European Heart Journal Advance Access originally published online on March 19, 2007
European Heart Journal 2007 28(12):1448-1453; doi:10.1093/eurheartj/ehm020
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cost-effectiveness of high-dose atorvastatin compared with regular dose simvastatin
1 European Health Economics, Vasagatan 38, 111 20 Stockholm, Sweden
2 Pfizer Inc., New York, USA
3 Linköping University Hospital, Linköping, Sweden
4 Ullevål University Hospital, Oslo, Norway
5 Stockholm School of Economics, Stockholm, Sweden
Received 9 August 2006; revised 8 February 2007; accepted 15 February 2007; online publish-ahead-of-print 19 March 2007.
* Corresponding author. Tel: +46 854 528540; fax: +46 854 528549. E-mail address: peter.lindgren{at}healtheconomics.se
See page 1405 for the editorial comment on this article (doi:10.1093/eurheartj/ehm168)
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Aims: The aim of the study was to evaluate the long-term cost-effectiveness of high-dose atorvastatin when compared with generic simvastatin for secondary prevention in Denmark, Finland, Norway, and Sweden based on the recently completed IDEAL trial.
Methods and results: The IDEAL trial showed that high-dose treatment with atorvastatin was associated with fewer non-fatal myocardial infarctions (MI) or coronary heart disease death (RR 0.89; 95% CI 0.78–1.01) and major cardiovascular events by (RR 0.87; 95% CI 0.77–0.98) or any coronary event (RR 0.84; 95% CI 0.76–0.91) than simvastatin with no significant difference in the number of serious adverse events. Costs during the trial period was estimated based on the trial data and a Markov model was constructed where the risk of MIs and revascularization procedures and the long-term costs, quality of life, and mortality associated with these events was simulated. Costs were based on resource consumptions recorded in the trial multiplied with recent unit costs from each country. Both direct health care costs and indirect costs (costs from lost production due to work absence) were included. Intervention lasted for the duration of the trial (4.8 years) while health-effects and costs are predicted for the lifespan of the patient. The main outcome was quality adjusted life-years (QALY) gained. High-dose treatment was predicted to lead to a mean increase in survival of 0.049 years per patient and 0.033 QALYs gained. The cost to gain one QALY was predicted to 47 197
(Denmark), 62 639 (Finland), 35 210 (Norway), and 43 667 (Sweden), with cost-effectiveness ratio decreasing with higher risk.
Conclusion: In the prevention of cardiovascular events among patients with a previous MI, high-dose atorvastatin appears to be a cost-effective strategy when compared with generic simvastatin 20–40 mg in Denmark, Norway, and Sweden. In Finland, it is cost-effective in high-risk patients. The key driver of the cost-effectiveness is the price-difference between 80 mg atorvastatin and generic simvastatin.
Key Words: Atorvastatin • Secondary prevention • Heart disease • Costs • Decision modelling
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
There is compelling economic evidence for the use of statins in secondary prevention of cardiovascular disease.1–4 The comparisons for all these studies was placebo, thus establishing the fact that treatment with statins is cost-effective compared with no intervention in these patients. With the reduction in price of simvastatin, earlier studies would show cost-savings in high-risk populations. Recently, clinical trials such as treating to new targets (TNT), Pravastatin, or Atorvastatin Evaluation and Infection Therapy–Thrombolysis in Myocardial Infarction 22 (PROVE IT–TIMI 22), and Incremental Decrease in Endpoints Through Aggressive Lipid Lowering (IDEAL) have established the clinical benefit of high-dose atorvastatin treatment when compared with previously established standard treatments.5–7 With the patent expiration of simvastatin in many markets, and the subsequent sharp decrease in prices, questions are being raised whether the increased clinical benefits are a good value despite the higher cost.
The recently completed IDEAL trial was designed to build upon the 4S trial and to demonstrate if incremental decreases in lipid lowering were associated with reductions in cardiovascular events. The IDEAL trial showed that high-dose treatment with atorvastatin was associated with fewer non-fatal myocardial infarctions (MI) or coronary heart disease death (RR 0.89; 95% CI 0.78–1.01) and major cardiovascular events by (RR 0.87; 95% CI 0.77–0.98) or any coronary event (RR 0.84; 95% CI 0.76–0.91) than simvastatin with no significant difference in the number of serious adverse events.7 The study, undertaken in Denmark, Finland, Iceland, the Netherlands, Norway, and Sweden, randomized 8888 patients with a history of acute MI to treatment with either 80 mg atorvastatin or 20–40 mg simvastatin. The timeframe of the study was March 1999–March 2005, with a median follow-up of patients of 4.8 years.
With the low price of generic simvastatin, it is natural to ask if the use of high-dose atorvastatin represents a rational use of resources in the patients studied, in other words is it cost-effective? The aim of this study was to estimate the cost-effectiveness of high-dose treatment with atorvastatin when compared with regular dose simvastatin in Denmark, Finland, Norway, and Sweden based on the IDEAL trial.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
The IDEAL trial was a prospective, randomized, open-labelled, blinded endpoint evaluation trial randomizing patients to either 20–40 mg simvastatin or 80 mg atorvastatin. The trial enrolled 8888 patients aged 80 years or less with a history of AMI and followed them for an average of 4.8 years.7 We performed the economic assessment in two steps: first by estimating the total cost and number of events avoided during the trial period, and then by extrapolating costs and effects for the lifetime of the patient in a modelling approach.
Within-trial estimation of costs and effects
As a first step to assess the cost-effectiveness, we calculated the costs and number of events (the total number of events of any type) during the trial period, aggregating data from all patients on an intention-to-treat basis. Endpoints were assigned a diagnosis-related group (DRG) based on the NordDRG classification logic, and were assigned a cost based on this DRG.8 In cases where DRG classification was ambiguous, a weighted average based on the frequency of the DRG in the general population was used. The cost of study drug was based on the number of days and dose taken. There was no significant difference in the use of concomitant medication between the treatment groups, and this cost was therefore excluded. In addition to these direct costs, loss of production (in the form of work absence) was also recorded on the case report forms. The work absence was evaluated according to the value of production (average salary + employer contribution). The unit costs used can be found in Table 1.
|
Cost-effectiveness model
In order to estimate the long-term cost-effectiveness, a Markov model was constructed.9 A Markov model divides the disease into discrete health states where patients spend a fixed amount of time (a cycle). After each cycle, the patient has certain probabilities of moving to other states or remaining in the present one, these are referred to as transition probabilities. The model for IDEAL consists of four health states: at risk, MI, revascularization procedure, and death. Patients start in the at-risk state, where they have an yearly risk of suffering from AMI, a revascularization procedure, or die from other causes. Patients suffering from AMI or a revascularization either die or remain in their respective states. This means that only the first event is modelled explicitly.
The risk of suffering an MI or revascularization procedure was estimated using Weibull regression models on the simvastatin-treated patients in the trial. The effect of atorvastatin was applied as a relative risk reduction during the trial period. After the end of the trial period, patients in both arms are assumed to have the same risk as the simvastatin-treated patients. To estimate the risk of dying each year, two Weibull models were estimated: one for patients not suffering an event (where the events where counted as censoring) and one for patients suffering an AMI. There were too few deaths after revascularization to model long-term survival after that event. In the model, we used the same mortality as in patients without an event in this case. We assumed no difference in mortality between the treatment arms, any gains in long-term survival in the model is thus only driven by the difference in MI event rates. To make sure that survival was not overestimated in older ages (where no observations was available in the trial data), the yearly mortality was programmed not to exceed the standard mortality rates in the general Swedish population.10 To be able to conduct pre-specified sub-group analyses on age, gender, diabetic status, and baseline LDL, these factors were included in all Weibull models when significant.
The costs associated with each health state were estimated based on the within-trial analysis described above. A yearly cost for the events was based on a comparison between the resource consumption prior to the event and the consumption in the years following. This cost thus included the impact of events other than those explicitly modelled, as there is a higher risk of suffering from them after the main events which will lead to a higher resource consumption in the following years. However, we conservatively assume that there is no difference in the event cost between the treatment arms, and that any difference in costs is thus driven by the lower incidence of the events in the atorvastatin arm.
Two measures of health outcomes were included in the model: life-years gained (LYG) based on the predicted survival in the two arms and quality adjusted life-years (QALY) gained. In the latter case each life-year is weighed according to the health status of the patients. The weights used (called utility weights) are normally between 1 and 0 where 1 represents a health state equal to perfect health and 0 to a health state equivalent to death. Patients are assumed to have the same utility weights as patients in the general Swedish population (adjusted for age and gender) based on a study by Burström utilizing the EQ-5D instrument and transforming the quality of life data to utilities using a tariff developed by Dolan.11–13 To account for the effect of events, data from a cross-sectional survey (also utilizing the EQ-5D instrument) on patients hospitalized at coronary-care units in Sweden was used to estimate the utility reduction 1 year after either a MI or a revascularization procedure.14 The study was also used to adjust the population figures downward to take into account that the patients in the study already have cardiovascular disease (a reduction of 0.035) resulting in a baseline utility at study beginning of 0.77. For patients with a revascularization the utility was reduced with 0.03 and for patients with an MI with 0.05. This is similar to what has been reported in Swedish patients in the ASCOT trial,15 but lower than what has been used in previous models.16 We therefore conducted a sensitivity analysis using a higher utility loss due to events.
Uncertainty was incorporated into the model through probabilistic analysis using second order Monte Carlo simulation. Thousand simulations where performed. In each of the simulations, each parameter was sampled from its underlying distribution. The distributions were estimated by performing non-parametric bootstrapping of each parameter.17 The results from the probabilistic simulation were reported in the form of cost-effectiveness acceptability curves based on the net-benefit statistic.18
In the base case, patients were treated for 5 years and were followed until a maximum of 100 years (when all patients are assumed to have died). All costs and effects were discounted at 3% per annum. Patients were assumed to have the same basic characteristics as in the trial: 62 years of age, 19% female, 12% diabetics, and a mean LDL of 121.5 mg/dL.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
Within-trial analysis
The simvastatin arm of the trial enrolled 4449 patients compared with 4439 patients in the atorvastatin arm. There was a difference of 446 events between the arms during the trial. On average, there was 0.46 (95% CI 0.43–0.49) events per patient in the atorvastatin arm compared with 0.56 (95% CI 0.53–0.59) in the simvastatin arm, a difference of 0.10 events per patient.
The total mean cost (SD) per patient in the atorvastatin arm was 31 046 (68 020), 17 138 (32 100), 26 489 (57 145), and 20 693 (42 318) in Denmark, Finland, Norway, and Sweden, respectively, the key difference in the cost being the value of production. The corresponding figures in the simvastatin arm was 29 773 (68 231), 15 042 (32 300), 25 664 (57 343), and 19 421 (42 614), giving a net cost between 2096 (95% CI 668–3467) in Finland and 824 (95% CI 2380–3152) in Norway for patients treated with atorvastatin. Between 61% (Norway) and 31% (Finland) of the atorvastatin cost was offset by cost savings from fewer event and higher work capacity. The per patient cost to avoid an event was 12 862, 21 177, 8330, and 12 856 in Denmark, Finland, Norway, and Sweden, respectively.
Cost-effectiveness model
Table 2 shows the coefficients from the Weibull regressions used to model the risk of events and mortality in the model. Table 3 shows the cost associated with events. After 3 years, there was no longer an increased direct cost among patients with MI. The events were associated with 10.2, 6.6, 6.9, and 6.9 (MI) and 9.8. 5.6, 4.5, and 4.2 (revascularization) weeks of work absence at 1, 2, 3, and 4 years after the event in question.
|
|
The mean discounted survival in the simvastatin arm was 13.80 years (10.44 QALYs) which can be compared with a predicted mean survival of 13.85 years (10.47 QALYs) in the atorvastatin arm. Table 4 shows the predicted incremental cost comparing the atorvastatin arm to the simvastatin arm. The drug cost is of course higher in the atorvastatin arm, but this is in part offset by saving fewer hospitalizations and lower work loss. The resulting cost-effectiveness ratios are similar in Denmark, Norway, and Sweden, and slightly higher in Finland.
|
Figure 1 shows the uncertainty around the cost-effectiveness in the form of a cost-effectiveness acceptability curve. The curve indicates what proportion of simulations fall below different threshold values when considering a treatment as cost-effective.
|
The subgroup analysis indicate that, as expected, a higher risk of subsequent event (male sex, high LDL-C, and in particular the presence of diabetes) is associated with lower cost-effectiveness ratios. A somewhat different pattern can be seen for age, which is due to the impact of indirect costs. The results of the subgroup analysis for Sweden can be found in Table 5. The other countries show a similar pattern. In Finland, the ICER (incremental cost-effectiveness ratio) for patients with diabetes was 47 767
/QALY.
|
To test the impact of other variables included in the model, a set of sensitivity analyses was conducted. The model was somewhat sensitive to variations in the discount rate. A discount rate of 0% gave an ICER of 29 206
/QALY gained in Sweden, compared with an ICER of 55 793 when applying a discount rate of 5%. The reason for this is that the higher cost of atorvastatin comes early (during the initial 5 years), while health benefits span a long timeframe. Doubling the utility loss from a revascularization procedure has little impact on the result (the ICER decreases to 42 952), the same is true for the utility reduction after an MI (ICER: 43 215). Increasing the treatment duration to 10 years leads to higher cost-effectiveness ratios (58 780/QALY gained), however, this needs to be interpreted carefully, as the adherence to the study drug is likely to fall over time, which is not captured in the model. |
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
To our knowledge, this is the first trial evaluating statin treatments after the introduction of generic simvastatin. In IDEAL, intensive atorvastatin reduced major coronary events by 11% and any cardiovascular event by 16% beyond that achieved with usual dose simvastatin. Our cost-effectiveness results indicate an ICER for atorvastatin vs. generic simvastatin ranged between 35 210
/QALY (23 261/LYG) gained and 62 639/QALY (41 381/LYG) gained in the four countries with Denmark, Norway, and Sweden showing similar ratios. The key explanatory factor of the difference between the settings is the price difference between atorvastatin and generic simvastatin in the different markets. The largest difference can be found in Finland which also has somewhat lower event costs, and weekly wages, and hence the cost-effectiveness ratio is the highest there.
No formally derived threshold for what is an acceptable cost-effectiveness ratio exists, although a threshold value of 50 000 is often used.19 In the guidelines for coronary care developed by the Swedish National board of health and Welfare divided studies into those with a low cost per QALY (<10 000), average cost per QALY (<50 000), a high cost per QALY (<100 000), and very high cost per QALY (>100 000).20 The results for Denmark, Norway, and Sweden all fall below 50 000.
Finland showed higher ratios overall, with a best ratio of 47 765 for diabetics. At a price of 1.7, the ICER would be similar to those of the other countries when looking at the trial population as a whole: 44 805.
A higher risk of events is associated with lower cost-effectiveness ratios. This is most evident when comparing patients with and without diabetes. This is natural, as this was the strongest factor influencing the risk of new events in the data. Female sex had less of an impact, as this was only a significant factor when predicting mortality in patients without an event in this dataset. This risk includes fatal coronary events for which men have higher risk, so it is natural that females should have a lower risk than males in this case.
Our model indicates a mean (undiscounted) survival of 18.9 years for patients treated with simvastatin which can be compared with an expected gender-adjusted survival of 20.1 years in the general Swedish population.21 It is natural that survival should be shorter in the modelled population, as they have a history of cardiovascular disease. Our model results are also consistent with those from the modelling study based on the 4S trial where the mean survival in a similar (simvastatin treated) population can be estimated to be about 17.5 years. With the improvement in cardiac care in the last 10 years, it is logical that mean survival would be somewhat longer today.
There is reason to believe that the estimated cost-effectiveness ratios are somewhat conservative. As we used trial data to estimate the cost associated with each MI and revascularization, it is possible that some potentially important cost-drivers has been omitted. One such item could be admission to nursing homes which is very expensive. Another factor could be rehabilitation and similar services, which was not recorded. The model also only takes first events into consideration, which could potentially lead to an underestimation of the total cost. However, when comparing results during the trial period to results from the model run for the same duration, the difference in the net undiscounted direct cost was only 18 (Sweden), which indicates that this is a minor issue in terms of costs. However, it is possible that there is some impact on the reduction in quality of life. Another difficulty that may arise when using data from an international trial is that resource consumption may vary between countries. We saw no evidence of this in terms of events or DRG costs, although there was a tendency for Dutch and Finnish patients to have a slightly lower number of weeks lost due to work absence, in spite of these countries having an average proportion of patients in the work-force at baseline (34 and 32%, compared with 28–37% for the other countries).
It has been shown that to accurately capture the true societal cost for an intervention, costs due to increased survival (production minus consumption) should be included in the analysis.22,23 This is the stance taken by, for example, the Swedish reimbursement authorities, however, it is not often reported in the literature. Such data are available for Sweden.24 Incorporating these costs gave a higher cost-effectiveness ratio (74 100/QALY gained), as the mean age of the population is quite high and the treatment was predicted to lead to an increase in survival.
|
Conclusions |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
The IDEAL trial showed that treatment with 80 mg atorvastatin reduced major cardiovascular events by (RR 0.87; 95% CI 0.77–0.98) or any coronary event (RR 0.84; 95% CI 0.76–0.91). In the prevention of cardiovascular events among patients with a previous MI, high-dose atorvastatin appears to be a moderately cost-effective strategy compared with generic simvastatin 20–40 mg in Denmark, Norway, and Sweden. In Finland, it is best used in high-risk patients at current prices. The key driver of the cost-effectiveness is the price-difference between 80 mg atorvastatin and generic simvastatin.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
This study was funded by Pfizer Inc., NY, USA.
Conflict of interest: P.L. has acted as a consultant to and has received research grants from Pfizer. J.G. is an employee of Pfizer. A.O. has received honoraria from Pfizer as a steering committee member. T.J.P. has received consultation fees and speakers honoraria from Pfizer. B.J. has acted as a consultant to and has received research grants from Pfizer.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionsAcknowledgementsReferences |
|---|
- Jonsson B, Johannesson M, Kjekshus J, Olsson AG, Pedersen TR, Wedel H. Cost-effectiveness of cholesterol lowering. Results from the Scandinavian Simvastatin Survival Study (4S). Eur Heart J (1996) 17:1001–1007.
[Abstract/Free Full Text] - Johannesson M, Jönsson B, Kjekshus J, Olsson AG, Pedersen TR, Wedel H. Cost effectiveness of simvastatin treatment to lower cholesterol levels in patients with coronary heart disease. N Engl J Med (1997) 336:332–336.
[Abstract/Free Full Text] - Tsevat J, Kuntz KM, Orav EJ, Weinstein MC, Sacks FM, Goldman L. Cost-effectiveness of pravastatin therapy for survivors of myocardial infarction with average cholesterol levels. Am Heart J (2001) 141:727–734.[CrossRef][Web of Science][Medline]
- Heart Protection Study Collaboration Group. Cost-effectiveness of simvastatin in people at different levels of vascular disease risk: economic analysis of a randomised trial in 20 536 individuals. Lancet (2005) 365:1779–1785.[CrossRef][Web of Science][Medline]
- Cannon CP, Braunwald E, McCabe CH, Rader DJ, Rouleau JL, Belder R, Joyal SV, Hill KA, Pfeffer MA, Skene AM. Intensive versus moderate lipid lowering with statins after acute coronary syndromes. N Engl J Med (2004) 350:1495–1504.
[Abstract/Free Full Text] - LaRosa JC, Grundy SM, Waters DD, Shear C, Barter P, Fruchart JC, Gotto AM, Greten H, Kastelein JJ, Shepherd J, Wenger NK. Intensive lipid lowering with atorvastatin in patients with stable coronary disease. N Engl J Med (2005) 352:1425–1435.
[Abstract/Free Full Text] - Pedersen TR, Faergeman O, Kastelein JJ, Olsson AG, Tikkanen MJ, Holme I, Larsen ML, Bendiksen FS, Lindahl C, Szarek M, Tsai J. High-dose atorvastatin vs. usual-dose simvastatin for secondary prevention after myocardial infarction. The IDEAL study: a randomized controlled trial. JAMA (2005) 294:2437–2445.
[Abstract/Free Full Text] - Nordic Centre for Classifications in Health Care. NordDRG Users' Manual 2006. http://www.nordclass.uu.se/verksam/norddrgmanual/NordDRG_2006_NC.
- Sonnenberg FA, Beck JR. Markov models in medical decision making: a practical guide. Med Decis Making (1993) 13:322–338.
[Abstract/Free Full Text] - National Board of Health and Welfare. Statistical Databases (2006) Socialstyrelsen, Stockholm.
- The EuroQol Group. EuroQol: a new facility for the measurement of health-related quality of life. Health Policy (1990) 16:199–208.[CrossRef][Web of Science][Medline]
- Dolan P. Modeling valuations for EuroQol health states. Med Care (1997) 35:1095–1108.[CrossRef][Web of Science][Medline]
- Burstrom K, Johannesson M, Diderichsen F. Swedish population health-related quality of life results using the EQ-5D. Qual Life Res (2001) 10:621–635.[CrossRef][Web of Science][Medline]
- Lindgren P, Stenestrand U, Hambraeus K, Wallentin L, Jönsson B. PCI reduces utility loss after MI in Sweden. Eur Heart J (2006) 27(Suppl_1):773.
[Abstract/Free Full Text] - Lindgren P, Kahan T, Poulter N, Buxton M, Svarvar P, Dahlöf B, Jönsson B. Loss of utility and indirect costs caused by cardiovascular events in hypertensive patients: a health economic substudy of ASCOT. J Clin Hypertens (2006) 8(Suppl. A):216A.
- Johannesson M. At what coronary risk level is it cost-effective to initiate cholesterol lowering drug treatment in primary prevention? Eur Heart J (2001) 22:919–925.
[Abstract/Free Full Text] - Efron B, Tibshirani R. An Introduction to the Bootstrap (1993) New York: Chapman & Hall/CRC.
- Lothgren M, Zethraeus N. Definition, interpretation and calculation of cost-effectiveness acceptability curves. Health Econ (2000) 9:623–630.[CrossRef][Web of Science][Medline]
- Eichler HG, Kong SX, Gerth WC, Mavros P, Jonsson B. Use of cost-effectiveness analysis in health-care resource allocation decision-making: how are cost-effectiveness thresholds expected to emerge? Value Health (2004) 7:518–528.[CrossRef][Web of Science][Medline]
- The National Board of Health and Welfare. Socialstyrelsens riktlinjer för hjärtsjukvård. Beslutsstöd för prioriteringar (2004).
- Statistics Sweden. Population Statistics. Statistiska centralbyrån, Stockholm, 2003; Available from: www.scb.se.
- Johannesson M, Meltzer D, O'Conor RM. Incorporating future costs in medical cost-effectiveness analysis: implications for the cost-effectiveness of the treatment of hypertension. Med Decis Making (1997) 17:382–389.
[Abstract/Free Full Text] - Meltzer D. Accounting for future costs in medical cost-effectiveness analysis. J Health Econ (1997) 16:33–64.[CrossRef][Web of Science][Medline]
- Ekman M. Two Essays in Health Economics. Consumption and Production by Age with Emphasis on Health Care Expenditures and Economic Evaluation of Beta-Blocker Therapy in Heart Failure (2001) Stockholm: Managerial Economics Section and Centre for Health Economics, Stockholm School of Economics.
CiteULike 
Connotea 
Del.icio.us What's this?
Related articles in EHJ:
- Cost-effectiveness of high-dose atorvastatin compared with usual-dose simvastatin: less than IDEAL?
- Padma Kaul
EHJ 2007 28: 1405-1406.[Extract] [Full Text]
This article has been cited by other articles:
|
|
 |
|
  R. J. Simpson Jr, J. Signorovitch, H. Birnbaum, J. Ivanova, C. Connolly, Y. Kidolezi, and A. Kuznik Cardiovascular and Economic Outcomes After Initiation of Lipid-Lowering Therapy With Atorvastatin vs Simvastatin in an Employed Population Mayo Clin. Proc., December 1, 2009; 84(12): 1065 - 1072. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Kaul Cost-effectiveness of high-dose atorvastatin compared with usual-dose simvastatin: less than IDEAL? Eur. Heart J., June 2, 2007; 28(12): 1405 - 1406. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||