Five major guidelines on statin use for primary prevention of atherosclerotic cardiovascular disease (ASCVD) have been published since 2014: the National Institute for Health and Care Excellence (NICE; 2014), US Preventive Services Task Force (USPSTF; 2016), Canadian Cardiovascular Society (CCS; 2016), European Society of Cardiology/European Atherosclerosis Society (ESC/EAS; 2016), and American College of Cardiology/American Heart Association (ACC/AHA; 2018).
To compare the sensitivity, specificity, and estimated number needed to treat (NNT10) to prevent 1 ASCVD event in 10 years according to statin criteria from the 5 guidelines.
Population-based contemporary cohort study. Analyses were performed in the Copenhagen General Population Study, with a mean follow-up time of 10.9 years. We included 45 750 individuals aged 40 to 75 years. The participants were enrolled between 2003 and 2009 and were all free of ASCVD at baseline. Data were analyzed between January 1, 2019, and August 4, 2019.
Statin treatment according to guideline criteria. We assumed a 25% relative reduction of ASCVD events per 38 mg/dL (to convert to millimoles per liter, multiply by 0.0259) reduction in low-density lipoprotein cholesterol.
Sensitivity and specificity for ASCVD events and the NNT10 to prevent 1 ASCVD event according to guideline criteria.
Median age at baseline examination was 56 years, and 43% of participants were men (n = 19 870 of 45 750). During follow-up, we observed 4156 ASCVD events. Overall, 44% of individuals in Copenhagen General Population Study were statin eligible with CCS (n = 19 953 of 45 750), 42% with ACC/AHA (n = 19 400 of 45 750), 40% with NICE (n = 19 400 of 45 750), 31% with USPSTF (n = 13 966 of 45 750), and 15% with ESC/EAS (n = 6870 of 45 750). Sensitivity and specificity for ASCVD events were 68% (n = 2815 of 4156) and 59% (n = 24 456 of 41 594) for CCS, 70% (n = 2889 of 4156) and 60% (n = 25 083 of 41 594) for ACC/AHA, 68% (n = 2815 of 4156) and 63% (n = 26 213 of 41 594) for NICE, 57% (n = 2377 of 4156) and 72% (n = 30 005 of 41 594) for USPSTF, and 24% (n = 1001 of 4156) and 86% (n = 35 725 of 41 594) for ESC/EAS. The NNT10 to prevent 1 ASCVD using moderate-intensity and high-intensity statin therapy, respectively, was 32 and 21 for CCS criteria, 30 and 20 for ACC/AHA criteria, 30 and 20 for NICE criteria, 27 and 18 for USPSTF criteria, and 29 and 20 for ESC/EAS criteria.
With similar NNT10 to prevent 1 event, the CCS, ACC/AHA, and NICE guidelines correctly assign statin therapy to many more of the individuals who later develop ASCVD compared with the USPSTF and ESC/EAS guidelines. Our results therefore suggest that the CCS, ACC/AHA, or NICE guidelines may be preferred for primary prevention.