Figure 3.

Effects of rosuvastatin on myocardial infarction/stroke/cardiovascular death in high-risk patients, according to baseline characteristics. The hazard ratios and 95% confidence intervals for rosuvastatin when compared with placebo are shown for patients with Framingham risk score >20% and systematic coronary risk evaluation risk ≥5% (extrapolated and capped models). Size of the point estimate rectangle is proportional to the number of clinical events. The dashed vertical line indicates the relative risk reduction for the entire trial cohort. Also shown are P-values for the test of an interaction between the composite endpoint and categories within each subgroup.