Is a CT (Computed Tomography) abdomen and pelvis a suitable first-line imaging choice for a suspected upper gastrointestinal (GI) bleed?

CT Abdomen and Pelvis for Suspected Upper GI Bleed

Yes, you can and should order a CT angiography (CTA) of the abdomen and pelvis without and with IV contrast for suspected upper GI bleeding when endoscopy cannot be performed or has failed, but you must specifically order CTA—not a standard CT with contrast—as these are distinct procedures with different diagnostic capabilities. 1, 2

Critical Distinction: CTA vs Standard CT

The most important caveat is understanding that CTA is fundamentally different from a standard "CT abdomen/pelvis with contrast." The ACR Appropriateness Criteria explicitly state there is no significant literature supporting standard CT with contrast alone for upper GI bleeding, rating it as "usually not appropriate" (rating 2-3) across all clinical scenarios. 1, 2

Why CTA is Required:

• CTA uses optimized contrast timing to capture arterial phase imaging, detecting active bleeding as slow as 0.3 mL/min (compared to 0.5-1.0 mL/min for conventional angiography). 1, 2
• Multiphasic acquisition is essential: non-contrast phase followed by arterial and portal venous phases, with the highest sensitivity of 92% achieved with this complete protocol. 1, 2
• Thin collimation (≤1 mm) enables high-quality 3D reconstructions that are required elements of CTA but not standard CT. 2
• Non-contrast images identify sentinel clots and baseline high-attenuation materials that could mimic bleeding on contrast-enhanced images. 1, 2

When to Order CTA for Upper GI Bleeding

Scenario 1: No Endoscopy Performed

CTA abdomen and pelvis without and with IV contrast is "usually appropriate" (rating 8-9) when:

• Large volume bleeding prevents initial endoscopy 2
• Patient is clinically unstable for endoscopy 2
• Endoscopy is contraindicated 1

CTA leads to faster triage in the emergency room compared to endoscopy and can guide subsequent interventional angiography. 1

Scenario 2: Endoscopy Confirms Bleeding But Source Unclear

CTA is equally appropriate as visceral arteriography (both rating 8-9) when endoscopy shows upper GI bleeding but cannot localize the source. 1, 2 In this scenario, CTA provides a roadmap for subsequent catheter angiography if intervention is needed. 2

Scenario 3: Endoscopy Identifies Source But Treatment Failed

Both CTA and visceral arteriography are "usually appropriate" when endoscopy confirms the bleeding source but treatment is not possible or bleeding continues after endoscopic therapy. 1, 2 Visceral arteriography may be preferred here as it allows simultaneous diagnosis and therapeutic embolization. 1

Diagnostic Performance

• Sensitivity: 79-85% for active GI bleeding 2
• Specificity: 92-95% 1, 2
• Positive predictive value: 86% 2
• Negative predictive value: 92% 2

However, one contradictory study found CTA had only 20% sensitivity when endoscopy was used as the criterion standard, with 67% of patients with negative CTA having bleeding lesions identified on subsequent endoscopy. 3 This highlights that CTA is best used when endoscopy is not feasible or has failed, not as a replacement for endoscopy when endoscopy is possible. 3

Critical Protocol Requirements

When ordering, specify these technical parameters:

• "CTA abdomen and pelvis WITHOUT AND WITH IV contrast" (not just "with contrast") 1, 2
• Multiphasic acquisition: non-contrast + arterial + portal venous phases 1, 2
• NO oral contrast (positive oral contrast renders the exam nondiagnostic and water dilutes intraluminal hemorrhage) 1
• Thin collimation ≤1 mm for multiplanar reformations 2

Common Pitfalls to Avoid

• Do not order "CT abdomen/pelvis with contrast"—this is insufficient and rated as usually not appropriate. 1, 2
• Do not order CTA of abdomen only—both abdomen and pelvis must be imaged because the bleeding site is unclear without endoscopy. 1
• Do not order CTA chest for suspected upper GI bleeding—even though esophageal bleeding can present as upper GI bleeding, the literature supports abdomen and pelvis imaging. 1
• Avoid ordering if endoscopy is feasible—endoscopy remains the gold standard when clinically appropriate, as CTA may miss lesions that endoscopy can identify and treat. 3

Limitations and Rebleeding Risk

• Intermittent bleeding may cause false negatives—nearly 60% of patients with suspected upper GI bleeding and negative CTA did not rebleed in 8-year follow-up, but 40% did rebleed despite negative imaging. 1
• Greater contrast extravasation volume correlates with need for hemostatic therapy and massive transfusion, but not with mortality. 1

Alternative Imaging Modalities Are Not Appropriate

• RBC scintigraphy: No significant literature supports its use for initial imaging of suspected upper GI bleeding. 1
• Upper GI fluoroscopy series: Has no role in acute upper GI bleeding in modern practice. 1
• MR enterography: No significant literature supports its use for overt upper GI bleeding. 1