As with any emergency, the medical history, physical examination, and diagnostics often must be accomplished simultaneously with resuscitation and stabilization. Factors associated with a high morbidity rate are hemodynamic instability, repeated hematochezia, gross blood on initial rectal examination, initial hematocrit <35%, syncope, nontender abdomen (predictive of severe bleeding), aspirin or nonsteroidal anti-inflammatory drug use (predictive of diverticular hemorrhage), and more than two comorbid conditions.2,3,13
Although most patients will volunteer complaints of hematochezia or melena, signs and symptoms of hypotension, tachycardia, angina, syncope, weakness, or altered mental status can all occur as a result of LGI bleeding.
Ask about previous GI bleeding as well as a history of pain, trauma, ingestion or insertion of foreign bodies, and recent colonoscopies. Weight loss and changes in bowel habits may suggest malignancy. A history of an aortic graft may suggest the possibility of an aortoenteric fistula. Medications, such as salicylates, nonsteroidal anti-inflammatory drugs, and warfarin, increase the risk of LGI bleeding.14,15,16 Ingestion of iron or bismuth can simulate melena, and certain foods, such as beets, can simulate hematochezia. However, stool guaiac testing in those cases will be negative.
Hypotension and tachycardia, or decreased pulse pressure or tachypnea, develop with significant bleeding. However, changes in vital signs may be masked by concurrent medications, such as β-blockers, or medical conditions such as poorly controlled hypertension. Thus, relative tachycardia and hypotension may represent subtle clues to ongoing bleeding. Some patients can tolerate substantial volume losses with minimal or no changes in vital signs.
Cool, pale skin and an increase in capillary refill can be signs of shock. Physical findings of liver disease, as well as petechiae and purpura, suggest an underlying coagulopathy. The abdominal examination may disclose tenderness, masses, ascites, or organomegaly. In patients with LGI bleeding, a lack of abdominal tenderness suggests bleeding from disorders involving the vasculature, such as diverticulosis or angiodysplasia. Inflammatory bowel disorders with LGI bleeding are associated with abdominal tenderness on examination.
Thorough examination of the rectal area may reveal an obvious source of bleeding, such as a laceration, masses, trauma, anal fissures, or external hemorrhoids. A vaginal or urinary source of bleeding mistaken for a GI source will be identified by examination and testing. Perform a digital rectal examination to detect gross blood (either bright red or maroon) and for guaiac testing. Rectal examination can also detect the presence of masses.
Anoscopy can also be performed at the bedside. A source of bleeding such as hemorrhoids can sometimes be elucidated by anoscopy. However, blood originating beyond the level of visualization should raise the suspicion for other causes.10
The most important laboratory tests are the CBC, coagulation studies, and typed and cross-matched blood. Coagulation studies, including prothrombin time, partial thromboplastin time, and platelet count, are of obvious benefit in patients taking anticoagulants or those with underlying hepatic disease. In addition, obtain blood urea nitrogen, creatinine, electrolytes, glucose, and liver function studies. In acute, brisk bleeding, the initial hematocrit level may not reflect the actual amount of blood loss. Bleeding from a source higher in the GI tract may elevate blood urea nitrogen levels through digestion and absorption of hemoglobin.
Obtain an ECG in patients with coronary artery disease. Silent ischemia can occur secondary to the decreased oxygen delivery accompanying significant GI bleeding.
Routine abdominal radiographs are of limited value without specific indications such as perforation, obstruction, or foreign bodies. Similarly, routine admission chest x-rays for patients with acute GI hemorrhage, even those admitted to the intensive care unit, are of limited utility in the absence of known pulmonary disease or abnormal findings on lung examination.17 Barium contrast studies are not helpful and can interfere with subsequent emergent endoscopy or angiography.
The initial diagnostic procedure of choice—angiography, scintigraphy, or endoscopy—depends upon resource ability and consultant preference.18,19,20,21 Angiography can sometimes detect the site of bleeding and help guide surgical management. Moreover, angiography permits therapeutic options such as transcatheter arterial embolization or the infusion of vasoconstrictive agents. However, angiographic diagnosis and therapy require a relatively brisk bleeding rate (at least 0.5 mL/min). Serious complications can also occur with angiography in up to 10% of cases.10
Technetium-labeled red cell scans can also localize the site of bleeding in obscure hemorrhage. Such localization can be used to help determine if angiography or surgery is the optimal approach. Scintigraphy appears more sensitive than angiography and can localize the site of bleeding at as low a rate as 0.1 mL/min. It also has potential value over angiography if bleeding occurs intermittently but requires a minimum of 3 mL of blood to pool.22
Multidetector CT angiography has a sensitivity and specificity of up to 100% and 99%, respectively, for detecting active or recent GI bleeding and is about 93% accurate in determining the site of bleeding.23,24 It can be a useful tool prior to treatment with conventional angiography.