GASTROENTEROLOGY 1988;95:1569-74
Diagnosis and Treatment of Severe
Hematochezia
The Role of Urgent Colonoscopy After Purge
DENNIS M. JENSEN and GUSTAVO A. MACHICADO
UCLA Center for the Health Sciences, Wadsworth Veterans Administration Medical Center,
Center for Ulcer Research and Education, Los Angeles and Valley Hospital Medical Certter, Van
Nuys, California
The purpose of this study was to prospectively
evaluate (a) the diagnosis and treatment of 80 con-
secutive patients with severe, ongoing hemato-
chezia from unknown source and (b) the effective-
ness and safety of urgent colonoscopy after oral
purge. Fifty-two men and 28 women (mean age, 64.5
yr) received a mean of 6.5 U of blood and had
negative anoscopy, rigid sigmoidoscopy, and naso-
gastric tube aspiration before our evaluation. Be-
cause of ongoing severe hematochezia in the inten-
sive care unit, urgent diagnosis and treatment was
recommended by the attending physicians and sur-
geons. Emergency panendoscopy was performed
before purge. Urgent colonoscopy was performed in
the intensive care unit after patients received oral
purge and their gut was cleared of blood, clots, and
stool. The final diagnosis in these patients was 74%
colonic lesions (30% angiomata, 17% diverticulosis,
11% polyps or cancer, 9% focal ulcers, 7% other),
11 % upper gastrointestinal lesions, and 9% pre-
sumed small bowel lesions. No lesion site was
identified in 6%. Clinically significant fluid reten-
tion (medically controlled) occurred in 4% of pa-
tients after purge. Sixty-four percent of patients had
intervention for control of bleeding: 39% had ther-
apeutic endoscopy, 24% surgery, and 1% therapeu-
tic angiography. For 22 patients who also had
emergency visceral angiography, the diagnostic
yield was 14% and the complication rate was 9%.
Our conclusions for patients with severe ongoing
hematochezia from an unknown site were as fol-
lows. (a) Oral purge was effective and safe for cleans-
ing the colon of stool, clots, and blood. Sulfate purge
appeared to be safer than saline purge. (b) Before
urgent colonoscopy and purge, emergency panen-
doscopy was indicated to exclude an upper gastro-
intestinal bleeding source. (c) Urgent colonoscopy
after purge was effective, safe, and often diagnostic.
(d) Compared with urgent colonoscopy, urgent vis-
ceral angiography was often nondiagnostic. Howev-
er, the examinations may be complementary. (e)
Hemostasis via colonoscopy has a definitive role in
the treatment of some focal colonic lesions such as
bleeding angiomata.
B leeding per rectum is a common medical-
surgical problem. Most patients do not have
severe bleeding and therefore require neither hospi-
talization nor intensive care management. Elective
outpatient evaluation and treatment are feasible in
this group of patients. Although less frequently seen,
patients with severe hematochezia are exceptions.
They are often elderly, have concomitant medical or
surgical problems, and frequently require intensive
care unit management and emergency intervention
for control of bleeding (1). Traditional interventions
for diagnosis and treatment of these patients are
urgent visceral angiography or emergency surgery
(1-4).
Based on the high prevalence of left-sided colonic
diverticulosis in the elderly, bleeding diverticulosis
is often assumed to be the most common cause of
severe hematochezia and emergency partial colec-
tomy is performed. Elderly patients treated with this
approach have a high mortality and high rates of
rebleeding (3,4). The recurrence of colonic bleeding
after blind hemicolectomy for severe hematochezia
is as high as 33% (4). Nondiverticular lesions, such
as mucosal angiomata of the ascending colon, are the
most prevalent bleeding site for elderly patients with
severe hematochezia (1-3,5-7). Such lesions are
difficult to diagnose but may account for the high
rebleeding rates after left hemicolectomy.
© 1988 by the American Gastroenterological Association
0016·5085/88/$3.50
1570 JENSEN AND MACHICADO
Most physicians have been taught that emergency
colonoscopy in patients with severe hematochezia is
dangerous, often nondiagnostic, and impractical.
However, urgent colonoscopy is feasible with ade-
quate prior colonic cleansing (5,6). Only a few others
have reported good results with emergency colonos-
copy for diagnosis, but they do not cleanse the colon
beforehand (1,7-9).
In the present study, our purposes were (a) to
prospectively evaluate the diagnosis and treatment
of patients with severe, ongoing hematochezia from
unknown sources and (b) to evaluate the effective-
ness and safety of urgent colonoscopy after oral
purge in these patients.
Materials and Methods
Patient Characteristics
Eighty consecutive inpatients who were referred
for evaluation and treatment of severe ongoing hemato-
chezia were studied. Fifty-two were male and 28 were
female, with a mean age of 64.5 yr (range, 21-93 yr).
Because of ongoing hematochezia in the hospital, all
patients were thought to be bleeding actively at the time of
referral and evaluation. Criteria for ongoing, severe hema-
tochezia were (a) admission to an intensive care unit for
resuscitation, (b) continued passage of bright red blood,
clots, or burgundy stools for at least 6 h in an intensive
care unit, and (e) clinical severity of bleeding such that the
primary care physician, surgeon, and gastroenterologist
agreed that some intervention (endoscopic, artgiographic,
or surgical) was required for control of bleeding. The mean
number of blood transfusions administered before our
evaluation was 6.5 U (range, 0-22 U; the patient not
transfused refused because he was a Jehovah's Witness,
although his hematocrit was 22%). Ninety-four percent of
the patients had concomitant, major medical or surgical
problems. Before referral, each patient had a negative
anoscopy, rigid sigmoidoscopy, and nasogastric tube aspi-
ration.
Equipment
Colonoscopy was performed using standard colon-
oscopes with either two suction channels or a single large
(;:0::3.7 mm diameter) suction channel. Endoscopic thera-
peutic devices included a monopolar polypectomy snare
(Olympus Corp., Lake Success, N.Y.), a bipolar electrode
with a 50-W power generator (BlCAP-American ACMI,
Stamford, Conn.), a heater probe (Olympus Corp.), and an
argon laser (model 770; Cooper Lasersonics, Santa Clara,
Calif.) with a maximal power output of 12 W.
Patient Preparation
After evaluation and stabilization in an intensive
care unit, each patient gave written informed consent for
purge, urgent colonoscopy, and hemostasis via colon-
oscopy. They received a saline (10) or sulfate purge (11)
administered orally or via a nasogastric tube to cleanse the
GASTROENTEROLOGY Vol. 95, No.6
colon. Saline purge was used in this study until the
development of the sulfate purge (11). The patients con-
sumed 4-15 L (mean, 5.5 L) of purge solution over a period
of 2-7 h (mean, 3.2 h) until the rectal effluent was clear of
stool and clots. Administration of the purge solution via a
nasogastric tube was necessary for only those patients who
(a) were unable to drink the solution, (b) had endotracheal
intubation, or (c) had a nasogastric tube already in place.
The solution was at room temperature or lukewarm so that
it did not decrease body temperature.
Once the rectal effluent was clear of clots and stool,
colonoscopy was performed at the bedside using intrave-
nous medication and standard techniques. Portable coag-
ulation equipment was brought to the bedside. All exam-
inations were performed by the authors, who are very
experienced endoscopists.
Criteria for Diagnoses
Our criteria for definitive diagnosis of a lesion as
the bleeding site were (a) active bleeding (for diverticula,
angiomata, or other focal lesions), (b) an adherent clot in a
single diverticulum or on an ulcerated lesion, resistant to
washing but with fresh blood nearby and other lesions
absent, or (c) nonbleeding visible vessel for an ulcer when
other lesions were absent. We presumed a small bowel
bleeding site when the panendoscopy and colonoscopy
were negative but fresh blood or clots, or both, were
coming through the ileocecal valve at colonoscopy.
This study was approved by the Human Research Re-
view Committee at our institutions in October 1978.
Results
Preparation of the Colon
In all patients the preparation of the colon was
good to excellent and the colon was clear of stool
and large clots during colonoscopy. Residual purge
solution could be easily suctioned during the
colonoscopy with the large suction channel. Some
patients had fresh (liquid) blood or small clots that
could be suctioned or washed from the area. No
colonic strictures were encountered that precluded
complete colonoscopy. The cecum was reached in
all patients and, whenever possible, the terminal
ileum was intubated. In patients with the diagnosis
of "presumed small bowel lesions," fresh blood or
clots could be seen coming through the ileocecal
valve.
Diagnosis
Initial observations. Initially, we observed
three unexpected results. First, several patients with
a negative history of peptic ulcer disease, no hema-
temesis, and negative nasogastric aspirates actually
had upper gastrointestinal lesions accounting for
their severe hematochezia. Along with these panen-
doscopy findings, subsequent negative colonoscopy
December 1988
confirmed the upper gastrointestinal lesions as the
definitive bleeding site. Although bile was not al-
ways present in the nasogastric aspirate, when it
was, those patients did not have an upper gastroin-
testinal bleeding site as the cause of their severe
hematochezia.
A second observation was that some patients with
difficult to diagnose rectal lesions had severe hema-
tochezia clinically, and on rigid anoscopy or rigid
sigmoidoscopy had fresh blood and clots above the
sigmoid colon. Initially their bleeding was thought
to be from proximal colon or small bowel lesions.
The rectal lesions in these patients were only found
by turnaround examination (retroversion) in the rec-
tum during urgent colonoscopy and were in the area
of the rectum blind to routine rigid instruments. All
patients with rectal lesions included in this study
had their diagnosis made by retroversion during
urgent colonoscopy, rather than by anoscopy or rigid
sigmoidoscopy.
A third initial observation was the high frequency
of intervention for hemostasis in these patients
(~64%; refer to the section on treatment).
Complications. Four patients (4% of the total
patients purged) had clinically significant fluid over-
load and early heart failure during purge. One pa-
tient with heart disease was treated with diuretics
and 3 patients in chronic dialysis had dialysis dur-
ing purge. Three of these overloads occurred with
saline purges (3 of 36 or 8% of saline purge patients)
and one was with sulfate purge (1 of 44 or 2% of
sulfate purge patients). All patients responded to
medical management and none had more severe
complications. All had successful urgent colon-
oscopy.
There were no complications of urgent panendos-
copy, colonoscopy, or endoscopic hemostasis. There
was a 9% complication rate for angiography (see
below).
Diagnosis of the bleeding site. Of the 80 pa-
tients with severe hematochezia, 74% had their
bleeding site in the colon (Table 1). Angiomata were
the most common lesions, accounting for 30% of the
patients' diagnosis. Diverticulosis accounted for
17% (8% had active bleeding and 9% had a single
adherent clot). Polyps or cancer were the bleeding
site in 11%. Focal colitis or ulcers were the diagnosis
in 9%. Bleeding rectal lesions, polyp stalks, and
endometriosis comprised the remainder of the co-
lonic lesions. Nonbleeding diverticulosis was a very
common finding (present in 60% of patients) at
colonoscopy.
Upper gastrointestinal bleeding sites were diag-
nosed in 11% of patients in this study. One patient
had actively bleeding gastric varices at endoscopy
and another an actively bleeding angioma of the
SEVERE HEMATOCHEZIA AND URGENT COLONOSCOPY 1571
Table 1. Final Diagnosis for 80 Patients With Severe
Hematochezia
Lesion site Number of patients
Colonic 59 (74%)
Angiomata 24 (30%)
Diverticulosis 13 (17%)
Active bleeding 6 (8%)
Adherent clot 7 (9%)
Polyps or cancer 9 (11%)
Focal colitis or ulcers 7 (9%)
Rectal lesions 3 (4%)
Bleeding polyp stalk 2 (2%)
Endometriosis 1 (1%)
Upper gastrointestinal 9 (11%)
Small bowelo 7 (9%)
No site found 5 (6%)
Q A diagnosis of presumed small bowel site of bleeding was made
when the panendoscopy and colonoscopy were negative, but
fresh blood or clots, or both, were coming through the ileocecal
valve.
third duodenum. The remainder had peptic ulcers
with active bleeding or nonbleeding visible vessels.
No bleeding sites were found at colonoscopy in any
of these patients. Eight of the 9 patients with an
upper gastrointestinal bleeding site were male.
Presumed small bowel bleeding was diagnosed
after panendoscopy and colonoscopy were negative
but fresh blood was coming through the ileocecal
valve at colonoscopy. This accounted for 9% of
bleeding sites. Actual small bowel lesions were
subsequently confirmed in 70% of these patients by
surgery or enteroclysis.
No diagnosis of the bleeding site was made in 6%
of patients. None of these patients had rebleeding
during the hospitalization nor required intervention
for hemostasis. During a mean follow-up of 2 yr,
none of these patients had rebleeding.
Angiography for diagnosis. Twenty-two pa-
tients of the total 80 patients with severe hemato-
chezia had emergency visceral angiography, panen-
doscopy, and urgent colonoscopy performed. The
overall diagnostic yield of urgent visceral angiogra-
phy for the 22 patients in the study was 14% and the
complication rate was 9%. Two patients had compli-
cations of emergency angiography (9%): 1 had tran-
sient renal failure from the contrast and 1 had
arterial embolization. The overall diagnostic yield of
urgent panendoscopy-colonoscopy in this group of
patients was 86% and the complication rate from
purge was 4%.
Seventeen patients with the final diagnosis of
colonic or small bowel lesions had both urgent
visceral angiography and urgent colonoscopy. Only
2 patients had a definitive diagnosis made by angi-
ography (see Table 2). One patient with negative
1572 JENSEN AND MACHICADO
Table 2. Urgent Angiography vs. Colonoscopy for
Diagnosis of Colonic and Small Bowel Lesions
Positive Positive
Final diagnosis angiogram colonoscopy
Angiomata 1/5 4/5
Diverticulosis 1/4 3/4
Small bowel lesion 0/3 2/3
Colon polyp or cancer 0/2 2/2
Rectal lesions 0/2 2/2
Endometriosis 0/1 1/1
Total 2/17 (12%) 14117 (82%)
Seventeen patients with the final diagnosis of colonic or small
bowel bleeding sites had both urgent visceral angiography and
urgent colonoscopy. Positive angiogram or colonoscopy means
the test gave a definitive or presumptive (e.g., for small bowel
sites) diagnosis. The numerators are positive examinations and
the denominators are the total examinations performed.
colonoscopy had a cecal angioma on angiography
and another patient had diverticular bleeding. In this
group of 17 patients, the diagnostic yield of urgent
visceral angiography was 12%. In these 17 patients,
colonoscopy provided a definitive diagnosis in 12
(70%) and a presumptive diagnosis of a small bowel
site in 2 (12%) for a total yield of 82%.
Urgent visceral angiography was also performed in
5 additional patients who ultimately had an upper
gastrointestinal bleeding source. One angiogram was
diagnostic for gastric varices and the remainder (1
patient with a bleeding duodenal angioma and 3
patients with peptic ulcers on endoscopy) were
negative. The diagnostic yield of urgent panendos-
copy in this group was 100% compared with 20% for
urgent visceral angiography.
Treatment
Interventions for control of bleeding. Sixty-
four percent of our patients had some form of inter-
vention to control continued bleeding or rebleeding.
Therapeutic endoscopy was performed in 39% (Ta-
ble 3). The lesions amenable to endoscopic therapy
included bleeding peptic ulcer disease, angiomata,
polyps, or bleeding polyp stalk. When the definitive
bleeding site was identified, all lesions amenable to
endoscopic therapy were treated at the time of the
diagnostic colonoscopy. A second colonoscopy for
treatment was not performed later. Successful ther-
apeutic angiography was performed in 1 patient who
was not considered a candidate for surgery.
Surgery was necessary in 24% of patients (Table
3). Surgeries included surgery for peptic ulcer dis-
ease, right hemicolectomy for multiple angiomata,
subtotal colectomy for diverticular bleeding, and
segmental resections for other focal lesions. When-
ever surgery was required, the diagnosis of the
GASTROENTEROLOGY Vol. 95, No.6
bleeding site had usually been made beforehand by
urgent endoscopy or colonoscopy. Therefore, resec-
tions tended to be limited rather than subtotal.
Medical management. Twenty-nine patients
(36% of the total) stopped bleeding spontaneously
and did not rebleed during the hospitalization.
Diverticulosis was the source in 10 (13% for the
total), focal colitis in 6 (7% of total), presumed small
bowel sources in 5 (6% of tota1) , and an upper
gastrointestinal lesion in 3 (4% of total). All patients
in this group left the hospital without intervention.
Discussion
Patients with severe hematochezia that per-
sists after admission to an intensive care unit often
require urgent diagnosis and treatment. This study
clearly demonstrates that urgent colonoscopy can be
performed safely and effectively in such patients.
Oral purge greatly facilitated safe and effective
colonoscopy and treatment. Our diagnostic ap-
proach to patients with severe hematochezia yielded
diagnoses in 94% of cases. Final diagnoses by loca-
tion included 74% colonic, 11% upper gastrointesti-
nal, and 9% small bowel.
Not only did colonoscopy help localize the bleed-
ing site, but it identified the specific type of bleeding
lesion, graded the bleeding rate, and, in many cases,
provided access for effective endoscopic hemostasis.
Moreover, colonoscopy could be performed at the
bedside under well-controlled conditions in an in-
tensive care unit with constant cardiopulmonary
monitoring and adequate nursing assistance.
Patients with rectal lesions can have severe lower
gastrointestinal bleeding yet rigid sigmoidoscopy
and anoscopy may be nondiagnostic. Retroflexion
with a flexible colonoscope or sigmoidoscope is
indicated in the rectal vault to diagnose bleeding
lesions that are in the blind area for rigid instrumen-
tation. Since the present study was completed, we
Table 3. Intervention for Control of Bleeding
Intervention Number of patients
Therapeutic endoscopy 31 (39%)
Endoscopic coagulation 24 (30%)
Argon laser 6 (7%)
BlCAP 14 (18%)
Heater probe 4 (5%)
Endoscopic polypectomy 5 (6%)
Hemorrhoid sclerosis 2 (3%)
Surgery 19 (24%)
Therapeutic angiography 1 (1%)
Total 51 (64%)
Intervention is tabulated by number of patients [and percentage of
the total (= 80 patients)].
December 1988
now recommend retroflexion in the rectum of all
patients with severe hematochezia before purge and
urgent colonoscopy.
In this study saline purge was used until the
development of the sulfate purge solution (11). Co-
lonic preparation before colonoscopy with saline or
sulfate purge was well tolerated by most patients.
With the exception of 3 patients on renal dialysis
and 1 patient with severe heart disease, there were
no episodes of clinically significant fluid overload.
These were seen more frequently early in the study
when saline purge was used (8% of cases) than later
with sulfate purge (2%). Sulfate purge appeared to be
safer than saline purge for urgent hemostasis pa-
tients as predicted by earlier studies (11). Fluid
overload is a potential complication of either saline
or sulfate purge, particularly in patients with severe
heart, liver, or renal diseases. Careful evaluation of
these patients is mandatory before and during the
purge.
A few patients vomited purge solution but only 1
of our initial patients had hematemesis from a bleed-
ing ulcer. Clinically, this is the only patient whose
pattern of bleeding (hematemesis in addition to
hematochezia) was changed by the purge. No patient
had worse bleeding during or after purge. No com-
plications such as Mallory-Weiss tear or induced
bleeding were observed. Metoclopramide hydro-
chloride (Reglan; A. H. Robins Company, Richmond,
Va.) decreases nausea and increases gastric emptying
(12). It diminishes the possibility of gastric overdis-
tention and regurgitation. Patients with nausea dur-
ing purge commonly were improved with this med-
ication. Early in this study, purge solution was
chilled and patients complained of feeling cold
while drinking the purge. Although no documented
hypothermia occurred, we now recommend using
lukewarm, rather than chilled, purge to prevent this
potential complication.
Although the incidence of very severe colonic
bleeding is low, it occurs most often in the older
population who suffer from concomitant medical
problems (1,3,4). In contrast to upper gastrointesti-
nal bleeding, the majority (64%) of our patients with
severe ongoing hematochezia required intervention.
In contrast, for severe upper gastrointestinal bleed-
ing from nonvariceal sources, intervention was re-
quired in <30% of our patients (13). Among the
interventions used for control of active lower gastro-
intestinal bleeding or rebleeding, endoscopic treat-
ment was the most common (39% of total), followed
by surgery (24%), and therapeutic angiography (1%).
Previously, this entire group of patients in our hos-
pitals would have been treated with emergency
surgery. Before the purges and colonoscopies, the
attending physicians agreed that some intervention
SEVERE HEMATOCHEZIA AND URGENT COLONOSCOPY 1573
(endoscopy, angiography, or surgery, alone or in
combination) was indicated for control of severe
hemorrhage.
If surgery was planned in patients with active
bleeding or rebleeding, urgent colonoscopy often
yielded a specific diagnosis and a segmental location
of the bleeding. Sometimes the results of colon-
oscopy obviated the need for urgent surgery. When
urgent surgery was necessary, the results of our
evaluation and urgent colonoscopy assisted our sur-
geons in planning the procedure and shortening the
time for actual surgery. This resulted because (a)
extra surgical time was not required for a diagnosis,
(b) the resection tended to be segmental or limited
rather than radical or extensive, and (c) the specific
surgery could be planned beforehand.
Overall, 17% of our patients had a final diagnosis
of diverticular bleeding (active or single adherent
clot). However, our study patients were elderly and
60% had diverticulosis at colonoscopy. For those
with diverticulosis, 73% had a nondiverticular site
of bleeding, whereas 27% had diverticular bleeding.
Tedesco et al. previously reported a high prevalence
of colonic nondiverticular sources of bleeding in
patients with self-limited rectal bleeding (14) or
unexplained melena (15). They also emphasized the
importance of colonoscopy for definitive diagnosis
and treatment (14,15).
Emergency visceral angiography reportedly local-
ized colonic bleeding sites in 23%-66% of patients
with severe hematochezia and the complication rate
was ~4% (2,16-19). In our patients who had both
emergency angiography and colonoscopy-panendos-
copy for ongoing hematochezia during hospitaliza-
tion, angiography identified bleeding sites in only
14%. In contrast, colonoscopy or panendoscopy
yielded a diagnosis in 86% of these patients. In our
opinion, there are several reasons for the failure of
angiography to yield a diagnosis compared with
colonoscopy. Colonic bleeding tends to be from
small vessels, such as arteriovenous malformations,
and tends to be intermittent. Angiography may be
done at a time when no active bleeding is occurring
or the bleeding rate is too low to show extravasation.
Small lesions, such as angiomata, may not even be
seen by magnified angiographic techniques or selec-
tive catheterization. On the other hand, urgent
colonoscopy is quite sensitive for small mucosal
lesions either with active bleeding or stigmata of
recent bleeding (visible vessel, affixed clot, and fresh
blood in the bowel segment). There are several other
disadvantages of emergency visceral angiography for
diagnosis of severe hematochezia in our experience:
(a) mobilization of the angiographic team was some-
times slow at odd hours when patients were bleed-
ing, (b) sick patients with active bleeding required
1574 JENSEN AND MACHICADO
transport from the intensive care unit to the angiog-
raphy suite where monitoring and nursing care may
have been suboptimal compared with the intensive
care unit, and (c) there was a significant complica-
tion rate of emergency angiography in these sick
patients compared with the diagnostic yield (9% vs.
14%, respectively, in our patients). None of these
limitations were evident for colonoscopy after purge.
Visceral angiography can detect bleeding lesions
in some patients with negative colonoscopy and
panendoscopy. We consider colonoscopy and angi-
ography complementary examinations for the evalu-
ation and treatment of some patients with severe
hematochezia. Two of our patients with negative
colonoscopy later had positive visceral angiograms.
Before colonoscopic hemostasis was available in
one hospital, 1 patient, whose colonoscopic diagno-
sis was angiomata, subsequently had successful ther-
apeutic angiography for rebleeding. Nevertheless,
after the present study, we recommend urgent pan-
endoscopy and colonoscopy before visceral angiog-
raphy for diagnosis of severe ongoing hematochezia.
Because urgent colonoscopy can be technically
more difficult than elective colonoscopy, we recom-
mend this procedure to experienced colonoscopists
but not neophytes. The effective and safe use of
hemostasis devices in the colon also requires more
than the usual skill and training (13).
Our conclusions for patients with severe, ongoing
hematochezia from an unknown site were as follows.
(a) Oral purge was effective and safe for cleansing the
colon of stool clots and blood. Sulfate purge ap-
peared to be safer than saline purge. (b) Before urgent
colonoscopy and purge, urgent panendoscopy is
indicated to exclude an upper gastrointestinal bleed-
ing source, particularly in male patients. (c) Urgent
colonoscopy after purge was effective, safe, and often
diagnostic. (d) Compared with urgent colonoscopy,
urgent visceral angiography is often nondiagnostic.
However, the examinations may be complementary.
(e) Hemostasis via colonoscopy has a definitive role
in the treatment of some focal colonic lesions such as
bleeding angiomata.
References
1. Colacchio TA. Forde KA, Patsos TJ, Nunez D. Impact of
diagnostic methods on the management of active rectal bleed-
ing. Am J Surg 1982;143:607-10.
2. Baum S, Athanasoulis CA, Waltman AC. Angiographic diag-
nosis and control of large bowel bleeding. Dis Colon Rectum
1974;17:447-53.
3. Wright HK. Massive colonic hemorrhage. Surg Clin North Am
1980;60:1297-304.
4. McGuire HH, Haynes BW. Massive hemorrhage from diver-
GASTROENTEROLOGY Vol. 95, No.6
ticulosis of the colon: guidelines for therapy based on bleed-
ing patterns observed in fifty cases. Ann Surg 1972;175:847-
55.
5. Jensen DM, Machicado GA. Urgent colonoscopy in patients
with severe gastrointestinal bleeding (abstr). Gastroenterology
1981;80:1184.
6. Jensen DM, Machicado GA, Tapia JI. Urgent colonoscopy in
patients with severe hematochezia. Gastrointest Endosc 1983;
29:177.
7. Forde KA. Colonoscopy in acute rectal bleeding. Gastrointest
Endosc 1981;27:219-20.
8. Deyhle P, Blum AL, Nuesch HI. Jenny S. Emergency colon-
oscopy in the management of acute perianal hemorrhage.
Endoscopy 1974;6:229.
9. Rossini FP, Ferrari A. Emergency colonoscopy. In: Hunt RH,
Waye JD, eds. Colonoscopy: techniques, clinical practice and
colour atlas. London: Chapman & Hall, 1981:289-99.
10. Levy AG, Benson JW, Hewlett EL, et al. Saline lavage: a rapid,
effective, and acceptable method for cleansing the gastroin-
testinal tract. Gastroenterology 1976;70:157-61.
11. Davis GR, Santa Ana CA, Morawski SG, Fordtran JS. Devel-
opment of a lavage solution associated with minimal water
and electrolyte absorption or secretion. Gastroenterology
1980;78:991-5.
12. Ramsbottom N, Hunt IN. Studies of the effect of metoclopra-
mide and apomorphine on gastric emptying and secretion in
man. Gut 1970;11:989-93.
13. Jensen DM. Endoscopic control of gastrointestinal bleeding
with non-laser devices. In: Fleischer D, Jensen D, Bright-
Asare P, eds. Therapeutic laser endoscopy in gastrointestinal
disease. Boston: Martinus Nijhoff, 1983:151-60.
14. Tedesco FJ, Waye JD, Raskin JB, Morris SJ, Greenwald RA.
Colonoscopic evaluation of rectal bleeding. A study of 304
patients. Ann Intern Med 1978;78:907-9.
15. Tedesco FJ, Pickens CA, Griffin JW, Sivak MV, Sullivan BH.
Role of colonoscopy in patients with unexplained melena:
analysis of 53 patients. Gastrointest Endosc 1981;27:221-3.
16. Wenz W. Abdominal angiography. New York, Springer-
Verlag, 1974.
17. Nusbaum M, Baum S, Blakemore WS. Clinical experience
with the diagnosis and management of gastrointestinal hem-
orrhage by selective mesenteric catheterization. Ann Surg
1969;170:506-14.
18. Eisenberg H, Laufer I, Skillman JJ. Arteriographic diagnosis
and management of suspected colonic diverticular hemor-
rhage. Gastroenterology 1973;64:1091-100.
19. Sigstedt B, Lunderquist A. Complications of angiographic
examinations. Am J Roentgenol 1978;130:455-60.
Received December 29, 1986. Accepted July 25, 1988.
Address requests for reprints to: Dennis M. Jensen, M.D., 44-133
CHS, GI Division, UCLA Center for the Health Sciences, Los
Angeles, California 90024.
This study was supported in part by the Veterans Administra-
tion Medical Research Service and by grant AM 17328 from the
National Institutes of Health to the Center for Ulcer Research and
Education.
The authors thank Mary Crump for secretarial assistance; the
many UCLA-Wadsworth GI fellows who assisted at all hours in
the evaluation and treatment of these patients; and the excellent
nurses in the intensive care units for their dedicated care of these
patients.