Sir, I am writing with regard to the recent Reviews in Medicine article 'Clinical pharmacology and therapeutics' by M.J. Kendall and R.C. Horton.' Several points are made in the section 'Peptic ulcer disease' (p. 174) about the prostaglandin analogue misoprostol (note correct spelling) which do not accurately reflect the available data. The authors incorrectly state that misoprostol does not prevent non-steroidal anti-inflammatory drug (NSAID) induced duodenal lesions. Several studies have shown its efficacy both in the prevention2-6 and healing5'7'8 of NSAID-induced duodenal lesions. It is important to note, as the authors correctly do, that NSAID-induced damage has been shown in most studies to be predominantly gastric. Thus from the clinician's viewpoint when prescribing an agent to protect against NSAID-induced damage it is important to use an agent such as misoprostol which will prevent both gastric and duodenal damage. The authors also incorrectly state that misoprostol is no better than placebo at preventing pain. Several studies have demonstrated that misoprostol is superior to placebo.9-" The authors state that 'how one identifies this population' (of patients likely to develop complications) is at present unknown. In fact there are now considerable data indicating various groups ofpatients who are at increased risk of developing complications if they take NSAIDs. Two large studies, one from the US'2 and one from Great Britain'3 have identified various 'at risk' groups, in particular the elderly and those with a past history of gastrointestinal problems. These groups have been confirmed in another recent study.'4 These data have recently been concisely reviewed.'5 In conclusion misoprostol has been shown to prevent NSAID-induced gastric and duodenal damage and certain 'at risk' groups of patients can be identified. Studies to address the question whether the use of misoprostol will prevent serious morbidity are underway. Meanwhile the conclusion in a recent review'6 regarding this aspect of the use of misoprostol would appear to be scientifically justified 'it seems likely (though unproven) that normalization of the gastric mucosa will reduce the likelihood of gastric ulcer complications'.