Gastrointestinal Bleeding Viva

Long Case Viva Defence

Candidate's opening statement (model answer)

"Mr Huang is a 62-year-old man who presents with two large episodes of haematemesis and melaena over the past eight hours. He has a background of alcohol-related cirrhosis (Child-Pugh B), type 2 diabetes, hypertension, and a prior variceal bleed two years ago for which he had endoscopic band ligation. [1]

On examination he is tachycardic at 112, blood pressure 96/60 with a postural drop, cool peripheries, and he has stigmata of chronic liver disease — spider naevi, palmar erythema, and a small amount of ascites. Rectal examination confirms melaena. [1]

His main problems are:

1. Acute upper GI haemorrhage — most likely variceal rebleed, but peptic ulcer remains in the differential
2. Haemodynamic shock requiring resuscitation
3. Decompensated cirrhosis (Child-Pugh B) with portal hypertension
4. Type 2 diabetes and hypertension [1]

My immediate priority is resuscitation, then risk-stratification, then definitive endoscopic therapy within 12 hours for suspected variceal bleeding." [1]

Examiner probing questions and model answers

Q1: "Walk me through your resuscitation and pharmacotherapy in the first hour." [1]

"Two large-bore cannulae, cross-match four units, FBC, U&E, LFTs, coagulation, and a crystalloid bolus — but I will resuscitate judiciously in cirrhosis, because over-resuscitation raises portal pressure and worsens variceal bleeding. I will transfuse red cells restrictively — threshold haemoglobin 70 g/L, target 70 to 90 g/L — based on Villanueva et al. (NEJM 2013), which showed improved survival with a restrictive strategy. Because this is a suspected variceal bleed, I will start a vasoactive drug immediately — terlipressin 2 mg IV every four hours — and prophylactic ceftriaxone 1 g IV daily. Antibiotics reduce mortality in cirrhotic GI bleeding (Bernard 1999). I will then arrange endoscopy within 12 hours." [1]

Q2: "Why terlipressin rather than octreotide?" [1]

"Terlipressin is the only vasoactive agent shown in meta-analysis to reduce mortality in acute variceal bleeding — the Ioannou Cochrane review (2003). It is a vasopressin analogue causing splanchnic vasoconstriction, reducing portal inflow and pressure. Octreotide is a reasonable alternative where terlipressin is unavailable or contraindicated (for example, in severe peripheral or mesenteric ischaemia), but the mortality signal is strongest for terlipressin. Both reduce portal pressure as a bridge to definitive endoscopic therapy." [1]

Q3: "How will you use the Glasgow-Blatchford and Rockall scores in this patient?" [1]

"The Glasgow-Blatchford Score is calculated before endoscopy and predicts the need for intervention. His will be high given the deranged observations and low blood pressure, confirming he needs admission and urgent endoscopic therapy. The Rockall score is calculated after endoscopy and predicts mortality and rebleeding — it incorporates age, shock, comorbidity, diagnosis, and stigmata of recent haemorrhage. With cirrhosis as a major comorbidity plus shock, his Rockall will be high regardless of endoscopic findings, which tells me he needs a high level of care and close monitoring for rebleeding." [1]

Q4: "Endoscopy shows active bleeding from an oesophageal varix. What is your endoscopic therapy, and what if it fails?" [1]

"Endoscopic variceal ligation (banding) is first-line — Stiegmann et al. (NEJM 1992) showed it is as effective as sclerotherapy with fewer complications. I would band the bleeding varix and any other large varices in the distal oesophagus. If bleeding is uncontrolled or recurs despite banding and ongoing vasoactive therapy, I would consider balloon tamponade (Sengstaken-Blakemore) strictly as a temporary bridge for up to 24 hours, and arrange an urgent transjugular intrahepatic portosystemic shunt (TIPS). In fact, in a high-risk cirrhotic like this — Child-Pugh B with active bleeding — early pre-emptive TIPS within 72 hours is recommended to reduce treatment failure and mortality." [1]

Q5: "He is on apixaban for atrial fibrillation. How does that change things?" [1]

"He did not mention AF in the brief, but if he is on apixaban, I would withhold it immediately and discuss reversal with haematology for this life-threatening bleed — andexanet alfa or prothrombin complex concentrate. I would weigh his thrombotic risk (CHA2DS2-VASc) against bleeding and plan to resume anticoagulation within 7 to 14 days once haemostasis is secure. Premature permanent cessation is a common and harmful error in cirrhotic patients who often have high thrombotic risk." [1]

Q6: "What is his prognosis and how will you discuss it?" [1]

"Variceal bleeding is a marker of decompensated cirrhosis with significant one-year mortality — often 30 to 40 percent or higher depending on Child-Pugh grade and whether rebleeding or complications like infection and renal failure occur. I would discuss this honestly but constructively with Mr Huang and his family — explaining that the acute bleed can be controlled, that secondary prophylaxis with a non-selective beta-blocker and repeat banding reduces rebleeding, but that the underlying liver disease is the determining factor. I would initiate a transplant referral conversation early, as variceal bleeding is often the event that prompts transplant assessment." [1]

Short Case Discussion

Scenario: "Examine this patient's abdomen and discuss"

Candidate presentation (model): [1]

"I examined Mr Chen's abdomen. He is alert but pale, comfortable at rest. There are spider naevi on the anterior chest wall and palmar erythema. There is no clubbing, jaundice, or lymphadenopathy. The pulse is regular at 96 per minute; blood pressure 110/70. [1]

On abdominal inspection there is no visible peristalsis or distension. There are caput medusae — dilated superficial abdominal veins indicating recanalisation of the umbilical vein from portal hypertension. On palpation the abdomen is soft and non-tender. The liver edge is firm and enlarged three centimetres below the costal margin; the spleen is palpable two centimetres below the costal margin, indicating splenomegaly from portal hypertension. There is shifting dullness and a fluid thrill, consistent with ascites. Bowel sounds are normal. Rectal examination reveals melaena. [1]

In summary, these findings are consistent with decompensated chronic liver disease with portal hypertension presenting with an acute upper gastrointestinal bleed, likely variceal." [1]

Examiner: "What is the significance of caput medusae?" [1]

"Caput medusae — dilated paraumbilical veins radiating from the umbilicus — indicates recanalisation of the umbilical vein due to portal hypertension. Portal pressure is transmitted through the ligamentum teres to the systemic circulation at the umbilicus. Its presence, with splenomegaly and ascites, strongly supports portal hypertension as the cause of the variceal bleed." [1]

Examiner: "Why does he have ascites, and what does that tell you about his prognosis?" [1]

"Ascites in cirrhosis results from portal hypertension (increased hydrostatic pressure) and hypoalbuminaemia (reduced oncotic pressure), compounded by splanchnic vasodilation activating the renin-angiotensin-aldosterone system causing sodium and water retention. Clinically apparent ascites indicates decompensated cirrhosis — it is one of the defining features and marks a poorer prognosis, with significantly reduced median survival compared to compensated disease. In the context of an acute variceal bleed, ascites places him in a higher Child-Pugh grade and higher risk category." [1]

Examiner: "Explain why urea rises disproportionately in upper GI bleeding." [1]

"When blood is digested in the upper gastrointestinal tract, the proteins are broken down and absorbed as amino acids in the small bowel, then converted to urea by the liver via the urea cycle. This produces a disproportionately high blood urea relative to creatinine — typically a urea-to-creatinine ratio above 30 in SI units. It is a useful biochemical discriminator confirming an upper GI source, particularly when the bleeding is occult or the history is unclear." [1]