Statin wars following coronary revascularization--evidence-based clinical practice?

Abstract

Background: Randomized clinical trials (RCTs) have shown that statins provide substantial heath benefits. Pharmaceutical companies spend enormous amounts of money on both clinical trials and marketing. The relative influence of information from clinical trials on physician prescription patterns for statins is unknown.

Objective: To examine the correlation between statin prescription patterns and the quality of evidence from RCTs.

Methods: Using the computerized administrative databases of the Quebec Health Insurance Board, the choice of statin for elderly patients (older than 65 years of age) following a coronary revascularization procedure (percutaneous coronary intervention or coronary artery bypass graft surgery) performed between January 1, 1994, and June 30, 2003, was examined. Prescriptions for each statin were compared with their evidence base obtained from a cumulative systematic literature review of RCTs that recorded mortality as an outcome and were published before December 31, 2002.

Results: The study cohort comprised 27,979 elderly revascularized patients who received at least one statin prescription. In 1996, the year atorvastatin was introduced, simvastatin and pravastatin had 38.3% and 37.1% of the market share, respectively. By 2003, atorvastatin had 44% of the market share, compared with 29.9% and 24.1% for simvastatin and pravastatin, respectively. In contrast, RCTs published up to the end of 2002 had culminated in 133,341 and 140,565 patient-years of follow-up for simvastatin and pravastatin, respectively, and only 1459 patient-years for atorvastatin.

Conclusions: Prescription patterns regarding the choice of statin do not appear to be determined uniquely from high-quality RCTs. Further research into the other possible determinants of physician prescription patterns is necessary.

Figure 1)

First statin prescribed after revascularization (percentage of each statin). Letters in the graph represent the publication of studies with more than 4000 patients and, as follows, are defined by the type of statin examined, the year of study publication, the name of the study and the number of patients examined: study A (simvastatin, 1994, Scandinavian Simvastatin Survival Study [4S], 4444 patients [7]); study B (simvastatin, 2002, Medical Research Council/British Heart Foundation Heart Protection Study, 20,536 patients [13]); study C (pravastatin, 1995, West of Scotland Coronary Prevention Study [WOSCOPS], 6595 patients [22]); study D (pravastatin, 1996, Cholesterol And Recurrent Events [CARE] study, 4159 patients [41]); study E (pravastatin, 1998, Long-Term Intervention with Pravastatin in Ischaemic Disease [LIPID] study, 9014 patients [28]); study F (pravastatin, 2000, Gruppo Italiano per lo Studio della Sopravvivenza nell’Infarto Miocardico [GISSI], 4271 patients [29]); study G (pravastatin, 2002, Prospective Study of Pravastatin in the Elderly at Risk [PROSPER], 5804 patients [31]); study H (lovastatin, 1998, Air Force/Texas Coronary Atherosclerosis Prevention Study [AFCAPS/TexCAPS], 6605 patients [19])

Figure 2)

Evidence base for each of the statins expressed as the number of patient-years of follow-up from double-blind, randomized controlled trials (RCTs). Data from references listed in Table 2