Statins for people at low risk of cardiovascular disease

摘要

The conclusions of the analysis of the effect of statins at different levels of risk are misleading and do not answer relevant questions facing clinicians, patients, and policy makers.The statistical models derived are based on data from populations with previous cardiovascular disease (CVD) or those enriched for risk-factors predisposing to CVD, and thus are not applicable to general primary prevention populations; risk prediction does not guide statin therapy in secondary prevention. Model 1 (derived from 22 mixed trials) included previous history of coronary heart disease and trial-specific covariants, and thus when applied to the same trials individually, observed and expected risk are unsurprisingly similar. Current clinical risk-prediction tools derived solely from general populations without prevalent disease or enrichment for risk factors are unlikely to do as well in the self-selected CTT dataset. Hence the predicted risks from Framingham and the CTT Collaborators are not equivalent and the benefits noted by the CTT Collaborators are likely to be a gross overstatement and should not be used to change guidelines or policy.