I claim no special knowledge or expertise in this area, but I do work with spatially oriented statistical data and I've learned to trust my intuition in these matters. An order of magnitude discrepancy compared to previous studies, particularly when those studies are from experienced and respected investigators with no incentive to back-pedal the numbers, leads me to suspect that the burden of proof here lies with the Lancet figures.

I suspect there is a methodological or ideological bias in these figures (probably the former, I'm not into conspiracies), and until shown otherwise by convincing alternative investigations, I am very skeptical about these figures. Numbers of this magnitude would have manifested themselves in other ways and would be very difficult to conceal in a theater filled with agents whose interest would be to publicize them as much as possible.

Until I'm shown otherwise, my money is with Tom on this one. BTW, I didn't mention it while the controversy was raging, but I feel the same way about the Israeli cluster-bomblet issue, and for much the same reasons.