This module will focus mainly on the evaluation and management of massive GIB which is defined for the purposes of this module as bleeding of the gastrointestinal tract associated with hemodynamic instability, acute anemia, and/or the need for blood transfusion. Bleeding from a source proximal to the ligament of Treitz is classified as upper gastrointestinal bleeding (UGIB); bleeding from a source distal to this ligament is classified as lower gastrointestinal bleeding (LGIB).
The classic presentation of GIB varies based on the location and rate of bleeding. UGIB classically presents as hematemesis, coffee-ground emesis, and/or melena. LGIB classically presents with hematochezia or bright red blood per rectum (BRBPR). However, it is important to be aware that approximately 10% of melena comes from a lower GI source and 10% of hematochezia comes from an upper GI source. The table below summarizes how the patient’s description of bleeding correlates to the likely location.
|Patient Description||Medical Terminology||Location of Bleeding|
|Dark or brown vomitus||Coffee-ground emesis||UGIB|
|Black or tarry stools||Melena||Probable UGIB/LGIB possible|
|Bloody stool (bright red or maroon)||Hematochezia||Probable LGIB /UGIB possible|
|Passing red blood, blood-streaked stool, blood on toilet paper, or blood in toilet bowl||BRBPR||LGIB
*UGIB is possible if BRBPR accompanied by hemodynamic instability
The specific lesion causing GIB may not be identifiable in the emergency department (ED). Patients presenting with GIB should be asked about any prior presentations as recurrent bleeding is often due to the same lesion. Statistically, the diseases to highly suspect for patients with massive UGIB are peptic ulcer diseases (PUD) and esophageal varices. Massive LGIB is typically due to diverticulosis or angiodysplasias. A guided history can help refine the differential diagnosis. The table below highlights historical items associated with specific diseases:
|Location of Bleeding||Cause of Bleeding||Historical Clues|
|UGIB||Peptic Ulcer||Known PUD, chronic epigastric pain, heavy alcohol use, gastrotoxic medications (e.g., NSAIDS, ASA, or steroids)|
|UGIB||Esophageal Varices||Severe liver disease, alcoholism, hemorrhoids on physical exam (external evidence of possible portal hypertension)|
|UGIB||Aortoentertic fistula||History of AAA repair|
|UGIB||Mallory-Weiss Tear||Vomiting or retching preceding hematemesis|
|LGIB||Diverticulosis||Known diverticular disease, massive and painless BRBPR|
All patients should be asked about symptoms suggestive of large blood loss such as dizziness, shortness of breath, or chest pain. It is also important to quickly determine whether a patient with GIB is on any anticoagulant medications so that they can be rapidly reversed if needed.
The initial goal of the physical exam is to rapidly identify patients in hemorrhagic shock as manifested by ill appearance, tachycardia, tachypnea, hypotension, depressed mental status, or cool and clammy skin. In hemodynamically stable patients, orthostatic vital signs can be obtained to assess for significant blood loss as suggested by a systolic blood pressure drop of more than 10mmHg or a pulse increase of more than 20 beats/minute. However, it is important to understand limitations of orthostatic vital signs in different age groups. The elderly may be on beta-blockers which can blunt a tachycardic response or conversely be orthostatic at baseline due to autonomic insufficiency. Young and healthy patients may have abundant cardiovascular reserve and thus may not drop their blood pressure even in the setting of significant blood loss.
The rectal exam plays an essential role in the evaluation of GIB in both characterizing the stool color (brown vs black vs maroon) and the presence of gross or occult blood. The rectal exam should also include an inspection for any masses, hemorrhoids, or fissures. It is also paramount that the physical exam includes close inspection for possible undiagnosed liver disease. Specific signs include scleral icterus, jaundice, hepatomegaly, ascitis, and spider angiomata.
A complete blood count should be obtained to evaluate for the degree of blood loss as well as for excluding thrombocytopenia as a contributing factor. As many patients may have chronic anemia, abnormal cell counts should be compared with previous values. Alternatively, it is important to remember that initial blood counts may be normal early in the disease course because bleeding involves the loss of whole blood and thus a proportionate loss of cells and plasma. The hematocrit subsequently declines due to hemodilution but this requires time for the extravascular fluid to move into the vascular space. For this reason, serial Hgb/HCT measurements are important in assessing the degree of acute blood loss.
Additional labs include coagulation studies to detect the presence of a coagulopathy. Liver function tests can be important in diagnosing or monitoring liver disease. GIB may also cause an elevated BUN due to degraded blood being absorbed in the gastrointestinal tract. A BUN/Cr ratio greater than or equal to 36 is suggestive of UGIB but volume depletion or renal insufficiency must also be considered.
A type and screen should be obtained on all patients with significant GIB. Crossmatched blood should be immediately ordered if the need for blood transfusion is clearly indicated or highly anticipated.
In an elderly patient or one with known coronary artery disease (CAD), cardiac markers are important in excluding a myocardial infarction (MI) precipitated by anemia. Patients who are having a MI in the setting of gastrointestinal bleeding may lack typical angina symptoms and be without chest pain or shortness of breath.
In the setting of massive GIB, nasogastric aspiration with saline lavage is used to assess the degree and location of bleeding. In a patient who presents with frank hematemesis, a persistently bloody aspirate despite saline lavage may help the physician anticipate the patient’s need for blood transfusion, emergent esophagoduoenoscopy (EGD), and ICU admission. In patients with brisk hematochezia, it can be difficult to ascertain whether the source is the upper or lower GI tract. If a bloody aspirate is obtained, then appropriate management for UGIB can be pursued. If the nasogastric aspirate contains copious bile but no blood, then an upper GI source is unlikely. However, if the aspirate is both non-bloody and non-bilious then the location remains undetermined as nasogastric aspiration may miss bleeding between a closed pylorus and the ligament of Treitz.
Due to the pain and iatrogenic vomiting associated with nasogastric tube insertion and its questionable accuracy, the routine use of nasogastric aspiration has been called into question. Opponents cite evidence that physicians are poor at determining the presence or absence of active bleeding when they rely on nasogastric aspiration. In patients with hematochezia or melena without hematemesis, nasogastric aspiration has been found to have poor diagnostic performance in excluding an upper GI source. In addition, there is evidence that challenges the traditional notion that gastric lavage is important in improving endoscopic visibility by evacuating blood in the stomach. Overall, opponents argue that nasogastric lavage is a painful procedure that provides unreliable information and that hemodynamic and laboratory values should serve as the basis of routine evaluation. (See references below for further reading on this debate)
In the case of bright red blood in the vomit or per rectum, the patient is clearly bleeding and occult testing is unnecessary. However, sometimes it can be unclear to the naked eye whether the reddish or dark appearance of the vomitus or stool is truly (digested) blood. In addition, iron pills, bismuth containing medications, and even blueberries can cause black stool or “false” melena. In these circumstances occult blood testing can help provide clarification of whether blood is truly present.
It is important to use the correct card and developer based on the sample: Gastroccult® developer comes in a bottle with an orange lid and is used to test gastric contents while the developer for the commonly used Hemoccult®Sensa® kit for stool testing comes in a bottle with a blue lid. The cards for each have a corresponding color stripe to make it easier to use the card with the appropriate developer.
An ECG should be obtained in patients with massive GIB to evaluate for myocardial ischemia. It is especially critical in any patient over the age of 50 who has a history of heart disease, chest pain, trouble breathing, or manifestations of shock.
How do I make the diagnosis?
In the context of massive GIB, the question isn’t whether the patient is bleeding but rather what is the source of bleeding. While definitive diagnosis of the underlying lesion typically occurs outside the ED by an endoscopic procedure, the main diseases to consider for patients presenting with massive UGIB are peptic ulcer disease or varices. Diverticulosis and angiodysplasias are the most common etiologies for massive LGIB.
EGD is the diagnostic test of choice to evaluate the upper GI tract in patients who present with hematemesis or melena. While hematochezia or BRBPR typically represent bleeding from a lower source, in the setting of hemodynamic compromise a brisk UGIB should also be considered as bleeding from the upper GI tract is overall much more common. A nasogastric aspirate of copious bile but no blood in an actively bleeding patient may direct further evaluation towards the lower GI tract. An EGD is the appropriate next diagnostic step if the aspirate contains blood and could still be considered as the next step if the aspirate is inconclusive due to the absence of bile.
Colonoscopy remains the diagnostic test of choice to evaluate LGIB. However, it may not be possible in the setting of massive bleeding. Under these circumstances, mesenteric angiography can be used to both localize and treat the source. However, it is an invasive procedure that can lead to complications including contrast allergy, contrast induced renal failure, and bleeding from arterial puncture. An alternative approach in a patient who continues to actively bleed but has been stabilized is to first proceed with nuclear imaging. A 99mTechnetium labeled red cell scan can detect bleeding at a slower rate than angiography: 0.1 ml/min versus 0.5 mL/min, respectively. It can help confirm that the bleeding is rapid enough to be detectable by angiography and thus merits the risks of an invasive procedure.
As with all emergency department patients, treatment of massive GIB begins with the primary survey and stabilization of the ABCs. This and other important initial actions including blood transfusion and the correction of clotting disorders are discussed in the “Approach to Gastrointestinal Bleeding” self-study module.
Treatment of massive GIB in the emergency department also involves disease-specific therapy. As the exact cause of bleeding is typically not known, these treatment decisions are based on the etiology of prior GIB episodes or a high clinical suspicion of disease. Patients with massive UGIB – especially those with cirrhosis – may be treated empirically for both PUD and varices until a definitive diagnosis can be made.
Patients with known or suspected varices (based on significant liver disease or history of alcoholism) who present with massive GIB should be started on octreotide. Octreotide is a vasoactive protein that causes selective vasoconstriction of the splanchnic vasculature. Blood flow to the upper GI tract is decreased thereby reducing blood loss. It is given as a bolus followed by a drip.
A major complication and cause of mortality following variceal bleeding is bacterial infection. Patients with cirrhosis who present with UGIB should receive prophylactic antibiotics– typically a floroquinolone such as ciprofloxacin- as it reduces the incidence of infection and decreases mortality.
The definitive treatment of choice for bleeding esophageal varices is endoscopic banding. However, in patients whose rate of hemorrhage exceeds efforts at volume resuscitation, esophageal tamponade can be attempted as a stabilizing maneuver. In this procedure a specialized gastric tube – typically the Sengstaken-Blakemore – which has two expanding balloons is placed. One balloon is expanded in the stomach and the other in the esophagus to achieve tamponade.
Peptic ulcer disease
Medical therapy for a bleeding ulcer involves high-dose IV infusion of a proton-pump inhibitor to decrease the acidity of the stomach and enhance clot stability. Pantoprazole and lansoprazole are available in the United States. Like octreotide, it is given as a bolus followed by a drip. EGD is the procedure of choice for confirming and treating a bleeder ulcer. Different endoscopic therapies are available to obtain hemostasis including injection therapy (e.g, saline, epinephrine, sclerosing agents, and tissue adhesives), thermal therapy (e.g., electrocautery or argon plasma coagulation), or mechanical therapy (e.g., endoscopic clips).
Diverticular bleeding – the most common cause of LGIB – stops spontaneously in the majority of patients. Colonoscopy is both the diagnostic and therapeutic procedure of choice and can stop bleeding through thermal or injection therapy. If massive bleeding precludes the use of colonoscopy, mesenteric angiography may achieve hemostasis through injection of intra-arterial vasopression or embolization. Surgery may be necessary when angiography is unavailable or unable to stop the bleeding. Preoperative or intraoperative localization of the bleeding should be attempted to allow limited resection which has better outcomes than a subtotal colectomy.
- Patients with continued brisk bleeding, hemodynamic instability, or significant comorbidities should be admitted to the ICU
- Patients who are hemodynamically stable and have minimal active bleeding may be candidates for a floor admission to a monitored bed
- Patients may go from the ED to interventional radiology for mesenteric angiography or the endoscopy suite for EGD and then on to a floor or ICU bed based on the results of the procedure.
Pearls and Pitfalls
- Orthostatic vital signs may help identify patients with significant blood loss but are limited by false-positives (e.g., autonomic insufficiency) and false-negatives (e.g., beta-blocker use).
- Nasogastric aspiration may miss UGIB occurring between a closed pylorus and the ligament of Treitz
- Patients who are having a myocardial infarction in the setting of GIB may lack typical angina symptoms. An ECG and cardiac markers should be obtained in older patients or those with known CAD who are having a massive GIB.
- Patient with cirrhosis and GIB should receive prophylactic antibiotics
- Endoscopic procedures are both the diagnostic and therapeutic modalities of choice for GIB.
- Written By: Sarah Farris, Duke University, Durham, North Carolina
- Edited By: David Gordon, MD, Duke University, Durham, North Carolina
- Hearnshaw S, Brunskill S, Doree C, Hyde C, Travis S, Murphy MF. Red cell transfusion for the management of upper gastrointestinal haemorrhage. Cochrane Database Syst Rev. 2009 Apr 15;(2):CD006613.
- Capell MS, Friedel D. Initial management of acute upper gastrointestinal bleeding: from initial evaluation up to gastrointestinal endoscopy. Med Clin N Am 2008;92:491-509.
- Gralnek IM, Barkun AN, Bardou M. Management of acute bleeding from a peptic ulcer. N Engl J Med. 2008;359:928-937.
- Lee J, Costantini TW, Coimbra R. Acute lower GI bleeding for the acute care surgeon: current diagnosis and management. Scand J Surg. 2009;98:135-142. free PDF
- Zuccaro G. Management of the adult patient with acute lower gastrointestinal bleeding. American College of Gastroenterology. Practice Parameters Committee. Am J Gastroenterol. 1998;93:1202-8.
- Das AM, Sood N, Hodgin K, Chang L, Carson SS. Development of a triage protocol for patients presenting with gastrointestinal hemorrhage: a prospective cohort study. Crit Care. 2008;12:R57. Epub 2008 Apr 22. free PDF
For additional reading on the debate over nasogastric aspiration:
- Anderson RS and Witting MD. Nasogastric aspiration: A useful tool in some patients with gastrointestinal bleeding. Ann Emerg Med. 2010;55:364-365.
- Pitera A, Sarko J. Just say no: gastric aspiration and lavage rarely provide benefit. Ann Emerg Med. 2010;55:365-366
- Palamidessi N, Sinert R, Falzon L, Zehtabchi S. Nasogastric aspiration and lavage in emergency department patients with hematochezia or melena without hematemesis. Acad Emerg Med. 2010;17:126-32.