Anti-inflammatory drug intake and the risk of ulcer complications.

Over the years conflicting data have been produced suggesting that the risk of upper gastrointestinal bleeding or perforation associated with anti-inflammatory drug intake could be negligible or substantial. Opinions have diverged for a number of reasons. Data concerning the use of aspirin have generally failed to include adequate controls or to distinguish analgesic intake which might be causal from that which might be consequential upon the presence of the bleeding lesion. When proper allowance is made, it seems likely that about one-third of aspirin intake in patients with bleeding is equivalent to that in controls and is by deduction non-causal; a further one-third by reference to parallel increases in paracetamol (acetaminophen) intake is non-causal but consequential upon the presence of the bleeding lesion. The remaining one-third is unexplained and likely to be causal. The importance of non-aspirin non-steroidal anti-inflammatory drug (NANSAID) intake has likewise been disputed. Opinions have conflicted because evidence has not generally been available from case-control studies but has been obtained either from simple case series or from the results of post-marketing surveillance. By formal case-control studies it has been possible to show that the risks of ulcer complications in individuals over the age of 60 are appreciable. These risks have not been detected during post-marketing surveillance because the apparently large case series studied during surveillance are dwarfed by the general extent of prescribing, and because elderly people may be particularly at risk and surveillance has not concentrated on this group.