Upper gastrointestinal bleeding (UGIB) is a common gastrointestinal emergency defined as bleeding with a source proximal to the ligament of Treitz.
The ligament of Treitz (suspensory ligament of the duodenum) marks the boundary between the upper and lower GI tracts. It arises from the right crus of the diaphragm and suspends the duodenojejunal flexure.
The incidence of acute UGIB is around 1 in 1,000 per year, with a mortality of 7-10%. UGIB occurs more frequently in males than females (♀ 1: 2 ♂).
Patients commonly present with haematemesis and/or melaena.
There are numerous causes of UGIB, which can be classified by anatomical location.
Peptic ulcer disease (PUD) is an umbrella term for the development of gastric and duodenal ulcers; it accounts for up to 50% of UGIB.
Peptic ulcer disease is strongly associated with Helicobacter pylori infection.
The classical description of bleeding in PUD is of a posterior duodenal ulcer eroding through into the gastroduodenal artery. However, most UGIB secondary to PUD occurs due to erosions through smaller-sized blood vessels within the submucosa.
For more information, see our notes on 'Peptic ulcer disease'.
Gastritis refers to inflammation of the lining of the stomach; erosive gastritis accounts for 15-20% of UGIB.
Gastritis refers to inflammation of the stomach associated with mucosal injury.
The term ‘gastritis’ is also used to describe the appearance of ‘inflamed’ mucosa on endoscopy. This can be rather misleading as many irritants (e.g. bile, NSAIDs, alcohol) can cause epithelial cell injury without resultant inflammation, which is termed ‘gastropathy’. Consequently, a biopsy is usually required during endoscopy to determine the extent of the damage.
Gastritis can be classified based upon location, timing or type of pathology:
Oesophageal varices are abnormal, dilated veins that occur at the lower end of the oesophagus; they account for 10-20% of UGIB.
Oesophageal varices occur secondary to portal hypertension. The most common cause being chronic liver disease / cirrhosis.
Increases in portal pressure lead to the development of a collateral circulation to overcome the obstruction to flow in the portal system. The lower end of the oesophagus forms an important ‘portacaval anastomosis’ which allows the flow of venous blood from the portal system to the systemic circulation.
Other portacaval anastomoses include:
Oesophageal varices can cause massive haemorrhage.
A Mallory-Weiss tear refers to a liner mucosal laceration; they account for 5-10% of UGIB.
They typically occur at the gastro-oesophageal junction or within the gastric cardia.
The classical description is of an episode of haematemesis (typically blood-streaked vomitus), which follows repeated episodes of retching / vomiting.
The lesion may be seen on upper GI endoscopy, but most cases resolve spontaneously.
Osler-Weber-Rendu syndrome, also called hereditary haemorrhagic telangiectasia (HHT), is a rare condition characterised by mucocutaneous telangiectasias (small, visible dilated blood vessels) and arteriovenous malformations (an abnormal connection between artery and vein). The condition is inherited in an autosomal dominant pattern.
Gastric antral vascular ectasia (GAVE) is a rare cause of severe acute and chronic gastrointestinal bleeding. This may lead to chronic iron-deficiency anaemia or an acute UGIB. The pathogenesis of GAVE is largely unknown.
GAVE is also termed “watermelon stomach” because of the classical endoscopic appearance - ‘red tortuous ectatic vessels along the folds of the antrum’.
Several risk factors increase the likelihood of a patient developing an UGIB.
NSAIDs inhibit the synthesis of prostaglandins, which are gastroprotective.
Prostaglandins work by inhibiting enterochromaffin-like cells, which are involved in the secretion of histamine. Histamine stimulates parietal cells to secrete hydrochloric acid. Therefore, inhibition of prostaglandins leads to excessive HCl secretion and damage to the underlying mucosa.
The two characteristic clinical findings of UGIB are haematemesis and melaena.
Haematemesis refers to vomiting blood. Haematemesis may present with bright red bleeding or as ‘coffee-ground’ vomitus.
Melaena is the passage of ‘black, tarry stool’, which has an offensive odour. The colour is due to partly digested blood.
Upper GI endoscopy is the main diagnostic test. It should be completed immediately in any unstable patient following initial resuscitation, or within 24 hours in all other patients.
During resuscitation, initial bedside and blood tests can be completed. This is important for the acquirement of blood products (if required).
All patients presenting with an UGIB should initially be resuscitated with respects to airway (A), breathing (B) and circulation (C).
Following resuscitation, unstable patients should be transferred for an immediate endoscopy. It is recommended that all other patients have endoscopy within 24 hours of admission.
Management can then be divided according to ‘non-variceal’ or ‘variceal' bleeding.
A number of techniques can be employed at the time of endoscopy to treat non-variceal causes of UGIB:
Proton pump inhibitor therapy should be reserved for patients with a non-variceal UGIB with evidence of recent haemorrhage during endoscopy.
A repeat endoscopy should be completed in patients who re-bleed, or are suspected to be high-risk of re-bleeding. Unstable patients who re-bleed post-endoscopy should be offered radiological or surgical intervention.
Early risk stratification helps identify high-risk patients & need for prompt intervention. Two scoring systems are used in patients presenting with UGIB.
The Rockall score is comprised of both a pre- and post-endoscopy score that can be added together to give an overall value.
The pre-endoscopy score is composed of three sections:
Patients with a score of 0 are at low risk of re-bleeding and death. This group of patients (approx. 15%) may be discharged early or not admitted.
The post-endoscopy score is composed of two sections:
This score can be remembered using the mnemonic ABCDe
The Blatchford score takes into account a number of different clinical findings, biochemical parameters and past medical history.
NICE recommends the use of the Blatchford score at first assessment, followed by the Rockall score post-endoscopy.
Similar to the Rockall score, a score of 0 on the Blatchford score is associated with a low risk of mortality and patients can be considered for early discharge and non-admission.
Have comments about these notes? Leave us feedback