Purpose: To determine the strength of clinical evidence for individual drugs as a cause of thrombocytopenia.
Data sources: All English-language reports on drug-induced thrombocytopenia.
Study selection: Articles describing thrombocytopenia caused by heparin were excluded from review. Of the 581 articles reviewed, 20 were excluded because they contained no patient case reports. The remaining 561 articles reported on 774 patients.
Data extraction: Two of the authors used a priori criteria to independently review each patient case report. Two hundred fifty-nine patient case reports were excluded from further review because of lack of evaluable data, platelet count of 100000 cells/microL or more, use of cytotoxic or nontherapeutic agents, occurrence of drug-induced systemic disease, or occurrence of disease in children. For the remaining 515 patient case reports, a level of evidence for the drug as the cause of thrombocytopenia was assigned. Data on bleeding complications and clinical course were recorded.
Data synthesis: The evidence supported a definite or probable causal role for the drug in 247 patient case reports (48%). Among the 98 drugs described in these reports, quinidine was mentioned in 38 case reports, gold in 11, and trimethoprim-sulfamethoxazole in 10. Of the 247 patients described in the case reports, 23 (9%) had major bleeding and 2 (0.8%) died of bleeding.
Conclusions: Many reports of drug-induced thrombocytopenia do not provide evidence supporting a definite or probable causal relation between the disease and the drug. Future patient case reports should incorporate standard criteria to clearly establish the etiologic role of the drug.