That Lancet article uses the figure "100,000" but that number is extrapolated from 79 violent deaths. Indeed, the magazine is respected, but that particular study is likely flawed. However, since the USA has no right to be in Iraq in the first place--although it probably is anticipating to do business in Iraq in the future--then even one death is wrong. And certainly there have been more than that. BTW, the study is about two years old now.