Prevention of coronary heart disease

Editor—Marshall commendably questions current trends in preventive medicine.¹ He looks at cost effectiveness of cardioprotective drug treatment, including simvastatin 40 mg, by assessing pooled data from different trials. Treatment effects are balanced with costs. The expenditure for simvastatin 40 mg is calculated as prescribing costs plus additional costs for administration and laboratory tests.

A close look at the heart protection study,²,³ Marshall's main reference for statins, quickly calls this equation into question. Simvastatin 40mg was less well tolerated and less effective than portrayed by the authors.

Firstly, tolerability: against all claims simvastatin was not well tolerated. A substantial number of patients did not enter the trial after a six week run in before randomisation (63 603 entered the original screening, 32 145 went forward to the run in phase and 11 609 patients dropped out then).

Secondly, independent treatment effects: Marshall assumes that statins work independently of other cardioprotective treatment as stated in the heart protection study. However, most of the patients in the study did not take appropriate non-statin drug treatment—for example, only 47% of all patients with peripheral vascular disease were taking aspirin.

Thirdly, reduced treatment effects for women: simvastatin in the heart protection study did not reduce overall mortality in women.⁴ This fact was buried by the researchers using composite end points.

Marshall questions the cost effectiveness of different cardiovascular treatments including statins. The properly interpreted results of the heart protection study lend even more weight to his scepticism.