When to Order Imaging for Nonvariceal Upper Gastrointestinal Bleeding: ACR Appropriateness Decoded
When to Order Imaging for Nonvariceal Upper Gastrointestinal Bleeding: ACR Appropriateness Decoded
It’s late in the evening, and a patient presents with hematemesis. Their vital signs are borderline, and you need to act quickly. While esophagogastroduodenoscopy (EGD) is the primary diagnostic and therapeutic tool for upper gastrointestinal (GI) bleeding, it may not be immediately available, may fail to identify a source, or may be unable to achieve hemostasis. In these critical moments, choosing the right imaging study is paramount to localize the bleed and guide further intervention. This article breaks down the American College of Radiology (ACR) Appropriateness Criteria for nonvariceal upper GI bleeding, providing clear, scannable guidance for clinicians to make informed decisions under pressure.
What Does ACR Nonvariceal Upper Gastrointestinal Bleeding Cover?
This ACR guideline focuses on adult patients with suspected or confirmed upper GI bleeding from a source other than esophageal or gastric varices. The recommendations apply to hemodynamically unstable patients who are not candidates for or have failed endoscopic treatment, as well as stable patients with occult or obscure bleeding after a negative endoscopy. Common causes of nonvariceal bleeding include peptic ulcer disease, Mallory-Weiss tears, arteriovenous malformations (AVMs), Dieulafoy lesions, and tumors. The guidance is structured around several distinct clinical variants, from the initial workup before endoscopy to scenarios involving failed endoscopic therapy or postsurgical complications. These criteria do not apply to patients with known or suspected variceal bleeding secondary to portal hypertension, as the management and imaging pathways for that condition are substantially different.
What Imaging Should I Order for Nonvariceal Upper Gastrointestinal Bleeding? Recommendations by Clinical Scenario
The optimal imaging strategy for nonvariceal upper GI bleeding depends heavily on the clinical context, particularly the results and feasibility of endoscopy. The ACR provides specific recommendations for each scenario to guide clinicians toward the most effective study.
For an adult with suspected nonvariceal upper gastrointestinal bleeding where no endoscopy has been performed, multiphase Computed Tomography Angiography (CTA) is the preferred initial imaging modality. The ACR rates CTA abdomen and pelvis without and with IV contrast as Usually appropriate. This study is highly sensitive for detecting active extravasation of contrast, can localize the bleeding site, and can identify the underlying cause, which is crucial for planning subsequent endoscopic or interventional radiology procedures. In this context, visceral arteriography is rated May be appropriate, typically reserved for when a CTA is nondiagnostic but clinical suspicion remains high, or when immediate therapeutic embolization is planned.
When endoscopy confirms a nonvariceal bleed with a clear source, but endoscopic treatment is not possible or bleeding continues, both diagnostic imaging and therapeutic intervention are needed. Here, both Arteriography visceral and CTA abdomen and pelvis without and with IV contrast are rated Usually appropriate. CTA serves to confirm the location of active bleeding and provide a vascular map for the interventional radiologist, while visceral arteriography allows for definitive treatment via embolization in the same session.
In cases where endoscopy confirms bleeding but cannot identify a clear source, the primary goal of imaging is localization. Again, CTA abdomen and pelvis without and with IV contrast is rated Usually appropriate. For intermittent or slower bleeding, a tagged Red Blood Cell (RBC) scan may be considered and is rated May be appropriate, as it can detect bleeding rates as low as 0.1 mL/min over a prolonged imaging period. CT enterography also May be appropriate to evaluate for small bowel sources that are beyond the reach of a standard endoscope.
For a patient with a negative endoscopy, the search for an obscure GI bleed often extends to the small bowel. In this scenario, both CT enterography and CTA abdomen and pelvis without and with IV contrast are rated Usually appropriate. CT enterography is optimized for evaluating the small bowel wall and lumen for subtle masses, inflammation, or vascular lesions. MR enterography, which avoids ionizing radiation, May be appropriate as an alternative.
Finally, in complex cases of postsurgical or traumatic nonvariceal upper GI bleeding where endoscopy is contraindicated, imaging is the primary diagnostic tool. Both Arteriography visceral and CTA abdomen and pelvis without and with IV contrast are rated Usually appropriate. CTA is excellent for a rapid, noninvasive survey of the anatomy to identify pseudoaneurysms, fistulas, or active extravasation, while arteriography offers the ability to immediately treat the identified source.
ACR Imaging Recommendations Table
| Clinical Scenario | Top Procedure(s) | ACR Rating | Adult RRL | Pediatric RRL |
|---|---|---|---|---|
| Adult. Suspected nonvariceal upper gastrointestinal bleeding; no endoscopy performed. Initial imaging. | CTA abdomen and pelvis without and with IV contrast | Usually appropriate | ☢ ☢ ☢ ☢ 10-30 mSv | |
| Adult. Endoscopy confirms nonvariceal upper gastrointestinal bleeding with a clear source, but treatment not possible or continued bleeding after endoscopic treatment. Initial imaging. | Arteriography visceral; CTA abdomen and pelvis without and with IV contrast | Usually appropriate | ☢ ☢ ☢ 1-10 mSv; ☢ ☢ ☢ ☢ 10-30 mSv | |
| Adult. Endoscopy confirms nonvariceal upper gastrointestinal bleeding without a clear source. Initial imaging. | CTA abdomen and pelvis without and with IV contrast | Usually appropriate | ☢ ☢ ☢ ☢ 10-30 mSv | |
| Adult. Nonvariceal upper gastrointestinal bleeding; negative endoscopy. Initial imaging. | CT enterography; CTA abdomen and pelvis without and with IV contrast | Usually appropriate | ☢ ☢ ☢ ☢ 10-30 mSv | ☢ ☢ ☢ ☢ 3-10 mSv [ped] |
| Adult. Postsurgical or traumatic causes of nonvariceal upper gastrointestinal bleeding. Endoscopy is contraindicated. Initial imaging. | Arteriography visceral; CTA abdomen and pelvis without and with IV contrast | Usually appropriate | ☢ ☢ ☢ 1-10 mSv; ☢ ☢ ☢ ☢ 10-30 mSv |
Adult vs. Pediatric Nonvariceal Upper Gastrointestinal Bleeding Imaging: Radiation Dose Tradeoffs
The ACR guidelines for this topic are primarily focused on adult patients, as reflected in the clinical variants. However, relative radiation level (RRL) information is provided for pediatric patients where applicable. The principle of As Low As Reasonably Achievable (ALARA) is paramount in pediatric imaging due to the increased radiosensitivity of developing tissues and the longer life expectancy over which potential stochastic effects of radiation could manifest. For modalities like CT enterography, the pediatric RRL (☢ ☢ ☢ ☢ 3-10 mSv) is noted alongside the adult dose (☢ ☢ ☢ ☢ 10-30 mSv), underscoring the use of dose-reduction techniques in children. When evaluating pediatric patients, clinicians should always consider non-ionizing alternatives like MR enterography, which is rated May be appropriate in the negative endoscopy scenario and carries an RRL of ‘O’ (0 mSv). The choice of imaging must always balance the diagnostic urgency against the long-term risks of cumulative radiation exposure.
Imaging Protocol Details for Nonvariceal Upper Gastrointestinal Bleeding
Once you’ve decided on the right study, the specific imaging protocol is critical for diagnostic accuracy. A GI bleed CTA, for instance, requires precise multiphase acquisition timing to capture arterial, portal venous, and delayed phases to maximize the chance of seeing contrast extravasation. Our protocol guides cover technique, contrast, and reading principles for key studies recommended in these guidelines.
Tools to Help You Order the Right Study
Navigating imaging guidelines during a busy clinical shift can be challenging. GigHz offers several tools designed to streamline this process, ensuring you can quickly access evidence-based recommendations and technical details.
The ACR Appropriateness Criteria Lookup provides a searchable interface to the complete ACR guidelines, allowing you to find recommendations for hundreds of clinical scenarios beyond nonvariceal GI bleeding. It’s designed to help you confirm the right study for any indication in moments.
For detailed technical specifications, the Imaging Protocol Library offers a comprehensive collection of institutional imaging protocols. You can review parameters for contrast timing, slice thickness, and reconstruction algorithms for the studies mentioned above, ensuring the exam is performed correctly.
Communicating radiation risk is a key part of shared decision-making. The Radiation Dose Calculator helps you estimate effective dose for various CT scans and compare them to background radiation levels. This tool can support conversations with patients and families about the benefits and risks of imaging.
Why is endoscopy the first-line investigation for most upper GI bleeds?
Endoscopy (specifically, esophagogastroduodenoscopy or EGD) is preferred as the initial step because it is both diagnostic and therapeutic. An endoscopist can directly visualize the mucosa to identify the bleeding source (e.g., an ulcer or vascular malformation) and simultaneously apply treatment, such as thermal coagulation, injection of epinephrine, or mechanical clipping, to achieve hemostasis in a single procedure.
What is the role of CTA in nonvariceal upper GI bleeding?
Computed Tomography Angiography (CTA) is a powerful, noninvasive imaging tool used to localize bleeding when endoscopy is unavailable, unsuccessful, or contraindicated. By acquiring images during multiple phases of IV contrast injection (non-contrast, arterial, and portal venous), CTA can detect active extravasation of contrast into the GI lumen, pinpointing the bleeding site with high accuracy. This information is critical for guiding subsequent treatment, whether it be repeat endoscopy or catheter-based embolization by interventional radiology.
When is a tagged RBC scan considered?
A technetium-99m-labeled red blood cell (RBC) scan is a nuclear medicine study that is highly sensitive for detecting intermittent or slow gastrointestinal bleeding. It is rated May be appropriate in scenarios where endoscopy is negative or cannot find a source. Because it involves imaging over several hours, it can detect bleeding rates as low as 0.1 mL/min, which may be too slow to be seen on a CTA. Its primary limitation is poor anatomic localization compared to CTA.
Why is a standard CT of the abdomen and pelvis with IV contrast usually not appropriate?
A standard, single-phase (typically portal venous) CT scan is not optimized for detecting active arterial bleeding. The specific multiphase protocol of a CTA is required to maximize the conspicuity of contrast extravasation, which is most evident on the arterial phase. A standard protocol may miss the bleed entirely or fail to characterize the source vessel, making it an inadequate first-line imaging test for this indication.
What is the difference between CT Enterography and a standard CT Abdomen/Pelvis?
CT Enterography is a specialized CT protocol designed specifically to evaluate the small bowel. It involves the patient drinking a large volume of a neutral oral contrast agent (like water or a polyethylene glycol solution) to distend the small bowel loops. This distention allows for much better visualization of the bowel wall to identify subtle inflammation, small tumors, or vascular lesions that could be sources of obscure GI bleeding. A standard CT uses positive oral contrast or no oral contrast and does not achieve the same level of small bowel distention, making it less sensitive for these findings.
Reviewed by Pouyan Golshani, MD, Interventional Radiologist — May 12, 2026
