Rebuttal

We very much agree with the guidelines' stated approach of assessing coronary artery disease (CAD) risk, as opposed to relying solely on lipid levels. However, Genest and his colleagues did not do a good job of estimating the individual and population impact of their guidelines.1 The 2003 guidelines contain no information on the benefit of nonpharmacological interventions and no estimates on the absolute benefit of statins or other drugs. In their response they have provided no data to refute our position that, compared with the 2000 guidelines, the 2003 guidelines will expand statin treatment recommendations to hundreds of thousands more people at relatively low risk and increase expenditures on statins by hundreds of millions of dollars, resulting in only small additional reductions in the number of CAD-related deaths. At the same time, the guidelines inappropriately disregard 193 000 high-risk people who potentially would have a large benefit from statins. The guidelines are both more costly and less effective than the New Zealand guidelines.2

Instead, Genest and colleagues3 quibble about the data and methods we used (the same data and methods that 3 of the authors have used themselves to assess screening recommendations4), quibbles that in no way change the overall results of the analysis. Most of their comments have already been addressed in the online appendix (www.cmaj.ca/cgi/content/full/172/8/1027/DC1).

Their only substantive comment relates to the target threshold for the low-risk group. Their “clearly stated” low-density lipoprotein cutoff point for the very-low risk group can be found in a small-print footnote in 1 table of the guidelines. It states that “treatment may be deferred” for people with a 10-year baseline risk of cardiovascular disease less than 5% and low-density lipoprotein cholesterol levels less than 5.0 mmol/L. Modifying our results to reflect no statin treatment in this group would result in a 6-fold instead of a 10-fold increase in the number of very-low-risk and low-risk people for whom statins are recommended (increasing from 61 000 people in the 2000 guidelines to 344 000 people in the 2003 guidelines).

Authors of guidelines for cardiovascular risk reduction must consider the population-based effectiveness and cost-effectiveness of their recommendations for both pharmacological and other interventions. To do otherwise will lead to poor public policy and patient outcomes.

𝛃 See related articles pages 1027 and 1033