Diagnosing a gastrointestinal (GI) bleed starts with identifying the visible signs, like vomiting blood or noticing blood in your stool, then narrowing down where in the digestive tract the bleeding is coming from. The process typically involves a physical assessment, blood tests, and one or more procedures to visualize the source directly. How quickly this happens and which tools doctors use depends on how severe the bleeding is and whether it’s coming from the upper or lower GI tract.
Recognizing Upper vs. Lower GI Bleeding
The first diagnostic step is figuring out whether the bleeding originates above or below a small ligament that connects the end of the small intestine to the diaphragm. This distinction matters because it determines which tests and procedures come next.
Upper GI bleeding, which involves the esophagus, stomach, or upper small intestine, typically shows up in two ways. Hematemesis is vomiting blood or material that looks like coffee grounds. Melena is black, tarry stool with a strong, distinctive odor, caused by blood being partially digested as it moves through the intestines. Both point to a source higher in the digestive system.
Lower GI bleeding, from the colon or rectum, usually presents as hematochezia: bright red or maroon-colored blood passed during a bowel movement. That said, a very heavy upper GI bleed can also produce bright red blood in the stool simply because the blood moves through the system too fast to be digested. So the color alone isn’t always definitive.
Blood Tests That Help Locate the Source
A basic set of blood tests is drawn early in the evaluation. A complete blood count checks your hemoglobin level, which reveals how much blood you’ve lost. A drop of 1 g/dL or more signals significant bleeding. Doctors also check your blood type in case a transfusion is needed and run clotting tests to see if a bleeding disorder is contributing.
One particularly useful lab value is the ratio of blood urea nitrogen (BUN) to creatinine. When you bleed into the upper GI tract, your body absorbs proteins from the digested blood, which raises BUN levels disproportionately. A BUN-to-creatinine ratio of 30:1 or higher has about 85% specificity for an upper GI source, meaning it’s a reasonably reliable signal that bleeding is coming from above the colon. A ratio of 20:1 is sometimes used as a lower threshold, but it’s less precise. This test helps guide whether an upper endoscopy should be the first procedure performed.
Risk Scoring Before Procedures
Before deciding how urgently to scope a patient, doctors often calculate a risk score. The most widely used tool for upper GI bleeds is the Glasgow-Blatchford Score, which combines blood pressure, heart rate, hemoglobin level, BUN, and clinical details like whether the person has fainted or has a history of liver disease or heart failure. Scores range from 0 to 23, with higher numbers indicating a greater risk of ongoing bleeding or death. A score of 0 or 1 often means you can be safely evaluated as an outpatient. A score of 12 or higher puts you in a high-risk category where hospital admission and closer monitoring are standard.
Upper Endoscopy for Upper GI Bleeds
The primary diagnostic tool for an upper GI bleed is an upper endoscopy (sometimes called an EGD), where a flexible camera is passed through the mouth and into the esophagus, stomach, and the first part of the small intestine. This lets doctors see the bleeding site directly and, in many cases, treat it during the same procedure by cauterizing a vessel, injecting medication, or placing clips.
International guidelines recommend performing this procedure within 24 hours of presentation. For years, there was debate about whether doing it sooner, within 6 hours, would save more lives in high-risk patients. A major randomized trial published in the New England Journal of Medicine settled this: in patients with high Glasgow-Blatchford scores, endoscopy within 6 hours did not reduce 30-day mortality compared to endoscopy performed between 6 and 24 hours. So while the procedure shouldn’t be unnecessarily delayed, rushing it within the first few hours doesn’t appear to improve survival.
Before the endoscopy, a nasogastric lavage is sometimes performed. This involves passing a thin tube through the nose into the stomach and flushing it with saline to clear out blood and clots. It can help confirm the stomach as a bleeding source and improve visibility during the scope, though its routine use has become less common.
Colonoscopy for Lower GI Bleeds
When the bleeding appears to come from the lower GI tract, colonoscopy is the recommended first-line test for patients who are hemodynamically stable, meaning their blood pressure and heart rate are not dangerously compromised. The American College of Gastroenterology and the American College of Radiology both endorse this approach. Colonoscopy requires bowel preparation to clear the colon, which takes time but dramatically improves the ability to identify the source.
Colonoscopy is also used when a patient has melena (black stool suggesting an upper source) but the upper endoscopy comes back normal. In that scenario, the bleed may actually be in the right side of the colon, which can produce darker stool because blood spends more time in the digestive tract before being passed.
For patients who are hemodynamically unstable with massive lower GI bleeding, waiting for bowel prep isn’t practical. In those cases, CT angiography or conventional angiography becomes the first choice.
CT Angiography for Active Bleeding
CT angiography is a contrast-enhanced scan that can detect active bleeding by showing contrast material leaking from a blood vessel into the GI tract. It’s fast, widely available, and doesn’t require bowel preparation, making it especially useful in emergencies. Studies using experimental models have shown that modern multi-detector CT can identify bleeding rates as low as 0.25 mL per minute with 97% sensitivity and 100% specificity. In practical terms, this means the bleeding needs to be actively happening at a modest rate for the scan to catch it. If the bleeding has temporarily stopped, the scan may come back clean.
The advantage of CT angiography is speed and its ability to pinpoint the location of the bleed before a more invasive procedure. If contrast extravasation is identified, the patient can go directly to angiographic embolization, where a radiologist threads a catheter to the bleeding vessel and blocks it.
Nuclear Medicine Scans for Slow Bleeding
When bleeding is too slow or too intermittent for CT angiography to detect, a tagged red blood cell scan can be helpful. In this test, a small sample of your red blood cells is labeled with a radioactive tracer and reinjected. A gamma camera then tracks where those cells accumulate over time. This scan can detect bleeding rates as low as 0.05 to 0.2 mL per minute, significantly lower than the threshold for CT angiography. Reported sensitivity is around 93% with 95% specificity, though results vary.
The key advantage of this scan is patience. Because the tracer stays in the bloodstream for hours, the camera can keep imaging over an extended period, catching intermittent bleeds that come and go. The drawback is that it tells you the general region of bleeding but not the precise vessel, so a positive result usually leads to angiography or endoscopy for definitive localization and treatment.
Capsule Endoscopy for Obscure Bleeding
Sometimes a patient continues to bleed, or has recurrent episodes, but both the upper endoscopy and colonoscopy are normal. This is classified as obscure GI bleeding, and it usually points to a source in the small intestine, the long stretch of bowel between the stomach and colon that standard scopes can’t fully reach.
The next step in this scenario is capsule endoscopy. You swallow a pill-sized camera that takes thousands of images as it travels through the entire small bowel over 8 to 12 hours, transmitting them wirelessly to a recorder you wear on a belt. Multiple international guidelines recommend capsule endoscopy as the first-line test for evaluating the small bowel once upper and lower endoscopy have come back negative. It’s noninvasive, doesn’t require sedation, and has a strong track record for identifying sources like small blood vessel malformations, small tumors, or Crohn’s disease lesions that other tests miss.
Capsule endoscopy isn’t appropriate for everyone. It can’t be used in patients with known bowel obstructions or strictures, because the capsule could get stuck. And unlike traditional endoscopy, it can only diagnose, not treat. If it identifies a bleeding source, a follow-up procedure is needed to address it.