Review

Continuing Medical EducationNature Clinical Practice Gastroenterology & Hepatology (2008) 5, 80-93
doi:10.1038/ncpgasthep1034  
Received 8 June 2007 | Accepted 18 October 2007

Gastrointestinal bleeding in the elderly

Patrick S Yachimski and Lawrence S Friedman*  About the authors

Correspondence *Department of Medicine, Newton-Wellesley Hospital, 2014 Washington Street, Newton, MA 02462, USA

Email
 lfriedman@partners.org

Medscape logo

Medscape Continuing Medical Education online
Medscape, LLC is pleased to provide online continuing medical education (CME) for this journal article, allowing clinicians the opportunity to earn CME credit. Medscape, LLC is accredited by the Accreditation Council for Continuing Medical Education (ACCME) to provide CME for physicians. Medscape, LLC designates this educational activity for a maximum of 1.0 AMA PRA Category 1 Credits™. 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 receive credit, please complete the post-test.

Learning objectives

Upon completion of this activity, participants should be able to:

  1. Describe the etiology and management of upper gastrointestinal bleeding in elderly adults.
  2. Describe conditions that can cause sudden upper gastrointestinal bleeding in elderly adults.
  3. Identify the most common cause of lower gastro-intestinal bleeding in the elderly.
  4. Describe the etiology and management of lower gastrointestinal bleeding in elderly adults.
  5. Specify diagnostic methods for obscure gastro-intestinal bleeding in the elderly.

Competing interests

The authors declared no competing interests. Charles P Vega, the CME questions author, declared that he has served as an advisor or consultant to Novartis, Inc.

To complete the questions online and earn continuing education credits, you must be a registered user on Medscape.com. If you are not registered on Medscape.com please click on the New Users: Free Registration link on the top left-hand side of the website to register. Registration is free. For questions regarding the content of this activity, contact the accredited provider for this CME activity: CME@medscape.net. For technical assistance, contact CME@webmd.net.

Summary

Gastrointestinal bleeding affects a substantial number of elderly people and is a frequent indication for hospitalization. Bleeding can originate from either the upper or lower gastrointestinal tract, and patients with gastrointestinal bleeding present with a range of symptoms. In the elderly, the nature, severity, and outcome of bleeding are influenced by the presence of medical comorbidities and the use of antiplatelet medication. This Review discusses trends in the epidemiology and outcome of gastrointestinal bleeding in elderly patients. Specific causes of upper and lower gastrointestinal bleeding are discussed, and recommendations for approaches to endoscopic diagnosis and therapy are given.

Review criteria

A PubMed search was performed using the terms "colonoscopy", "elderly", "gastrointestinal bleeding", "gastrointestinal endoscopy", and "gastrointestinal hemorrhage" to identify English-language articles related to gastrointestinal bleeding and gastrointestinal endoscopy in both the elderly and the general population. No time limits were specified for the search. Selected articles, but no abstracts, are referenced in this article. The reference list was updated in August 2007.

Top

Introduction

More than 1% of people aged 80 years and older are hospitalized each year because of gastrointestinal bleeding.1 Gastrointestinal bleeding in such elderly people can originate from lesions common to all age-groups or from lesions associated specifically with aging. In elderly people, morbidity and mortality from gastrointestinal bleeding is determined by both the nature of the bleeding lesion and the presence of comorbid medical conditions. The incidence and outcome of gastrointestinal bleeding in elderly people can also be influenced by the use of aspirin and other antiplatelet and anticoagulant agents.

Gastrointestinal bleeding can be classified as acute (presenting as hematemesis, melena, or hematochezia), or chronic, suspected because of the detection of occult gastrointestinal blood loss or anemia. Initial assessment of a patient with suspected gastrointestinal bleeding must include determination of the acuity and pace of blood loss. Subsequent clinical evaluation is undertaken to identify the source of the bleeding. Once a bleeding source has been identified, directed pharmacologic, endoscopic, and non-endoscopic therapy can be offered.

This Review provides an approach to the evaluation and management of gastrointestinal bleeding in the elderly. Although the approach to the diagnosis and management of gastrointestinal bleeding is not specific to elderly people, elderly people with gastrointestinal bleeding differ from younger people with the same condition with respect to several aspects of clinical presentation and outcome (Box 1). For the purposes of this Review, we define 'elderly' somewhat arbitrarily as older than 60 years of age. The term 'very elderly' is defined as an age greater than 85 years, a cutoff that might be more clinically relevant as the health of the aging population improves. This article also reviews the epidemiology and pathogenesis of the main causes of gastrointestinal bleeding in the elderly—peptic ulcer and diverticular hemorrhage, in particular, are discussed in detail with special reference to the elderly population. Attention is also focused on the utility and safety of endoscopic techniques for the diagnosis of gastrointestinal bleeding in the elderly—including reference to emerging techniques for small-bowel evaluation—and on the role of aspirin and other NSAIDs in gastrointestinal bleeding in the elderly.

Box 1 Clinical features of upper gastrointestinal bleeding in elderly versus younger patients.

 

Similarities

  • Presenting manifestations of bleeding: hematemesis (50%); melena (30%); hematemesis and melena (20%)
  • Peptic ulcer disease most common etiology
  • Safety and efficacy of endoscopic therapy

Differences (in elderly patients)

  • Fewer antecedent symptoms (abdominal pain, dyspepsia, heartburn)
  • Prior aspirin and NSAID use
  • Presence of comorbid conditions
  • Higher rates of hospitalization
  • Higher rates of rebleeding (peptic ulcer, see Table 2)
  • Higher mortality rate

With permission from Elsevier Ltd © Farrell JJ and Friedman LS (2000) Gastroenterol Clin North Am 29: 1–36.

Top

Upper gastrointestinal bleeding

Upper gastrointestinal bleeding (UGIB) is defined as gastrointestinal blood loss that originates proximal to the ligament of Treitz. Although the overall incidence of nonvariceal UGIB seems to have declined since the 1990s,2 time trend analyses suggest that people aged over 60 years and over 80 years each constitute an increasing proportion of those presenting with acute UGIB.3 As many as 70% of acute UGIB episodes occur in patients older than 60 years,4 and the incidence of UGIB increases with increasing age.5, 6 Advanced age has been consistently identified as a risk factor for mortality among patients presenting with UGIB,3, 5, 7, 8 presumably because of the high prevalence of comorbid illnesses—including pulmonary9 and cardiovascular10 disease—in elderly as compared with younger patients with UGIB.

Clinical presentation, initial evaluation and immediate management

Patients with UGIB can present with vomiting of fresh blood (hematemesis) or of coffee-ground material, or with rectal passage of blood. Melena is suggestive of bleeding proximal to the ligament of Treitz but can be due to bleeding from more distal gastrointestinal sources, as far as the cecum. Rapid or high-volume UGIB can result in hematochezia. The initial manifestations of UGIB can include syncope, light-headedness, or postural hypotension, even in the absence of overt bleeding. Nasogastric intubation that returns a bloody aspirate can confirm the presence of UGIB. A clear or bilious nasogastric lavage, however, does not definitively exclude UGIB.

Medical history taking should include details of prior gastrointestinal bleeding, previous abdominal surgery and current medication use, particularly aspirin and other NSAIDs. The physical examination should begin with an assessment of hemodynamics, including postural changes. Physical findings of chronic liver disease are suggestive of underlying portal hypertension. Laboratory evaluation should include serum electrolyte and coagulation parameter measurements, liver biochemical tests and a complete blood count.

Hypotension at presentation is associated with high mortality among elderly people who present with peptic ulcer hemorrhage,11 and early aggressive volume resuscitation decreases mortality.12 Volume resuscitation in patients with acute UGIB should be aimed at maintaining adequate hemodynamics and end-organ perfusion. The patient's cardiopulmonary status should be monitored repeatedly during the course of aggressive volume resuscitation to assess for pulmonary edema or signs of right-sided heart failure. Prophylactic endotracheal intubation might be necessary in patients who have an altered mental status and ongoing hematemesis, to protect against aspiration.

Peptic ulcer disease

Peptic ulcer disease is the most common cause of UGIB, including in elderly people (Table 1).3, 5, 13, 14 The discovery of Helicobacter pylori, the recognition of the importance of gastric acid suppression in the early treatment of peptic ulcer hemorrhage, and the development of endoscopic hemostatic techniques have meant that since the 1980s the diagnosis and treatment of bleeding caused by peptic ulcer disease have improved. Data from the early 1980s, which cite high rates of rebleeding15, 16 and mortality16, 17 in older people with peptic ulcer hemorrhage, must, therefore, be interpreted with caution. Although the number of hospital admissions and operations for complications of peptic ulcer disease has declined for the overall population (including younger people) since the 1980s,18, 19, 20 rates of hemorrhage and hospital admissions have increased among elderly people with peptic ulcers.18, 19, 20 The trends among elderly patients might reflect the increasing rates of NSAID and anticoagulant use among the elderly.18, 20

Table 1 Causes of upper gastrointestinal bleeding in the elderly.
Table 1 - Causes of upper gastrointestinal bleeding in the elderly.
Full tableFigures & Tables indexDownload PowerPoint slide (90K)

Platelet inhibition is impaired at an acidic gastric pH.21 Both platelet inhibition and mucosal fibrinolytic activity can be mitigated by gastric acid suppression.22 Treatment with a PPI is now the standard of care for patients with peptic ulcer hemorrhage and has been shown to decrease ulcer rebleeding rates independent of endoscopic therapy.23 Long-term PPI use can be associated with impaired absorption of supplemental calcium24 and with vitamin B12 deficiency,25 and one case–control analysis reported an increased risk of hip fracture.26 There are, however, no data to suggest that elderly patients who receive short-term PPI therapy for acute UGIB experience these adverse effects.

Esophagogastroduodenoscopy (EGD) should be performed once a patient with UGIB has been hemodynamically resuscitated; our practice is to perform EGD within 12h of presentation. EGD can confirm the presence of peptic ulcer disease, identify stigmata that connote a risk of recurrent or continued hemorrhage, and permit endoscopic therapy to reduce the risk of rebleeding. Endoscopic therapy with an epinephrine injection, thermocoagulation, hemostatic clip placement, or a combination of these modalities, decreases rebleeding rates in patients with a spurting vessel, adherent clot (after removal of the clot), or visible vessel (Table 2).27 Endoscopic therapy for peptic ulcer hemorrhage seems to be well tolerated in the elderly (Table 3).10, 28 Combining endoscopic therapy with PPI administration results in lower rebleeding rates than with endoscopic therapy alone.29 Among patients who experience rebleeding, there are no data to suggest that age influences the timing of rebleeding.

Table 2 Peptic ulcer: endoscopic stigmata of rebleeding and mortality, with and without endoscopic therapy.
Table 2 - Peptic ulcer: endoscopic stigmata of rebleeding and mortality, with and without endoscopic therapy.
Full tableFigures & Tables indexDownload PowerPoint slide (76K)

Table 3 Peptic ulcer rebleeding and mortality by patient age.
Table 3 - Peptic ulcer rebleeding and mortality by patient age.
Full tableFigures & Tables indexDownload PowerPoint slide (72K)

Hypovolemic shock at initial presentation, which is suggestive of massive ulcer hemorrhage, is associated with a poor outcome in the elderly.11 Elderly patients who have ulcers of 2cm or more in diameter are at a particularly high risk of rebleeding and mortality.30 Surgical therapy for peptic ulcer hemorrhage is reserved for patients who have hemorrhage that is refractory to pharmacologic and endoscopic therapy. Sepsis and multiorgan failure are leading causes of postoperative mortality in elderly people who undergo surgery for peptic ulcer hemorrhage. In one retrospective series of 136 elderly patients with peptic ulcer hemorrhage (median age 77 years), mortality was 31% among the 42 patients who underwent surgery; sepsis accounted for nearly 50% of these deaths.31

Esophageal varices and portal hypertensive gastropathy

Esophageal and gastric varices are caused by increased venous collateral flow from the portal circulation through the gastric coronary veins, usually because of portal hypertension. Variceal hemorrhage can occur when the hepatic venous pressure gradient exceeds 12mmHg. Features predictive of variceal hemorrhage include large variceal size and the presence of red 'wale' marks on varices.32

The mortality of patients who present with variceal hemorrhage has historically exceeded 30%.33 There are data to suggest that the mortality associated with variceal hemorrhage has improved over time, ostensibly as a result of advances in medical and endoscopic therapy; however, these data come from cohorts of patients with a mean age well below 65 years.33, 34 Available data suggest that mortality after variceal hemorrhage correlates with the Child–Turcotte–Pugh score (Child class) and not with advanced age.35

Pharmacologic therapy for acute variceal hemorrhage generally includes intravenous administration of a somatostatin analog to reduce splanchnic circulation inflow. Cardiovascular effects can be observed in patients who take these medications; these effects include a decrease in heart rate (with terlipressin)36 and peripheral vasoconstriction (with octreotide).37 Although these effects have not been shown to correlate with adverse outcomes, elderly patients who are taking these medications should be carefully monitored.

Endoscopic variceal band ligation has supplanted injection sclerotherapy as the endoscopic therapy of choice because of the lower rate of complications.38 Nonselective beta-blockers are effective for both primary and secondary prophylaxis of variceal hemorrhage; however, elderly patients should be monitored closely for adverse effects, which include orthostasis, fatigue, and affective disturbance.

Patients with portal hypertensive gastropathy (PHG) can present with or without endoscopically apparent gastroesophageal varices. If gastrointestinal bleeding occurs in a patient with PHG, it is more commonly chronic and occult than overt and hemodynamically significant.39 Reduction of portal venous pressure with a nonselective beta-blocker might be beneficial for patients with PHG and gastrointestinal bleeding; however, the data to support this indication are limited as compared with those for variceal bleeding prophylaxis.

Other causes

Esophagitis and gastropathy

Esophagitis and gastropathy describe mucosal injury to the esophagus and stomach, respectively, and are most commonly caused by gastric acid or specific medications, such as NSAIDs. Although overt hemorrhage is an uncommon manifestation of esophagitis or gastropathy, these lesions are implicated as bleeding sources more frequently in elderly than in younger people.14, 40, 41, 42 Moreover, severe esophagitis might not be heralded by heartburn in elderly people.42 Esophagitis and gastropathy are not typically amenable to endoscopic therapy unless a single ulcer with hemorrhagic stigmata is identified. Therapy for these two conditions consists of agents to suppress gastric acid secretion and the avoidance of offending medications, particularly aspirin and other NSAIDs. Stress-related gastric mucosal injury can result in UGIB in critically ill patients in an intensive care unit; risk factors include mechanical ventilation, multiorgan failure, and coagulopathy, but not advanced age.

Gastric antral vascular ectasia

Patients with gastric antral vascular ectasia (also known as 'watermelon stomach') typically present with occult or subacute blood loss and transfusion-dependent anemia. The prevalence of gastric antral vascular ectasia does not increase with age per se, but the condition can be associated with certain medical comorbidities, including end-stage renal disease43 and cirrhosis. Ablation therapy with argon plasma coagulation can stabilize hemoglobin levels and ameliorate transfusion requirements;44 multiple treatment sessions may be required.

Aortoenteric fistula

An aortoenteric fistula (AEF) develops in 0.5% of patients who have undergone aortoiliac bypass surgery.45 An AEF can also develop after endovascular aortic repair, a procedure that is being performed with increasing frequency in the elderly.46 AEF has been described in patients with native anatomy and after enteral stent placement.47 Both mechanical and inflammatory factors have a role in the development of an AEF.

In a large retrospective series, the median time interval from an aortic intervention to the development of an AEF was 90 months.48 The classic presentation is with a herald (sentinel) bleed followed by exsanguinating hemorrhage. An AEF typically involves the third portion of the duodenum but can occur anywhere throughout the gastrointestinal tract, including the esophagus and colon.48 The median time to diagnosis of an AEF after presentation is approximately 10 days.45 Perhaps because of this delay, short-term mortality exceeds 30% even after accurate diagnosis and repair.48 Noninvasive imaging with CT angiography can enable the diagnosis of an AEF, whereas mesenteric angiography is often not helpful and is not generally indicated. Endoscopy can also be diagnostic and should include careful examination of the third portion of the duodenum for an ulcer, erosion or hematoma, or for visualization of graft material. Biopsy samples should not be taken, and there is no role for endoscopic therapy.

Dieulafoy lesion

A Dieulafoy lesion is a dilated submucosal artery that can rupture and result in overt gastrointestinal hemorrhage. These lesions can occur in, but are not specific to, elderly people. They are often found in the cardia of the stomach but can be found throughout the gastrointestinal tract.49 As the lesions are surrounded by normal musosa, endoscopic diagnosis is challenging unless performed at the time of active bleeding. Patients can, therefore, present with recurrent hemorrhage before a diagnosis is made.49 Either endoscopic or angiographic therapies can be applied.

Top

Lower gastrointestinal bleeding

The incidence of lower gastrointestinal bleeding (LGIB) is less than that of UGIB but also increases with age. One retrospective study estimated a greater than 200-fold increase in the incidence of LGIB from the third to the ninth decade of life.50 The mean age of people with acute LGIB ranges from 63 to 77 years.51 This trend is driven by the age-associated increase in the incidence of diverticular hemorrhage, which is the most common cause of LGIB in elderly people.1, 52 Compared with younger patients with LGIB, elderly people with LGIB incur longer hospital stays and greater health-care costs.53

Clinical presentation, initial evaluation, and immediate management

In patients with LGIB, taking a detailed history is critical. The history can distinguish among possible etiologies of LGIB in elderly patients (Table 4). Characterization of the color, volume, and consistency of blood passed per rectum can help to localize a bleeding site. Passage of formed stool mixed with blood is suggestive of an anorectal bleeding source. The passage of melena is not exclusive to UGIB and can be seen when patients have proximal colonic bleeding sources. Cramping abdominal pain or abdominal tenderness suggests colitis.

Table 4 Causes of lower gastrointestinal bleeding in the elderly.
Table 4 - Causes of lower gastrointestinal bleeding in the elderly.
Full tableFigures & Tables indexDownload PowerPoint slide (76K)

The onset and pace of LGIB can provide additional clues as to the underlying etiology. The abrupt onset of painless, hemodynamically significant hematochezia (passage of bright red blood per rectum) is characteristic of diverticular hemorrhage. Bleeding from angiodysplasia, on the other hand, can vary in presentation from overt hemorrhage to chronic occult blood loss.54

Anoscopy is recommended, in addition to a digital rectal examination, as a standard component of the initial physical examination in patients with LGIB to assess for anorectal pathology.55 Nasogastric intubation and lavage can be useful, not only to evaluate for UGIB presenting as rectal hemorrhage, but also to rapidly purge the bowel in preparation for urgent colonoscopy.

The initial management of patients with LGIB should, in the same way as for patients with UGIB, emphasize hemodynamic stabilization. In contrast to UGIB, however, in which EGD is typically the diagnostic and therapeutic modality of choice, the range of diagnostic modalities available for patients with LGIB includes endoscopic, noninvasive radiologic, and angiographic options.

Diverticulosis

Most people who have colonic diverticulosis are asymptomatic and have an uncomplicated clinical course, but a small number of patients experience diverticular hemorrhage, which results from rupture of a vasa recta artery located in the wall of the diverticulum.56 Angiographic data, now several decades old, have fueled the clinical belief that hemorrhage originates from right-sided diverticula;57 however, diverticula in any colonic segment can be a source of bleeding. In most cases, and in more than 80% of cases in elderly people, diverticular hemorrhage ceases spontaneously.58 Most cases of diverticular hemorrhage in elderly people can be managed nonsurgically, and overall mortality is less than 5%.58 Massive or recurrent hemorrhage, however, can require a targeted intervention.

Flexible sigmoidoscopy or colonoscopy can be used to deliver endoscopic therapy, which includes epinephrine injection, thermal application,59 or band ligation.56 Skill and familiarity with epinephrine injection and a thermal technique should be in the armamentarium of all practicing endoscopists, whereas certain techniques, such as inversion and band ligation of a bleeding diverticulum, might be best reserved for experienced endoscopists at referral centers. In addition to the achievement of therapeutic hemostasis, the bleeding site should be tattooed to enable its ready localization in the event that repeat endoscopy or surgery is required.

Successful endoscopic therapy is contingent on visualization and identification of an actively bleeding diverticulum. This is especially difficult in patients with numerous diverticula throughout the colon, as is often the case. Additional hurdles to endoscopic therapy include impaired visualization caused by ongoing hemorrhage or inadequate bowel preparation. Moreover, hemorrhage can cease spontaneously before endoscopy. It is advisable to give a rapid purgative bowel preparation to patients who present with suspected acute diverticular hemorrhage to maximize the diagnostic and therapeutic yield.

Radiologic studies can be used as first-line investigative tools, or to evaluate patients who have ongoing or recurrent hemorrhage when endoscopic evaluation is unsuccessful. Technetium-labeled red blood cell scintigraphy can theoretically detect active gastrointestinal bleeding that occurs at rates as low as 0.1ml/min, although the sensitivity of the test may vary among institutions.55 Use of multidetector CT angiography is an option available at selected centers. Some institutions require a positive scintigraphy result as a prelude to the performance of mesenteric angiography, a technique that can both locate a bleeding source and deliver targeted therapy (including intra-arterial vasopressin infusion or embolization). The often intermittent nature of diverticular hemorrhage means that scintigraphy and angiography can fail to detect the culprit lesion in patients with diverticular hemorrhage. Among patients with LGIB and a negative scintigraphy result, up to 25% will experience recurrent bleeding.60

In a randomized, prospective study that has compared investigative modalities for acute LGIB, 100 patients with LGIB were randomly allocated to urgent colonoscopy or to technetium-labeled red blood cell scintigraphy followed by angiography; 67% of the patients were classified as having definite or presumptive diverticular hemorrhage.61 Colonoscopic therapy was delivered to 17 patients, whereas angiographic therapy was performed in 10 patients. There were no differences between the groups for key outcomes, which included early or late rebleeding rates, transfusion requirements, or the need for emergency surgery.61

Surgery is necessary in a minority of elderly patients who have persistent or refractory diverticular hemorrhage.58 A critical decision is whether a presumptive bleeding source can be identified with sufficient confidence to permit segmental colonic resection. This identification can be a challenge in patients who have pan-colonic diverticular disease, in whom subtotal colectomy is an option. Emergency colonic resection confers higher mortality than elective surgery.62 An advanced age also seems to be associated with increased mortality in patients who undergo emergency colectomy for LGIB. One retrospective study reported 37% mortality for patients 70 years of age or older, compared with 21% for patients aged less than 70 years, among a subset of patients with LGIB who underwent emergency colectomy with primary anastomosis.63

Ischemic colitis

Colitis of any etiology, including infectious colitis and idiopathic IBD, can be a source of gastrointestinal blood loss as a result of mucosal injury and sloughing.64 The principal mechanism that underlies acute ischemic colitis—nonocclusive mesenteric ischemia—occurs primarily in the elderly. Episodes of ischemic colitis result from transient colonic hypoperfusion and can be precipitated by dehydration. Additional risk factors include small-vessel vascular disease (including vasculitis) and the use of certain medications, such as diuretics or vasoactive agents. The clinical presentation often includes cramping abdominal pain. Abdominal CT can demonstrate colonic wall thickening, and colonoscopy can demonstrate edema, friable mucosa, and submucosal hemorrhage in affected colonic segments.

Ischemic colitis is distinct from both chronic mesenteric ischemia (intestinal angina) and acute mesenteric ischemia caused by mesenteric arterial or venous occlusion. The incidence of ischemic colitis increases with age, and the disorder is associated with the presence of medical comorbidities, including hypertension64 and dialysis-dependent chronic kidney disease.65

The differential diagnosis of ischemic colitis includes infectious colitis, IBD, and drug-associated colitis.64 Bleeding from ischemic colitis is rarely hemodynamically significant, and supportive care leads to full recovery in most patients. Patients with recurrent episodes can develop subacute colitis or stricture in the affected colonic segment. The presence of tachycardia, hypotension, or fever might suggest the presence of transmural colonic injury and impending bowel infarction. Death from ischemic colitis is uncommon, but mortality rate seems to increase with age and is associated with the presence of cardiovascular or cerebrovascular disease.66

Other causes

Angiodysplasias

Angiodysplasias (also referred to as angioectasias) are a common source of gastrointestinal bleeding in the elderly. Although angiodysplasias are commonly located in the small intestine and colon,67 and most are small and clinically inconsequential, an important minority located in the right side of the colon can cause massive hemorrhage.68 An association between bleeding caused by angiodysplasias and acquired von Willebrand disease associated with aortic stenosis (Heyde syndrome) has been proposed but is controversial.69 Affected individuals can experience either chronic occult blood loss or massive hemorrhage.54, 61 Endoscopic therapy can be targeted to actively bleeding lesions, but effective endoscopic therapy is difficult to achieve in patients who have numerous lesions or small-bowel lesions that are inaccessible to standard endoscopy. Single, large, bleeding angiodysplasias in the cecum can be particularly amenable to endoscopic, angiographic, or surgical therapy.

Hemorrhoids

Although the prevalence of hemorrhoids seems to decrease after age 65 years,70 the association between hemorrhoids and constipation that is seen in younger patients persists in the elderly.71 Hemorrhoids can be detected during retroflexion in the rectum during an endoscopic examination, although anoscopy, which can easily be performed at the bedside in the early evaluation of rectal bleeding, might be required for optimal visualization.

Stercoral ulcer and solitary rectal ulcer syndrome

Stercoral ulcer and solitary rectal ulcer syndrome can result in massive rectal hemorrhage. Most patients with a hemorrhagic rectal ulcer are older than 60 years of age.72 Stercoral ulcers result from mucosal trauma caused by hard or impacted stool in the rectum or from a foreign body such as a rectal tube in a hospitalized patient. Solitary rectal ulcer syndrome is thought to result from rectal mucosal prolapse, also as a result of constipation and straining. When endoscopic stigmata of hemorrhage are identified in a rectal ulcer, hemostatic techniques analogous to those used in the endoscopic treatment of peptic ulcers can be applied.

Colonic neoplasms and post-polypectomy hemorrhage

Colonic neoplasms usually cause occult gastrointestinal blood loss, but patients with colonic neoplasms can also present with overt bleeding. Premalignant neoplasms (tubular adenomas) can be removed by polypectomy, and post-polypectomy hemorrhage, either early or delayed, occurs after a small fraction of polypectomies. One prospective analysis of nearly 10,000 polypectomies identified age greater than 65 years as a risk factor for post-polypectomy bleeding.73

Radiation proctitis

Radiation proctitis occurs in people who have undergone radiation therapy for prostate, genitourinary, or gynecologic malignancies. It can develop years after treatment has ended, and can result in either overt rectal hemorrhage or chronic transfusion-dependent blood loss. Argon plasma coagulation,74 formalin application,75 sucralfate enemas,76 and hyperbaric oxygen therapy77 have been described as effective treatments for hemorrhagic radiation proctitis; however, comparative controlled data are limited, and it is not known which of these modalities is the most effective.

Top

Obscure gastrointestinal bleeding

Obscure gastrointestinal bleeding is defined as bleeding from a source that remains undetermined after EGD and colonoscopy. Sources of obscure overt hemorrhage can include lesions that were missed78 or not actively bleeding at the time of initial endoscopy. In instances of acute, overt hemorrhage, repeat endoscopy, scintigraphy, or angiography can be considered. Alternatively, bleeding can originate from a small bowel source beyond the reach of standard endoscopic examination. New endoscopic technologies have an increasing role to play in the investigation of small bowel bleeding. Two such techniques are video capsule endoscopy and double-balloon endoscopy.

Video capsule endoscopy

Video capsule endoscopy (VCE, also known as wireless capsule endoscopy) has a diagnostic yield of 58–80% in patients with obscure gastrointestinal bleeding.79, 80, 81 In head-to-head comparisons, the yield of VCE is superior to push enteroscopy, small bowel enteroclysis, and mesenteric angiography.79 Angiodysplasia is the culprit lesion most commonly identified by VCE.79 In elderly people with impaired swallowing or delayed gastric emptying, the video capsule can be placed beyond the pylorus during EGD. VCE seems to be safe to conduct in people with permanent pacemakers82 and implantable defibrillators.83

Double-balloon enteroscopy

Double-balloon enteroscopy (DBE) makes use of a modified enteroscope and overtube; both parts are equipped with a balloon that is inflated and deflated in sequential fashion to enable deep small-bowel exploration via either oral or rectal intubation. The diagnostic yield of DBE is 60–76% in patients with obscure gastrointestinal bleeding.80, 84, 85, 86 By use of DBE, the most commonly identified small bowel lesions in older patients are angiodysplasias; small bowel tumors are relatively more common in younger people.84, 86 Even though the procedure is time consuming and is more invasive than VCE, it allows for the possibility of therapeutic intervention. The age of patients in published studies of DBE ranges from 48 to 57 years,84, 85, 86 and the technique has not been performed extensively, nor its safety established, in elderly patients.

Top

Special topics

NSAIDs and gastrointestinal bleeding

Aspirin increases the risk of UGIB in a dose-dependent fashion.87 The addition of a non-aspirin antiplatelet agent or NSAID to regular aspirin use has at least an additive impact on this increased risk.88, 89, 90 An increasing proportion of patients with UGIB report taking aspirin or other antithrombotic drugs.91 Taking aspirin and other NSAIDs also seems to confer an increased risk for LGIB, in particular diverticular hemorrhage.92, 93

The increased risk of UGIB and LGIB associated with the use of aspirin and other NSAIDs is particularly pertinent to elderly people, many of whom require one or more of these medications for management of chronic cardiovascular or rheumatologic conditions. In elderly patients who require aspirin and have risk factors for gastrointestinal bleeding—including a prior history of gastrointestinal bleeding—aspirin should be prescribed in the lowest possible dose needed to achieve cardioprotection. The use of low-dose aspirin (e.g. 100mg per day) does not, however, eliminate the risk of bleeding (adjusted relative risk for UGIB=2.7 in aspirin users compared with nonusers88). The long-term risk of UGIB in elderly patients who take aspirin can be reduced by concurrent administration of a PPI.94

NSAID use might also be a risk factor for nosocomial gastrointestinal bleeding in elderly people. In a study of elderly patients (mean age 82.1 years) hospitalized for hip fracture, prior use of aspirin or a selective cyclo-oxygenase 2 inhibitor, and a history of peptic ulcer disease, were among the risk factors associated with the development of perioperative UGIB.95 Prophylactic gastric acid suppression (with a histamine H2-receptor antagonist or PPI) seems to reduce the frequency of nosocomial UGIB (0.72% in those receiving prophylaxis versus 13.4% in those not receiving prophylaxis) among patients with identified risk factors for perioperative UGIB.95

Endoscopic safety

The relative safety of gastrointestinal endoscopy has been established in an average-risk population, and this has enabled the widespread acceptance of colonoscopy as a screening tool. Acute hemorrhage and intercurrent illness influence the safety of endoscopy in an individual patient. For instance, the mortality associated with EGD has been estimated to be 0.0004%,96 but increases to 0.1% in patients undergoing EGD for evaluation of gastrointestinal hemorrhage.97

Studies of EGD98 or colonoscopy98, 99, 100 carried out in elderly patients report overall procedural success and morbidity similar to that reported for the general population; however, the indication for endoscopy in most patients in these studies was screening or surveillance, and a relatively small percentage underwent endoscopy for evaluation of acute gastrointestinal bleeding.

Factors likely to influence the morbidity and mortality associated with endoscopy in an elderly cohort with acute gastrointestinal bleeding include the acuity and severity of hemorrhage and the presence of comorbid conditions. The presence of certain anatomic lesions such as a Zenker's diverticulum or cervical spine osteophytes can increase the risk of perforation during upper gastrointestinal intubation in elderly people.101

Tolerance of sedation is an important consideration in elderly patients, who require lower doses of benzodiazepines than non-elderly patients of equivalent weight and body habitus.102, 103 Elderly patients prescribed benzodiazepines for sedation can be prone to oxygen desaturation during endoscopy.104, 105 Drugs used for sedation and analgesia should be administered at low doses to an elderly patient, with subsequent slow, careful titration. Continuous hemodynamic monitoring, use of pulse oximetry, and suction of oral secretions are necessary to safeguard against aspiration.106

Cognition and the elderly patient

The prevalence of dementia increases with age.51 Even in the absence of clinical dementia, sensory impairment can affect cognition, and in an individual patient these factors can influence both the course of illnesses and interactions with health-care professionals. The clinician must pay particular attention to these issues during the evaluation, management, and aftercare phases of illness.

Although initial evaluation begins with clinical history taking, several factors can influence an elderly patient's ability to provide an accurate history at presentation. Relative cerebral hypoperfusion as a result of hemorrhage and hypovolemia can impair cognition in elderly patients. In other cases, sensory impairment can limit an individual's ability to provide a detailed history. The ability to describe the onset, acuity, and volume of rectal bleeding, for instance, requires intact vision and attention to the physical environment. When dementia or impaired memory prevents direct history taking from an elderly patient, family members or the patient's primary care physician can serve as valuable sources of information.

Dementia or a clouded sensorium can similarly influence the individual's ability to participate in decision making, provide informed consent for endoscopy, and generate a plan and goals of care. In these instances, either advanced directives or a formal health-care proxy can be instrumental. As with all patients, the risk–benefit analysis of an invasive medical procedure such as endoscopy must be performed on an individual basis for each elderly patient.

After stabilization and acute control of hemorrhage, an elderly patient's cognitive status must be carefully monitored; this approach is particularly relevant for patients who have received sedative medications for endoscopy. Certain elderly people depend heavily on regular environmental cues and can be prone to delirium during a hospital stay. An elderly patient's ability to care independently for himself or herself, or otherwise access appropriate support, will determine whether the patient can safely return home immediately after hospitalization.

Top

Conclusions

Gastrointestinal bleeding occurs frequently in elderly people, and aspirin and antiplatelet therapy are associated with an increased risk of both UGIB and LGIB in this population. Knowledge of the gastrointestinal lesions likely to affect elderly patients, thorough history taking, and a complete physical examination should help to determine whether the bleeding source is from the upper or lower gastrointestinal tract. In a patient with acute gastrointestinal bleed-ing, hemodynamic stabilization should always precede endoscopic evaluation. In elderly patients with acute hemorrhage, urgent endoscopic evaluation can be undertaken, provided a risk-benefit assessment has been performed and informed consent has been obtained.

The most common source of UGIB in elderly people is peptic ulcer disease. EGD is the investigative study of choice for the evaluation of UGIB, and can be performed safely in elderly people. EGD can identify a bleeding source, deliver targeted hemostatic therapy, and provide estimates of the likelihood of recurrent bleeding in patients with peptic ulcer disease. Pharmacologic gastric acid suppression promotes ulcer healing and reduces the risk of rebleeding.

Diverticular hemorrhage is the most common source of LGIB in the elderly. In an elderly patient with LGIB, urgent colonoscopy should be performed after administration of a purgative bowel preparation. If colonoscopy is suggestive of active bleeding but fails to isolate a bleeding source, mesenteric angiography can be performed. If colonoscopy does not detect active bleeding, the patient should be maintained on a clear liquid diet. If there is evidence of acute (within 24–48h) recurrent hemorrhage, colonoscopy can be repeated or angiography performed. Early surgical consultation is advisable in any patient who has massive or recurrent LGIB, irrespective of age.

In patients with a suspected small-bowel bleeding source, VCE has emerged as the initial diagnostic study of choice. Review and interpretation of VCE findings should be feasible within 24h of image acquisition, thereby enabling timely management in cases of ongoing acute hemorrhage.

The care of elderly people with gastrointestinal bleeding can involve internists and intensivists, as well as gastrointestinal, surgical, and radiologic consultants. A coordinated approach to diagnosis and management should serve to optimize favorable outcomes in this vulnerable patient population.

Key points

  • In elderly patients who have gastrointestinal bleeding, immediate attention should focus on hemodynamic stabilization, followed by diagnostic evaluation to identify the bleeding source
  • The outcome of gastrointestinal bleeding in elderly patients is influenced by the nature of the bleeding lesion, presence of medical comorbidities, and use of anticoagulant or antiplatelet therapy
  • Peptic ulcer disease is the most frequent source of upper gastrointestinal bleeding in elderly patients; gastric acid suppression with a PPI and endoscopic hemostatic therapy reduce the rates of rebleeding and morbidity
  • Diverticulosis is the most frequent source of lower gastrointestinal bleeding in elderly patients, and urgent colonoscopy can be undertaken to identify a bleeding site and provide hemostasis; emergency colectomy is associated with high morbidity and mortality in elderly patients who have persistent or recurrent diverticular bleeding
  • Video capsule endoscopy and double-balloon enteroscopy offer expanded options for diagnosis and therapy of small-bowel bleeding sources
  • Endoscopy can be performed safely and effectively in elderly patients with gastrointestinal bleeding; however, risks and benefits must be considered carefully in each case, with particular attention paid to the tolerance of procedural sedation

Acknowledgments

Charles P Vega, 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 Medscape-accredited continuing medical education activity associated with this article.

References

  1. Kaplan RC et al. (2001) Risk factors for hospitalized gastrointestinal bleeding among older persons. J Am Geriatr Soc 49: 126–133 | Article | PubMed | ISI | ChemPort |
  2. Targownik LE and Nabalamba A (2006) Trends in management and outcomes of acute nonvariceal upper gastrointestinal bleeding: 1993–2003. Clin Gastroenterol Hepatol 4: 1459–1466 | Article | PubMed |
  3. Thomopoulos KC et al. (2004) Changes in aetiology and clinical outcome of acute upper gastrointestinal bleeding during the last 15 years. Eur J Gastroenterol Hepatol 16: 177–182 | Article | PubMed |
  4. Van Leerdam ME et al. (2003) Acute upper GI bleeding: did anything change? Time trend analysis of incidence and outcome of acute upper GI bleeding between 1993/1994 and 2000. Am J Gastroenterol 98: 1494–1499 | Article | PubMed | ChemPort |
  5. Rockall TA et al. (1995) Incidence of and mortality from acute upper gastrointestinal haemorrhage in the United Kingdom. Steering Committee and members of the National Audit of Acute Upper Gastrointestinal Hemorrhage. BMJ 311: 222–226 | PubMed | ChemPort |
  6. Longstreth GF (1995) Epidemiology of hospitalization for acute upper gastrointestinal hemorrhage: a population-based study. Am J Gastroenterol 90: 206–210 | PubMed | ISI | ChemPort |
  7. Katschinski B et al. (1994) Prognostic factors in upper gastrointestinal bleeding. Dig Dis Sci 39: 706–712 | Article | PubMed | ChemPort |
  8. Clason AE et al. (1986) Clinical factors in the prediction for further haemorrhage or mortality in acute upper gastrointestinal haemorrhage. Br J Surg 73: 985–987 | Article | PubMed | ChemPort |
  9. Cappell MS and Nadler SC (1995) Increased mortality of acute upper gastrointestinal bleeding in patients with chronic obstructive pulmonary disease. A case controlled, multiyear study of 53 consecutive patients. Dig Dis Sci 40: 256–262 | Article | PubMed | ChemPort |
  10. Yamaguchi Y et al. (2003) Endoscopic hemostasis: safe treatment for peptic ulcer patients aged 80 years or older? J Gastroenterol Hepatol 18: 521–525 | Article | PubMed |
  11. Hasselgren G et al. (1998) Risk factors for rebleeding and fatal outcome in elderly patients with acute peptic ulcer bleeding. Eur J Gastroenterol Hepatol 10: 667–672 | Article | PubMed | ChemPort |
  12. Baradarian R et al. (2004) Early intensive resuscitation of patients with upper gastrointestinal bleeding decreases mortality. Am J Gastroenterol 99: 619–622 | Article | PubMed |
  13. Cooper BT et al. (1988) Acute gastrointestinal haemorrhage in patients aged 80 years or more. Q J Med 69: 765–774
  14. Segal WN and Cello JP (1997) Hemorrhage in the upper gastrointestinal tract in the older patient. Am J Gastroenterol 92: 42–46 | PubMed | ChemPort |
  15. Macleod IA and Mills PR (1982) Factors affecting the probability of further haemorrhage after acute upper gastrointestinal haemorrhage. Br J Surg 69: 256–258 | Article | PubMed | ChemPort |
  16. Permutt RP and Cello JP (1982) Duodenal ulcer disease in the hospitalized elderly patient. Dig Dis Sci 27: 1–6 | Article | PubMed | ChemPort |
  17. Antler AS et al. (1981) Gastrointestinal bleeding in the elderly. Morbidity, mortality, and cause. Am J Surg 142: 271–273 | Article | PubMed | ChemPort |
  18. Ohmann C et al. (2005) Time-trends in the epidemiology of peptic ulcer bleeding. Scand J Gastroenterol 40: 914–920 | Article | PubMed |
  19. Higham J et al. (2002) Recent trends in admission and mortality due to peptic ulcer in England: increasing frequency of hemorrhage among older subjects. Gut 50: 460–464 | Article | PubMed | ISI | ChemPort |
  20. Kang JY et al. (2006) Recent trends in hospital admission and mortality rates for peptic ulcer in Scotland 1982–2002. Aliment Pharmacol Ther 24: 65–79 | Article | PubMed | ChemPort |
  21. Green FW Jr et al. (1978) Effect of acid and pepsin on blood coagulation and platelet aggregation. A possible contributor to prolonged gastrointestinal hemorrhage. Gastroenterology 74: 38–43 | PubMed | ISI | ChemPort |
  22. Vreeburg EM et al. (2001) Enhanced mucosal fibrinolytic activity in gastroduodenal ulcer haemorrhage and the beneficial effect of acid suppression. Aliment Pharmacol Ther 15: 639–646 | Article | PubMed | ChemPort |
  23. Khuroo MS et al. (1997) A comparison of omeprazole and placebo for bleeding peptic ulcer. N Engl J Med 336: 1054–1058 | Article | PubMed | ISI | ChemPort |
  24. O'Connell MB et al. (2005) Effects of proton pump inhibitors on calcium carbonate absorption: a randomized crossover trial. Am J Med 118: 778–781 | Article | PubMed | ISI | ChemPort |
  25. Valuck RJ and Ruscin JM (2004) A case-control study on adverse effects: H2 blocker or proton pump inhibitor use and risk of Vitamin B12 deficiency in older adults. J Clin Epidemiol 57: 422–428 | Article | PubMed |
  26. Yang YX et al. (2006) Long-term proton pump inhibitor therapy and risk of hip fracture. JAMA 296: 2947–2953 | Article | PubMed | ISI | ChemPort |
  27. Rockey DC (2006) Gastrointestinal bleeding. In Sleisenger and Fordtran's Gastrointestinal and Liver Disease: Pathophysiology/Diagnosis/Management, edn 2, 264 (Eds Feldman M et al.) Philadelphia: Elsevier Saunders
  28. Choudari CP et al. (1995) Age-related mortality in patients treated endoscopically for bleeding peptic ulcer. Gastrointest Endosc 41: 557–560 | Article | PubMed | ChemPort |
  29. Lau JY et al. (2000) Effect of intravenous omeprazole on recurrent bleeding after endoscopic treatment of bleeding peptic ulcers. N Engl J Med 343: 310–306 | Article | PubMed | ISI | ChemPort |
  30. Chow LW et al. (1998) Risk factors for rebleeding and death from peptic ulcer in the very elderly. Br J Surg 85: 121–124 | Article | PubMed | ChemPort |
  31. Bulut OB et al. (1996) Acute surgical treatment of complicated peptic ulcers with special reference to the elderly. World J Surg 20: 574–577 | Article | PubMed | ChemPort |
  32. Merkel C et al. (2000) Prognostic indicators of risk for first variceal bleeding in cirrhosis: a multicenter study in 711 patients to validate and improve the North Italian Endoscopic Club (NIEC) index. Am J Gastroenterol 95: 2915–2920 | Article | PubMed | ISI | ChemPort |
  33. Chalasani N et al. (2003) Improved patient survival after acute variceal bleeding: a multicenter, cohort study. Am J Gastroenterol 98: 653–659 | Article | PubMed |
  34. El-Serag HB and Everhart JE (2000) Improved survival after variceal hemorrhage over an 11-year period in the Department of Veterans Affairs. Am J Gastroenterol 95: 3566–3573 | Article | PubMed | ChemPort |
  35. Hosking SW et al. (1989) Management of bleeding varices in the elderly. BMJ 298: 152–153 | PubMed | ChemPort |
  36. Villaneuva C et al. (2005) Hemodynamic effects of terlipressin and high somatostatin dose during acute variceal bleeding in nonresponders to the usual somatostatin dose. Am J Gastroenterol 100: 624–630 | Article | PubMed |
  37. Chatila R et al. (2000) Local arterial vasoconstriction induced by octreotide in patients with cirrhosis. Hepatology 31: 572–576 | Article | PubMed | ISI | ChemPort |
  38. Villanueva C et al. (2006) A randomized trial comparing ligation and sclerotherapy as emergency endoscopic treatment added to somatostatin in acute variceal bleeding. J Hepatol 45: 560–567 | Article | PubMed |
  39. Primignani M et al. (2000) Natural history of portal hypertensive gastropathy in patients with liver cirrhosis. The New Italian Endoscopic Club for the study and treatment of esophageal varices (NIEC). Gastroenterology 119: 181–187 | Article | PubMed | ChemPort |
  40. Nankhonya JM et al. (1997) Acute upper gastrointestinal hemorrhage in older people: a prospective study in two neighboring districts. J Am Geriatr Soc 45: 752–754 | PubMed | ChemPort |
  41. Zimmerman J et al. (1997) Esophagitis is a major cause of gastrointestinal hemorrhage in the elderly. Scand J Gastroenterol 32: 906–909 | PubMed | ChemPort |
  42. Johnson DA and Fennerty MB (2004) Heartburn severity underestimates erosive esophagitis severity in elderly patients with gastroesophageal reflux disease. Gastroenterology 126: 660–664 | Article | PubMed |
  43. Stefanidis I et al. (2006) Gastric antral vascular ectasia (watermelon stomach) in patients with ESRD. Am J Kidney Dis 47: e77–e82 | Article | PubMed |
  44. Kwan V et al. (2006) 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 | Article | PubMed | ISI | ChemPort |
  45. Bergqvist D et al. (1996) Secondary aortoenteric fistulate—changes from 1973 to 1993. Eur J Endovasc Surg 11: 425–428 | Article | ChemPort |
  46. Baril DT et al. (2006) Experience with endovascular abdominal aortic aneurysm repair in nonagenarians. J Endovasc Ther 13: 330–337 | Article | PubMed |
  47. Verhey P et al. (2006) Successful endovascular treatment of aortoenteric fistula secondary to eroding duodenal stent. J Vasc Interv Radiol 17: 1345–1348 | PubMed |
  48. Cendan JC et al. (2004) Twenty-one cases of aortoenteric fistula: lessons for the general surgeon. Am Surg 70: 583–587 | PubMed |
  49. Reilly HF III and al-Kawas FH (1991) Dieulafoy's lesion. Diagnosis and management. Dig Dis Sci 36: 1702–1707 | Article | PubMed |
  50. Longstreth GF (1997) Epidemiology and outcome of patients hospitalized with acute lower gastrointestinal hemorrhage: a population-based study. Am J Gastroenterol 92: 419–424 | PubMed | ChemPort |
  51. Farrell JJ and Friedman LS (2000) Gastrointestinal bleeding in older people. Gastroenterol Clin North Am 29: 1–36 | Article | PubMed | ChemPort |
  52. Rios A et al. (2005) Acute lower gastrointestinal hemorrhages in geriatric patients. Dig Dis Sci 50: 898–904 | Article | PubMed |
  53. Comay D and Marshall JK.(2002) Resource utilization for acute lower gastrointestinal hemorrhage: the Ontario GI bleed study. Can J Gastroenterol 16: 7677–7682
  54. Richter JM et al. (1989) Angiodysplasia: natural history and efficacy of therapeutic intervention. Dig Dis Sci 34: 1542–1546 | Article | PubMed | ChemPort |
  55. Rockey DC (2006) Lower gastrointestinal bleeding. Gastroenterology 130: 165–171 | Article | PubMed |
  56. Farrell JJ et al. (2003) Treatment of bleeding colonic diverticula by endoscopic band ligation: an in vivo and ex vivo pilot study. Endoscopy 35: 823–829 | Article | PubMed | ChemPort |
  57. Casarella WJ et al. (1972) Right-sided colonic diverticula as a cause of acute rectal hemorrhage. N Engl J Med 286: 450–453 | PubMed | ChemPort |
  58. Bokhari M