Current guidelines and performance measures concerning risk factors for coronary artery disease emphasise the importance of achieving target cholesterol levels without strong guidance about the strategies used to achieve them. Some health programmes also pay doctors whose patients achieve specific lipid targets.1 The target approach opened the door for marketing campaigns to promote drugs that have not been shown to affect patient outcomes. For example, ezetimibe was heavily promoted in the United States, with the manufacturers spending more than $200m (£130m; €160m) on direct to consumer advertising of Vytorin (ezetimibe plus simvastatin) in 2007. Millions of prescriptions were written, even though no clinical trial of the drug had focused on patient outcomes.2 3 The focus on targets is out of alignment with evidence from clinical trials and discounts the importance of testing whether the benefits of a treatment exceed its risks, instead suggesting that we can rely on its effects on biomarkers.