Gastrointestinal haemorrhage and calcium antagonists: a pharmacoepidemiological case non-case study

Calcium antagonists were found to be associated with an increased risk of g astrointestinal haemorrhage (GIH) in hypertensive patients over 67 years ol d (Pahor et al. Lancet 1996; 347 : 1061). This unexpected result led us to investigate this question using the French pharmacovigilance system databas e. We use the case/non case methodology (Moore et al. Br J Pharmacol 1997; 44: 513) where cases and non cases were both identified from the spontaneous a dverse drug reaction (ADR) reporting database. Cases were reports of the re action of interest (i.e. GIH as recorded in the database). Non cases were a ll reports of reactions other than being studied. Exposure was considered a s the presence in a report of the drug of interest (calcium antagonists), w hether or not it was suspected of causing the reaction. We calculated Odds ratios (OR) as the ratio of the Odds of the association of reports of GIH w ith calcium antagonists in cases and in non cases. Calcium antagonists incl uded in the present study were dihydropyridines, diltiazem, verapamil and b epridil. Salicylates and non steroidal antiinflammatory drugs were used as positive controls. Among the 112,792 ADRs recorded in the database between January 1985 and De cember 1996, 864 (0.8%) were GIH. There was no association between GIH and the exposure to calcium antagonists whatever the class of the drugs (OR=1.2 , 95% CI: [0.9; 1.6]. A subgroup analysis among the GIH reported in patient s over 65 years old (470 GIH from 37,462 ADRs) also failed to find any asso ciation (OR=0.7, 95% Cl : [0.5-1.0%]). The present results failed to confirm the hypothesis of an association betw een GIH and use of caldium antagonists.