Abstract

The efficacy of cimetidine in the prevention of gastrointestinal haemorrhage in a general intensive care unit was evaluated in 221 patients by a placebo controlled double blind randomised study. Criteria for bleeding were (i) haematemesis or gastric aspirate greater than 50 ml fresh blood, (ii) melaena or fresh blood per rectum with an upper source verified by endoscopy if the gastric aspirate was clear, (iii) a fall in haemoglobin level greater than 2 g/dl in a 24 hour period associated with either 4+ occult blood in stools, or coffee ground gastric drainage of at least 100 ml. The drug and placebo groups were similar for age, sex, duration of study and risk factors. One hundred and fourteen received cimetidine and 107 placebo. Only 8% of the patients bled with no significant difference between the two groups (6/114 cimetidine, 11/107 placebo; p = 0.16). There was no correlation between the frequency of bleeding and either the number of risk factors per patient or the duration of intensive care unit stay. Thirteen patients died in each study group, resulting in overall mortality of 12%. The low incidence of haemorrhage, the lack of statistical benefit from cimetidine and the similar mortality all argue against the routine use of this drug in intensive care unit patients.