Well, here it is:
When including Falluja, it is impossible to have a precise estimate of post-invasion mortality. [p 3]The goal it seems is to show that the study is wrong. And what is the motivation to show that it is wrong? Yummy soup indeed!
Yet excluding Falluja is not “conservative.” In fact, including this cluster — i.e., using all the available data — generates a result with such a wide confidence interval that the reported increase in Iraqi mortality becomes statistically insignificant. [p 5]
From the calculations above, it is impossible to be 95% confident that there was an increase in mortality. The lower bound of the confidence interval for the relative risk can not be 1.6, as reported in L1. It must be much lower. [last paragraph]