Cardiovascular disease (CVD) is the leading cause of morbidity and mortality in the United States. The estimated direct and indirect cost of CVD for 2008 is $448.5 billion.1 CVD ranked highest among all disease categories for hospital discharges and was the most frequent primary diagnosis coded for physician, outpatient, and emergency department visits.1 Hypercholesterolemia, particularly elevated lowdensity lipoprotein cholesterol (LDL-C), is strongly associated with an increased risk of CVD, including coronary heart disease (CHD) events (ie, myocardial infarction [MI], angina, coronary revascularization) and stroke.2-4 Landmark statin studies targeting LDL-C lowering have shown dramatic reductions in heart attacks, stroke, and cardiac death.5-14 More recent studies have shown that intensive reduction of LDL-C levels in high-risk individuals with a history of CHD is associated with an even greater reduction in CVD events than conventional LDL-C lowering.7-9
Although aggressive LDL-C therapy resulting in positive patient outcomes has been well documented, reports continue to find inadequate LDL-C therapy and poor compliance with statin therapy.15-17 Employers and commercial insurers bear significant medical cost and productivity burdens associated with undertreatment of hypercholesterolemia and poor compliance with statins, particularly for the high-risk population. Although studies have reported cost-effectiveness associated with statin therapy,18-23 none has reported the cost burden in terms relevant to an employer or a health plan, such as per member per month costs.
Specifically, reports lack outcomes adjusted to a working-age population. Many studies report cost-effectiveness utilizing quality-adjusted life-years, which provides information of limited utility for employers and payers. A more meaningful approach may be to provide costeffectiveness in terms that would allow employers and payers to project short-term impact on medical utilization and costs for budgetary purposes. Other studies report cost-effectiveness using LDL-C reduction and unit drug costs across statins, without considering the impact of CVD event reduction cost offsets.
Our analysis models the cost impact of aggressive statin therapy for a commercially insured cohort of members at high risk for secondary events from CVD as a result of a history of CHD. This group represents an important subset of working-age health plan members for a commercial insurer. Using this newly developed model, we compare the CVD event and cost burden under conventional current therapy to represent the results of intensive statin therapy, which is the current standard of care for high-risk patients.
Disease management efforts by employers and other payers that are focused on commercial members with CHD can represent a substantial financial investment. Such efforts seek to identify and provide early clinical intervention to improve outcomes in populations at risk for adverse events. Therefore, individuals with a history of CHD are targeted for disease management programs, because payers believe that such interventions can avoid various complications, including stroke and secondary events (eg, acute MI). However, the claims of improved outcomes of disease management vendors have received significant criticism.24,25 In particular, the range of improved outcomes demonstrated by applying the results of intensive statin therapy to this population can suggest plausible maximum improvements from disease management programs.
In addition, employers and insurers are experimenting with benefit designs, such as lower copayments or other financial incentives to maximize drug compliance for those with chronic diseases. Studies consistently report decreased statin compliance with higher copayments.26-29 The present article offers guidance to such experiments on the potential health and cost improvements possible through various initiatives.
Using the National Health and Nutrition Survey (NHANES) 2003-2004 and 2005-2006 survey data, we identified the target population of people at high risk for CVD who had a history of CHD. NHANES is a survey of healthcare information for US residents readily available from the National Center for Health Statistics and the Centers for Disease Control and Prevention. It is designed to provide reliable estimates of the health and nutritional status of the US civilian noninstitutionalized population.30
Our target population intentionally closely resembles the sample in the Treating to New Targets (TNT) study,8 which consisted of individuals with a history of CHD who are at high risk for CVD events. High risk for CVD as used here is based on the high-risk National Cholesterol Education Program (NCEP) III criteria,31 which include a history of CHD (ie, MI, unstable angina, stable angina, coronary artery procedures), diabetes, or stroke, and patients with >20% 10-year Framingham risk for CHD, combined with 2 or more of the following risk factors:
- Cigarette smoking
- Hypertension (blood pressure =140 mm Hg systolic or =90 mm Hg diastolic, or taking an antihypertensive medication)
- High-density lipoprotein cholesterol <40 mg/dL
- Men =45 years
- Women =55 years
- Family history of premature CHD.
The Appendix contains the NHANES survey examination fields used to identify individuals with high-risk criteria.
The population identified includes 2386 people aged 35 to 69 years in NHANES 2003-2006. Of these, 133 met TNT-like criteria for high CVD risk and a history of CHD. Because the NHANES data do not allow precise selection based on the TNT criteria, some NHANES participants who would have matched TNT criteria were likely omitted. We applied the NHANES portions of treatment target lives by quinquennial agegroups and sex to a population of 10,000 employees and 4721 spouses who had large employer-sponsored health coverage. The demographics of a working-age population are taken from Milliman's 2008 Health Cost Guidelines,32 which is designed to represent the demographics of a typical large, employer-sponsored health benefit program.
Extrapolating this target population to an employer with 10,000 employees yields 256 men and 187 women (aged 35-69 years) at high risk for CVD who have a history of CHD. The Table in the Appendix displays the individual distributions by age and sex. The target population, with an average age of 53.5 years, is about 7 years older than the cohort of all 35- to 69-year-old persons in a working-age population (plus dependents). The target population is more likely to have a history of diabetes (32% vs 9%) or hypertension (32% vs 18%) than the total population of that age-group. Of the target population, 55% was receiving statins according to the NHANES drug file.
Using the Framingham risk scoring system,33-35 we calculated the probability of MI and stroke for each patient in the target population. The 3-year period used in this study reflects a compromise between a long-term study and the short-term horizon of many employee benefit managers.
To model the impact of intensive statin therapy for the target population, we applied the reduction in CHD and stroke events reported in the TNT study8 and in the Heart Protection Study (HPS).7 The TNT study compares the results of 80-mg and 10-mg atorvastatin treatment arms. It was assumed that those in the target population who were receiving statin therapy were currently being treated with conventional statin therapy and were achieving outcomes similar to the 10-mg atorvastatin arm. For those currently receiving statins, an additional event reduction shown for the 80-mg arm compared with the 10-mg arm was applied: 22% (7%-34%, 95% confidence interval [CI]) for CHD and 25% (4%-41%, 95% CI) for stroke.
To model the impact of intensive statin therapy for those not receiving a statin, the event reduction of the 80-mg arm over the 10-mg arm in the TNT study was applied. In addition, the results of the HPS, which compared 40-mg simvastatin to placebo,7 was applied. The HPS reductions of 38% (30%-46%, 95% CI) for CHD and 25% (15%-34%, 95% CI) for stroke, combined with the TNT study reductions, produced reductions of 52% for CHD and 44% for stroke, and were applied for individuals not currently receiving statin therapy.
The HPS population closely resembled that of the TNT study, and we assumed that the event reductions in the TNT study for the 10-mg atorvastatin arm would be similar to the event reductions in the HPS 40-mg simvastatin arm, based on LDL-C reductions reported in each study, as well as other results.36
To quantify the cost impact of intensive statin therapy for a working-age population, we applied data queries to Medstat MarketScan 2003-2007 claims data.37 Medstat contains paid claims generated by more than 14 million commercially insured lives across the United States. Because our model uses the reduction in CVD events among target high-risk individuals as the driver of reduced costs, we identified costs of people with CVD events who also had diabetes, because other characteristics of the target population are not readily identifiable in claims. The following International Classification of Diseases, Ninth Revision (ICD-9) and Current Procedural Terminology (CPT-4) codes were used to identify individuals with a history of CVD events: stroke (ICD-9 code: 430.x, 431.x, 432.x, 433.x1, 434.x1); MI (ICD-9 code: 410.x0, 410.x2); coronary artery bypass graft (CPT code: 33510-33514, 33516-33519, 33521-33523); coronary angioplasty (CPT code: 33533-33536, 35600, 33572, and ICD-9 procedure code 36.1, 36.2x); and percutaneous transluminal coronary angioplasty (CPT code: 33140, 92980-92982, 92984, 92995, 92996, and ICD-9 procedure code: 36.01, 36.02, 36.05, 36.09).
Costs were applied to each individual based on average annual historic per-person medical claim costs, developed separately by sex by 5-year age bands for each CVD event. The event costs (costs during the year of an event) and costs incurred in the 2 years after an event populated a longitudinal cost model for people having an event. Baseline costs for individuals in the year before a CVD event served as a "background" cost. All costs were projected to 2008 levels, using industry trends from the 2008 Milliman Medical Index.38
Distinct costs were not applied to events that resulted in death. The costs include contributions from a small number of people with multiple events. However, the model assumes that a person could have, at most, one type of CVD event (ie, stroke, MI without revascularization, MI with revascularization, or revascularization without MI) over the 3-year model period.
Two distinct scenarios were modeled: a "current status" and a "therapy" scenario. The current status assumes that the current risk, treatment status, and costs continue throughout the 3-year time period modeled, while the therapy scenario assumes individuals with a history of CHD and a high risk for CVD events and receive the additional therapies specified in the TNT and HPS studies.
Analysis of the NHANES 2003-2006 data identifies 4.2% of patients in the 35- to 69-year-old age-group at high risk for CVD and with a history of CHD. Based on statin studies subsequent to the 2002 NCEP III guidelines, updated guidelines have set a therapeutic option of LDL-C <70 mg/dL for high-risk individuals, and some experts recommend the use of intensive statin therapy for all high-risk patients, even when LDL is <100 mg/dL.31
Only 55% of patients in the high-risk cohort with a history of CHD were taking a statin, although 95% were above LDL-C goal (considering the <70 mg/dL therapeutic option). Table 1 provides the distribution of 35 to 69 years old by risk category, NCEP III–updated LDL-C goals, and portion of each cohort above NCEP III–updated goals.
Using the Framingham risk score and NHANES data, the probability of a CHD event or a stroke during 3 years was calculated for each patient. Table 2 presents the number of CHD events and strokes that occurred in the study cohort drawn from a typical working-age population of 10,000 employees and their 4721 dependents. The risk for CHD and stroke is concentrated disproportionately in the target population, which comprises just 4.2% of the 35- to 69-year-old commercial population but is responsible for 22% of the CHD and stroke risk.
Based on the Framingham risk scoring system, the 3-year risk for stroke and CHD in the target population is 2.3% and 9.9%, respectively. If all high-risk patients with CHD were to receive intensive lipid-lowering therapy, stroke and CHD events in an employer population of 10,000 employees and 4721 dependents would be reduced from 49 events to 33 events, a 33% overall reduction. Table 3 illustrates the reduction under both current status and therapy scenarios and presents a range of event reduction based on the CIs reported in the TNT and HPS studies.
To estimate the cost impact of the reduced number of events, we applied event costs calculated from our Medstat analysis. The applied CHD costs used the database's cost and distribution of MI with revascularization, MI without revascularization, and revascularization without MI identified in Medstat. Table 4 presents the costs in the year before the event, the event year, and the costs in the 2 years after the event. These costs reflect payer paid claims dollars and exclude member cost-sharing.
Costs before the event vary by event type, which we postulate reflects differential patient risk characteristics. Costs in particular geographic regions vary from the national averages. Milliman's 2008 Group Health Insurance Survey of preferred provider organization plans with highest and lowest cost regions vary by up to ¬± 11% to 12% from the national average.39 For a particular payer, costs can fall above or below this range, depending on negotiated reimbursement or local patterns, and costs for particular patients can be even more variable.
Incremental costs net the costs from the year before the event, from the year of the event, and the 2 subsequent years. Table 5 presents estimated incremental event costs for the 3 years after the events, based on Table 4, using equal probability of an event in all 3 years. Incremental costs were applied to the number of events reduced over 3 years and calculated as an average $58 (95% CI, $29-$80) per target member per month reduction, or an annual cost reduction of $696 per target member, assuming intensive statin therapy. This cost reduction does not consider the cost of statin therapy, nor does it consider the potential impact on nonmedical costs, such as disability costs or lost work time.
To estimate how this cost reduction per target member compares against the cost of intensive statin therapy, Medstat 2006 and the first 3 quarters of Medstat 2007 were analyzed to identify claims for atorvastatin 80 mg/day, the dosage used in the TNT study. The average allowed cost for 2006 and for 2007 was $97 for a 30-day supply. Assuming a $25 copayment (2nd tier) and a 15% rebate to the payer/employer, the cost to a payer for a 30-day supply of intensive statin therapy with atorvastatin would be $57. Thus, compared with average medical cost reduction of $58 per target member per month, intensive statin therapy may be costneutral to the payer in this model (Table 5).
The literature on the cost–benefit relationship of statin therapy includes mention of some cost-effective nonstatin treatments to prevent CVD events, such as aspirin therapy.20,40 In addition, some critics of intensive statin treatment point out that the TNT study did not find a reduction in overall mortality in the higher-dose group, and did find persistent elevations in liver aminotransferase levels, raising questions about whether the side-effect risks exceed reported benefits of aggressive treatment for patients with stable CHD.41,42 Of note, potential risk reduction for individuals at low baseline risk may be necessarily small; however, the focus in this analysis is on patients at high CVD risk who have a history of CHD, which intends to provide additional perspective on this controversy. In addition, the target population is defined by its CVD risk, not solely by its LDL-C level, suggesting a new perspective for consideration.
Patients with a history of CHD at high risk for CV events can be identified in claims data and frequently are identified through routine disease management algorithms. Specific strategies, including data such as dosage, patient prescription fills, and medication possession ratios, can be monitored through claims data and gaps targeted for intervention.
The reported medical cost reduction in our analysis is driven largely by fewer hospital admissions for stroke and CHD events. Disability, lost work time, and replacement costs for workers suffering from these events were not considered. According to the UnumProvident report, CVD was reported as the fifth leading cause of long-term disability, accounting for 8% of disability claims incidence.43 A 2007 report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee indicates that 15% to 30% of stroke patients become permanently disabled, and that stroke is a leading cause of long-term disability in the United States.44 By not considering disability costs, our modeling underestimates the potential cost-savings of intensive statin therapy for stroke prevention.
Landmark statin studies have shown that treatment of hypercholesterolemia with statin therapy significantly reduces the probability of a stroke or a CHD event. This report, although not the first to attempt to quantify the value of intensive therapy,18 presents this information in the context of employer-sponsored benefit plans. This present analysis did not consider that some individuals in the model cohort generated from NHANES may already be receiving intensive statin therapy, and potential risk reduction would therefore be reduced. But, this concern is at least partially offset by the assumption that patients in the modeled comparator group who were receiving lower-dose statin therapy were indeed compliant and received the full benefit of the HPS study, which is not likely to be the case.
A challenge to the applicability of this model is whether clinical trial results can be replicated in the real-world setting of a health plan. Compliance in the real world may be lower and patients in a clinical trial may receive more attention and be managed more carefully than those enrolled in a health plan.
Another potential limitation is that the target population identified in NHANES may not precisely match the TNT population for characteristics such as age (our population was ages 35-69 years, while TNT was ages 35-75 years) or rates of comorbidities other than CHD. Consequently, this model may reflect the "upper limit" of various CVD event reductions associated with intensive statin therapy and may be used to estimate this maximum potential benefit, with the additional advantage of being practical to implement in the real world of health plan operations.
In addition, the present analysis did not consider that during the 3-year period modeled, aging would increase the size of the target population. Although the TNT and HPS outcomes were based on 4.9- and 5.0-year treatment periods, respectively, our analysis uses a 3-year period. And whereas the TNT and HPS studies are well-regarded, our assumption that the results can be used in sequence and additively is not based on clinical trial results.
Finally, the cost figures presented in this article are point estimates of national average costs, although we offer ranges that indicate regional variation. However, a particular payer's costs may fall outside this range, depending on contractual arrangements. This variation in business practices is distinct from random variation.
The present analysis quantifies risk, events, and cost avoidance, with a focus on one of the widely and easily available tools-intensive statin therapy-that superior patient management efforts could bring to the population at high CVD risk. The range of improved outcomes demonstrated by applying intensive statin therapy to this population can suggest plausible maximum improvements in CVD risk reduction.
Information regarding specific program efforts (identification and therapies), outcomes (reduction in CVD events), and net costs (medical and prescription drugs) that can be measured demonstrates that intensive statin therapy for patients with established CHD may ultimately be cost-neutral to the payer. These results offer potential targets to determine the value of disease management programs aimed at populations at high CVD risk.
This research was supported by funding from Pfizer.
Ms Fitch, Ms Goldberg, and Mr Pyenson are consultants to Pfizer. Ms Fitch is also a consultant to Boehringer Ingelheim. Dr Kuznik and Mr Solomon are employed by Pfizer.
- Rosamond W, Flegal K, Furie K, et al, for the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics-2008 update. Circulation. 2008;117: e25-e146.
- Castelli WP. Epidemiology of coronary heart disease: the Framingham study. Am J Med. 1984;76:4-12.
- Kurth T, Everett BM, Buring JE, et al. Lipid levels and the risk of ischemic stroke in women. Neurology. 2007;68:556-562.
- Greenland P, Knoll M, Stamler J, et al. Major risk factors as antecedents of fatal and nonfatal coronary heart disease events. JAMA. 2003;290: 891-897.
- Shepherd J, Cobbe SM, Ford I, et al, for the West of Scotland Coronary Prevention Study Group. Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. N Engl J Med. 1995; 333:1301-1307.
- Downs JR, Clearfield M, Weis S, et al. Primary prevention of acute coronary events with lovastatin in men and women with average cholesterol levels: results of AFCAPS/TexCAPS. Air Force/Texas Coronary Atherosclerosis Prevention Study. JAMA. 1998;279:1615-1622.
- Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals. Lancet. 2002;360:7-22.
- LaRosa JC, Grundy SM, Waters DD, et al. Intensive lipid lowering with atorvastatin in patients with stable coronary disease. N Engl J Med. 2005;352:1425-1435.
- Cannon CP, Braunwald E, McCabe CH, et al, for the Pravastatin or Atorvastatin Evaluation and Infection Therapy–Thrombolysis in Myocardial Infarction 22 Investigators. Intensive versus moderate lipid lowering with statins after acute coronary syndromes. N Engl J Med. 2004; 350:1495-1504.
- Shepherd J, Blauw GL, Murphy MB, et al, for the PROSPER Study Group. Pravastatin in elderly individuals at risk of vascular disease (PROSPER). Lancet. 2002;360:1623-1630.
- The ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group. Major outcomes in moderately hypercholesterolemic, hypertensive patients randomized to pravastatin vs usual care: the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT-LLT). JAMA. 2002;288:2998-3007.
- Sever PS, Dahl√∂f B, Poulter NR, et al. Prevention of coronary and stroke events with atorvastatin in hypertensive patients who have average or lower-than-average cholesterol concentrations, in the Anglo-Scandinavian Cardiac Outcomes Trial-Lipid Lowering Arm (ASCOTLLA). Lancet. 2003;361:1149-1158.
- Colhoun HM, Betteridge DJ, Durrington PN, et al. Primary prevention of cardiovascular disease with atorvastatin in type 2 diabetes in the Collaborative Atorvastatin Diabetes Study (CARDS). Lancet. 2004; 364:685-696.
- Pedersen TR, Faergeman O, Kastelein JJ, et al, for the Incremental Decrease in End Points Through Aggressive Lipid Lowering (IDEAL) Study Group. High-dose atorvastatin vs usual-dose simvastatin for secondary prevention after myocardial infarction. JAMA. 2005;294:2437-2445.
- Blackburn DF, Dobson RT, Blackburn JL, et al. Adherence to statins, beta-blockers and angiotensin-converting enzyme inhibitors following a first cardiovascular event. Can J Cardiol. 2005;21:485-488.
- Parris ES, Lawrence DB, Mohn LA, Long LB. Adherence to statin therapy and LDL cholesterol goal attainment by patients with diabetes and dyslipidemia. Diabetes Care. 2005;28:595-599.
- Ellis JJ, Erickson SR, Stevenson JG, et al. Suboptimal statin adherence and discontinuation in primary and secondary prevention populations. J Gen Intern Med. 2004;19:638-645.
- Chan PS, Nallamothu BK, Gurm HS, et al. Incremental benefits and cost-effectiveness of high-dose statin therapy in high-risk patients with coronary artery disease. Circulation. 2007;115:2398-2409.
- Ohsfeldt RL, Ghandi SK, Fox KM, et al. Effectiveness and costeffectiveness of rosuvastatin, atorvastatin, and simvastatin among highrisk patients in usual clinical practice. Am J Manag Care. 2006;12(15 suppl):S412-S423.
- Armstrong EP, Zachry WM 3rd, Malone DC. Cost-effectiveness analysis of simvastatin and lovastatin/extended-release niacin to achieve LDL and HDL goal using NHANES data. J Manag Care Pharm. 2004;10: 251-258.
- Lindgren P, Graff J, Olsson AG, et al. Cost-effectiveness of high-dose atorvastatin compared with regular dose simvastatin. Eur Heart J. 2007; 28:1448-1453.
- Lindgren P, Buxton M, Kahan T, et al. Cost-effectiveness of atorvastatin for the prevention of coronary and stroke events: an economic analysis of the Anglo-Scandinavian Cardiac Outcomes Trial-lipid lowering arm (ASCOT-LLA). Eur J Cardiovasc Prev Rehabil. 2005;12:29-36.
- Russell MW, Huse DM, Miller JD, et al. Cost effectiveness of HMGCoA reductase inhibition in Canada. Can J Pharmacol. 2001;8:9-16.
- Motheral BR. 2008: A tipping point for disease management? J Manag Care Pharm. 2008;14:643-649.
- Mattke S, Seid M, Ma S. Evidence for the effect of disease management: Is $1 billion a year a good investment? Am J Manag Care. 2007; 13:670-676.
- Thiebaud P, Patel BV, Nichol MB. The demand for statin: the effect of copay on utilization and compliance. Health Econ. 2008;17:83-97.
- Schneeweiss S, Patrick AR, Maclure M, et al. Adherence to statin therapy under drug cost sharing in patients with and without acute myocardial infarction: a population-based natural experiment. Circulation. 2007;115:2128-2135.
- Goldman DP, Joyce GF, Karaca-Mandic P. Varying pharmacy benefits with clinical status: the case of cholesterol-lowering therapy. Am J Manag Care. 2006;12:21-28.
- Gibson TB, Mark TL, Axelsen K, et al. Impact of statin copayments on adherence and medical care utilization and expenditures. Am J Manag Care. 2006;12:SP11-SP19.
- Centers for Disease Control and Prevention. NHANES 2003-2006 public data release file documentation. www.cdc.gov/nchs/nhanes.htm. Accessed October 17, 2008.
- Grundy SM, Cleeman JI, Merz NB, et al. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation. 2004;110:227-239.
- Milliman. Health Cost Guidelines: Commercial Rating Structures. Seattle, WA; 2007.
- D'Agostino RB, Wolf PA, Belanger AJ, Kannel WB. Stroke risk profile: adjustment for antihypertensive medication: the Framingham Study. Stroke. 1994;25:40-43.
- Wilson PW, D'Agostino RB, Levy D, et al. Prediction of coronary heart disease using risk factor categories. Circulation. 1998;97:1837-1847.
- D'Agostino RB, Russell MW, Huse DM, et al. Primary and subsequent coronary risk appraisal: new results from the Framingham Study. Am Heart J. 2000;139:272-281.
- Rogers SL, Magliano DJ, Levison DB, et al. A dose-specific metaanalysis of lipid changes in randomized controlled trials of atorvastatin and simvastatin. Clin Ther. 2007;29:242-252.
- Thompson Financial. Medstat MarketScan commercial claims database, 2003-2007. Licensed to Milliman, Inc.
- Milliman. 2008 Milliman Medical Index. Seattle, WA; 2008. www.milliman.com/expertise/healthcare/products-tools/mmi/pdfs/milli manmedical- index-2008.pdf. Accessed August 6, 2009.
- Milliman. Milliman's 2008 Group Health Insurance Survey. www.milliman.com/expertise/healthcare/products-tools/group-health-insurancesurve y/. Accessed August 6, 2009.
- Curtiss FR. Methods to attain optimal outcomes with lipid-lowering drug therapy. J Manag Care Pharm. 2004;10:251-258.
- Pitt B. Low-density lipoprotein cholesterol in patients with stable coronary heart disease-is it time to shift our goals? N Engl J Med. 2005;352:1483-1484.
- Auer J, Lamm G, Eber B. Intensive lipid lowering with atorvastatin in coronary disease [letter]. N Engl J Med. 2005;353:93.
- UnumProvident Corp. Unum says cancer still no. 1 cause of disability, but survival and return-to-work rates are increasing. www.unumprovident.com/newsroom/news/corporate/default.aspx. Accessed September 7, 2007.
- American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statistics–2007 update. Circulation. 2007;115:e69-e171.