Do you remember that contentious paper in the Lancet in October, claiming that the invasion of Iraq had led to the deaths of 655,000 Iraqis? After its publication world leaders queued up to denounce the methodology. Away from the headlines, scepticism continues to harden.

Peter Lynn, Professor of Survey Methodology at the Institute of Social and Economic Research, University of Essex, has been quietly investigating and, despite several e-mails to the researchers, has been unable to get answers: “I can’t come out and say I support the paper, or say that it’s rubbish, because I haven’t been able to find out enough. But the fact that they are dragging their feet on answering my questions makes me sceptical.” Had the paper been submitted to a statistics journal instead, Professor Lynn says, it might not have survived peer review.

So, back to the original paper, by Professor Gilbert Burnham and colleagues, from Johns Hopkins University in Maryland. A team of Iraqi doctors visited 1,849 households. They knocked on doors, and asked the person who answered whether anyone in that household had died recently, and what the cause was. Death certificates were obtained for 92 per cent of reported deaths, confirming the interviewees’ stories. The prewar death rate was 5.5 people per thousand per year; the postwar death rate, the researchers found, was 13.3. The excess deaths — the number of deaths above what would have been expected normally — amounted to 655,000 between 2003 and 2006, of which 601,000 were due to violence. This equates to 2.5 per cent of the Iraqi population. The new figure was ten times higher than any other estimate, even by human rights organisations.

But how were the 1,849 households chosen? Professor Lynn says: “The researchers made a list of all the roads intersecting the main road, and took one of those at random. They then went to 40 adjacent addresses going up one side. But these were all near the main road, so streets away from the main road may not have been represented.” Iraq is plagued by car bombs, which, the professor speculates reasonably, are more likely to affect those living near main roads. He also adds that it is difficult to guarantee the continuity of households over several years. If bereaved relatives separate, for example, double-counting may result.

Professor Lynn muses that, especially in a war zone, incomplete information is often better than none at all. Is that the case here? “It is difficult to say. Sometimes, misleading information is worse than none at all.”

• To the British Museum, where Sir John Sulston, the biologist who shared a Nobel prize for his contribution the Human Genome Project, was holding forth on the evils of the market, Big Pharma and private health insurance. How his tirade must have delighted his hosts: BP (stock market value £119 billion) and the Financial Times.