Communicating risks at the population level: application of population impact numbers

Richard F Heller and colleagues have reported an error in their Education and Debate article (15 November, pp 1162-5) that was due to a little recognised problem in calculating population attributable risk for multiple levels of exposure. This led to an overestimation of population attributable risk in table 2, which shows the impact of blood cholesterol concentration on premature death from coronary heart disease (p 1164). However, this does not alter the general conclusion drawn from the table (that the population impact of cholesterol concentrations of 5.2-6.5 mmol/l and of 6.5-7.8 mmol/l is larger than that of concentrations above 7.8 mmol/l) or the substance of the article. Full details of the correct calculations and the corrected table 2 appear on bmj.com (http://bmj.bmjjournals.com/cgi/content/full/327/7424/1162/DC1).