Search

Background— The diagnostic utility of N-terminal pro-B-type natriuretic peptide (NT-proBNP) in heart failure has been documented. However, most of the data were derived from countries with high healthcare resource use, and randomized evidence for utility of NT-proBNP was lacking. Methods and Results— We tested the hypothesis that NT-proBNP testing improves the management of patients presenting with dyspnea to emergency departments in Canada by prospectively comparing the clinical and economic impact of a randomized management strategy either guided by NT-proBNP results or without knowledge of NT-proBNP concentrations. Five hundred patients presenting with dyspnea to 7 emergency departments were studied. The median NT-proBNP level among the 230 subjects with a final diagnosis of heart failure was 3697 compared with 212 pg/mL in those without heart failure ( P P =0.031), the number of patients rehospitalized over 60 days by 35% (51 to 33; P =0.046), and direct medical costs of all ED visits, hospitalizations, and subsequent outpatient services (US $6129 to US $5180 per patient; P =0.023) over 60 days from enrollment. Adding NT-proBNP to clinical judgment enhanced the accuracy of a diagnosis; the area under the receiver-operating characteristic curve increased from 0.83 to 0.90 ( P Conclusions— In a universal health coverage system mandating judicious use of healthcare resources, inclusion of NT-proBNP testing improves the management of patients presenting to emergency departments with dyspnea through improved diagnosis, cost savings, and improvement in selected outcomes.

Dyslipidemia has been recognized as an important risk-factor for the development of cardiovascular disease. The current, available therapies of dyslipidemia, their effectiveness, costs, cost-effectiveness and healthcare implications are discussed. At the present time, the lipid-lowering therapies are dominated by statins. Despite a variety of assumptions regarding modeling cardiovascular disease risks and costs, statin therapy is generally cost-effective for secondary prevention and for primary prevention in individuals with additional risk-factors. The costs of drug therapy and the absolute risk of developing future cardiovasular events are the dominant factors determining the cost-effectiveness. When developing clinical guidelines, the cost-effectiveness and proportion of the population to be treated must be considered as well as the total population costs of treatment.

Background. Although exercise training improves cardiovascular disease (CVD) risk factors, few studies have evaluated its potential long-term cost-effectiveness. Methods. Using the Cardiovascular Disease Life Expectancy Model, a validated disease simulation model, we calculated the life expectancy of average 35- to 74-year-old Canadians found in the 1992 Canadian Heart Health Survey. The impacts of exercise training on cardiovascular risk factors were estimated as a 4% decrease in low-density lipoprotein (LDL) cholesterol, a 5% increase in high-density lipoprotein (HDL) cholesterol, and a 6 mm Hg decrease in both systolic and diastolic blood pressure. Exercise adherence was estimated at 50% for the first year and 30% for all additional years. Costs for a supervised exercise program determined from Canadian sources and converted to US dollars were estimated at $605 for the first year (medical evaluation, stress test, exercise prescription, and program costs) and $367 for all additional years (program costs). For an unsupervised program, the costs were estimated at $311 for the first year and $73 for all additional years. Results. The cost-effectiveness (CE) of an unsupervised exercise program (1996 U.S. dollars) was less than $12,000 per year of life saved (YOLS) for all individuals. The CE of a supervised exercise program was less than $15,000/YOLS for men with CVD, and between $12,000 and $43,000 for women with CVD and men without CVD. Conclusions. Given the relatively few risks, substantial long-term benefits, and modest costs, an unsupervised exercise training program represents good value for all. A more expensive supervised exercise program is also cost-effective for most individuals with CVD.

Figure 1: Model of CDI transmission across three major subpopulations Objectives: The transmission of C. difficile infection (CDI) has recently changed, resulting in a five-fold increase in the incidence in the general population and an eight-fold increase among the elderly. We examined the relationship between three major subpopulations of CDI transmission: hospitals, long-term care facilities (LTCF), and communities, to evaluate treatment effectiveness and costs. Methods: We developed a decision analytic computer simulation model to compare various strategies for the management of CDI, across three major subpopulations of CDI transmission: hospitals, communities, and LTCFs. To estimate CDI rates in Canada, stratified by age and sex, by subpopulation (hospital, community, and LTCFs), and by primary vs. recurrent infections, we conducted a systematic analysis of all provincial and federal CDI data in Canada. Based on the reports, the incidence and recurrence rates, stratified by age, sex and the three subpopulations were estimated for Canada as a whole. We performed rigorous validation analyses to demonstrate that the estimated CDI rates are a reasonable representation of the selected actual rates for Canada. Results: We estimated the annual number of C. difficile infections at 21,632 cases, 75% in terms of new infections and the remaining in terms of recurrences. The rate per 100,000 population is estimated at 79, with 86 per 100,000 population among females, and 72 per 100,000 population among males. The rates per 100,000 population for ages 80 years and older are four-fold higher than for the 60-79 years old group. The recurrence rates are most pronounced among the elderly. We estimated that 69% of CDI cases occur in hospitals, 26% in community and 5% in LTFCs. The numbers of CDI cases among the elderly are expected to grow as the Canadian population ages. Conclusions: Our age-specific model allows to project and to quantify the impact of a CDI outbreak in terms of clinical burden and costs. Using a scenario-based approach, comparisons of current treatments with the novel approach of duodenal infusion (fecal transplant) are carried out. Our validated CDI model is also capable of estimating net clinical and economic benefits of new disinfection systems currently being evaluated and applied in various health care institutions across Canada. Clostridium difficile infection (CDI) is the leading cause of health care associated infectious diarrhea in hospitals. The disease spectrum caused by CDI ranges from a mild to a severe, life threatening, colitis. Until recently, CDI was considered to be a hospital based infection. The prevalence and severity of CDI are increasing; and CDI is now appearing in LTCFs, and the community. There are considerable gaps in the current understanding of CDI transmission between hospital, community, and LTCF settings. Advanced age has increasingly been identified as an important risk factor for CDI but little information is available on the effects of age and sex on clinical responses to CDI. Efforts to model CDI to date have been limited. Most models do not address the contribution of asymptomatic carriers as sources of new infections and are restricted to hospital acquired CDI. In the general population 5% -20% of adults are estimated to be asymptomatic carriers of CDI but up to 80% of the elderly are colonized in LTCFs. In 2005, 28% of CDI in Canada were among the population aged 80 and older. In 2011, 40% of infections occurred among patients aged 80+. Elderly patients experience more severe episodes of CDI. Age-specific recurrence rates of CDI are important characteristics of CDI. Patients over 65 experienced almost twice the recurrence rate compared with younger populations. We developed a decision analytic computer simulation model to compare various strategies for the management of CDI. Our objectives were (1) to examine systematically the dynamic relationship between three major subpopulations of CDI transmission: hospitals, communities, and LTCFs, and (2) to provide extensive model validation based on available provincial databases. We developed a decision analytic computer simulation model to compare various strategies for the management of CDI, between and within three major subpopulations of CDI transmission: hospitals, communities, and LTCFs (figure 1). To estimate CDI rates in Canada, stratified by age and sex, by subpopulation (hospital, community, and LTCFs), and by primary vs. recurrent infections, we conducted an extensive systematic analysis of all provincial and federal CDI data in Canada. The most recent data from four provinces, Quebec, Ontario, British Columbia, and Manitoba were considered to be relevant for the estimation of stratified CDI rates for Canada, as their combined populations represent over 75% of Canada’s population. We used piecewise polynomial interpolation to estimate annual stratified CDI rates across age, gender, subpopulation, and province. The fitted data consisted of a piecewise polynomial spline with three change points identified from the observed pattern of CDI for three age groups, 18-59 years, 60-79 years, and 80 years and over. Based on the provincial reports, the incidence and recurrence rates, stratified by age, sex and the three subpopulations were estimated for Canada as a whole. We calibrated the results of the analysis using selected provincial data, and goodness of fit measures were applied to all sets of predictive equations. We performed validation analyses to demonstrate that the CDI rates as predicted by the synthesized regression analyses are a reasonable representation of the selected actual rates for Canada. Predicted versus observed CDI rates were compared by age, gender, and subpopulation, using the following measures; number of cases, cases per 100,000 population, per 1,000 hospital admissions, and per 10,000 patient-days. We estimated the annual number of C. difficile infections at 21,632 cases, 75% in terms of new infections and the remaining in terms of recurrences. The rate per 100,000 population is estimated at 79, with 86 per 100,000 population among females, and 72 per 100,000 population among males. Figure 2 demonstrates that the peak number of cases occurs between ages 80-84 years. However, on a per population basis the rates are rapidly increasing starting from age 65 onward (figure 3). The rates per 100,000 population for ages 80 years and older are four-fold higher than for 60-79 years old group. The recurrence rates are most pronounced among the elderly. We estimated that 69% of CDI cases occur in hospitals, 26% in community and 5% in LTFCs. The number of CDI cases among the elderly are expected to grow as the Canadian population will age. The present distributions of CDI cases by various ages are 19% among 18-59 years old, 41% among 60-79, and 40% among those over 80 years old. We performed validation analyses to demonstrate that the predicted CDI rates were a reasonable representation of the selected actual rates for Canada. The CDI estimates were compared to selected data reported by the Canadian Nosocomial Infection Surveillance Program (CNISP), as a part of an independent validation. CNISP data was not used in the development of the model. Another external validation was performed using the Institute for Clinical Evaluative Sciences (ICES) and the Ontario Ministry of Health and Long-Term Care selected databases. Goodness of fit tests across nine age groups show a high degree of association between estimated and observed CDI rates. Our age-specific model of CDI transmission intends to (1) advance understanding of CDI transmission between the three subpopulations (hospital, community, and LTCF); (2) to improve the capacity to project and to quantify the impact of CDI in terms of clinical burdens and costs among these three subpopulations. Given the extensive validation, the model can be used for forecasting purposes using numerous scenarios, and will help identify patients at high risk for CDI. Currently, we are in the process of comparing clinical benefits and costs of current treatment regimens with promising novel approaches (e.g. fecal transplant). Our validated CDI model is also capable of estimating net benefits and costs of new disinfection systems currently being evaluated and applied in various health care institutions across Canada. Background

We are grateful of the comments provided by Chung et al about the observation in our study1 that N-terminal pro–B-type natriuretic peptide (NT-proBNP) levels were increased in patients with prior history of heart failure and/or left ventricular dysfunction even when dyspnea was not due to heart failure. Our findings were consistent with those from previous …

Recent studies suggest that the benefit of lipid-lowering treatment for the primary and secondary prevention of cardiovascular disease (CVD) extends to individuals with average cholesterol levels, to women and to the elderly. However, the proportion of the general population for which treatment is cost-effective has not been evaluated.Using data provided by the Canadian Heart Health Survey, the level of CVD risk was estimated for a random sample of the total population. A cost-effectiveness ratio for simvastatin was then calculated for each individual in the sample. Lastly, the proportion of the total population for which lipid-lowering therapy would be cost-effective for primary and secondary prevention of CVD was estimated according to total cholesterol (TC) levels.Among the surveyed individuals who were 30 to 74 years of age, 2212 had CVD and 12,982 did not. Among those with a TC level higher than 6.2 mmol/L, the proportions of individuals for which lipid-lowering therapy was cost-effective (at a level of less than 50,000 dollars per year of life saved) were 85.6% of men and 28.7% of women for primary prevention, and 99.8% of men and 86.1% of women for secondary prevention. The estimated cost of one year of lipid-lowering treatment for all individuals in the population with a TC level higher than 6.2 mmol/L and for all individuals regardless of TC levels for whom treatment would be cost-effective was $1 billion and 3.9 billion dollars, respectively.Lipid-lowering treatment for CVD prevention is cost-effective for a high proportion of the population, even for primary prevention. As a result, the cost of population-wide treatment for only one year is high even among individuals with a TC level higher than 6.2 mmol/L. Such costs should be considered in health care policy decisions.

To estimate the cost-effectiveness ratio of highly-active antiretroviral therapy (HAART) in Canada.A before-and-after analysis to calculate incremental cost of life year gained (LYG) between 1991 and 1995 (pre-HAART period) and between 1997 and 2001 (HAART period) for non-AIDS and AIDS groups (CDC stage of HIV infection).For two Quebec HIV hospital clinics, mean inpatient (IP) days, outpatient (OP) visits and direct health care costs per patient-year (PPY) were calculated. Cox's proportional hazards models calculated disease progression, stratified by study periods and adjusted for gender, age at cohort entry, sexual orientation, injecting drug use and baseline CD4 cell count.For non-AIDS patients, mean IP days was 1.6 (pre-HAART period) compared with 0.8 PPY (HAART period); mean OP visits increased from 2.8 to 5.5 PPY. Total cost was US$ 4265 (pre-HAART period) and US$ 9445 PPY (HAART period) of which 66 and 84%, respectively were spent on antiretroviral drugs. Median progression time was 6.3 years in the pre-HAART period compared with 12.5 years in HAART period (log rank chi = 270, P0.0001). Incremental cost per LYG between periods was US$ 14 587. For AIDS patients, mean IP days decreased from 13.3 to 4.4 PPY between periods; OP visits increased from 8.3 to 9.2 PPY. Total costs increased from US$ 9099 to US$ 11 754 PPY, while expenditure on antiretroviral drugs increased from 29 to 72% of total cost. Median progression time was 3.8 years in the pre-HAART period, which increased to 13.3 years in the HAART period (log rank chi = 158, P0.0001); incremental cost per LYG between periods was US$ 12 813.HAART appeared a cost-effective intervention in Canada.

Osteoarthritis and hypertension are highly prevalent among older Americans. Anti-inflammatory medications can destabilize blood pressure control. We estimated the decreased cardiovascular risk, premature mortality, and direct health care costs that could be avoided if blood pressure control is not destabilized among hypertensive Americans taking cyclooxygenase-2 (COX-2)–specific inhibitors for osteoarthritis. Data from the Third National Health and Nutrition Examination Survey (NHANES III) provided the distribution of cardiovascular risk factors among American adults with osteoarthritis and hypertension. The Cardiovascular Disease Life Expectancy Model was used to estimate the impact of a 2.26% increase in systolic blood pressure on the basis of results of a randomized trial comparing COX-2–specific inhibitors. A similar analysis was completed for American adults with osteoarthritis and untreated hypertension (≥140/90 mm Hg). Among 7.3 million Americans with treated hypertension, maintaining blood pressure control would avoid >30 000 stroke deaths and 25 000 coronary deaths resulting in >449 000 person years of life saved and $1.4 billion in direct health care cost savings. When an additional 3.8 million Americans with untreated hypertension are considered, maintaining blood pressure control could prevent >47 000 stroke deaths, 39 000 coronary deaths, and result in 668 000 person years of life saved and >$2.4 billion in direct health care cost savings. We conclude that even a small increase in systolic blood pressure among hypertensive Americans with osteoarthritis may substantially increase the clinical and economic burden of cardiovascular disease. Maintaining blood pressure control may be associated with substantial benefits.

Purpose Diabetes mellitus is associated with an increased risk of cardiovascular disease. We compared the clinical effects of treating dyslipidemia in patients who had diabetes mellitus but no diagnosed cardiovascular disease with the effects of similar treatment in patients who had cardiovascular disease but no diabetes mellitus. Methods We estimated the number of adults (ages 30 to 74 years) requiring lipid therapy using data from the third National Health and Nutrition Examination Survey and current lipid treatment guidelines. Using the Cardiovascular Life Expectancy Model, we estimated the mean increase in life expectancy that would result from lowering low-density lipoprotein cholesterol levels by 35% and increasing high-density lipoprotein cholesterol levels by 8% based on results from the Scandinavian Simvastatin Survival Study. Results The mean number of years of life saved ranged from 3 to 3.4 years for men with diabetes versus 2.4 to 2.7 years for men with cardiovascular disease. In women, the estimated benefits were 1.6 to 2.4 years for those with diabetes versus 1.6 to 2.1 years for those with cardiovascular disease. Total population benefits were also substantial for patients with diabetes (25.4 million person-years of life saved) and those with cardiovascular disease (16.0 million person-years of life saved). Conclusion The clinical benefits of treating dyslipidemia in patients with diabetes should be at least equivalent to, if not more substantial than, the benefits observed among those with cardiovascular disease.

OBJECTIVE— Epidemiological studies have shown that the risk of myocardial infarction (MI) in diabetic patients without cardiovascular disease(CVD) is comparable to the risk of MI in patients with CVD. We used a validated Markov model to compare the long-term costs and benefits of treating dyslipidemia in diabetic patients without CVD versus treating CVD patients without diabetes in the U.S. The generalizability and robustness of these results were also compared across six other countries (Canada, France,Germany, Italy, Spain, and the U.K.). RESEARCH DESIGN AND METHODS— With use of the Cardiovascular Disease Life Expectancy Model, cost effectiveness simulations of simvastatin treatment were performed for men and women who were 40-70 years of age and had dyslipidemia. We forecast the long-term risk reduction in CVD events after treatment. On the basis of the Scandinavian Simvastatin Survival Study results, we assumed a 35% reduction in LDL cholesterol and an 8% rise in HDL cholesterol. RESULTS— In the U.S., treatment with simvastatin for CVD patients without diabetes was cost-effective, with estimates ranging from$8,799 to $21,628 per year of life saved (YOLS). Among diabetic individuals without CVD, lipid therapy also appeared to be cost-effective, with estimates ranging from $5,063 to $23,792 per YOLS. In the other countries studied, the cost effectiveness of treating diabetes in the absence of CVD was comparable to the cost effectiveness of treating CVD in the absence of diabetes. CONCLUSIONS— Among diabetic men and women who do not have CVD,lipid therapy is likely to be as effective and cost-effective as treating nondiabetic individuals with CVD.

* CNISP (Canadian Nosocomial Infection Surveillance Program) † BC (Provincial Infection Control Network of British Columbia) Objectives: Efforts to model Clostridium difficile infection (CDI) to date have been limited. Most models do not address the contribution of asymptomatic carriers as sources of new infections and are restricted to hospital acquired CDI. We aim to develop a simulation model to examine systematically the dynamic relationship between three major subpopulations of CDI transmission: hospitals, communities, and long-term facilities, and to evaluate treatment effectiveness and costs. Methods: We conducted a systematic investigation to determine the key epidemiological factors influencing CDI transmission according to the three major subpopulations: hospitals, communities, and long-term care facilities. We have developed a stochastic agent-tracking meta-population network model of CDI transmission, and identified parameters that would capture transmission from symptomatic and asymptomatic carriers to uninfected individuals among the subpopulations. Results: We identified eight infection states: susceptible, gastrointestinal exposure, colonized, diseased, deceased, clinically resolved colonized, relapse of CDI, and cleared. Key parameters include; health outcomes of target populations, time horizon, diagnostic characteristics, treatment effectiveness, transmission rates, susceptibility rates, recurrence rates, and costs. In the general population 5% -20% of adults are estimated to be asymptomatic carriers of CDI but up to 80% of the elderly in long-term care facilities are considered to be colonized. Age-specific recurrence rates of CDI are important parameters of the model. Major predictors of recurrences were advanced age and duration of initial hospitalizations. The probability of recurrent CDI increases with the number of recurrences experienced. Recurrences were associated with major increases in hospital LOS and costs. Conclusions: Using a scenario-based approach we are well positioned to compare clinical benefits and costs of current treatment regimens (Metronidazole, Vancomycin, Fidaxomicin) with the potential novel approach of duodenal infusion (fecal transplant).

Objective: To forecast the long-term benefits and costeffectiveness of lipid modification in the secondary prevention of cardiovascular disease. Methods: A validated model based on data from the Lipid Research Clinics cohort was used to estimate the benefits and cost-effectiveness of lipid modification with 3-hydroxy-3-methylglutaryl–coenzyme A reductase inhibitors (statins) based on results from the Scandinavian Simvastatin Survival Study (4S), including a 35% decrease in low-density-lipoprotein (LDL)–cholesterol levels and an 8% increase in high-density-lipoprotein (HDL)– cholesterol levels. After comparing the short-term outcomes predicted for the 4S with the results actually observed, we forecast the long-term risk of recurrent myocardial infarction, congestive heart failure, transient ischemic attacks, arrhythmias, and strokes and the need for surgical procedures such as coronary artery bypass grafting, catheterization, angioplasty, and pacemaker insertions. Outpatient follow-up care costs were estimated, as were the costs of hospital care and drug therapy. All costs were expressed in 1996 US dollars. Results: The short-term outcomes predicted for the 4S were consistent with the observed results. The longterm benefits of lipid modification among low-risk subjects (normotensive nonsmokers) with a baseline LDL/ HDL ratio of 5 but no other risk factors ranged from $5424 to $9548 per year of life saved for men and $8389 to $13 747 per year of life saved for women. In high-risk subjects (hypertensive smokers) with an LDL/HDL ratio of 5, the estimated costs ranged from $4487 to $8532 per year of life saved in men and $5138 to $8389 per year of life saved in women. Assuming that lipid modification h as n o e ffect o n t he r isk o f s troke, c osteffectiveness increased by as much as 100%. Conclusions: These long-term cost estimates are consistent with the short-term economic analyses of the published 4S results. The long-term treatment of hyperlipidemia in secondary prevention is forecasted to be cost-effective across a broad range of patients between 40 and 70 years of age. Recognizing the additional effects of lipid changes on cerebrovascular events can substantially improve the cost-effectiveness of treating hyperlipidemia.

We thank Dr. Egede (1) for his letter in this issue of Diabetes Care and for taking the time to read our study (2) and provide us with his thoughtful comments. He raises several important points. The first point concerns the choice of years of life saved rather than quality-adjusted life years in the cost-effectiveness analysis. Dr. Egede’s point that both diabetes and cardiovascular disease may have a significant impact on quality of life is well taken. We agree with him that it would be desirable to include quality-of-life issues in …

Background —The objective of this study was to estimate the long-term costs and benefits of treating hyperlipidemia among diabetic patients with and without known cardiovascular disease after validating the Cardiovascular Life Expectancy Model. Methods and Results —The model estimates were compared with the Scandinavian Simvastatin Survival Study (4S) and used to estimate the long-term costs and benefits of treatment with simvastatin. Simulations were performed for men and women, 40 to 70 years of age, having pretreatment LDL cholesterol values of 5.46, 4.34, and 3.85 mmol/L (211, 168, and 149 mg/dL). We forecasted the long-term risk of cardiovascular events, the need for medical and surgical interventions, and the associated costs in 1996 US dollars. The model validated well against the observed results of the of the 4S diabetic patients. In this validation, the model estimates fell within the 95% confidence interval of the observed results for 7 of the 8 available end points (coronary deaths, total deaths, and so forth). Treatment with simvastatin for patients with cardiovascular disease is cost-effective for men and women, with or without diabetes. Among diabetic individuals without cardiovascular disease, the benefits of primary prevention were also substantial and the cost-effectiveness ratios attractive across a wide range of assumptions (≈$4000 to $40 000 per year of life saved). These conclusions were robust even among diabetics with lower baseline LDL values and smaller LDL reductions as observed in the Cholesterol and Recruitment Events (CARE) trial. Conclusions —Among adults with hyperlipidemia, the presence of diabetes identifies men and women among whom lipid therapy is likely to be effective and cost-effective even in the absence of other risk factors or known cardiovascular disease.

Background: The losses in productivity due to cardiovascular disease (CVD) are substantial but rarely considered in health economic analyses. We compared the costeffectiveness of lipid level modification in the primary prevention of CVD with and without these indirect costs. Methods: We used the Cardiovascular Life Expectancy Model to estimate the long-term benefits and costeffectiveness of lipid level modification with atorvastatin calcium, including 28% and 38% reductions in total cholesterol and low-density lipoprotein cholesterol levels, respectively, and a 5.5% increase in high-density lipoprotein cholesterol level. The direct costs included all medical care costs associated with CVD. The indirect costs represented the loss of employment income and the decreased value of housekeeping services after different manifestations of CVD. All costs were expressed in 2000 Canadian dollars. Results: When only direct medical care costs were considered, the incremental cost-effectiveness ratios for lifelong therapy with atorvastatin calcium, 10 mg/d, were generally positive, ranging from a few thousand to nearly $20000 per year of life saved. When the societal point of view was adopted and indirect costs were included, the total costs were generally negative, representing substantial cost savings (up to $50000) and increased life expectancy for most groups of individuals. Conclusions: Lipid therapy with statins can reduce CVD morbidity and mortality as demonstrated in a number of clinical trials. Adding the indirect CVD costs associated with productivity losses at work and home can result in forecasted cost savings to society as a whole such that lipid therapy could potentially save lives and money. Arch Intern Med. 2003;163:333-339

Objective. —To evaluate the lifetime cost-effectiveness of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors for treatment of high blood cholesterol levels. Design. —We added cost data to a validated coronary heart disease (CHD) prevention computer model that estimates the benefits of lifelong risk factor modification. The updated model takes into account the costs of cholesterol reduction, the savings in CHD health care costs attributable to intervention, the additional non-CHD costs resulting from patients' living longer, and the beneficial effects of reducing CHD risk by reducing total cholesterol and increasing high-density lipoprotein cholesterol (HDL-C). Patients. —Men and women aged 30 to 70 years who were free of CHD, had total cholesterol levels equal to the 90th percentile of the US distribution in their age and sex group, had HDL-C levels equal to the mean of the US distribution in their age and sex group, and were either with or without additional CHD risk factors. Intervention. —Use of 20 mg of lovastatin per day, which on average reduces total serum cholesterol by 17% and increases HDL-C by 7%. Main Outcome Measures. —Cost per year of life saved after discounting benefits and costs by 5% annually. Results. —The increase in HDL-C associated with lovastatin lowered cost-effectiveness ratios by approximately 40%, such that the treatment of hypercholesterolemia was relatively cost-effective for men (as low as $20 882 per year of life saved at age 50 years) and women ($36 627 per year of life saved at age 60 years) with additional risk factors. Non-CHD costs resulting from longer life expectancy after intervention added at most 23% to the cost-effectiveness ratios for patients who began treatment at age 70 years, and as little as 3% for patients at age 30 years. Conclusion. —The cost-effectiveness of HMG-CoA reductase inhibitors varied widely by age and sex and was sensitive to the presence of non-lipid CHD risk factors. The additional non-CHD costs due to increased life expectancy may be significant for the elderly. Accounting for the drug effects of raising HDL-C levels increased the proportion of the population for which medication treatment was relatively cost-effective. (JAMA. 1995;273:1032-1038)