LETTERS

摘要

J. Bohannon, in his news of the week story "author of iraqi deaths study sanctioned" (6 March, p. 1278), quotes Les Roberts, a coauthor of the controversial Lancet survey (1) that estimated 601,000 violent deaths during the first 3.3 years of the Iraq war. Roberts emphasizes that the key to verifying the study's findings lies in replication.rnThe Lancet survey has already failed a replication test (2): The World Health Organization (WHO) published the results of its Iraq Family Health Survey (IFHS) in 2008 (3). This was a rigorous, well-supervised, and much larger survey than the Lancet study, and it estimated 151,000 violent deaths, compared with 601,000 violent deaths estimated by the Lancet survey for almost precisely the same time period.