k
so
30% non-response and you have 80/1000 babies dying at birth in Fallujah
"There were 34 deaths in the age group 0–1 in this period giving a rate of 80 deaths per 1,000 births."
So
Say that the 30% non-respondents all had NO baby deaths, thus giving you the maximum possible bias against the study's conclusion that the Fallujah death rate is elevated
(if I do something wrong please point it out here)
So (1000/0.7)=1428
So 80 dead babies per 1428 births
So your new ratio per thousand (including the 30% perfect health fallujahns) is 56/1000
From the study:
"This may be compared with a rate of 19.8 in Egypt (RR = 4.2 p < 0.00001) 17 in Jordan in 2008 and 9.7 in Kuwait in 2008."
It seems (to me) that even if you had no dead babies in the 30% of surveyed Fallujahns that didn't respond, you'd still have an elevated dead baby rate in Fallujah.
Compared to Egypt, Jordan and Kuwait, sure. But since we don't know what was the death rate of infants in Fallujah before the war and we don't know how other effects of the war affected them, we can't conclude it was an environmental hazard caused by DU.
McHrozni