The London Times’ coverage of the Medicines Partnership report cites Professor Rory Collins on the need to educate patients so that they know that statins have to be taken long term. Professor Collins should know. He is a lead investigator on the Heart Protection Study, the massive (20,536 patients) trial which constitutes Hot Paper #3 this time. Compliance is, of course, easier to achieve within the discipline of a well-planned and well-executed trial, which this surely is, and for simvastatin compliance was about 85%. Compliance with placebo was less impressive because, as one might expect, a participating patient’s physician would often decide to prescribe an off-study statin.

Compliance is very important clinically and for healthcare budgeting, too, but it was not the main endpoint of the Heart Protection Study and is probably not the sort of variable that attracts numerous citations. This paper (#3) records findings for the simvastatin part of a controlled trial that also studied antioxidant vitamins, and the main object was to look at mortality and fatal or nonfatal vascular events—and to do so in a large number of patients at high risk by virtue of a medical history of heart disease, for example, or diabetes. The active drug lowered the rates of heart attacks, revascularization procedures, and strokes by about a quarter, but if non-compliance is allowed for the reduction would be about one-third. This adjustment confirms the importance of compliance and permits the estimate that in these high-risk patients "5 years of simvastatin would prevent about 70-100 people per 1000 from suffering at least one of these major vascular events." Interestingly, the benefits did not seem to depend strongly on the patient’s serum cholesterol at the start of the trial.

Triallists working in the cardiovascular area are well aware of the need to record total mortality and not just cardiovascular deaths lest a benefit in the latter be offset or worse by an increase in non-cardiovascular mortality. That did not happen here, and mortality from all causes was 12.9% in those allocated to simvastatin and 14.7% in those allocated to placebo. Nor was myopathy, which has been a concern with this drug class, a real problem in this or other clinical trials. A recent review of statin-associated myopathy (see P.D. Thomson, et al., JAMA, 289[13]: 1681-90, 2 April 2003) mentions only 3,339 cases of rhabdomyolysis in the U.S. Food and Drug Administration’s MEDWATCH database in more than 12 years from January, 1990, to March, 2002. The word "only" is justified because of the millions of patients now taking one statin or another.