Margaret A.WinkerMD, Senior EditorIndividualAuthorPhil B.FontanarosaMD, Senior EditorIndividualAuthor
In Reply.—Dr Fremont-Smith and Dr Kravitz both raise important concerns about whether the point estimate for ADR mortality by Dr Lazarou and colleagues1 is accurate and note that most of the data allowing estimation of this rate are more than 20 years old. Both suggest that the true rate of fatal ADRs is likely lower than the final point estimate reached by Lazarou et al, and I agree. One study that can serve as an external standard to address this issue and has the methodological advantage of being population based is the Harvard Medical Practice Study (MPS).2 The MPS estimated the magnitude of iatrogenic injury in New York in 1984; when its data are extrapolated, the national mortality rate due to all types of adverse events was 180,000.3 Drugs represented only 14.1% of serious injuries (those resulting in death, permanent disability, or disability lasting longer than 6 months), although the percentage resulting in death was not reported. Nonetheless, it would be surprising if there were actually 110,000 deaths annually from drugs alone; based on the MPS data, the expected figure would be approximately 25,000.
Bates DW. Adverse Drug Reactions in Hospitalized Patients—Reply. JAMA. 1998;280(20):1741. doi:10-1001/pubs.JAMA-ISSN-0098-7484-280-20-jac80013