GI Bleeding (GIB)

• Luminal signature 👀 : the observation of hematemesis/coffee-ground emesis vs. melena/hematochezia (i.e. generally obvious)

• Hematologic signature 🩸: tachycardia → hypotension*, ↑ BUN/Cr ratio (UGIB) → ↓ hemoglobin*, ↓ iron stores (both requiring time i.e. not hyperacute findings)

• Risk profile 💀 : conditions (e.g. NSAID/EtOH exposure, cirrhosis, recent bowel ischemia) vs. coagulation (anticoagulants > anti-platelets)

• Physical Exam ✋ : abdominal exam is generally unproductive, as GI bleeding generally denotes luminal disease that does not touch the layers of bowel that produce pain, thus making the abdominal exam generally normal. (Notable exceptions to this include intramural hematoma & perforation due to ischemia vs. PUD.) More helpful than the abdominal exam is general assessment of the patient for findings concerning for impending GIB (e.g. new hypoactive delirium, poor capillary refill, new subtle paleness of hypoperfusion)

⚪ The absence of tachycardia does not rule-out active GIB, & can usually be accounted for by age-related chronotropic incompetetence (i.e. simply advanced age) or rate-controlling medications (e.g. β-blocker)

⚪ The presence of normal iron stores in patients with “suspected GI bleed” can be extremely helpful in dramatically reducing that suspicion, as the vast majority of bleeds are truly subacute-chronic, intermittent bleeds that necessitate blood & iron loss

*When a patient bleeds, the body rapidly replaces the lost content, but doesn't do so at the same pace: (a) crystalloid is rapidly mobilized from interstitium (b) WBCs & platelets are also replaced rapidly from pools in the marrow & spleen (c) RBCs (half life = 120 days), are made MUCH more slowly, thus resulting in ↓ Hgb. For this reason, hyperacute, massive hemorrhage will not lead to observation of ↓ Hgb

Epidemiology & Mortality 🌎

• Hematemesis & Coffee-ground emesis: unless the patient has bleeding due to a Mallory-Weiss tear related to repetitive vomiting, the observation of hematemesis implies a massive volume of blood loss beyond the stomach’s capacity, thereby resulting in reflexive vomiting, thus making hematemesis an emergency. Coffee-ground emesis implies the upper bleeding remained stagnant in the stomach for a short duration, but still represents an emergency given the amount of blood loss necessary to trigger the vomiting reflex

• Melena: majority is upper → small bowel in origin, but rarely right-sided colonic sources can present with melena. In descending frequency, most likely sources of UGIB include PUD, severe gastritis/duodenitis, severe esophagitis, EV, PHG, angioectasia, but notably the epidemiology of source varies widely depending on the clinical presentation (e.g. most unstable patients have bleeding artery vs. varix, while most stable patients are more likely to have a source that “oozes” rather than “hemorrhages”)

• Hematochezia: ~85% is lower, ~15% is upper. ~80% of lower GIB will stop spontaneously with a mortality rate of 2-4%, but in the acutely unstable patient, consideration of massive UGIB resulting in spillover hematochezia is always warranted due to need for immediate control of the source of bleeding. In descending frequency, most likely sources of LGIB include diverticulosis, anorectal sources, ischemia, & other colitis sources, but the epidemiology of source varies widely depending on context (e.g. hemorrhoids overtake diverticulosis if age < 50)

• Recurrent, EGD ⊖, Colonoscopy ⊖ (”obscure”): a small bowel source is generally uncommon during the initial evaluation, but becomes the most common source for GIB that persists/recurs without an obvious source (~5-10% of overall GIB, ~75% of obscure GIB). In the young, small bowel sources are more likely to be due to IBD, anatomic defect (e.g. Meckel’s diverticulum, Dieulafoy lesion), or small bowel neoplasm (e.g. GIST), whereas in elders, small bowel sources are more likely to be due to vascular lesions (e.g. angiodysplasia), or erosions/ulcers (e.g. NSAIDs). Small bowel oozing should be considered in the work-up of occult IDA given most small bowel oozing is actually clinically silent (i.e. no melena)

Acute on Chronic ❓

• Refers to the clinical syndrome of acute bleeding (e.g. hematemesis, melena) accompanied by laboratory evidence of subacute-chronic bleeding (i.e. iron deficiency, ↓ Hgb), which can be explained by erosive mucosal disease leading to arterial compromise (”visible vessel” of PUD) vs. pop-off valve physiology of already exposed artery (e.g. varix)

Scopes 📷

• Lower (Colonoscopy): unfortunately more often than not is limited by stool (lack of prep) & blood (recent hemorrhage) precluding clear visualization of the GI tract (sensitivity for source = ~45-90%). Additionally, even if vision is clear, the success rate of local bleed control is also not very high, & the identification of >1 potential bleeding sites is common (e.g. diverticulosis + hemorrhoids). For this reason, GI often resorts to CTA abdomen for demonstration of active extravasation prior to directing patient to IR suite for embolization

• Upper (EGD): generally, visualization is much clearer, & EGD allows effective local control of highly morbid conditions (i.e. varices, ulcer, Dieulafoy lesion)

Rx > Dx 💊

Endoscopy ⊖ GIB (@e_chiu17)

• @rabihmgeha

• @AndreMansoor’s Frameworks for Internal Medicine

• Is This Patient Hypovolemic?

• Do patients assess rectal bleeding accurately?

• Approach to acute lower gastrointestinal bleeding in adults (UpToDate)

• Evaluation of suspected small bowel bleeding (formerly obscure gastrointestinal bleeding) (UpToDate)

• Angiodysplasia of the gastrointestinal tract (UpToDate)

• Angiodysplasia image