Bleeding from stress-induced mucosal lesions continues to be a potential pr
oblem in critically ill patients. although its incidence has decreased dram
atically over the past decade. Patients considered to be at risk are those
with respiratory failure, coagulopathy, severe burns or tetraplegia. The mo
st important cause of stress ulcer bleeding is tissue hypoxia. Provided tha
t appropriate dosage regimens are administered, all agents approved for str
ess ulcer prophylaxis may reduce the incidence of overt as well as clinical
ly important bleeding. However, the efficacy of stress ulcer prophylaxis do
es not correlate with the efficacy of gastric acid inhibition. Although num
erous studies have demonstrated that an alkaline gastric juice is associate
d with gastric Gram-negative bacterial overgrowth, controversy remains over
whether the pharmacological suppression of gastric acid in critically ill
patients facilitates nosocomial pneumonia. The reasons for these divergent
results are discussed, as is a possible association between gastric acid su
ppression and other systemic infections. Finally, several cost-effectivenes
s analyses performed over recent years have demonstrated that, in properly
selected critically ill patients, stress ulcer prophylaxis is cost-effectiv
e.