Haematemesis is simply defined as “vomiting blood”. It is caused by bleeding from part of the upper portion of the gastrointestinal tract.
It has a wide range of possible causes, depending on the site of blood loss and the tissue that is actively bleeding. Indeed, patients with haematemesis can present in a number of clinical states.
In this article, we shall look at the causes, investigations and management of haematemesis.
Oesophageal varices refer to dilations of the porto-systemic anastomoses in the oesophagus. These dilated anastomoses are prone to rupture, with the potential to cause a catastrophic haemorrhage.
The most common underlying cause for oesophageal varices is alcoholic liver disease, causing portal hypertension, and any haematemesis in a patient with known history of alcohol abuse should be investigated with an urgent OGD.
Gastric ulceration is responsible for about 60% of haematemesis cases. Ulceration can result in erosion into the blood vessels supplying the upper GI tract (most commonly the lesser curvature (20%) or posterior duodenum (40%)) and can result in significant haemorrhage.
Patients may present with known active ulcer disease / H. Pylori positive, a history of NSAID or steroid use, or previous epigastric symptoms suggesting peptic ulceration, all of which may aid your initial assessment and diagnosis.
A Mallory-Weiss tear is a relatively common phenomenon, typified by episodes of severe or recurrent vomiting, then followed by minor haematemesis. Such forceful vomiting causes a tear in the epithelial lining of the oesophagus, resulting in a small bleed.
Most cases are benign and will resolve spontaneously, therefore providing the patient reassurance and monitoring is usually all that is required. Any prolonged or worsening haematemesis in a suspected Mallory-Weiss tear may warrant further investigation.
Oesophagitis is a condition that describes inflammation of the intraluminal epithelial layer of the oesophagus, most often due to either gastric acid reflux (GORD) or less commonly from infections (typically Candida Albicans), medication (such as bisphosphonates), radiotherapy, ingestions of toxic substances, or Crohn’s disease.
Other causes of haematemesis that may not require such immediate resuscitation and intervention include gastritis, gastric malignancy, Meckel’s diverticulum, or vascular malformations (e.g. Dieulafoy lesion)
The key facts to ascertain from a history of haematemesis are:
- Timing, frequency, and the volume of bleeding
- History of dyspepsia, dysphasia or odynophagia
- Past medical history and smoking and alcohol status
- Use of steroids, NSAIDs, anticoagulants, or bisphosphonates
On examination, it is important to assess specifically for epigastric tenderness, evidence of varices or liver stigmata, or a point of bleeding in the mouth or nose.
Investigations are obviously tailored to the specific presentation, but a generic outline would involve a structured approach as discussed below.
A full set of blood tests should be taken, including:
- Useful in bleeding or septic patients, especially for the pH, pO2, pCO2, and lactate, for signs of tissue hypoperfusion.
- Routine bloods (FBC, U&Es, LFTs, and clotting)
- Any acute bleed may not initially show an anaemia in the FBC, whereas LFTs may reveal underlying liver damage as a potential cause.
- All patients with haematemesis should have a Group and Save; those with significant haematemesis (especially suspected variceal bleed) should have at least 4 units of blood cross-matched.
The definitive investigation in most cases of haematemesis is via an oesophagogastroduodenoscopy (OGD); which also forms part of the management in cases of ongoing unstable bleeding. This should be performed within 12hrs in most cases of acute haematemesis, or as soon as possible if the patient is unstable.
An erect CXR may also be required if a perforated peptic ulcer is suspected as the underlying cause. In such a case, air may be visible underneath the diaphragm (pneumoperitoneum).
Glasgow-Blatchford Bleeding Score
The Glasgow-Blatchford bleeding score (GBS) is a scoring system used to risk stratify patients admitted with an upper GI bleed, based purely on clinical and biochemical parameters. This allows for appropriate management of further investigations, especially as the score can be calculated prior to any OGD.
The score is determined by:
|Known Hepatic Failure||Present|
Interpretation will vary across endoscopy departments, but scores ≥6 have been associated with a >50% risk of needing an intervention. Other risk scores are used in the clinical setting, such as the AIMS65 Score (risk score for in-hospital mortality from upper GI bleeding) or Rockall Score (severity score for GI bleeding post-endoscopy)
Any acute upper GI bleed warrants careful resuscitation and urgent endoscopy. Begin with a standard ABCDE approach, gaining 2 large bore cannulae, IV fluid, and blood products (if required).
During OGD, a range of therapeutic options are available depending on the underlying causes suspected or confirmed:
- Peptic ulcer disease – requires injections of adrenaline and cauterisation of the bleeding. High dose intravenous PPI therapy should be administered (e.g. IV 40mg omeprazole) to reduce acid secretion +- H. Pylori eradication therapy if necessary
- Oesophageal varices management should be swift and performed at the same time as active resuscitation, including the use of blood products.
- Endoscopic banding is the most definitive method of management however can be technically difficult.
- Prophylactic antibiotic therapy should be initiated.
- Somatostatin analogues (e.g. terlipressin or octreotide) should also be started, acting to reduce splanchnic blood flow and hence reduce bleeding.
- Sengstaken-Blakemore tube can be used in severe cases. It is inserted to the level of the varices and inflated to compress the bleeding.
Long term management warrants repeat banding of the varices and long-term beta-blocker therapy.
- Regardless of the underlying cause, adequate initial resuscitation should always be priority
- Most cases of haematemesis should be investigated with an OGD
- Various scoring systems are available to allow for the risk stratify of patients admitted with an upper GI bleed