acute lower GI bleeding
predicts shorter hospital stay: a retrospective
study of experience in a single center.
Gastrointest. Endosc. 58, 841–846 (2003).
Strate, L. L. & Syngal, S. Timing of colonoscopy:
impact on length of hospital stay in patients with
acute lower intestinal bleeding. Am. J.
Gastroenterol. 98, 317–322 (2003).
Green, B. T. et al. Urgent colonoscopy for
evaluation and management of acute lower
gastrointestinal hemorrhage: a randomized
controlled trial. Am. J. Gastroenterol. 100,
2395–2402 (2005).
Grisolano, S. W., Darrell, S. P., Petersen, B. T. &
Dierkhising, r. A. Stigmata associated with
recurrence of lower gastrointestinal hemorrhage
[abstract]. Gastrointest. Endosc. 57, 117 (2003).
Kollef, M. H., Canfield, D. A. & Zuckerman, G. r.
Triage considerations for patients with acute
gastrointestinal hemorrhage admitted to a
medical intensive care unit. Crit. Care Med. 23,
1048–1054 (1995).
Kollef, M. H., O’Brien, J. D., Zuckerman, G. r. &
Shannon, W. BLEED: a classification tool to
predict outcomes in patients with acute upper
and lower gastrointestinal hemorrhage. Crit. Care
Med. 25, 1125–1132 (1997).
Strate, L. L., Orav, E. J. & Syngal, S. Early
predictors of severity in acute lower intestinal
tract bleeding. Arch. Intern. Med. 163,
838–843 (2003).
Strate, L. L., Saltzman, J. r., Ookubo, r.,
Mutinga, M. L. & Syngal, S. Validation of a clinical
prediction rule for severe acute lower intestinal
bleeding. Am. J. Gastroenterol. 100, 1821–1827
(2005).
Das, A. et al. Prediction of outcome in acute
lower‑gastrointestinal haemorrhage based on an
artificial neural network: internal and external
validation of a predictive model. Lancet 362,
1261–1266 (2003).
646 | NOVEMBER 2009 | VOluME 6
79. Powell, N. & McNair, A. Gastrointestinal
evaluation of anaemic patients without evidence
of iron deficiency. Eur. J. Gastroenterol. Hepatol.
20, 1094–1100 (2008).
80. Kepczyk, T. & Kadakia, S. C. Prospective
evaluation of gastrointestinal tract in patients
with iron‑deficiency anemia. Dig. Dis. Sci. 40,
1283–1289 (1995).
81. McIntyre, A. S. & Long, r. G. Prospective survey
of investigations in outpatients referred with iron
deficiency anaemia. Gut 34, 1102–1107 (1993).
82. rockey, D. C. & Cello, J. P. Evaluation of the
gastrointestinal tract in patients with iron‑
deficiency anemia. N. Engl. J. Med. 329,
1691–1695 (1993).
83. Gordon, S. r., Smith, r. E. & Power, G. C. The
role of endoscopy in the evaluation of iron
deficiency anemia in patients over the age of 50.
Am. J. Gastroenterol. 89, 1963–1967 (1994).
84. Norman, D. A., Newton, r. & Nicholas, G. V.
Direct current electrotherapy of internal
hemorrhoids: an effective, safe, and painless
outpatient approach. Am. J. Gastroenterol. 84,
482–487 (1989).
85. Witte, J. T. Band ligation for colonic bleeding:
modification of multiband ligating devices for use
with a colonoscope. Gastrointest. Endosc. 52,
762–765 (2000).
86. Jensen, D. M., Machicado, G. A., Cheng, S.,
Jensen, M. E. & Jutabha, r. A randomized
prospective study of endoscopic bipolar
electrocoagulation and heater probe treatment of
chronic rectal bleeding from radiation
telangiectasia. Gastrointest. Endosc. 45, 20–25
(1997).
87. Wong Kee Song, L. M. & Baron, T. H. Endoscopic
management of acute lower gastrointestinal
bleeding. Am. J. Gastroenterol. 103, 1881–1887
(2008).
88. Beejay, U. & Marcon, N. E. Endoscopic treatment
of lower gastrointestinal bleeding. Curr. Opin.
Gastroenterol. 18, 87–93 (2002).
89. Hirana, H. et al. A case of ulcerative colitis with
local bleeding treated by endoscopic injection of
absolute etanaol and 1% polidocanol.
Gastroenterol. Endosc. 4, 969–973 (1999).
90. Yoshida, Y. et al. Endoscopic treatment of
massive lower GI hemorrhage in two patients
with ulcerative colitis. Gastrointest. Endosc. 54,
779–781 (2001).
91. Jensen, D. M. et al. Prospective randomized
comparative study of bipolar electrocoagulation
versus heater probe for treatment of chronically
bleeding internal hemorrhoids. Gastrointest.
Endosc. 46, 435–443 (1997).
92. Johanson, J. F. & rimm, A. Optimal nonsurgical
treatment of hemorrhoids: a comparative
analysis of infrared coagulation, rubber band
ligation, and injection sclerotherapy. Am. J.
Gastroenterol. 87, 1600–1606 (1992).
93. Eisen, G. M. et al. Endoscopic therapy of
anorectal disorders. Gastrointest. Endosc. 53,
867–870 (2001).
94. Zinberg, S. S., Stern, D. H., Furman, D. S. &
Wittles, J. M. A personal experience in comparing
three nonoperative techniques for treating
internal hemorrhoids. Am. J. Gastroenterol. 84,
488–492 (1989).
acknowledgments
Désirée Lie, University of California, Irvine, CA, is the
author of and is solely responsible for the content of
the learning objectives, questions and answers of the
MedscapeCME‑accredited continuing medical
education activity associated with this article.
www.nature.com/nrgastro
© 2009 Macmillan Publishers Limited. All rights reserved