"This is not some trivial academic disagreement," says Andrew Mack, director of the Simon Fraser University-based Human Security Report Project (HSRP), which published a detailed critique of the BMJ's claims in the December issue of the Journal of Conflict Resolution (JCR).

"Accurate statistics on the health impacts of war are critically important not just for researchers but also for humanitarian organizations whose assistance programs save millions of lives around the world."

The HSRP article, "Estimating War Deaths: An Arena of Contestation," takes issue with a 2008 BMJ piece by Harvard researcher Ziad Obermeyer and colleagues, which rejected widely accepted battle-death statistics compiled by the International Peace Research Institute, Oslo (PRIO).

The PRIO data reveal that global war deaths declined by more than 90 per cent between 1946 and 2002.

But the BMJ claimed in a release publicizing the Obermeyer article that "war has killed three times more people than previously estimated and there is no evidence to support claims of a recent decline in war deaths."

Mack and HSRP study co-author Michael Spagat at the University of London, both leading experts on the subject, determined that the BMJ article failed to substantiate either assertion, was marred by serious mistakes and showed little understanding of the research it critiqued.

So Mack, the former strategic planning director to UN Secretary-General Kofi Annan, emailed the BMJ identifying six major methodological and factual errors in the Obermeyer article "Fifty Years of Violent War Deaths from Vietnam to Bosnia."

"The problems we point out would have been obvious to any of the many conflict researchers familiar with the research being criticized," says Spagat.

The U.K. journal acknowledged none of the errors, he says, and it ruled out publishing a detailed rebuttal, claiming the Obermeyer study was very thoroughly reviewed by appropriate experts.

"But the BMJ is well aware that its peer review process is flawed," says Spagat. "A recent study, whose authors include the journal's current editor, revealed that, on average, only a third of the 'major errors' deliberately inserted in a BMJ article were picked up by reviewers."

Adds Mack: "There appears to be no way of effectively rebutting BMJ articles that contain unwarranted -- and damaging -- critiques of the work of other scholars.

"This makes the journal effectively unaccountable by shielding it from serious criticism."

Note: The JCR paper and technical appendix, PRIO battle deaths data, the BMJ paper and technical appendix, and a paper on the BMJ peer review are all available at http://www.hsrgroup.org/index.php?option=content&task=view&id=469.

Note: If no author is given, the source is cited instead.

• Today's Healthcare
• Accident and Trauma
• Chronic Illness

• Conflict
• Scientific Conduct
• Social Issues

• Physical trauma
• Sudden infant death syndrome
• Tobacco smoking
• Evidence-based medicine
• Epidemiology
• Mediterranean diet