EBM Upper GI Haemorrhage

Pedagogical disambiguation: Emergency Medicine Lecture Notes and Evidence Based emergency medicine principles from Professor A.F.T Brown and the Life in the Fast Lane team.


  • Commonest causes: peptic ulcer (35-50%); oesophagitis (20-30%);  duodenitis/gastritis/erosions (10-20%); varices (5-12%); Mallory-Weiss tear (2-5%); tumour (2-5%); angiodysplasia (2-3%); aorto-enteric fistula (<1%).
  • Mortality 10 – 14%. Majority are over 65 years or variceal.
    • Mortality / morbidity risk factors include: cause of the bleeding particularly varices; advanced age; shock; fresh red blood; low Hb; co-morbid disease; re-bleed; endoscopic findings.
    • Endoscopic stigmata that predict re-bleeding, need for surgery and death include active arterial bleeding, adherent clot, non-bleeding but visible vessel, ulcer size and location. Scoring systems exists eg. Rockall (max post-endoscopy score 11); score 8 = mortality 41%.
  • Low-risk group who may be managed as outpatient without early endoscopy can be predicted by Glasgow Blatchford score (GBS), or the pre-endoscopic Rockall score.
    • GBS of 0 safely allows discharge without endoscopy, although up to half may then not present for OP endoscopy! Note age is not scored.

Stanley A, Ashley D, Dalton H et al. Outpatient management of patients with low-risk upper GI haemorrhage: multicentre validation and prospective evaluation. Lancet 2009;373:42-7. [Reference]


Medical therapy

  • Resuscitation, ABC including look for orthostatic hypotension etc. Consider transfusion for shock or acute fall in Hb below 7.0 g/dL (below 10.0 if IHD, PVD etc). Also FFP if INR >1.5 and or platelets if low < 50 x 109/L.
  • Gastric lavage – of no proven benefit, but NGT may indicate ongoing bleeding.
  • Proton pump inhibitor by infusion ie. omeprazole / pantoprazole 80 mg stat and 8 mg/hr for 72 hours. Reduces high-risk stigmata and need for endoscopic therapy if given pre-endoscopy (OR 0.67). Reduces risk of rebleeding, surgery and death in high-risk patients if given after endoscopy (RR 0.4 / 0.43 / 0.41 respectively). Overall cost-effective and safe, and often now used routinely.

Lau J, Leung W, Wu J et al. Omeprazole before endoscopy in patients with gastrointestinal bleeding. NEJM 2007; 356:1631- 40. [Reference]

Leontiadis GI, Sharma VK, Howden CW. Proton pump inhibitor treatment for acute peptic ulcer bleeding. Cochrane Database Syst Rev 2006;1:CD002094. [2006 Reference] [2010 Reference]

  • H2 blocker IV. Cheap, safe but poor ability to consistently maintain a high intragastric pH >6. No evidence for effect in acute bleeding.
  • Somatostatin or octreotide reduce rebleeding, need for transfusion and surgery, but with no improvement in mortality. Also not routine.

Endoscopy within first 24 hrs of admission:

  • Early endoscopy provides diagnosis, prognosis and allows immediate therapy. Reduces overall LOS. Bleeding source found in over 90%, and most (>80%) will need no more than supportive therapy initially.
  • Otherwise may need injection therapy with adrenaline first line for active bleeding +/- other procedure such as second injectate, thermal contact, clips etc.
  • Other sclerosants, thrombin, tissue glue, heater probe, multipolar electrocoagulation, laser, mechanical endoclips (haemoclips) are alternatives, as adjuncts or as monotherapy.

Sung J, Tsoi K, Lai L et al. Endoscopic clipping versus injection and thermo-coagulation in the treatment of non-variceal upper gastrointestinal haemorrhage bleeding: a meta-analysis. Gut 2007;56:1364-73. [Reference]

Surgery / angiography

  • Surgery if endoscopy fails, and for high risk of re-bleed in the elderly.
  • Angiography for severe, persistent bleeding in high risk patient unsuitable for surgery.  May then use intra-arterial gelatin, springs or tissue adhesive.

Barkun A, Bardou M, Kuipers E et al. International consensus recommendations on the management of patients with non-variceal upper gastrointestinal bleeding. Ann Intern Med 2010;152:101-13. [Reference]

Scottish Intercollegiate Guidelines Network (SIGN). Management of acute upper and lower gastrointestinal haemorrhage. Sept 2008. A national clinical guideline. [PDF Reference]

Print Friendly


  1. Brian O'Riordan says

    Terlipressin has been shown to reduce mortality versus placebo unlike somatostatin. Although no difference shown in studies comparing somatostatin and terlipressin. Regardless, many guidelines have terlipressin as first line, and somatostatin if terlipressin unavailable.