IM15.1-18 | GI Bleeding — Graded Quiz

Correct. A liberal transfusion strategy targeting Hb 10 g/dL is harmful in variceal UGIB. The TRANSFUSE trial (Villanueva et al., NEJM 2013) demonstrated that a restrictive strategy (transfuse when Hb <7 g/dL, target 7-9 g/dL) reduced rebleeding and 45-day mortality in cirrhotic patients. Liberal transfusion raises portal pressure, worsening variceal bleeding.

Restrictive transfusion (Hb threshold 7 g/dL, target 7-9 g/dL) reduces 45-day mortality and rebleeding in variceal UGIB compared to liberal transfusion (Villanueva NEJM 2013). Overtransfusion raises portal venous pressure and directly worsens variceal haemorrhage.

A restrictive transfusion strategy (threshold Hb <7 g/dL, target 7-9 g/dL) is the evidence-based approach for variceal UGIB. Targeting Hb 10 g/dL is the incorrect 'liberal' strategy: it raises portal pressure, worsening variceal haemorrhage. IV access, cross-matching, and fluid resuscitation are all appropriate immediate actions.

Correct. Prophylactic antibiotics — ceftriaxone 1 g IV daily for 5-7 days (or norfloxacin 400 mg BD orally if lower risk) — are an essential component of the variceal bleeding bundle. Bacterial infections (particularly spontaneous bacterial peritonitis and bacteraemia from gut translocation) are a major driver of early rebleeding and death in cirrhotic patients with variceal haemorrhage. Failure to start antibiotics at admission is the most likely omission here.

The four-component variceal bundle: (1) vasoactive drug (terlipressin or octreotide) — start on suspicion; (2) EVL — within 12 hours; (3) prophylactic antibiotics (ceftriaxone 1 g/day IV) — start at admission, not after infection develops; (4) restrictive transfusion. All four components independently reduce mortality.

Prophylactic antibiotics are a mandatory component of the variceal bleeding bundle: ceftriaxone 1 g IV daily for 5-7 days. Bacterial translocation in cirrhosis causes infection, raises portal pressure, and promotes rebleeding. Band ligation is preferred over sclerotherapy. Terlipressin dosing (2 mg IV every 4-6 hours) is correct. Early re-endoscopy is not routine without rebleeding evidence.

Correct. For Forrest IIa lesions after endoscopic haemostasis, high-dose IV PPI reduces gastric acid, stabilises the fibrin-platelet clot, and prevents re-digestion of the haemostatic clot. The standard regimen is omeprazole 80 mg IV bolus followed by 8 mg/hr continuous infusion for 72 hours, then transition to oral PPI. H2 blockers are inferior. If H. pylori is truly negative (confirmed with two tests), focus is on stopping NSAIDs and PPI maintenance.

High-dose IV PPI after endoscopic haemostasis for high-risk peptic ulcer (Forrest Ia, Ib, IIa): omeprazole 80 mg IV bolus then 8 mg/hr for 72 hours. Mechanism: raises intragastric pH >6, stabilises platelet aggregation and fibrin clot. Reduces rebleeding by ~50% compared to no PPI.

Forrest IIa requires high-dose IV PPI post-endoscopy (omeprazole 80 mg bolus then 8 mg/hr for 72 hours): this raises intragastric pH to >6, stabilising the haemostatic clot over the visible vessel. H2 blockers do not achieve adequate acid suppression. H. pylori eradication is only indicated when infection is confirmed (false-negative urease tests can occur acutely — consider stool antigen or UBT after 4-6 weeks).

Correct. In actively bleeding patients on anticoagulants, coagulopathy must be corrected before endoscopic intervention. INR 3.8 indicates supratherapeutic anticoagulation: FFP provides clotting factors for rapid reversal (vitamin K IV works over 6-24 hours — too slow for emergency endoscopy). Platelet threshold for active bleeding with endoscopic procedures is >50 × 10^9/L, so platelet transfusion is also needed here. PRBCs replace red cell mass.

Blood product transfusion triggers in active GI bleeding: PRBCs if Hb <7 g/dL (or <8 in CVD); FFP if INR >1.5 (4-unit rule: 4 FFP per 4 PRBCs in massive transfusion); platelets if <50 × 10^9/L in active bleeding. Cryoprecipitate if fibrinogen <1.5 g/L. Vitamin K IV for warfarin reversal works over 6-24 hours.

In supratherapeutic anticoagulation with active GI bleeding: (1) PRBCs for Hb <7 g/dL; (2) FFP for INR >1.5 in active bleeding (vitamin K too slow for emergencies); (3) platelet transfusion when platelets <50 × 10^9/L for active bleeding. This is the evidence-based component replacement strategy. Whole blood transfusion is not the standard for individual component deficits.

Correct. For rebleeding after endoscopic haemostasis, guidelines recommend a SECOND endoscopic attempt at haemostasis before escalating to surgery or interventional radiology. Repeat endoscopy with rescue therapy (additional clipping, thermal therapy, haemostatic powder) is successful in approximately 70% of rebleed cases. Surgery is indicated if second endoscopy fails. Angiographic embolisation is an alternative for patients unfit for surgery.

Management of rebleeding post-endoscopy: repeat endoscopy with rescue therapy first (80% success rate overall). Surgical referral when repeat endoscopy fails (Forrest Ia posterior duodenal ulcer = highest surgical risk = gastroduodenal artery). Angiographic embolisation is the alternative for unfit surgical patients.

Rebleeding after endoscopic haemostasis: step 1 = repeat endoscopy with rescue haemostasis (second attempt succeeds ~70%); step 2 = surgery (oversewing ± vagotomy) or angiographic embolisation if repeat endoscopy fails. Going directly to surgery after first rebleed is not current practice. IV PPI adjustments do not stop active rebleeding.

Correct. CT angiography (CTA) detects active bleeding when the rate exceeds 0.3-0.5 mL/min. In patients with acute severe lower GI bleeding where colonoscopy is too risky, CTA provides anatomical localisation for targeted therapy — either endoscopic (if a discrete bleeding point is found after stabilisation) or angiographic embolisation. Capsule endoscopy is used for occult obscure GI bleeding, not acute massive haemorrhage. Barium studies are contraindicated in acute GI bleeding.

Investigation hierarchy for acute LGIB: (1) colonoscopy after prep if haemodynamically stable; (2) CT angiography if too unstable for colonoscopy or bleeding too rapid; (3) nuclear RBC scan if bleeding intermittent (detects 0.1 mL/min). Barium contrast studies = contraindicated in acute GI bleeding.

For acute severe LGIB when colonoscopy is risky: CT angiography (detects bleeding at >0.3 mL/min) or radionuclide scan (Technetium 99m RBC scan, detects 0.1 mL/min) can localise the source. CT angiography is preferred for immediacy and ability to guide embolisation. Barium enema is absolutely contraindicated in acute GI bleeding.

Correct. The Rockall Score is a post-endoscopy tool predicting rebleeding and mortality. A Forrest III lesion (clean base) has the lowest rebleed risk (<5%). Post-endoscopy Rockall Score ≤2 = low risk = early discharge with oral PPI is appropriate. Even though the score is 3 here, the Forrest III finding dominates clinical decision-making. High-dose IV PPI infusion is only indicated for high-risk stigmata (Forrest Ia, Ib, IIa).

Forrest classification rebleed risk: Ia spurting (>90%), Ib oozing (>50%), IIa visible vessel (~50%), IIb adherent clot (~25%), IIc flat spot (<10%), III clean base (<5%). High-dose IV PPI post-endoscopy = Forrest Ia/Ib/IIa only. Low-risk = Forrest IIc/III = oral PPI, early discharge if Rockall low.

The Rockall Score is used post-endoscopy. A Forrest III lesion (clean base) carries <5% rebleed risk — the lowest stigmata category. High-dose IV PPI is only indicated for Forrest Ia, Ib, and IIa (high-risk stigmata). Clean-base ulcers with low Rockall Score are candidates for early discharge with oral PPI. The Rockall and Glasgow-Blatchford scores serve different purposes and are not interchangeable.

Correct. In cirrhotic patients with variceal bleeding, the Child-Pugh Score (albumin, bilirubin, INR, encephalopathy, ascites) and MELD Score (bilirubin, creatinine, INR) quantify liver reserve and predict medium-term mortality. Child-Pugh C (score 10-15) carries 30-40% 1-year mortality from a variceal bleed. These scores guide decisions about TIPS (transjugular intrahepatic portosystemic shunt) and transplant listing. Glasgow-Blatchford and Rockall predict immediate intervention need and short-term rebleed risk, not 3-month cirrhosis-related mortality.

Child-Pugh Score: A (5-6) = good reserve, B (7-9) = moderate, C (10-15) = poor reserve, ~35% 1-year mortality post-variceal bleed. MELD Score: used for transplant prioritisation. After variceal bleed: Child-Pugh C or MELD >15 = consider TIPS referral or transplant listing. All survivors require secondary prophylaxis: non-selective beta-blocker + EVL program.

In cirrhotic variceal bleeding, Child-Pugh and MELD scores predict liver-function-driven medium-term mortality and guide escalation (TIPS, transplant). Glasgow-Blatchford = pre-endoscopy intervention need. Rockall = post-endoscopy rebleed risk. CURB-65 = community-acquired pneumonia severity. The patient's parameters (Child-Pugh C territory) carry ~30-40% 1-year mortality.

Correct. In an NSAID-using patient with melaena, the key branch point is: peptic ulcer disease (non-variceal) versus variceal bleeding (portal hypertensive). If he has chronic liver disease, the management shifts completely — vasoactive drugs, EVL, antibiotics, restrictive transfusion, Child-Pugh assessment, and secondary prophylaxis. NSAID duration, family history of CRC, and dyspepsia would all influence management, but none bifurcates the management pathway as fundamentally as the presence or absence of portal hypertension.

Structured GI bleeding history priorities: (1) type of bleeding — characterise route; (2) haemodynamic features — volume status; (3) portal hypertension history — bifurcates entire management; (4) medications — NSAIDs, anticoagulants, aspirin; (5) prior GI disease, procedures, or malignancy; (6) alcohol/liver history. The portal hypertension question is the highest-yield single item.

The single most important history item in suspected UGIB after melaena is confirmed is whether portal hypertension / chronic liver disease is present. This single item changes the entire management pathway: vasoactive drugs, EVL, prophylactic antibiotics, restrictive transfusion, and Child-Pugh risk stratification are all specific to variceal aetiology. NSAID-related PUD has an entirely different treatment protocol.

Correct. In patients with significant cardiovascular disease — including heart failure (HFrEF, LVEF ≤40%), ischaemic heart disease, or haemodynamically significant valvular disease — the transfusion threshold is relaxed to Hb <8 g/dL. The rationale is that these patients have limited cardiac reserve and may develop myocardial ischaemia at lower haemoglobin concentrations. The target remains restrictive (not liberal), however: target Hb 8-10 g/dL rather than 10-12 g/dL.

Transfusion thresholds in GI bleeding: no CVD = Hb <7 g/dL; significant CVD (HFrEF, IHD) = Hb <8 g/dL. This distinction is explicit in the NICE GI Bleeding guideline (CG141) and BSG guidelines. Document the clinical indication for every transfusion unit administered.

For patients with cardiovascular disease (HFrEF, IHD, valvular disease), the PRBCS transfusion threshold in GI bleeding is Hb <8 g/dL (vs <7 g/dL in those without CVD). This reflects the reduced cardiac reserve and risk of demand-supply myocardial ischaemia at lower Hb. Targeting Hb 10 g/dL remains too liberal and is not evidence-based.

Correct. Urgent colonoscopy after bowel preparation is the investigation of choice for acute LGIB when the patient is haemodynamically stable (or stabilised). It has the highest diagnostic yield (70-90%) and allows simultaneous endoscopic therapy (haemostasis of diverticular bleeding, polypectomy site bleeding). The clear nasogastric aspirate makes an upper GI source unlikely. Barium enema is contraindicated in acute GI bleeding. CT without contrast does not reliably identify bleeding sources.

LGIB investigation pathway: exclude upper GI source (OGD or NGT aspirate) → colonoscopy (first choice if stable, yields 70-90%) → CT angiography (if unstable or rapid bleeding, detects >0.3 mL/min) → angiography + embolisation (therapeutic) → surgery (last resort). Colonoscopy is preferred as it combines diagnosis and therapy.

For haemodynamically stable (or stabilised) acute LGIB with upper GI source excluded: urgent colonoscopy after rapid bowel prep is the investigation of choice, with 70-90% diagnostic yield and therapeutic capability. Barium enema is absolutely contraindicated in acute bleeding. CT angiography is preferred over plain CT and is the alternative to colonoscopy when haemostasis is inadequate for bowel prep.

Correct. Secondary prophylaxis after variceal haemorrhage combines: (1) repeat EVL sessions every 1-2 weeks until variceal obliteration (typically 3-6 sessions); plus (2) a non-selective beta-blocker — propranolol (titrate to resting HR 55-60 bpm) or carvedilol. The combination reduces rebleed risk more than either alone. TIPS is reserved for Child-Pugh B-C patients with refractory rebleeding.

Variceal bleeding prophylaxis: primary (first bleed prevention) = non-selective beta-blocker or EVL for large varices. Secondary (prevent rebleed) = combination EVL + non-selective beta-blocker. TIPS for Child-Pugh B-C with refractory rebleeding. Non-selective beta-blockers: propranolol (target HR 55-60/min) or carvedilol (6.25-12.5 mg OD).

Secondary prophylaxis for variceal rebleeding = combination of repeat EVL (every 1-2 weeks to obliteration) + non-selective beta-blocker (propranolol titrated to HR 55-60, or carvedilol 6.25-12.5 mg). Beta-blocker alone is used for primary prophylaxis of first bleed. Combination therapy is superior for secondary prophylaxis. Isosorbide mononitrate has been studied but is not standard first-line. Antibiotics prevent SBP but are not secondary rebleed prophylaxis.