Professor Faculty of Pharmacy, Université Laval, Quebec City, QC, Canada., Canada
Background: Evidence concerning the effect of statins in primary prevention of cardiovascular disease (CVD) among older adults is lacking.
Objectives: To assess the the risk of a first CVD event or mortality among new users of statins
Methods: Using Quebec population-wide administrative data, we emulated a hypothetical randomized trial including older adults ≥66 years on April 1st, 2013, with no CVD history and no statin use in the previous year. We included individuals who initiated statins and followed them until the occurrence of coronary disease (myocardial infarction, coronary bypass, percutaneous coronary intervention), stroke, all-cause death or until March 31, 2018. Individuals who persisted with statins at least three months after initiation were compared to those non-persistent. Primary outcome was the composite endpoint of coronary disease, stroke, and mortality. The intention-to-treat effect was estimated with adjusted Cox models, and per-protocol effect with inverse probability of censoring weighting. Several sensitivity analyses (e.g., varying definitions of persistent users, excluding or not early events) were conducted.
Results: A total of 61,656 individuals were included (mean age=70.96±5.54, female=55.4%) and 57,812 (93.8%) were persistent users. There was no association between persistence with statins and the composite outcome (hazard ratio [HR]: 0.99; 95% confidence interval[CI]: 0.87-1.12). While we observed a 13% reduction in mortality among persistent users (HR:0.87;95%CI:0.76-0.99), coronary disease was higher among persistent users than non-persistent users (HR: 1.83;95%CI: 1.36-2.47). Results for the per-protocol and sensitivity analyses were consistent with the intention-to-treat analysis.
Conclusions: The results for coronary disease do not seem plausible and suggest residual biases. Emulating a trial may minimize but does not preclude bias and caution is required when interpreting results.
