with thrombocytopenia. Evidence from the study by Oski et al. (1966), individual reports in the literature, and VAERS suggests a causal relation between measles vaccine and thrombocytopenia. However, the measles vaccine strain (the live attenuated Edmonston B strain) studied by Oski is no longer used in the United States. The evidence concerning the live, more attenuated measles vaccine strain currently used in the United States is scarce. Although transient decreases in platelet counts following measles vaccination or from other nonvaccine causes may be common, clinically significant thrombocytopenia is extremely rare, especially considering the very large number of doses of measles vaccine that have been administered.
The reports of thrombocytopenia following mumps diseases provide biologic plausibility that mumps vaccine could be associated with thrombocytopenia. The evidence bearing on a causal relation between mumps vaccine and thrombocytopenia consists of one report of a child who had thrombocytopenia as part of hemolytic-uremic syndrome following receipt of mumps vaccine (Dosik and Tricarico, 1970).
Published reports of passive surveillance systems from several countries provide evidence that MMR is associated with clinically significant thrombocytopenia within two months of vaccination. On the basis of data from Finland and Sweden, the incidence appears to be on the order of 1 per 30,000 to 40,000 vaccinated children. This is a sixfold higher incidence than that reported in the only study of background incidence of thrombocytopenia identified by the committee (Cohn, 1976). The committee could not identify the component of MMR responsible for the thrombocytopenia, but the data from Oski and from the experience with wild-type measles virus suggest that the measles vaccine component of MMR might be responsible for the thrombocytopenia that occurs after MMR.
The evidence establishes a causal relation between MMR and thrombocytopenia. On the basis of data from Finland and Sweden, the incidence appears to be on the order of 1 per 30,000 to 40,000 vaccinated children.
The evidence is inadequate to accept or reject a causal relation between monovalent measles and mumps vaccines and thrombocytopenia.
Because so little information is available, the committee does not have the means to recommend any precautions to prevent clinically significant thrombocytopenia from occurring after administration of live attenuated measles vaccine or MMR. One child in the series by Nieminen and colleagues (1993) had had acute idiopathic thrombocytopenic purpura 9 months prior to the episode following MMR administration. Children with a prior his-