The relation of acute thrombocytopenic purpura (TP) to the use of drugs was investigated in a case-control study conducted in eastern Massachusetts, Rhode Island, and the Philadelphia region; 62 cases over the age of 16 years with acute onset and with a rapid recovery were compared with 2,625 hospital controls. After control for confounding by multiple logistic regression, use of the following drugs in the week before the onset of symptoms was significantly associated: trimethoprim/sulfamethoxazole (relative risk [RR] estimate, 124), quinidine/quinine (101), dipyridamole (14), sulfonylureas (4.8), and salicylates (2.6). The overall annual incidence of acute TP was estimated to be 18 cases per million population. The excess risks for the associated drugs were estimated to be 38 cases per million users of trimethoprim/sulfamethoxazole per week, 26 per million for quinidine/quinine, 3.9 per million for dipyridamole, 1.2 per million for sulfonylureas, and 0.4 per million for salicylates. Associations with sulfonamides, quinidine/quinine, sulfonylureas, and salicylates have been previously reported, but the present study has provided the first quantitative measures of the risk. The association with dipyridamole was unexpected. In general, despite large RRs, the incidence rates attributable to the drugs at issue (excess risks) were low, suggesting that TP is not an important consideration in the use of the various drugs.
