Nonsteroidal anti-inflammatory drugs: add an anti-ulcer drug for patients at high risk only. Always limit the dose and duration of treatment with NSAIDs.

In addition to their cardiac, renal, hepatic, cutaneous and neuropsychological adverse effects, nonsteroidal anti-inflammatory drugs (NSAIDs) can have severe effects on the entire gastrointestinal tract, including bleeding, perforation and occlusion. Which anti-ulcer drugs reduce the risk of the severe gastrointestinal adverse effects of NSAIDs, and which patients should receive them? To answer these questions, we conducted a review of the literature, using the standard Prescrire methodology. The main risk factors for severe gastrointestinal adverse effects during NSAID therapy are: a high dose regimen; age over 65 years; a history of gastric or duodenal ulcer or gastrointestinal bleeding; heavy use of both alcohol and tobacco; and concomitant treatment with a corticosteroid, antiplatelet drug, anticoagulant, or selective serotonin reuptake inhibitor (SSRI) antidepressant. Gastrointestinal symptoms and ulceration (on endoscopy) are poor predictors of severe gastrointestinal reactions. A meta-analysis examined randomised placebo-controlled trials of misoprostol in more than 11 000 patients. The results were mainly based on a large trial including about 9000 rheumatoid arthritis patients with an average age of 68 years. Misoprostol (400 microg to 800 microg/day, in 4 doses) prevented about 4 severe gastroduodenal events when 1000 patients over 60 years of age were treated for 6 months. Diarrhoea and other mild gastrointestinal disorders were frequent. There are no randomised trials comparing proton pump inhibitors (PPIs) and histamine H2 receptor antagonists versus misoprostol or versus placebo therapy for the prevention of severe adverse effects associated with NSAIDs. PPIs and H2 antagonists both reduce the incidence of gastric or duodenal ulceration detected by routine endoscopy. A randomised trial compared an H2 antagonist (famotidine) versus a PPI (pantoprazole) in 128 patients with an average age of 69 years who had a very high risk of serious gastrointestinal adverse effects while taking low-dose aspirin. After 48 weeks of treatment, pantoprazole was more effective than famotidine for the prevention of overt gastrointestinal bleeding. The symptomatic effects of PPIs and H2 antagonists may create a false sense of security, leading some patients to increase their NSAID use and resulting in a paradoxical increase in severe gastrointestinal effects. In practice, anti-ulcer drugs are not sufficiently effective to warrant their use by NSAID-treated adults who are not at high risk of severe gastrointestinal events. Misoprostol has proven efficacy in patients with risk factors for NSAID-induced severe gastroduodenal adverse effects, especially patients over 65 years of age, but it also has frequent adverse effects and necessitates 4 daily doses. Omeprazole is an alternative when the adverse effects or dosing frequency of misoprostol are unacceptable, provided patients are warned not to increase their NSAID consumption.