HAEMATEMESIS AND MELAENA

HAEMATEMESIS AND MELAENA

Upper gastrointestinal haemorrhage remains a major medi - cal problem with an incidence of over 100/100 000 per year in Western practice. The incidence increases with age. Haemorrhage is strongly associated with NSAID use. Despite improvements in diagnosis and the proliferation in tr eatment modalities over the last few decades, an in-hospital mortality of 5–10% can be expected. This rises to 33% when bleeding is first observed in patients who are hospitalised for other reasons. In patients in whom the cause of bleeding can be found, the most common causes are peptic ulcer, erosions, Mallory–Weiss tear and bleeding oesophageal varices ( Table 67.4 ). Whatever the cause, the principles of management are identical. First, the patient should be adequately resuscitated and, following this, should be investigated urgently to deter mine the cause of the bleeding. Intravenous access should be established and, for those with severe bleeding, central venous pressure monitoring should be established and bladder catheterisation performed. Blood should be cross-matched and the patient transfused as clinically indicated, usually when >30% of blood volume has been lost (see Chapter 2 ). There is no evidence for the use of intravenous PPI prior to endos copy . As a general rule, most gastrointestinal bleeding will stop, albeit temporarily , but there are sometimes instances when this is not the case. In these circumstances, resuscita tion, diagnosis and treatment should be carried out simultaneously . There are occasions when life-saving manoeuvres have to be undertaken without the benefit of an absolute diagnosis. In some patients, bleeding is secondary to a coagulopathy . The most important current causes are liver disease and anticoagulation therapy . In /uni00A0 these circumstances the coagulopathy should be corrected, Timothy Alexander Rockall , contemporary , Royal Surrey County Hospital, Guilford, Surrey , UK. factors with haematology advice. Upper gastrointestinal endoscopy should be carried out by an experienced operator as soon as practicable after the patient has been stabilised. In patients in whom the bleeding is r ela - tively mild, endoscopy may be carried out on the morning after admission; this is usually guided by local policy . In all cases of severe bleeding, it should be carried out immediately . A num - ber of scoring systems have been advocated f or the assessment of rebleeding and death after upper gastrointestinal haemor - rhage. The Rockall score ( Table 67.5 ) can be used in a pre - endoscopy f ormat to stratify patients to safe early discharge and, post endoscopy , it can relatively accurately predict rebleeding and death.

Condition Percentage Ulcers 60 Oesophageal 6 Gastric 21 Duodenal 33 Erosions 26 Oesophageal 13 Gastric 9 Duodenal 4 Mallory–Weiss tear 4 Oesophageal varices 4 Tumour 0.5 Vascular lesions 0.5 Others 5 TABLE 67.5 The Rockall scoring system of bleeding severity. Score 0 1 2 3 Age <60 60–79 Pulse >100 /uni00A0 bpm Shock Pulse <100 /uni00A0 bpm Systolic BP <100 /uni00A0 mmHg Systolic BP

100 /uni00A0 mmHg Comorbidities None Endoscopic signs None/dark spot of bleeding All other diagnoses Malignancy of the upper Diagnosis Mallory–Weiss syndrome/no pathology BP , blood pressure; bpm, beats per minute.