---

Upper GI Bleeding - Management Protocol

---

Causes (by frequency)

---

Phase 1 - Immediate Resuscitation

Airway

• In hemodynamically unstable patients with active hematemesis, consider prophylactic endotracheal intubation before endoscopy to prevent aspiration. Use smaller doses of induction agents to minimize peri-intubation hypotension.

Access & Volume

• Two large-bore IV lines (at least 18G)
• IV crystalloid resuscitation
• Type and cross-match; consider massive transfusion protocol if large-volume blood products anticipated
• ICU or monitored bed for hemodynamically unstable patients

Transfusion Thresholds

• Restrictive strategy is recommended for most patients: transfuse when Hb < 7 g/dL
• In older patients or those with significant comorbidities: transfuse when Hb < 9 g/dL
• Liberally transfusing all patients (high threshold Hb < 9 g/dL for all) causes harm

Coagulopathy Correction

• Reverse coagulopathy if: INR elevated, platelets < 50,000/µL, or life-threatening bleeding
• Exception: caution with reversal in patients with cardiac/vascular stents or prosthetic valves
• INR ≥ 1.5 is a significant predictor of mortality in anticoagulated UGIB patients
• Newer oral anticoagulants (direct thrombin inhibitors, Xa inhibitors): manage per institutional protocol
• Reversal must not delay time to endoscopy

Nasogastric Lavage

• A positive aspirate (bloody, maroon, coffee-ground) confirms upper GI source
• A negative aspirate does not exclude UGIB (false negative due to intermittent bleeding, pyloric spasm)
• Only 23% of patients without hematemesis who have UGIB will have a positive aspirate
• If bright red blood or clots: gentle gastric lavage with room-temperature water on mild intermittent suction

---

Phase 2 - Risk Stratification

Pre-endoscopy Scoring Systems

• Blood pressure, BUN, hemoglobin, heart rate, syncope, melena, liver disease, heart failure
• Score = 0: < 1% chance of requiring intervention → consider outpatient management
• Any score > 0 in the context of overt bleeding: hospital admission indicated

• Rockall Score 0-2 after endoscopy → low risk, consider early discharge
• Full Rockall (post-endoscopic) adds findings/diagnosis/SRH - correlates better with mortality

• Each = 1 point; higher scores predict in-hospital mortality

• Advanced age, comorbidities, red hematemesis, hematochezia, positive NG aspirate, hemodynamic instability, abnormal labs, prior variceal banding/cauterization/TIPS procedure

---

Phase 3 - Medical Therapy

For Non-Variceal UGIB (especially peptic ulcer)

• 80 mg IV bolus followed by 8 mg/hour infusion (e.g., omeprazole, pantoprazole)
• Pre-endoscopic PPI reduces stigmata of recent hemorrhage (SRH) and need for endoscopic therapy
• Post-endoscopic high-dose PPI for 72 hours in high-risk stigmata (active bleeding, non-bleeding visible vessel, adherent clot)

• Erythromycin 250 mg IV (or metoclopramide) 30-90 minutes before endoscopy
• Clears blood from stomach, improves visualization, may reduce need for repeat endoscopy

For Suspected Variceal UGIB

• Octreotide: 50 µg IV bolus, then 25-50 µg/hour infusion (elderly: start at 25 µg bolus/infusion)
• Somatostatin or terlipressin (long-acting vasopressin analog) are equivalent alternatives
• Meta-analyses show vasoactive drugs are as effective as sclerotherapy for controlling variceal bleeding and cause fewer adverse events

• Up to 50% of cirrhotic patients develop infection during hospitalization for GI bleeding
• Ceftriaxone 1 g IV every 24 hours for 7 days (preferred, esp. in high quinolone-resistance areas)
• Alternatives: Ciprofloxacin 400 mg IV every 12 hours, or levofloxacin 500 mg IV every 24 hours, or norfloxacin 400 mg PO BID (not available in USA)
• Antibiotics reduce mortality and bacterial infection rates in cirrhotic UGIB

---

Phase 4 - Endoscopy

Timing

• Within 24 hours for most patients with overt UGIB after adequate resuscitation (ACG/ESGE standard)
• Within 12 hours for high-risk patients (persistent hemodynamic instability, suspected variceal source)
• Urgent (< 6 hours): reserved for patients with ongoing massive hemorrhage
• Very low-risk patients (Blatchford 0): may be managed as outpatients

Preparation

• Use a therapeutic endoscope (large accessory channel, foot-pump jet irrigation)
• Administer prokinetic pre-procedure
• Consider intubation for aspiration protection in active bleeders

Forrest Classification (peptic ulcer SRH)

• Classes Ia, Ib, IIa (and IIb after clot removal): endoscopic hemostasis indicated
• Classes IIc and III: medical therapy only, safe for early discharge

Endoscopic Hemostasis Methods

• Through-the-scope clips (TTSC) - first-line for many lesions
• Over-the-scope clips (OTSC) - recent meta-analyses show benefit for refractory/recurrent UGIB
• Band ligation - preferred for esophageal varices

• Epinephrine 1:10,000 - always combine with another modality; monotherapy insufficient
• Tissue adhesives (cyanoacrylate - especially for gastric varices)

• Contact: electrocoagulation, thermocoagulation (bipolar/gold probe)
• Non-contact: argon plasma coagulation (APC), laser photocoagulation

• Detects residual arterial blood flow after hemostasis
• UCLA CURE RCT: rebleeding significantly reduced with DEP-guided vs. standard visually guided hemostasis

Variceal-Specific Endoscopic Treatment

• Esophageal varices: Endoscopic band ligation (EBL) preferred over sclerotherapy (lower complication rate)
• Gastric varices: Cyanoacrylate injection or EBL; more technically challenging
• Sclerotherapy (sclerosants: ethanolamine oleate, sodium tetradecyl sulfate, sodium morrhuate): used when banding unavailable

---

Phase 5 - Rescue/Salvage Therapies

Balloon Tamponade (bridge therapy only)

• Sengstaken-Blakemore tube (gastric + esophageal balloons)
• Minnesota tube (adds esophageal aspiration port)
• Linton-Nachlas tube (single large gastric balloon)
• Controls bleeding initially in 85-98% of cases; rebleeding recurs in 21-60% after deflation
• 30% serious complication rate (aspiration pneumonia, esophageal rupture, airway obstruction)
• Always intubate before insertion
• Used only to bridge to definitive therapy (TIPS or repeat endoscopy)

TIPS (Transjugular Intrahepatic Portosystemic Shunt)

• For variceal bleeding uncontrolled by endoscopy
• Considered within 72 hours (early TIPS) in high-risk patients (Child-Pugh B with active bleeding, or Child-Pugh C < 14 points)
• Also used when endoscopy fails to localize or control variceal source

Angiography & Embolization

• When endoscopy fails to localize or control non-variceal bleeding
• Celiac and superior mesenteric angiograms obtained; CO2 contrast useful (lower viscosity, better sensitivity)
• Selective arterial embolization for peptic ulcer bleeding (gastric bleeds: left gastric artery; duodenal bleeds: gastroduodenal artery)
• Prophylactic embolization acceptable when massive bleeding has ceased at time of angiography
• Percutaneous transhepatic coronary vein embolization for varices when endoscopy unavailable or fails

Surgery

• Reserved for: failure of all endoscopic/angiographic therapy, massive ongoing hemorrhage, or specific lesions (aortoenteric fistula)
• Less frequently required given advances in endoscopic and IR therapies

---

Phase 6 - Post-Hemostasis & Secondary Prevention

• H. pylori testing and eradication in all peptic ulcer bleeders (reduces recurrence markedly)
• PPI maintenance in NSAID users; switch to COX-2 inhibitor + PPI if NSAID continuation required
• Beta-blocker prophylaxis (propranolol or nadolol) + repeat EBL sessions until variceal obliteration for secondary prevention of variceal bleeding
• Anticoagulation restart: individualize risk-benefit; most patients can safely resume anticoagulation within 7 days if indication is strong (e.g., prosthetic valve, AF)
• Aspirin/antiplatelet restart: for high CV risk patients, restart aspirin as soon as hemostasis confirmed (typically within 3-5 days)

---

Special Populations

---

Recent Evidence Updates (PubMed 2024-2026)

• Over-the-scope clips (OTSC) for non-variceal UGIB: A 2025 systematic review and meta-analysis (Ni et al., PMID 39500365) supports OTSC as superior to standard therapy for refractory/recurrent non-variceal UGIB.
• Tranexamic acid (TXA): A 2025 meta-analysis (Calderon Martinez et al., PMID 40029534) - evidence remains mixed for TXA in UGIB; current ACG guidelines do not routinely recommend it (the HALT-IT trial showed no mortality benefit and possible harm).
• A 2025 CAG guideline specifically addresses non-variceal non-peptic ulcer bleeding (Dieulafoy lesions, GAVE, Mallory-Weiss tears, malignant lesions), with GRADE-based recommendations for each lesion type.

---