[Skip to Content]
[Skip to Content Landing]
Citations 0
September 8, 1999

Higher Standards for Epidemiologic Studies—Replication Prior to Publication?—Reply

Author Affiliations

Margaret AWinkerMD, Deputy EditorIndividualAuthorPhil B.FontanarosaMD, Interim CoeditorIndividualAuthor

JAMA. 1999;282(10):937-938. doi:10-1001/pubs.JAMA-ISSN-0098-7484-282-10-jbk0908

In Reply: Dr Weed is critical of my suggestion that epidemiologists consider not reporting small relative risks (perhaps 2 or 3) until the observation is replicated in a different environment. The replication, in addition to showing that the observation can be repeated, would be more stringently designed, with fully detailed prospectively described protocols and a single hypothesis. Weed gives 4 reasons for believing that "mandating replication prior to publication is a bad idea." None seems persuasive.