Families could have reported deaths that did not occur, although this seems unlikely, since most reported deaths could be corroborated with a certificate. However, certificates might not be issued for young children, and in some places death certificates had stopped being issued; our 92% confirmation rate was therefore deemed to be reasonable.
Although interviewers used a robust process for identifying clusters, the potential exists for interviewers to be drawn to especially affected houses through conscious or unconscious processes. Although evidence of this bias does not exist, its potential cannot be dismissed.
Burhnam et al. (2006)
The rate found before the war (5.4 per 1000) is very close to other data for that period (around 5.5 per 1000). It suggests that the sample was representative of that period, was it representative of post-war Iraq?
ABE: I see modulous never linked this, it's a companion article to the actual lancet study...
Home | MIT Center for International Studies
Edited by melatonin, : added link