|
|
||||||||||
Staff Pharmacist, Department of Pharmacy, Good Samaritan Hospital, 2222 Philadelphia Dr., Dayton, OH 45406, fax 937/431-9048, Rbelloto{at}woh.rr.com
Reprints: Dr. Belloto
A comment on the conduct of a meta-analysis analyzing the effects of statins on cardiovascular and cerebrovascular outcomes is presented in this editorial. Competing risk and risk-stratified analysis models that may be more appropriate for the analysis of clinical trials in general and statin trials in particular are introduced. The avoidance of these more complicated models seems to be paradoxical when other models that clinicians have used for some time and are using today are seen to contain more patient variables than the proposed models. Certainly, one way these competing risk and general mathematical models will be used is to have the data readily available for other researchers to duplicate the calculations or use in entirely new ways. This aids in the learning process and helps students, reviewers, editors, and others who wish to research and judge the clinical trials.
Key Words: competing risks, meta-analysis, risk-stratified analysis, statins, stroke
Published Online, November 6, 2007. www.theannals.com, DOI 10.1345/aph.1K528
Related articles in The Annals: