BMJ  2006;333 (16 December), doi:10.1136/bmj.333.7581.0-b

This week in the BMJ

Journals omit absolute risks

Absolute risks should be routinely included in abstracts, adjacent to any reported risk ratio, to allow readers to discern the meaning of ratio measures. In a structured review of the accessibility of absolute risk data in six leading journals, including the BMJ, Schwartz and colleagues (10.1136/bmj.38985.564317.7C) examined 222 articles with study designs that allow absolute risks to be calculated. They found 68% of articles failed to report absolute risks in the abstract, and half of these did not report them anywhere in the article.

Related Article

Ratio measures in leading medical journals: structured review of accessibility of underlying absolute risks

Lisa M Schwartz, Steven Woloshin, Evan L Dvorin, and H Gilbert Welch
BMJ 2006 333: 1248. [Abstract] [Full Text] [PDF]