Cost-Effectiveness of Cholesterol-Lowering Therapies according to Selected Patient Characteristics

Several large-scale, long-term clinical trials evaluating statin drugs (3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors) have confirmed the beneficial effect of reducing cholesterol levels on coronary event rates and related mortality (1-5). Statin drugs are expensive, especially considering the large number of persons who could potentially benefit from cholesterol-lowering therapies. As a result, many analyses have focused on the costs, resource use, and cost-effectiveness of using statins to lower cholesterol levels (6-15). In this analysis, the cost-effectiveness of primary and secondary prevention with cholesterol-lowering therapies was evaluated in separate risk subgroups to assess how cost-effectiveness varies with individual patient characteristics. This analysis extends the results of previous analyses by examining a greater number of specific patient subgroups, particularly with respect to primary prevention. It improves on previous analyses by including updated costs and epidemiologic estimates and by following recommendations from the U.S. Panel on Cost-Effectiveness in Health and Medicine. The incremental cost-effectiveness of statins compared with diet therapy was explicitly modeled. Finally, we compared the cost-effectiveness results with the treatment guidelines recommended by the National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel II) (16). Methods The analysis used a previously validated computer simulation model, the Coronary Heart Disease Policy Model (17-19), to estimate the effects and costs of each cholesterol-lowering strategy in each risk group. The assumptions and design of the Coronary Heart Disease Policy Model are described in detail elsewhere (17, 18). The model consists of three integrated submodels: the demographic-epidemiologic submodel, the bridge submodel, and the disease history submodel. The demographic-epidemiologic submodel predicts coronary heart disease incidence and noncoronary heart disease mortality among people 35 to 84 years of age without coronary heart disease. The risk function for incidence of coronary heart disease is based on age, sex, diastolic blood pressure, smoking status, low-density lipoprotein (LDL) cholesterol level, and high-density lipoprotein (HDL) cholesterol level. The risk function for noncoronary heart disease mortality includes age, sex, diastolic blood pressure, and smoking status. Noncoronary heart disease mortality is assumed to be unaffected by serum cholesterol levels (1, 2). After a person in the model develops coronary heart disease, he or she moves into the bridge submodel, which characterizes the initial coronary heart disease event (cardiac arrest, myocardial infarction, or angina) and the sequelae in the first 30 days after the event. The disease history submodel tracks the subsequent development of coronary heart disease events, revascularization procedures (coronary artery bypass grafting and angioplasty), coronary heart disease mortality, and noncoronary heart disease mortality among patients with coronary heart disease. Target Population The base-case analysis evaluated the cost-effectiveness of primary and secondary prevention in all persons with LDL cholesterol levels of 4.1 mmol/L or greater ( 160 mg/dL). Specific subgroup analyses examined how the cost-effectiveness changed according to patient age (35 to 44, 45 to 54, 55 to 64, 65 to 74, or 75 to 84 years), sex, smoking status (yes or no), diastolic blood pressure (<95 mm Hg or 95 mm Hg), HDL cholesterol level (<0.9, 0.9 to 1.3, or>1.3 mmol/L [<35, 35 to 49, or 50 mg/dL]), and LDL cholesterol level (4.2 to 4.9 or 4.9 mmol/L [160 to 189 or 190 mg/dL]). These risk factors closely correspond to but are not exactly the same as the National Cholesterol Education Program risk factor definitions (16). Effectiveness Lipid Levels Results from five studies were pooled to estimate the effects of a low-cholesterol diet (step I diet) on cholesterol levels (20-24). Data used to model the effectiveness of primary prevention with a statin came from three long-term studies of pravastatin, 40 mg/d, because the quality of effectiveness data was high for this dosage (2, 25, 26). Effectiveness estimates for secondary prevention with a statin was based on results from the Scandinavian Simvastatin Survival Study (Appendix Table 1) (1). A 2-year time lag between the start of treatment and the effects of treatment on coronary events was assumed (1, 2). Quality of Life Quality-of-life weights in the general population without coronary heart disease were based on data from the Beaver Dam Health Outcomes Study according to age and sex (27). Additional quality-of-life adjustments for coronary heart diseaserelated morbidity were made for persons in the disease history submodel; because community preferences were not available, these adjustments were based on a survey of Medicare patients with a history of coronary heart disease (28, 29). Costs In the Coronary Heart Disease Policy Model, total costs were calculated as the sum of intervention costs, costs of coronary heart disease care, and costs of noncoronary heart disease health care. All costs were converted to 1997 U.S. dollars by using the Medical Care Component of the Consumer Price Index. Intervention costs included the costs of medication, physician visits (including the associated patient time), and laboratory tests. National Cholesterol Education Program guidelines were used to guide estimates of the number of physician visits and laboratory tests each year (16). Medication Medication costs for primary and secondary prevention with statins were calculated by using the average wholesale prices of pravastatin and simvastatin, respectively (30). The base-case analysis does not include adjustment of future medication costs resulting from loss of patent protection for statin drugs. To adjust for compliance, it was assumed that patients receiving statins take 95% of the suggested regimen (31); this average compliance rate is reflected in the pool of studies from which the estimates of effectiveness were derived (Appendix Table 1). Primary Prevention Patients receiving diet therapy were assumed to have two physician visits per year. Patients receiving primary prevention with a statin were assumed to have five physician visits in the first year and two physician visits in each year after the first year. A cost of $34.34 was associated with each office visit (32, 33). In addition, the value of patient time associated with each visit was estimated by multiplying the average time per visit (including travel, waiting, and encounter times) by age- and sex-specific average hourly wages. These age- and sex-adjusted patient time costs range from approximately $12 to $26 per physician visit (34, 35). Patients receiving diet therapy were assumed to have one chemical profile, one HDL measurement, and one mid-year measurement of total cholesterol, for a total annual laboratory cost of $39.49 (36). Patients receiving drug therapy were assumed to receive five sets of tests in the first year (five chemical profiles and five HDL measurements for a cost of $151.55) and two sets of tests in each subsequent year ($60.62) (36). The resulting cost estimate of primary prevention with step I diet was $108 per year. For primary prevention with a statin, cost estimates were $1512 in the first year and $1318 in subsequent years. Annual age- and sex-specific patient time costs of $25 to $59 were added to these estimates. Secondary Prevention For secondary prevention, it was assumed that patients would already be seeing their physicians for reasons related to their previous coronary heart disease event. Patient time costs, additional laboratory costs of measuring cholesterol, and costs of cholesterol-lowering medication were the only additional intervention-related costs associated with secondary prevention. The annual cost of secondary prevention with a statin was estimated to be $1329 (excluding patient time costs). Coronary Heart Disease Costs of treating coronary heart disease are also included in the model and are described in detail elsewhere (8). These include the costs of hospitalization; physician visits; laboratory tests; pharmaceuticals associated with a coronary heart disease event; and any associated procedures, such as coronary artery bypass grafting, coronary angioplasty, and cardiac catheterization. NonCoronary Heart Disease Health Care The analysis included annual costs related to noncoronary heart disease health care that were developed from the 1987 National Medical Expenditure Survey. Details of the calculation are described elsewhere (8). Cost-Effectiveness Analysis An incremental cost-effectiveness ratio is the difference in time-discounted costs between the evaluated strategy and the comparison strategy divided by the difference in time-discounted quality-adjusted life-years (QALYs) between the two strategies. For each incremental cost-effectiveness comparison, we examined the cost-effectiveness of extending therapy to a new indication, assuming that the decision had been made to use it in situations with more favorable cost-effectiveness ratios. For example, for primary prevention, statin therapy was compared with step I diet, and step I diet was compared with no primary prevention. For the age-specific analyses, we assumed that once treatment was initiated at a given age, it would be continued through older age ranges, in which treatment typically had more favorable cost-effectiveness results. All analyses of primary prevention assumed a backdrop of secondary prevention with a statin. A separate analysis of the cost-effectiveness of secondary prevention with a statin compared with no secondary prevention was also performed within each risk factor group. Additional Assumptions A discontinuation rate of 6% because of adverse affects was assumed for