Yesterday, the Lancet published a report stating that the number of people killed in Iraq "since the invasion" was 650,000. The figure was arrived at by a linear extrapolation calculated from a random sampling of 1,800 Iraqi households in 47 areas across the country.
I don't there is much more to say than the last sentence really. Calculating the number of people killed by taking a random sampling data is not exactky sound practice. What's more the extrapolation claims to be "since" the invasion, even if it were right it does not necessarily mean the deaths are "because" of the invasion.
The Lancet's report has however been picked up on by the likes of MPAC and will be milked for every last drop.