Free
Original Contribution  |   August 2010
Cost-Effectiveness of Rosuvastatin for Primary Prevention of Cardiovascular Events According to Framingham Risk Score in Patients With Elevated C-Reactive Protein
Gary P. MacDonald, DO, MS
Author Notes
  • Dr MacDonald holds a master of science degree in health economics. 
  • Address correspondence to Gary P. MacDonald, DO, MS, Rapid City Regional Hospital, Family Medicine Residency Program, 502 E Monroe St, Rapid City, SD 57701-1400. E-mail: gary.p.macdonald@gmail.com 
Article Information
Cardiovascular Disorders / Preventive Medicine
Original Contribution   |   August 2010
Cost-Effectiveness of Rosuvastatin for Primary Prevention of Cardiovascular Events According to Framingham Risk Score in Patients With Elevated C-Reactive Protein
The Journal of the American Osteopathic Association, August 2010, Vol. 110, 427-436. doi:https://doi.org/10.7556/jaoa.2010.110.8.427
The Journal of the American Osteopathic Association, August 2010, Vol. 110, 427-436. doi:https://doi.org/10.7556/jaoa.2010.110.8.427
Abstract

Context: The Food and Drug Administration (FDA) recently approved rosuvastatin calcium for prevention of cardiovascular events in patients who have elevated levels of high-sensitivity C-reactive protein (hs-CRP) but not overt hyperlipidemia. The FDA's decision was based primarily on research reported by the JUPITER (Justification for the Use of Statins in Prevention: An Intervention Trial Evaluating Rosuvastatin) Study Group. The cost-effectiveness of such treatment is unknown.

Objective: To compare the cost-effectiveness of treatment with rosuvastatin vs standard management, according to Framingham Risk Score (FRS), for the primary prevention of cardiovascular events in patients who have hs-CRP levels of 2.0 mg/L or higher and low-density lipoprotein cholesterol (LDL-C) levels of less than 130 mg/dL.

Methods: A Markov-type model was used to calculate the incremental cost-effectiveness ratio of rosuvastatin (20 mg daily) vs standard management for the primary prevention of cardiovascular events in patients over a 10-year period. Cost data were obtained from the Centers for Medicare & Medicaid Services and the Red Book drug reference. Health utility measures were obtained from the literature. Cardiovascular event data were obtained directly from the JUPITER Study Group. One-way sensitivity analysis and probabilistic sensitivity analysis were conducted.

Results: Treating patients with rosuvastatin to prevent cardiovascular events based on a hs-CRP level greater than 2.0 mg/L and an LDL-C level of 130 mg/dL or lower would result in estimated incremental cost-effectiveness ratios of $35,455 per quality-adjusted life year (QALY) in patients with an FRS greater than 10% and $90,714 per QALY in patients with an FRS less than or equal to 10%. Results of probabilistic sensitivity analysis suggested that in patients with an FRS greater than 10%, the probability that rosuvastatin is considered cost-effective at $50,000 per QALY is approximately 98%. In patients with an FRS less than or equal to 10%, the probability that rosuvastatin is considered cost-effective at $50,000 per QALY is 0%.

Conclusions: Compared with standard management, treatment with rosuvastatin is a cost-effective strategy over a 10-year period for preventing cardiovascular events in patients with FRS greater than 10%, elevated hs-CRP levels, and normal LDL-C levels.

Complications from cardiovascular disease (CVD) account for more than one-third of all deaths in the United States—more deaths than from any other single disease.1 Rates of first cardiovascular events (eg, myocardial infarction, ischemic stroke, unstable angina) range from 3 per 1000 men at ages 35 to 44 years to 74 per 1000 men at ages 85 to 94 years.1 Comparable rates occur for women 10 years later in life.1 The monetary impact of CVD in the United States in 2009 was an estimated $475.3 billion, including direct and indirect costs.1 These cost are almost certain to increase in the future. 
Current guidelines for the primary prevention of cardiovascular events center on the reduction of risk factors.2 These risk factors, which form the basis for calculating the 10-year risk of a cardiovascular event, are derived from results of the Framingham Heart Study.2 Many of the risk factors, such as blood pressure, cholesterol level, and smoking status, are modifiable. However, age, sex, and family history are not modifiable. In most cases, drug therapy is instituted after an attempt to modify or reduce risk has failed. Treatment with HMG-CoA reductase inhibitors (ie, statins) is the mainstay of pharmacotherapy for lowering low-density lipoprotein cholesterol (LDL-C) levels and cardiovascular event risk.2 
In recent years, inflammation has been recognized as having a significant impact on the development and progression of atherosclerosis.3-5 In an effort to isolate a marker for inflammation, high-sensitivity C-reactive protein (hs-CRP) has become the biomarker of choice.6,7 The hs-CRP level is a precise indicator of an ongoing inflammatory process. However, this biomarker is nonspecific for CVD. Elevated hs-CRP levels are also associated with cancer, diabetes mellitus, high blood pressure, increased age, infection, metabolic syndrome, obesity, obstructive sleep apnea, rheumatologic disorders, and smoking.8-12 
Statin therapy has been shown to lower hs-CRP levels and to reduce the risk of cardiovascular events in patients with elevated hs-CRP levels.13,14 Routine measurement of hs-CRP levels for the purpose of intervention is not currently recommended, though elevated levels are recognized to play an important role in the development of CVD.2 
In February 2010, the US Food and Drug Administration (FDA) approved rosuvastatin calcium (Crestor; AstraZeneca, Wilmington, Delaware) for the prevention of cardiovascular events in women aged 60 years or older and in men aged 50 years or older who have hs-CRP levels greater than or equal to 2 mg/L and one additional CVD risk factor—without hyperlipidemia.15 Implementation of hs-CRP monitoring for means of risk stratification and treatment will lead to expansion of rosuvastatin treatment as millions of Americans will be prescribed a statin for the first time.16 Although previous studies have suggested that statin therapy is cost-effective in both primary and secondary prevention of cardiovascular events, there have been no previous statin cost-effectiveness studies examining a patient population with elevated levels of hs-CRP and relatively low levels of LDL-C.17,18 
The purpose of the present study was to design a model that would estimate the cost-effectiveness of rosuvastatin therapy according to Framingham Risk Score (FRS) for the primary prevention of cardiovascular events in patients with elevated levels of hs-CRP and normal levels of LDL-C over a period of 10 years. 
Methods
For the present study, cardiovascular event data were obtained directly from JUPITER (Justification for the Use of Statins in Prevention: An Intervention Trail Evaluating Rosuvastatin) Study Group (Paul M. Ridker, MD, MPH, personal communication, September 2009), allowing for the separation of JUPITER patients into two categories—patients with an FRS greater than 10% and patients with an FRS less than or equal to 10%. The information received was in a similar format to that reported in table 3 of the JUPITER Study Group's previous publication.14 
Patients in JUPITER14 had a median age of 66 years. More than one-third of the patients were women. Initial LDL-C levels for patients were within normal ranges for their particular health states, and initial hs-CRP levels were greater than or equal to 2.0 mg/L. All patients were statin naive and without previous history of cardiovascular events.14 
Design of the Markov Model
TreeAge Pro 2009 software (TreeAge Software, Inc, Williamstown, Massachusetts) was used to design a Markov-type model from a healthcare system perspective (Figure 1). A Markov model is a type of decision-analytic model that is dynamic over time. It allows for variation in disease states, costs, and health utility over a specified time frame. This model was applied to the group of patients with FRS greater than 10% and the group with FRS less than or equal to 10%, mirroring the design of JUPITER14 in terms of treatment arms and outcomes assessment. The treatment arms used in the model were rosuvastatin therapy and standard management, which was defined as no statin therapy, with annual monitoring of lipid profiles and hs-CRP levels. 
The model applied the same exclusion criteria for patient eligibility and hs-CRP testing as established in JUPITER.14 The rationale for using these exclusion criteria was based on the lack of specificity of hs-CRP testing. Clinicians should recognize that patients who have disease states that would elicit an elevated hs-CRP level are unlikely to benefit from such testing or treatment. Whether there are outcome benefits in such patients remains to be seen. 
The model was designed for primary prevention insofar that only those parameters up to and including a cardiovascular event were considered. A 10-year timeframe was used in an effort to coordinate the timeline of cost-effectiveness with that of disease risk models, such as the FRS model and the Reynolds Risk Score model.19-21 
The outcome used for the design of the model was a primary endpoint event, defined as a patient having experienced a nonfatal myocardial infarction, nonfatal ischemic stroke, arterial revascularization, hospitalization for unstable angina, or death from a cardiovascular cause. The model was designed to move patients through 40 cycles of 3 months per cycle, for a total of 10 years, using constant probabilities for each cycle. 
The patients in the statin group entered the model as being assigned to a “healthy/no event” state, an adverse event state, or a primary endpoint event state, based on the probability of such events (Table 1). An adverse event was defined as any event requiring discontinuation of treatment.22 Any discontinuation of therapy led the patient to cycle through the model without treatment, as if he or she were in the standard management group. Patients in the standard management group entered the model as being assigned to a “healthy/no event” state or a primary endpoint event state, based on the probability of such events (Table 1). 
Table 1
Model Parameters and Base Case Values Used in 1-Way Sensitivity Analysis
Model Parameters
Weight*
Base Case Value (Range)†
Value Reference (Range Reference)
▪ Probabilities of Primary Endpoint Event‡
□ FRS >10%
– Probability of primary endpoint event
    Statin group ... 0.0031 (0.0024-0.0037) Ridker§ (Author assumption)
    Standard management group...0.0054 (0.0043-0.0065)Ridker§ (Author assumption)
□ FRS ≤10%
– Probability of primary endpoint event
    Statin group ... 0.0011 (0.0009-0.0013) Ridker§ (Author assumption)
    Standard management group...0.0020 (0.0016-0.0024)Ridker§ (Author assumption)
– Probability of adverse event∥ ... 0.014 (0.008-0.02) AstraZeneca23 (Author assumption)
▪ Costs, mean
□ Outpatient costs, statin group
– Office visit (CPT 99213)...$61.31CMS24
– Lipid panel ... $19.01 CMS25
– Alanine aminotransferase...$10.44CMS25
– Asparate aminotransferase ... $10.20 CMS25
– hs-CRP...$25.55CMS25
– Venipuncture ... $3.00 CMS25
–Total initial outpatient cost...$129.51 ($103.61-$155.41) (Author assumption)
– Total incremental outpatient cost# ... $32.38 ($25.90-$38.85) # (Author assumption)
□ Outpatient costs, standard management group
– Office visit (CPT 99213) ... $61.31 CMS24
– Lipid panel...$19.01CMS25
– hs-CRP ... $25.55 CMS25
– Venipuncture...$3.00CMS25
– Total initial outpatient cost ... $108.87 ($87.10-$130.64) (Author assumption)
– Total incremental outpatient cost#...$27.22 ($21.77-$32.66)# (Author assumption)
□ Other costs
– Rosuvastatin, 20 mg/tablet...$4.11 ($0.13-$5.00)Red Book23 (Author assumption)
– Primary endpoint event ... $13,123 ($5397-$26,770) ** (Author estimate)
– Nonfatal myocardial infarction0.14$6891.39 ($4459.66-$9951.16)CMS26,27 (Author estimate)
– Nonfatal stroke 0.19 $7973.14 ($4327.86-$14,727.25) CMS26,27 (Author estimate)
– Arterial revascularization0.45$20,581.76 ($9809.11-$39,200.54)CMS26,27 (Author estimate)
– Hospitalization for unstable angina 0.10 $2549.85 ($2039.88-$3059.82) CMS26,27 (Author estimate)
– Death from cardiovascular causes0.12$7599.23 ($0.00-$39,200.54)†† (Author estimate)
▪ Health Utilities‡‡
□ Primary endpoint event...0.606 (0-1) **
□ Nonfatal myocardial infarction 0.14 0.704 Sullivan and Ghushchyan28
□ Nonfatal stroke0.190.650Sullivan and Ghushchyan28
□ Arterial revascularization 0.45 0.6995 Author estimate
□ Hospitalization for unstable angina0.100.695Sullivan and Ghushchyan28
□ Death from cardiovascular causes 0.12 0 Author assumption
□ Healthy/no event...1.00 (0.90-1.00)Author assumption
 Abbreviations: CMS, Centers for Medicare & Medicaid Services; CPT, Current Procedural Terminology; FRS, Framingham Risk Score; hs-CRP, high-sensitivity C-reactive protein; NA, not applicable.
 *Weight estimates based on event numbers in JUPITER study.14 Weights apply only to parameters used to calculate weighted averages.
 †Ranges specified only where appropriate for data analysis.
 ‡Probabilities—on a 0-1 scale—calculated for 3-month period. The primary endpoint event was defined as a nonfatal myocardial infarction, nonfatal stroke, arterial revascularization, hospitalization for unstable angina, or death from cardiovascular causes.
 §According to the JUPITER Study Group (Paul M. Ridker, MD, MPH, personal communication, September 2009).
 ∥Adverse event modelled as a one-time event during the initial 3 months of statin therapy. Probability is on a 0-1 scale.
 Costs calculated for initial 3-month period as sum of all outpatient costs.
 #Costs calculated for incremental 3-month period as sum of all outpatient costs divided by 2. One office visit per year was assumed (ie, one-quarter of costs of initial office visit for each incremental 3-month period).
 **Weighted average of primary endpoint events. Range explained in “Utility Data” section of text.
 ††Estimated cost calculated as average of the other events. Low range of $0 was assumed because a patient could expire without being admitted to a hospital.
 ‡‡Utility is defined on a 0-1 scale, in which 0 represents death and 1 represents perfect health.
View Large
Table 1
Model Parameters and Base Case Values Used in 1-Way Sensitivity Analysis
Model Parameters
Weight*
Base Case Value (Range)†
Value Reference (Range Reference)
▪ Probabilities of Primary Endpoint Event‡
□ FRS >10%
– Probability of primary endpoint event
    Statin group ... 0.0031 (0.0024-0.0037) Ridker§ (Author assumption)
    Standard management group...0.0054 (0.0043-0.0065)Ridker§ (Author assumption)
□ FRS ≤10%
– Probability of primary endpoint event
    Statin group ... 0.0011 (0.0009-0.0013) Ridker§ (Author assumption)
    Standard management group...0.0020 (0.0016-0.0024)Ridker§ (Author assumption)
– Probability of adverse event∥ ... 0.014 (0.008-0.02) AstraZeneca23 (Author assumption)
▪ Costs, mean
□ Outpatient costs, statin group
– Office visit (CPT 99213)...$61.31CMS24
– Lipid panel ... $19.01 CMS25
– Alanine aminotransferase...$10.44CMS25
– Asparate aminotransferase ... $10.20 CMS25
– hs-CRP...$25.55CMS25
– Venipuncture ... $3.00 CMS25
–Total initial outpatient cost...$129.51 ($103.61-$155.41) (Author assumption)
– Total incremental outpatient cost# ... $32.38 ($25.90-$38.85) # (Author assumption)
□ Outpatient costs, standard management group
– Office visit (CPT 99213) ... $61.31 CMS24
– Lipid panel...$19.01CMS25
– hs-CRP ... $25.55 CMS25
– Venipuncture...$3.00CMS25
– Total initial outpatient cost ... $108.87 ($87.10-$130.64) (Author assumption)
– Total incremental outpatient cost#...$27.22 ($21.77-$32.66)# (Author assumption)
□ Other costs
– Rosuvastatin, 20 mg/tablet...$4.11 ($0.13-$5.00)Red Book23 (Author assumption)
– Primary endpoint event ... $13,123 ($5397-$26,770) ** (Author estimate)
– Nonfatal myocardial infarction0.14$6891.39 ($4459.66-$9951.16)CMS26,27 (Author estimate)
– Nonfatal stroke 0.19 $7973.14 ($4327.86-$14,727.25) CMS26,27 (Author estimate)
– Arterial revascularization0.45$20,581.76 ($9809.11-$39,200.54)CMS26,27 (Author estimate)
– Hospitalization for unstable angina 0.10 $2549.85 ($2039.88-$3059.82) CMS26,27 (Author estimate)
– Death from cardiovascular causes0.12$7599.23 ($0.00-$39,200.54)†† (Author estimate)
▪ Health Utilities‡‡
□ Primary endpoint event...0.606 (0-1) **
□ Nonfatal myocardial infarction 0.14 0.704 Sullivan and Ghushchyan28
□ Nonfatal stroke0.190.650Sullivan and Ghushchyan28
□ Arterial revascularization 0.45 0.6995 Author estimate
□ Hospitalization for unstable angina0.100.695Sullivan and Ghushchyan28
□ Death from cardiovascular causes 0.12 0 Author assumption
□ Healthy/no event...1.00 (0.90-1.00)Author assumption
 Abbreviations: CMS, Centers for Medicare & Medicaid Services; CPT, Current Procedural Terminology; FRS, Framingham Risk Score; hs-CRP, high-sensitivity C-reactive protein; NA, not applicable.
 *Weight estimates based on event numbers in JUPITER study.14 Weights apply only to parameters used to calculate weighted averages.
 †Ranges specified only where appropriate for data analysis.
 ‡Probabilities—on a 0-1 scale—calculated for 3-month period. The primary endpoint event was defined as a nonfatal myocardial infarction, nonfatal stroke, arterial revascularization, hospitalization for unstable angina, or death from cardiovascular causes.
 §According to the JUPITER Study Group (Paul M. Ridker, MD, MPH, personal communication, September 2009).
 ∥Adverse event modelled as a one-time event during the initial 3 months of statin therapy. Probability is on a 0-1 scale.
 Costs calculated for initial 3-month period as sum of all outpatient costs.
 #Costs calculated for incremental 3-month period as sum of all outpatient costs divided by 2. One office visit per year was assumed (ie, one-quarter of costs of initial office visit for each incremental 3-month period).
 **Weighted average of primary endpoint events. Range explained in “Utility Data” section of text.
 ††Estimated cost calculated as average of the other events. Low range of $0 was assumed because a patient could expire without being admitted to a hospital.
 ‡‡Utility is defined on a 0-1 scale, in which 0 represents death and 1 represents perfect health.
×
Patients who experienced a primary endpoint event would transition to the primary endpoint event state. This transition was a temporary state in the model, allowing for appropriate accumulation of costs and reduced quality-adjusted life year (QALY) measures before terminating into a secondary prevention state. Because the analysis evaluated primary prevention only, a secondary prevention state was built into the model as the final absorbing state. This secondary prevention state was used for accounting purposes only, and it was not populated with costs or utility data. 
Patients without an event in each cycle were allowed to continue cycling through the model as healthy patients with no event—either until an event occurred or until the end of the 10-year study period. 
Figure 1.
Markov-type model used to calculate cost-effectiveness estimation.
Markov-type model used to calculate cost-effectiveness estimation.
View Original  |  Slide (.ppt)
Figure 1.
Markov-type model used to calculate cost-effectiveness estimation.
Markov-type model used to calculate cost-effectiveness estimation.
View Original  |  Slide (.ppt)
Clinical Parameters
Outcomes data associated with the primary endpoint event were obtained directly from the JUPITER Study Group (Dr Ridker, personal communication, September 2009). Constant event probabilities over the 10-year timeline were used in the model and were not adjusted for age-associated risk. Actual event data are not reported with any measure of certainty. The uncertainty in the event data was compensated in 1-way sensitivity analysis by setting the standard deviation at an arbitrary 20% to achieve robustness. 
The 3-month probabilities of a primary endpoint are summarized in Table 1. An adverse event rate of 1.4% was applied within the first 3 months of statin therapy. 
Cost Data
Costs analyzed in the present study included costs of the study drug (rosuvastatin, 20 mg daily), costs associated with outpatient management, and costs associated with cardiovascular events (Table 1). The average wholesale price of rosuvastatin 20 mg was obtained from the 2009 edition of the Red Book drug reference published by Thomson Reuters.23 Because this price does not accurately reflect costs that the typical patient or insurer might pay now or in the future, the price of the drug was arbitrarily adjusted in 1-way sensitivity analysis between $4.00 per month (the price of a generic statin at many retail pharmacies) and $150.00 per month (a price that exceeds the common retail price of rosuvastatin). 
Costs associated with outpatient management were calculated at national levels for Current Procedural Terminology (CPT) codes for a typical office visit and laboratory tests associated with statin therapy.24,25 All outpatient management costs were adjusted by 20% for 1-way sensitivity analysis to account for uncertainty in the estimates, for variability of regional and institutional reimbursement rates, and for clinicians' outpatient management practices. 
Costs associated with cardiovascular events were calculated according to Diagnosis Related Groups (DRGs), which were available from the Centers for Medicare & Medicaid Services.26,27 A composite cost for the primary endpoint event was estimated by assigning weights to each of the events. These weights were estimated based on the event numbers published in JUPITER and were assumed to be constant over the 10-year period for the present analysis.14 All outcomes contained within the primary endpoint had multiple DRGs associated with them. To arrive at a point estimate for each event, the costs of all DRGs within a given event were averaged, and the range (lowest cost to highest cost) was used for 1-way sensitivity analysis (Table 1). 
Costs associated with adverse reactions were omitted under the assumption that the majority of these reactions would subside upon discontinuation of therapy. 
Utility Data
Utilities of the various events in the model were obtained from the literature.28 Event utilities—more appropriately termed health utilities—are quantitative assessments of a particular health state. Health utility is commonly defined on a 0-to-1 scale, in which 0 represents death and 1 represents perfect health. A composite health utility was calculated using event weights. 
The utility of arterial revascularization was estimated as the average utility of nonfatal myocardial infarction and unstable angina. The reasoning for using this method was that patients who receive a revascularization procedure typically experience a myocardial infarction or unstable angina before the procedure is performed. A state of healthy/no event was included in the model to represent all individuals within a cycle who did not have a primary endpoint event. The utility for this health state was assumed to have a range of 0.9 to 1 to account for uncertainty in the health state of each patient as he or she cycled through the model. 
Analyses
An incremental cost-effectiveness ratio (ICER) was calculated as a cost per QALY, which was obtained after the model was populated with the cost and utility estimates and event probabilities. Costs and utilities were discounted at 3% per year (0.75% per cycle) and were adjusted between 0% and 7% for sensitivity analysis. 
All parameters in the model, except for those associated with the probability of no event, were analyzed using 1-way sensitivity analysis and were allowed to vary between the high and low values shown in Table 1. The utility (ie, QALY) associated with a primary endpoint event was allowed to vary between 0 and 1. The rationale behind this decision was to allow the range in which the utility varied to act as a proxy for the amount of time that a patient remained in a primary endpoint event state. 
A probabilistic sensitivity analysis was conducted with a sample size of 10,000 and a sampling rate of once per Markov stage. The results were plotted on a cost-effectiveness acceptability curve. All model inputs were assigned a distribution, and distribution parameters (ie, alpha, beta) were estimated using the point estimate and a 20% standard deviation. A gamma distribution was assigned to all cost estimates, and a beta distribution was assigned to all probabilities and health utility estimates. 
The distribution parameters for the probability of a primary endpoint event were calculated from the event data obtained from the JUPITER Study Group (Dr Ridker, personal communication, September 2009) by using the TreeAge Pro 2009 software. 
Results
Base Case Analysis
Based on the data obtained from the JUPITER Study Group (Dr Ridker, personal communication, September 2009) and other sources, it was estimated that using rosuvastatin to prevent a cardiovascular event in patients with FRS greater than 10% would cost $11,700 more per patient than would standard management over a 10-year period. Using rosuvastatin to treat patients with FRS less than or equal to 10% would cost an estimated $12,700 more than standard management. These estimates include the cost of the statin, the costs of outpatient medical management, and the cost associated with a cardiovascular event. 
Over the 10-year study period, patients treated with rosuvastatin accrued an estimated 0.33 more QALYs compared to their standard management counterpart if they had an FRS greater than 10%, and an estimated 0.14 more QALYs compared to their standard management counterpart if they had an FRS less than or equal to 10%. The resulting ICER was estimated at $35,455 per QALY gained in those patients with an FRS greater than 10%—but $90,714 per QALY gained in patients with an FRS less than or equal to 10% (Table 2). 
Table 2
Cost-Effectiveness of Statin Therapy vs Standard Management *
Parameter
Value
▪ FRS >10%
□ Total cost
– Statin group$14,900
– Standard management group $3200
□ Total QALY
– Statin group 8.37
– Standard management group8.04
□ Incremental cost $11,700
□ Incremental QALY0.33
□ ICER $35,455/QALY
▪ FRS ≤10%
□ Total cost
– Statin group$14,600
– Standard management group $1900
□ Total QALY
– Statin group 8.68
– Standard management group8.54
□ Incremental cost $12,700
□ Incremental QALY0.14
□ ICER $90,714/QALY
 Abbreviations: FRS, Framingham Risk Score; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year.
 *Data based on JUPITER study,14 Red Book,23 Centers for Medicare & Medicaid Services,24-27 and other sources.
View Large
Table 2
Cost-Effectiveness of Statin Therapy vs Standard Management *
Parameter
Value
▪ FRS >10%
□ Total cost
– Statin group$14,900
– Standard management group $3200
□ Total QALY
– Statin group 8.37
– Standard management group8.04
□ Incremental cost $11,700
□ Incremental QALY0.33
□ ICER $35,455/QALY
▪ FRS ≤10%
□ Total cost
– Statin group$14,600
– Standard management group $1900
□ Total QALY
– Statin group 8.68
– Standard management group8.54
□ Incremental cost $12,700
□ Incremental QALY0.14
□ ICER $90,714/QALY
 Abbreviations: FRS, Framingham Risk Score; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year.
 *Data based on JUPITER study,14 Red Book,23 Centers for Medicare & Medicaid Services,24-27 and other sources.
×
1-Way Sensitivity Analysis
In the 1-way sensitivity analysis of patients with an FRS greater than 10%, the ICER was most sensitive to the following variables: the probability of a primary endpoint event in the standard management group, the cost of the statin, the probability of a primary endpoint event in the statin group, and the utility of a primary endpoint event (Figure 2). In the 1-way sensitivity analysis of patients with an FRS less than or equal to 10%, the ICER was most sensitive to the following variables: the cost of the statin, the probability of a primary endpoint event in the standard management group, the probability of a primary endpoint event in the statin group, the utility of a primary endpoint event, the discount rate, and the utility of no event (Figure 3). 
The maximum incremental cost-effectiveness ratio observed during 1-way sensitivity analysis was $69,000 per QALY for patients with an FRS greater than 10% and $165,000 per QALY for patients with an FRS less than or equal to 10% (Table 3). These ratios indicate that if all other parameter estimates are unbiased, and if the true value of the probability of a primary endpoint event in the standard management group is 20% less than that reported in JUPITER,14 then the resulting ICER in both Framingham cohorts would be as stated. 
Table 3
Results of 1-Way Sensitivity Analysis
US $/QALY
Parameter*
 Low Value, High Value
 Biased Toward Statin
 Biased Against Statin
▪ FRS >10%
□ Probability of primary endpoint event†
– Statin group0.0024, 0.003723,00069,000
– Standard management group 0.0043, 0.0065 27,000 48,000
□ Discount rate0%, 7%33,00038,000
□ Cost of statin per tablet $0.13, $5.00 -1000 43,000
□ Cost of primary endpoint event$5397, $26,77033,00037,000
□ Incremental outpatient costs
– Statin group$25.90, $38.8535,00036,000
– Standard management group $21.77, $32.66 35,000 36,000
□ Utility of primary endpoint event‡0, 1§34,00043,000
□ Utility of no event‡ 0.9, 1 35,000 39,000
▪ FRS ≤10%
□ Probability of primary endpoint event†
– Statin group0.0024, 0.003774,000118,000
– Standard management group 0.0043, 0.0065 63,000 165,000
□ Discount rate0%, 7%86,00099,000
□ Cost of statin per tablet $0.13, $5.00 2000 111,000
□ Cost of primary endpoint event$5397, $26,77088,00093,000
□ Incremental output costs
– Statin group$25.90, $38.8590,00093,000
– Standard management group $21.77, $32.66 90,000 93,000
□ Utility of primary endpoint event‡0, 1§88,000111,000
□ Utility of no event‡ 0.9, 1 91,000 102,000
 Abbreviations: FRS, Framingham Risk Score; QALY, quality-adjusted life year.
 *Only those parameters found to be sensitive are included in table.
 †Probabilities are on a 0-1 scale.
 ‡Utility is defined on a 0-1 scale, in which 0 represents death and 1 represents perfect health.
 §Range explained in “Utility Data” section of text.
View Large
Table 3
Results of 1-Way Sensitivity Analysis
US $/QALY
Parameter*
 Low Value, High Value
 Biased Toward Statin
 Biased Against Statin
▪ FRS >10%
□ Probability of primary endpoint event†
– Statin group0.0024, 0.003723,00069,000
– Standard management group 0.0043, 0.0065 27,000 48,000
□ Discount rate0%, 7%33,00038,000
□ Cost of statin per tablet $0.13, $5.00 -1000 43,000
□ Cost of primary endpoint event$5397, $26,77033,00037,000
□ Incremental outpatient costs
– Statin group$25.90, $38.8535,00036,000
– Standard management group $21.77, $32.66 35,000 36,000
□ Utility of primary endpoint event‡0, 1§34,00043,000
□ Utility of no event‡ 0.9, 1 35,000 39,000
▪ FRS ≤10%
□ Probability of primary endpoint event†
– Statin group0.0024, 0.003774,000118,000
– Standard management group 0.0043, 0.0065 63,000 165,000
□ Discount rate0%, 7%86,00099,000
□ Cost of statin per tablet $0.13, $5.00 2000 111,000
□ Cost of primary endpoint event$5397, $26,77088,00093,000
□ Incremental output costs
– Statin group$25.90, $38.8590,00093,000
– Standard management group $21.77, $32.66 90,000 93,000
□ Utility of primary endpoint event‡0, 1§88,000111,000
□ Utility of no event‡ 0.9, 1 91,000 102,000
 Abbreviations: FRS, Framingham Risk Score; QALY, quality-adjusted life year.
 *Only those parameters found to be sensitive are included in table.
 †Probabilities are on a 0-1 scale.
 ‡Utility is defined on a 0-1 scale, in which 0 represents death and 1 represents perfect health.
 §Range explained in “Utility Data” section of text.
×
When the cost of rosuvastatin was set at a retail price of $4.00 per month, the ICER was -$1000 per QALY for patients with an FRS greater than 10% and $2,000 per QALY for patients with an FRS less than or equal to 10% (Table 3). 
Probabilistic Sensitivity Analysis
When event probabilities, costs, and utilities were simultaneously varied according to their respective probability distributions, the cost-effectiveness of rosuvastatin therapy compared with standard management in patients with an FRS greater than 10% at the 10th, 50th, and 90th percentiles were, respectively, $34,400, $37,000, and $39,100 per QALY. The probability that rosuvastatin therapy, compared to standard management, was considered cost-effective at a $50,000 per QALY threshold was approximately 98% (Figure 4). 
In patients with an FRS less than or equal to 10%, the cost-effectiveness at the 10th, 50th, and 90th percentiles were, respectively, $87,400, $90,500, and $93,800 per QALY. The probability that rosuvastatin therapy, compared to standard management, was considered cost-effective at a $50,000 per QALY threshold was 0% (Figure 4). 
Comment
The main limitations of the present analysis are the failure to account for the increased probability of a cardiovascular event over time and the extrapolation of event results beyond the timeline of JUPITER.14 Although the model used in this analysis held event probabilities constant over the 10-year time-frame, any increase in event probability associated with age was similar in both arms of the study. Such similar increases are “differenced out” for the most part during calculation of the ICER. 
A statistically significant difference (P=.01) was detected in JUPITER between rosuvastatin treatment and standard management in the number of patients who had diabetes mellitus. In addition, a recent meta-analysis reported a 9% increased risk of incident diabetes mellitus among patients on statin therapy for a period of 4 years (odds ratio, 1.09; 95% confidence interval, 1.02-1.17).29 Costs associated with the possible development of diabetes mellitus were not considered in the present analysis. 
The present analysis was also limited in that it took into account only the outcomes data from the single existing study—JUPITER14—examining the outcomes benefit of statins in patients who have elevated hs-CRP levels. This limitation subjected the current analysis to many of the same biases that are present in JUPITER.14 Further research is needed to evaluate whether patients will have comparable outcomes from using other statins besides rosuvastatin. If results of such studies are similar to that of JUPITER,14 a generic statin could be substituted, leading to a large cost savings and a much lower ICER. 
Cost Questions
The price of statins has decreased in recent years as a result of a proliferative generic market and fierce competition in the branded market. Reduced price, combined with the relative safety of statins, has led to the question of whether expansion of statin treatment as a whole is cost-effective. A recent study by Pletcher et al30 suggests that when the National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) guidelines are fully implemented for the primary prevention of cardiovascular events, statin therapy will be cost-effective, with an estimated ICER of $42,000 per QALY given a statin price of $2.11 or less per tablet. 
AstraZeneca's patent for rosuvastatin will expire in 2016,31 allowing generic drug manufacturers to seek profits from rosuvastatin's indication for hyperlipidemia and elevated hs-CRP levels. As the price for rosuvastatin decreases, its ICER will decrease. When the price of rosuvastatin equals $8 for a 30-day supply, the ICER will be $0 per QALY in patients with an FRS greater than or equal to 10%. That will represent the breakeven point at which there is no net cost for treatment. 
An estimated 17.9 million people in the United States older than 35 years would require testing for hs-CRP levels if results of JUPITER14,16 were applied broadly. Patients in JUPITER had a median age of 66 years.14 When stratifying patients based on FRS and considering only those patients with an FRS greater than 10% based on the favorable cost-effectiveness established in the present analysis, approximately 10.2 million Americans would require hs-CRP testing.16 If after hs-CRP testing, treatment was restricted to patients with a hs-CRP level greater than 3.0 mg/L, 2.1 million Americans would require treatment with a statin.16 This restriction could result in a lower ICER than was established in this analysis. 
A large number of myocardial infarctions and cardiovascular events occur in individuals who have normal cholesterol levels and no history of atherosclerotic disease. The most commonly used CVD risk model—the Framingham Risk Score model—does not account for hs-CRP levels in risk stratification, whereas the Reynolds Risk Score model does. If we assume that the Reynolds Risk Score is a validated and acceptable method for calculating the risk of a cardiovascular event, the following two questions appear obvious: (1) are statins a cost-effective way of lowering cardiovascular event risk in patients with elevated hs-CRP levels?, and (2) is measurement of hs-CRP levels for the purposes of preventing disease cost-effective? 
Figure 2.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score greater than 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the probability of a primary endpoint event in the standard management group, the cost of rosuvastatin, the probability of a primary endpoint event in the statin group, and the utility of an aggregated primary endpoint event.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score greater than 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the probability of a primary endpoint event in the standard management group, the cost of rosuvastatin, the probability of a primary endpoint event in the statin group, and the utility of an aggregated primary endpoint event.
View Original  |  Slide (.ppt)
Figure 2.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score greater than 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the probability of a primary endpoint event in the standard management group, the cost of rosuvastatin, the probability of a primary endpoint event in the statin group, and the utility of an aggregated primary endpoint event.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score greater than 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the probability of a primary endpoint event in the standard management group, the cost of rosuvastatin, the probability of a primary endpoint event in the statin group, and the utility of an aggregated primary endpoint event.
View Original  |  Slide (.ppt)
Figure 3.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score less than or equal to 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the cost of rosuvastatin, the probability of a primary endpoint event in the standard management group, the probability of a primary endpoint event in the statin group, the utility of a primary endpoint event, the discount rate, and the utility of no event.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score less than or equal to 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the cost of rosuvastatin, the probability of a primary endpoint event in the standard management group, the probability of a primary endpoint event in the statin group, the utility of a primary endpoint event, the discount rate, and the utility of no event.
View Original  |  Slide (.ppt)
Figure 3.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score less than or equal to 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the cost of rosuvastatin, the probability of a primary endpoint event in the standard management group, the probability of a primary endpoint event in the statin group, the utility of a primary endpoint event, the discount rate, and the utility of no event.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score less than or equal to 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the cost of rosuvastatin, the probability of a primary endpoint event in the standard management group, the probability of a primary endpoint event in the statin group, the utility of a primary endpoint event, the discount rate, and the utility of no event.
View Original  |  Slide (.ppt)
Conclusion
In the present study, a Markov-type model was designed to examine the cost-effectiveness of rosuvastatin therapy vs standard management for the primary prevention of cardiovascular events in patients with elevated hs-CRP levels according to FRS. The results of this analysis indicate that rosuvastatin is cost-effective for patients with normal LDL-C levels, elevated hs-CRP levels, and an FRS greater than 10%, given a threshold of $50,000 per QALY. The ICER was favorable across the entire range of costs and health utilities in this cohort and exceeded the $50,000 per QALY threshold only when event probabilities were set at extreme values. In patients with an FRS less than or equal to 10%, the cost-effectiveness of rosuvastatin is considered favorable only when this drug's price is less than $2.35 per tablet. 
Although the present study does not address the much broader question of whether hs-CRP monitoring is cost-effective, it does answer the more limited question of whether rosuvastatin is cost-effective in patients with elevated hs-CRP levels. It is the opinion of this author that the present study represents one of many necessary steps needed to answer that broader question. Thus, further research is essential to assess whether hs-CRP monitoring for the purposes of primary prevention is cost-effective. 
Figure 4.
Cost-effectiveness acceptability curve, comparing both Framingham Risk Score (FRS) groups (FRS >10%, FRS ≥10%). The graph shows the probability that rosuvastatin is cost-effective given different willingness-to-pay thresholds. Such thresholds vary according to payer (ie, third-party or government).
Cost-effectiveness acceptability curve, comparing both Framingham Risk Score (FRS) groups (FRS >10%, FRS ≥10%). The graph shows the probability that rosuvastatin is cost-effective given different willingness-to-pay thresholds. Such thresholds vary according to payer (ie, third-party or government).
View Original  |  Slide (.ppt)
Figure 4.
Cost-effectiveness acceptability curve, comparing both Framingham Risk Score (FRS) groups (FRS >10%, FRS ≥10%). The graph shows the probability that rosuvastatin is cost-effective given different willingness-to-pay thresholds. Such thresholds vary according to payer (ie, third-party or government).
Cost-effectiveness acceptability curve, comparing both Framingham Risk Score (FRS) groups (FRS >10%, FRS ≥10%). The graph shows the probability that rosuvastatin is cost-effective given different willingness-to-pay thresholds. Such thresholds vary according to payer (ie, third-party or government).
View Original  |  Slide (.ppt)
 Editor's Note: Corrections to this article were published in the September 2010 issue of JAOA—The Journal of the American Osteopathic Association (2010;110[9]:508). The corrections have been incorporated in this online version of the article, which was posted January 2011. An explanation of these changes is available at http://www.jaoa.org/cgi/content/full/110/9/508.
 
 Financial Disclosures: None reported.
 
I thank Lehigh University in Bethlehem, Pennsylvania; Philadelphia College of Osteopathic Medicine in Pennsylvania; and Carl Asche, PhD, of the University of Utah College of Pharmacy in Salt Lake City. 
American Heart Association. Heart Disease & Stroke Statistics. 2009 Update At-A-Glance. Dallas, Texas: American Heart Association; 2009. http://www.americanheart.org/downloadable/heart/1240250946756LS-1982%20Heart%20and%20Stroke%20Update.042009.pdf. Accessed July 15, 2010.
Expert Panel on the Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on the Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA. 2001;285(19):2486 –2497.
Scirica BM, Morrow DA. Is C-reactive protein an innocent bystander or proatherogenic culprit? The verdict is still out. Circulation. 2006;113(17):2128-2151. http://circ.ahajournals.org/cgi/reprint/113/17/2128. Accessed July 17, 2010.
Danesh J, Wheeler JG, Hirschfield GM, Eda S, Eiriksdottir G, Rumley A, et al. C-reactive protein and other circulating markers of inflammation in the prediction of coronary heart disease. N Engl J Med. 2004;350(14):1387-1397. http://content.nejm.org/cgi/reprint/350/14/1387.pdf. Accessed July 17, 2010.
Ridker PM, Cushman M, Stampfer MJ, Tracy RP, Hennekens CH. Inflammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med. 1997;336(14):973-979. http://content.nejm.org/cgi/reprint/336/14/973.pdf. Accessed July 17, 2010.
Levinson S. Inflammatory and long-term risk markers. Clin Lab Med. 2006;26(3):553-570.
Koenig W, Sund M, Frohlich M, et al. C-reactive protein, a sensitive marker of inflammation, predicts future risk of coronary heart disease in initially healthy middle-aged men: results from the MONICA (Monitoring Trends and Determinants in Cardiovascular Disease) Augsburg cohort study, 1984 to 1992. Circulation. 1999;99(2):237-242. http://circ.ahajournals.org/cgi/reprint/99/2/237.pdf. Accessed July 17, 2010.
Yudkin JS, Stehouwer CD, Emeis JJ, Coppack SW. C-reactive protein in healthy subjects: associations with obesity, insulin resistance, and endothelial dysfunction: a potential role for cytokines originating from adipose tissue? Arterioscler Thromb Vasc Biol. 1999;19(4):972-978. http://atvb.ahajournals.org/cgi/reprint/19/4/972. Accessed July 17, 2010.
Pradhan AD, Manson JE, Rifai N, Buring JE, Ridker PM. C-reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus. JAMA. 2001;286(3):327-334.
Shamsuzzaman ASM, Winnicki M, Lanfranchi P, Wolk R, Kara T, Accurso V, et al. Elevated C-reactive protein in patients with obstructive sleep apnea. Circulation. 2002;105(21):2462-2464. http://www.circ.ahajournals.org/cgi/reprint/105/21/2462. Accessed July 17, 2010.
Sesso HD, Buring JE, Rifai N, Blake GJ, Gaziano JM, Ridker PM. C-reactive protein and the risk of developing hypertension. JAMA. 2003;290(22):2945-2951.
Rutter MK, Meigs JB, Sullivan LM, D'Agostino RB Sr, Wilson PWF. C-reactive protein, the metabolic syndrome, and prediction of cardiocardiovascular events in the Framingham Offspring Study. Circulation. 2004;110(4):380-385. http://www.circ.ahajournals.org/cgi/reprint/110/4/380. Accessed July 17, 2010.
Ridker PM, Cannon CP, Morrow D, Rifai N, Rose LM, McCabe CH, et al; Pravastatin or Atorvastatin Evaluation and Infection Therapy-Thrombolysis in Myocardial Infarction 22 (PROVE IT-TIMI 22) Investigators. C-reactive protein levels and outcomes after statin therapy. N Engl J Med. 2005;352(1):20-28. http://content.nejm.org/cgi/reprint/352/1/20.pdf. Accessed July 17, 2010.
Ridker PM, Danielson E, Fonseca FAH, Genest J, Gotto AM, Kastelein JJP, et al; for the JUPITER Study Group. Rosuvastatin to prevent cardiovascular events in men and women with elevated C-reactive protein. N Engl J Med. 2008;359(21):2195-2207. http://content.nejm.org/cgi/reprint/359/21/2195.pdf. Accessed July 17, 2010.
FDA approves new indication for Crestor [press release]. Washington, DC: US Food and Drug Administration; February 9, 2010. http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm200128.htm. Accessed July 17, 2010.
Woloshin S, Schwartz LM, Kerin K, Welch HG. Estimating the impact of adding C-reactive protein as a criterion for lipid lowering treatment in the United States. J Gen Intern Med. 2007;22(2):197-204.
Ganz DA, Kuntz KM, Jacobson GA, Avorn J. Cost-effectiveness of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor therapy in older patients with myocardial infarction. Ann Intern Med. 2000;132(10):780-787. http://www.annals.org/content/132/10/780.1.full.pdf+html. Accessed July 17, 2010.
Van Hout BA, Simoons ML. Cost-effectiveness of HMG coenzyme reductase inhibitors. Eur Heart J. 2001;22(9):751-761. http://eurheartj.oxfordjournals.org/cgi/reprint/22/9/751. Accessed July 17, 2010.
D'Agostino RB, Vasan RS, Pencina MJ, Wolf PA, Cobain M, Massaro JM, et al. General cardiovascular risk profile for use in primary care: the Framingham Heart Study. Circulation. 2008;117(6):743-753. http://circ.ahajournals.org/cgi/content/full/117/6/743. Accessed July 17, 2010.
Ridker PM, Paynter NP, Rifai N, Gaziano M, Cook NR. C-reactive protein and parental history improve global cardiovascular risk prediction: the Reynolds Risk Score for men. Circulation. 2008;118(22):2243-2251. http://www.circ.ahajournals.org/cgi/reprint/118/22/2243. Accessed July 17, 2010.
Ridker PM, Buring JE, Rifai N, Cook NR. Development and validation of improved algorithms for the assessment of global cardiovascular risk in women: the Reynolds Risk Score. JAMA. 2007;297(6):611-619.
Crestor (rosuvastatin calcium). Highlights of prescribing information. AstraZeneca Web site, revised June 2010. http://www1.astrazeneca-us.com/pi/crestor.pdf. Accessed June 17, 2010.
Red Book 2009: Pharmacy's Fundamental Reference. 113th ed. New York City, NY: Thomson Reuters;2009 .
Physician fee schedule relative value files, 2009. Centers for Medicare & Medicaid Services Web site. http://www.cms.hhs.gov/PhysicianFeeSched/PFSRVF/list.asp. Accessed December 3, 2009.
Clinical laboratory fee schedule, 2009. Centers for Medicare & Medicaid Services Web site. http://www.cms.hhs.gov/ClinicalLabFeesched/02_clinlab.asp. Accessed May 20, 2009.
Acute inpatient prospective payment system—overview, March 1, 2009. Centers for Medicare & Medicaid Services Web site. http://www.cms.gov/AcuteInpatientPPS/01_overview.asp#TopOfPage. Accessed March 5, 2009.
Centers for Medicare & Medicaid Services Medicare Program. Hospital Inpatient Prospective Payment Systems and Fiscal Year 2009 Rates: Final Fiscal Year 2009 Wage Indices and Payment Rates Including Implementation of Section 124 of the Medicare Improvement for Patients and Providers Act of 2008. Notice. College Park, MD: National Archives and Records Administration, Federal Register; October(2008). : 130. FR document no. E8-23083.
Sullivan PW, Ghushchyan V. Preference-based EQ-5D index scores for chronic conditions in the United States. Med Decis Making. 2006;26(4):410-420. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2634296/pdf/nihms-81406.pdf. Accessed July 17, 2010.
Sattar N, Preiss D, Murray HM, Welsh P, Buckley BM, de Craen AJ, et al. Statins and risk of incident diabetes: a collaborative meta-analysis of randomized statin trials [published online ahead of print February 16, 2010]. Lancet. 2010;375(9716):735-742.
Pletcher MJ, Lazar L, Bibbins-Domingo K, Moran A, Rodondi N, Coxson P, et al. Comparing impact and cost-effectiveness of primary prevention strategies for lipid-lowering. Ann Intern Med. 2009;150(4):243-254. http://www.annals.org/content/150/4/243.full.pdf+html. Accessed July 17, 2010.
CRESTOR patent upheld by US court [press release]. Wilmington, DE: AstraZeneca; June 30, 2010. http://www.astrazeneca.com/search/?itemId=9833512. Accessed July 26, 2010.
Figure 1.
Markov-type model used to calculate cost-effectiveness estimation.
Markov-type model used to calculate cost-effectiveness estimation.
View Original  |  Slide (.ppt)
Figure 1.
Markov-type model used to calculate cost-effectiveness estimation.
Markov-type model used to calculate cost-effectiveness estimation.
View Original  |  Slide (.ppt)
Figure 2.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score greater than 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the probability of a primary endpoint event in the standard management group, the cost of rosuvastatin, the probability of a primary endpoint event in the statin group, and the utility of an aggregated primary endpoint event.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score greater than 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the probability of a primary endpoint event in the standard management group, the cost of rosuvastatin, the probability of a primary endpoint event in the statin group, and the utility of an aggregated primary endpoint event.
View Original  |  Slide (.ppt)
Figure 2.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score greater than 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the probability of a primary endpoint event in the standard management group, the cost of rosuvastatin, the probability of a primary endpoint event in the statin group, and the utility of an aggregated primary endpoint event.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score greater than 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the probability of a primary endpoint event in the standard management group, the cost of rosuvastatin, the probability of a primary endpoint event in the statin group, and the utility of an aggregated primary endpoint event.
View Original  |  Slide (.ppt)
Figure 3.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score less than or equal to 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the cost of rosuvastatin, the probability of a primary endpoint event in the standard management group, the probability of a primary endpoint event in the statin group, the utility of a primary endpoint event, the discount rate, and the utility of no event.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score less than or equal to 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the cost of rosuvastatin, the probability of a primary endpoint event in the standard management group, the probability of a primary endpoint event in the statin group, the utility of a primary endpoint event, the discount rate, and the utility of no event.
View Original  |  Slide (.ppt)
Figure 3.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score less than or equal to 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the cost of rosuvastatin, the probability of a primary endpoint event in the standard management group, the probability of a primary endpoint event in the statin group, the utility of a primary endpoint event, the discount rate, and the utility of no event.
One-way sensitivity analysis, displayed as tornado diagram, for patients with a Framingham Risk Score less than or equal to 10%, rosuvastatin group vs standard management group. Wider bars represent greater sensitivity. Ranges of base case values for various model parameters are shown to the right of the diagram (see also Table 1). The diagram shows that the incremental cost-effectiveness ratio was most sensitive to the cost of rosuvastatin, the probability of a primary endpoint event in the standard management group, the probability of a primary endpoint event in the statin group, the utility of a primary endpoint event, the discount rate, and the utility of no event.
View Original  |  Slide (.ppt)
Figure 4.
Cost-effectiveness acceptability curve, comparing both Framingham Risk Score (FRS) groups (FRS >10%, FRS ≥10%). The graph shows the probability that rosuvastatin is cost-effective given different willingness-to-pay thresholds. Such thresholds vary according to payer (ie, third-party or government).
Cost-effectiveness acceptability curve, comparing both Framingham Risk Score (FRS) groups (FRS >10%, FRS ≥10%). The graph shows the probability that rosuvastatin is cost-effective given different willingness-to-pay thresholds. Such thresholds vary according to payer (ie, third-party or government).
View Original  |  Slide (.ppt)
Figure 4.
Cost-effectiveness acceptability curve, comparing both Framingham Risk Score (FRS) groups (FRS >10%, FRS ≥10%). The graph shows the probability that rosuvastatin is cost-effective given different willingness-to-pay thresholds. Such thresholds vary according to payer (ie, third-party or government).
Cost-effectiveness acceptability curve, comparing both Framingham Risk Score (FRS) groups (FRS >10%, FRS ≥10%). The graph shows the probability that rosuvastatin is cost-effective given different willingness-to-pay thresholds. Such thresholds vary according to payer (ie, third-party or government).
View Original  |  Slide (.ppt)
Table 1
Model Parameters and Base Case Values Used in 1-Way Sensitivity Analysis
Model Parameters
Weight*
Base Case Value (Range)†
Value Reference (Range Reference)
▪ Probabilities of Primary Endpoint Event‡
□ FRS >10%
– Probability of primary endpoint event
    Statin group ... 0.0031 (0.0024-0.0037) Ridker§ (Author assumption)
    Standard management group...0.0054 (0.0043-0.0065)Ridker§ (Author assumption)
□ FRS ≤10%
– Probability of primary endpoint event
    Statin group ... 0.0011 (0.0009-0.0013) Ridker§ (Author assumption)
    Standard management group...0.0020 (0.0016-0.0024)Ridker§ (Author assumption)
– Probability of adverse event∥ ... 0.014 (0.008-0.02) AstraZeneca23 (Author assumption)
▪ Costs, mean
□ Outpatient costs, statin group
– Office visit (CPT 99213)...$61.31CMS24
– Lipid panel ... $19.01 CMS25
– Alanine aminotransferase...$10.44CMS25
– Asparate aminotransferase ... $10.20 CMS25
– hs-CRP...$25.55CMS25
– Venipuncture ... $3.00 CMS25
–Total initial outpatient cost...$129.51 ($103.61-$155.41) (Author assumption)
– Total incremental outpatient cost# ... $32.38 ($25.90-$38.85) # (Author assumption)
□ Outpatient costs, standard management group
– Office visit (CPT 99213) ... $61.31 CMS24
– Lipid panel...$19.01CMS25
– hs-CRP ... $25.55 CMS25
– Venipuncture...$3.00CMS25
– Total initial outpatient cost ... $108.87 ($87.10-$130.64) (Author assumption)
– Total incremental outpatient cost#...$27.22 ($21.77-$32.66)# (Author assumption)
□ Other costs
– Rosuvastatin, 20 mg/tablet...$4.11 ($0.13-$5.00)Red Book23 (Author assumption)
– Primary endpoint event ... $13,123 ($5397-$26,770) ** (Author estimate)
– Nonfatal myocardial infarction0.14$6891.39 ($4459.66-$9951.16)CMS26,27 (Author estimate)
– Nonfatal stroke 0.19 $7973.14 ($4327.86-$14,727.25) CMS26,27 (Author estimate)
– Arterial revascularization0.45$20,581.76 ($9809.11-$39,200.54)CMS26,27 (Author estimate)
– Hospitalization for unstable angina 0.10 $2549.85 ($2039.88-$3059.82) CMS26,27 (Author estimate)
– Death from cardiovascular causes0.12$7599.23 ($0.00-$39,200.54)†† (Author estimate)
▪ Health Utilities‡‡
□ Primary endpoint event...0.606 (0-1) **
□ Nonfatal myocardial infarction 0.14 0.704 Sullivan and Ghushchyan28
□ Nonfatal stroke0.190.650Sullivan and Ghushchyan28
□ Arterial revascularization 0.45 0.6995 Author estimate
□ Hospitalization for unstable angina0.100.695Sullivan and Ghushchyan28
□ Death from cardiovascular causes 0.12 0 Author assumption
□ Healthy/no event...1.00 (0.90-1.00)Author assumption
 Abbreviations: CMS, Centers for Medicare & Medicaid Services; CPT, Current Procedural Terminology; FRS, Framingham Risk Score; hs-CRP, high-sensitivity C-reactive protein; NA, not applicable.
 *Weight estimates based on event numbers in JUPITER study.14 Weights apply only to parameters used to calculate weighted averages.
 †Ranges specified only where appropriate for data analysis.
 ‡Probabilities—on a 0-1 scale—calculated for 3-month period. The primary endpoint event was defined as a nonfatal myocardial infarction, nonfatal stroke, arterial revascularization, hospitalization for unstable angina, or death from cardiovascular causes.
 §According to the JUPITER Study Group (Paul M. Ridker, MD, MPH, personal communication, September 2009).
 ∥Adverse event modelled as a one-time event during the initial 3 months of statin therapy. Probability is on a 0-1 scale.
 Costs calculated for initial 3-month period as sum of all outpatient costs.
 #Costs calculated for incremental 3-month period as sum of all outpatient costs divided by 2. One office visit per year was assumed (ie, one-quarter of costs of initial office visit for each incremental 3-month period).
 **Weighted average of primary endpoint events. Range explained in “Utility Data” section of text.
 ††Estimated cost calculated as average of the other events. Low range of $0 was assumed because a patient could expire without being admitted to a hospital.
 ‡‡Utility is defined on a 0-1 scale, in which 0 represents death and 1 represents perfect health.
View Large
Table 1
Model Parameters and Base Case Values Used in 1-Way Sensitivity Analysis
Model Parameters
Weight*
Base Case Value (Range)†
Value Reference (Range Reference)
▪ Probabilities of Primary Endpoint Event‡
□ FRS >10%
– Probability of primary endpoint event
    Statin group ... 0.0031 (0.0024-0.0037) Ridker§ (Author assumption)
    Standard management group...0.0054 (0.0043-0.0065)Ridker§ (Author assumption)
□ FRS ≤10%
– Probability of primary endpoint event
    Statin group ... 0.0011 (0.0009-0.0013) Ridker§ (Author assumption)
    Standard management group...0.0020 (0.0016-0.0024)Ridker§ (Author assumption)
– Probability of adverse event∥ ... 0.014 (0.008-0.02) AstraZeneca23 (Author assumption)
▪ Costs, mean
□ Outpatient costs, statin group
– Office visit (CPT 99213)...$61.31CMS24
– Lipid panel ... $19.01 CMS25
– Alanine aminotransferase...$10.44CMS25
– Asparate aminotransferase ... $10.20 CMS25
– hs-CRP...$25.55CMS25
– Venipuncture ... $3.00 CMS25
–Total initial outpatient cost...$129.51 ($103.61-$155.41) (Author assumption)
– Total incremental outpatient cost# ... $32.38 ($25.90-$38.85) # (Author assumption)
□ Outpatient costs, standard management group
– Office visit (CPT 99213) ... $61.31 CMS24
– Lipid panel...$19.01CMS25
– hs-CRP ... $25.55 CMS25
– Venipuncture...$3.00CMS25
– Total initial outpatient cost ... $108.87 ($87.10-$130.64) (Author assumption)
– Total incremental outpatient cost#...$27.22 ($21.77-$32.66)# (Author assumption)
□ Other costs
– Rosuvastatin, 20 mg/tablet...$4.11 ($0.13-$5.00)Red Book23 (Author assumption)
– Primary endpoint event ... $13,123 ($5397-$26,770) ** (Author estimate)
– Nonfatal myocardial infarction0.14$6891.39 ($4459.66-$9951.16)CMS26,27 (Author estimate)
– Nonfatal stroke 0.19 $7973.14 ($4327.86-$14,727.25) CMS26,27 (Author estimate)
– Arterial revascularization0.45$20,581.76 ($9809.11-$39,200.54)CMS26,27 (Author estimate)
– Hospitalization for unstable angina 0.10 $2549.85 ($2039.88-$3059.82) CMS26,27 (Author estimate)
– Death from cardiovascular causes0.12$7599.23 ($0.00-$39,200.54)†† (Author estimate)
▪ Health Utilities‡‡
□ Primary endpoint event...0.606 (0-1) **
□ Nonfatal myocardial infarction 0.14 0.704 Sullivan and Ghushchyan28
□ Nonfatal stroke0.190.650Sullivan and Ghushchyan28
□ Arterial revascularization 0.45 0.6995 Author estimate
□ Hospitalization for unstable angina0.100.695Sullivan and Ghushchyan28
□ Death from cardiovascular causes 0.12 0 Author assumption
□ Healthy/no event...1.00 (0.90-1.00)Author assumption
 Abbreviations: CMS, Centers for Medicare & Medicaid Services; CPT, Current Procedural Terminology; FRS, Framingham Risk Score; hs-CRP, high-sensitivity C-reactive protein; NA, not applicable.
 *Weight estimates based on event numbers in JUPITER study.14 Weights apply only to parameters used to calculate weighted averages.
 †Ranges specified only where appropriate for data analysis.
 ‡Probabilities—on a 0-1 scale—calculated for 3-month period. The primary endpoint event was defined as a nonfatal myocardial infarction, nonfatal stroke, arterial revascularization, hospitalization for unstable angina, or death from cardiovascular causes.
 §According to the JUPITER Study Group (Paul M. Ridker, MD, MPH, personal communication, September 2009).
 ∥Adverse event modelled as a one-time event during the initial 3 months of statin therapy. Probability is on a 0-1 scale.
 Costs calculated for initial 3-month period as sum of all outpatient costs.
 #Costs calculated for incremental 3-month period as sum of all outpatient costs divided by 2. One office visit per year was assumed (ie, one-quarter of costs of initial office visit for each incremental 3-month period).
 **Weighted average of primary endpoint events. Range explained in “Utility Data” section of text.
 ††Estimated cost calculated as average of the other events. Low range of $0 was assumed because a patient could expire without being admitted to a hospital.
 ‡‡Utility is defined on a 0-1 scale, in which 0 represents death and 1 represents perfect health.
×
Table 2
Cost-Effectiveness of Statin Therapy vs Standard Management *
Parameter
Value
▪ FRS >10%
□ Total cost
– Statin group$14,900
– Standard management group $3200
□ Total QALY
– Statin group 8.37
– Standard management group8.04
□ Incremental cost $11,700
□ Incremental QALY0.33
□ ICER $35,455/QALY
▪ FRS ≤10%
□ Total cost
– Statin group$14,600
– Standard management group $1900
□ Total QALY
– Statin group 8.68
– Standard management group8.54
□ Incremental cost $12,700
□ Incremental QALY0.14
□ ICER $90,714/QALY
 Abbreviations: FRS, Framingham Risk Score; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year.
 *Data based on JUPITER study,14 Red Book,23 Centers for Medicare & Medicaid Services,24-27 and other sources.
View Large
Table 2
Cost-Effectiveness of Statin Therapy vs Standard Management *
Parameter
Value
▪ FRS >10%
□ Total cost
– Statin group$14,900
– Standard management group $3200
□ Total QALY
– Statin group 8.37
– Standard management group8.04
□ Incremental cost $11,700
□ Incremental QALY0.33
□ ICER $35,455/QALY
▪ FRS ≤10%
□ Total cost
– Statin group$14,600
– Standard management group $1900
□ Total QALY
– Statin group 8.68
– Standard management group8.54
□ Incremental cost $12,700
□ Incremental QALY0.14
□ ICER $90,714/QALY
 Abbreviations: FRS, Framingham Risk Score; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life year.
 *Data based on JUPITER study,14 Red Book,23 Centers for Medicare & Medicaid Services,24-27 and other sources.
×
Table 3
Results of 1-Way Sensitivity Analysis
US $/QALY
Parameter*
 Low Value, High Value
 Biased Toward Statin
 Biased Against Statin
▪ FRS >10%
□ Probability of primary endpoint event†
– Statin group0.0024, 0.003723,00069,000
– Standard management group 0.0043, 0.0065 27,000 48,000
□ Discount rate0%, 7%33,00038,000
□ Cost of statin per tablet $0.13, $5.00 -1000 43,000
□ Cost of primary endpoint event$5397, $26,77033,00037,000
□ Incremental outpatient costs
– Statin group$25.90, $38.8535,00036,000
– Standard management group $21.77, $32.66 35,000 36,000
□ Utility of primary endpoint event‡0, 1§34,00043,000
□ Utility of no event‡ 0.9, 1 35,000 39,000
▪ FRS ≤10%
□ Probability of primary endpoint event†
– Statin group0.0024, 0.003774,000118,000
– Standard management group 0.0043, 0.0065 63,000 165,000
□ Discount rate0%, 7%86,00099,000
□ Cost of statin per tablet $0.13, $5.00 2000 111,000
□ Cost of primary endpoint event$5397, $26,77088,00093,000
□ Incremental output costs
– Statin group$25.90, $38.8590,00093,000
– Standard management group $21.77, $32.66 90,000 93,000
□ Utility of primary endpoint event‡0, 1§88,000111,000
□ Utility of no event‡ 0.9, 1 91,000 102,000
 Abbreviations: FRS, Framingham Risk Score; QALY, quality-adjusted life year.
 *Only those parameters found to be sensitive are included in table.
 †Probabilities are on a 0-1 scale.
 ‡Utility is defined on a 0-1 scale, in which 0 represents death and 1 represents perfect health.
 §Range explained in “Utility Data” section of text.
View Large
Table 3
Results of 1-Way Sensitivity Analysis
US $/QALY
Parameter*
 Low Value, High Value
 Biased Toward Statin
 Biased Against Statin
▪ FRS >10%
□ Probability of primary endpoint event†
– Statin group0.0024, 0.003723,00069,000
– Standard management group 0.0043, 0.0065 27,000 48,000
□ Discount rate0%, 7%33,00038,000
□ Cost of statin per tablet $0.13, $5.00 -1000 43,000
□ Cost of primary endpoint event$5397, $26,77033,00037,000
□ Incremental outpatient costs
– Statin group$25.90, $38.8535,00036,000
– Standard management group $21.77, $32.66 35,000 36,000
□ Utility of primary endpoint event‡0, 1§34,00043,000
□ Utility of no event‡ 0.9, 1 35,000 39,000
▪ FRS ≤10%
□ Probability of primary endpoint event†
– Statin group0.0024, 0.003774,000118,000
– Standard management group 0.0043, 0.0065 63,000 165,000
□ Discount rate0%, 7%86,00099,000
□ Cost of statin per tablet $0.13, $5.00 2000 111,000
□ Cost of primary endpoint event$5397, $26,77088,00093,000
□ Incremental output costs
– Statin group$25.90, $38.8590,00093,000
– Standard management group $21.77, $32.66 90,000 93,000
□ Utility of primary endpoint event‡0, 1§88,000111,000
□ Utility of no event‡ 0.9, 1 91,000 102,000
 Abbreviations: FRS, Framingham Risk Score; QALY, quality-adjusted life year.
 *Only those parameters found to be sensitive are included in table.
 †Probabilities are on a 0-1 scale.
 ‡Utility is defined on a 0-1 scale, in which 0 represents death and 1 represents perfect health.
 §Range explained in “Utility Data” section of text.
×