Clinical evidence of arterial GI hemorrhage
Hematemesis, melena, hematochezia, hemobilia
Hypotension, tachycardia, transfusion requirement, pressor requirement despite conservative (and sometimes endoscopic therapy)
Imaging evidence of arterial GI hemorrhage
CTA
Endoscopy
Tagged RBC scan
Absolute: None, when intervention anticipated to be life-saving
Relative:
Uncorrectable bleeding diathesis
Contrast allergy (premedicate, as time allows)
Renal impairment
Prior surgery altering vascular anatomy (increases risk of gut ischemia)
CTA (3-phase: noncon, arterial, delay) required for most angio/embos, unless a source has already been localized by another method.
Tagged RBC scan (scintigraphy) is more sensitive (90%) but takes hours and poor at localizing bleed to a vascular territory. Therefore, not a good test when there is active bleeding clinically.
Upper GI bleed:
What did endoscopy show? CTA, if source not yet identified?
Duodenal ulcers bleeding from the GDA are the most common, and we will often be asked to “embolize GDA empirically” even if it looks normal angiographically.
Lower GI bleed:
What did CTA show? Endoscopy not 1st line due to lack of colon prep
Diverticulitis is most common
Is coagulopathy being reversed?
Is the patient unstable to the point of requiring anesthesia? NPO?
Consent: Risk of non-target embolization, ischemic bowel, contrast nephropathy, inability to localize bleed, access site complications