|
Cost-effectiveness of primary prevention of cardiovascular disease with atorvastatin in type 2 diabetes: results from the Collaborative Atorvastatin Diabetes Study (CARDS) |
Raikou M, McGuire A, Colhoun H M, Betteridge D J, Durrington P N, Hitman G A, Neil H A, Livingstone S J, Charlton-Menys V, Fuller J H |
|
|
Record Status This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn. Health technology The present study compared atorvastatin treatment, at doses of 10 mg daily, with placebo for the primary prevention of cardiovascular disease (CVD) in patients with Type 2 diabetes. Data from the Collaborative Atorvastatin Diabetes Study (CARDS) were used (Colhoun et al. 2004, see 'Other Publications of Related Interest' below for bibliographic details). Besides evaluating global differences, the study also evaluated the effects in three patient risk groups within the trial.
Economic study type Cost-effectiveness analysis and cost-utility analysis.
Study population The study population comprised adult patients with Type 2 diabetes and no documented prior history of CVD, but with at least one of the following CVD risk factors: retinopathy, micro- or macro-albuminuria, current smoking or hypertension. The inclusion criteria were LDL cholesterol of 4.14 mmol/L (160 mg/dL) or less and a fasting triacylglycerol level of 6.78 mmol/L (600 mg/dl) or less. The mean age of the CARDS population was 62 years and the majority of the patients were white (94%) and male (68%).
Setting The care level was primary care, but the setting was inpatient and outpatient care. The economic study was carried out in 32 centres in the UK and Ireland.
Dates to which data relate The dates when the data were collected for the CARDS were not reported in this paper (see Colhoun et al. 2004). The price year was 2003/04.
Link between effectiveness and cost data The effectiveness and cost data came from the same sample of patients and were prospectively evaluated.
Study sample The study sample consisted of 2,838 patients with Type 2 diabetes and no prior documented history of CVD and without elevated LDL cholesterol. The patients had been recruited from 32 centres and were randomly allocated to atorvastatin 10 mg daily (n=1,428) or placebo, (n=1,410). Power calculations were not reported.
Study design This was a multi-centre, randomised placebo-controlled trial. The randomisation method was not reported. The median length of follow-up was 3.91 years in the placebo group and 3.94 years for actively treated patients. Survival was extrapolated beyond the end of the trial.
Analysis of effectiveness The primary outcome was defined as the time to one of the following events:
acute coronary heart disease (CHD) (myocardial infarction including silent infarction, unstable angina, and acute CHD death);
coronary revascularisation procedure;
resuscitated cardiac arrest; or
stroke.
Clinical and sociodemographic variables were similar at baseline between the groups. The difference in the mean time to event between the two arms of the study was the basis of the effectiveness measure.
Effectiveness results The results showed that the within-trial gain in effectiveness was 0.0875 years (0.0685 years discounted) for the primary end point, with a mean time to the event of 5.19 years and 127 events in the placebo arm and 5.28 years and 83 events, respectively, in the atorvastatin arm.
For the broadest end point definition (any study end point), the values were a mean time to the event of 4.89 years and 229 events in the placebo arm, and 5.053 years and 167 events, respectively, in the atorvastatin arm. This led to a gain of 0.159 years (0.126 years discounted).
The study also extrapolated results on life expectancy, which was estimated to be 17.29 years (12.48 years discounted) on average in the placebo population and 18.20 years (12.97 years discounted) in the atorvastatin population. This accounted for a gain of 0.9091 years (0.49 years discounted) attributable to atorvastatin therapy over a patient's lifetime.
In the within-trial subgroup analysis, the effect differences per cardiovascular event free year were 0.109 for the low risk group, 0.164 for the medium risk group, and 0.260 for the high risk group.
Clinical conclusions The use of atorvastatin in the primary prevention of CVD in patient with Type 2 diabetes was clinically effective. Atorvastatin therapy was shown to increase the mean time to the event and to reduce the number of events in the trial horizon. It also increased life-years when considering a patient's lifetime.
Modelling Survival was extrapolated beyond the end of the trial to the lifetime of the patients. The main assumptions were that the benefit of atorvastatin continued after the end of the trial for patients who kept taking the drug (the continued treatment benefit was represented by the difference in cumulative survival between the two arms at the end of the trial), and that patients in the placebo arm would start taking the drug after any nonfatal cardiovascular event. A non-parametric approach, based on a longitudinal study, was used to derive life expectancy for the placebo arm.
Measure of benefits used in the economic analysis The measures of benefit used were the quality-adjusted life-years (QALYs), event-free years and life-years gained. The utility weights for diabetics were estimated from the UK Prospective Diabetes Study (Clarke et al. 2002, see 'Other Publications of Related Interest' below for bibliographic details). The European Quality of Life 5-dimensions (EQ-5D) tariff score for the diabetic population was used. No further details were provided. The authors also expressed their results as estimated survival-years gained. The benefits were discounted at an annual rate of 3.5%.
Direct costs Health-related direct costs included atorvastatin dosage, any additional statin therapy, number of clinic visits, the tests performed at each visit and the hospitalisations for end points. The costs were appropriately discounted at an annual rate of 3.5%. The resource use data were collected from the trial. The unit costs were taken from UK price publications and NHS cost estimates. Resource use and cost data are available from supplementary material in the online version of the paper. The price year was 2003/04.
Statistical analysis of costs A mean cost per patient over the study period was calculated for each arm of the study. No further details were reported.
Indirect Costs No productivity costs were included.
Currency UK pounds sterling (). The conversion rate to euros (EUR) was 1.00 = EUR 1.43 (January 2004 values).
Sensitivity analysis One-way sensitivity analyses were performed on the unit costs of atorvastatin (15% decrease assumed in order to match the price of a generic statin) and the event rates (decreased to match the level of effectiveness of a generic statin for the placebo population). Confidence intervals of cost-effectiveness ratios were calculated. A probabilistic sensitivity analysis and a cost-effectiveness acceptability curve were also used to address the uncertainty associated with the survival probabilities used in the extrapolation of the cumulative survival curves. Probability distributions were adequately described. In addition, a structural sensitivity analysis was undertaken to be less conservative. This included the long-term benefits of reducing nonfatal CVD events with atorvastatin.
Estimated benefits used in the economic analysis The results showed 13.28 QALYs for the placebo arm and 13.99 for the atorvastatin arm. The discounted values were 9.59 QALYs (placebo) and 9.97 QALYs (atorvastatin), respectively, showing a gain of 0.3871 QALYs attributable to atorvastatin therapy.
In the within-trial sub-group analysis, the difference in effect was 0.109 for the low-risk group, 0.164 for the medium-risk group and 0.260 years for the high-risk group.
See the 'Effectiveness Results' section for the results in terms of the life-years gained.
Cost results The total costs for the primary end point over the trial period were 1,003.75 for the placebo arm and 1,524.98 for the atorvastatin arm.
Over the total study period, the difference in average discounted treatment cost was 521 per patient.
The treatment costs were dominated by the cost of atorvastatin, which was calculated to be 3,117 per patient over their lifetime.
In the within-trial sub-group analysis, the cost-difference was 655 for the low-risk group, 564 for the medium-risk group and 540 for the high-risk group.
Synthesis of costs and benefits Based on the within-trial analysis, the discounted incremental cost-effectiveness ratios (ICERs) for the major end point categories over the trial period were 7,608.39 per primary end point-free year, 4,896.48 per cardiovascular end point-free year and 4,119.85 per year free of any end point. In addition, the estimated cost per CARDS primary end point averted was 16,916.
The extrapolated results on life expectancy showed that the ICER was 5,107 per discounted life-year gained and the incremental cost per discounted QALY gained was 6,471.
When the CARDS population was classified according to predicted risk of heart disease into low-, medium- and high-risk tertile groups, the ICERs per event-free year were 5,983 (low-risk), 3,436 (medium-risk) and 2,077 (high-risk), respectively, in the within-trial analysis.
The extrapolated results on life expectancy showed that the ICERs per life-year gained were 17,424 for the low-risk group, 6,781 for the medium-risk group and 3,124 for the high-risk group.
The corresponding ICERs per QALY gained were 20,781 (low-risk), 8,523 (medium-risk) and 4,001 (high-risk), respectively.
Authors' conclusions Most patients with diabetes, even those without raised cholesterol or established cardiovascular disease (CVD), would benefit from statin treatment. The study supported the use of atorvastatin in the primary prevention of CVD in patients with Type 2 diabetes by showing that this intervention, apart from being clinically effective, was also cost-effective under a range of assumptions. The calculated incremental cost-effectiveness ratios (ICERs) lay within the National Institute for Health and Clinical Excellence cost-effectiveness threshold of 20,000 per quality-adjusted life-year (QALY).
CRD COMMENTARY - Selection of comparators A justification was given for the comparator used. It reflected the standard practice in the authors' setting. You should decide if the comparator represents current practice in your own setting.
Validity of estimate of measure of effectiveness Although the authors stated that this was a randomised controlled study, no details of the randomisation method were described as they had been published elsewhere (Colhoun et al. 2004). Nevertheless, the authors reported an adequate balance between the groups in relation to their clinical and sociodemographic characteristics. In addition, although the sample size seems to have been adequate, no power calculations were reported. An advantage of the study was that the authors not only performed a within-trial analysis, but also extrapolated the results to a lifetime horizon, which is a more valid horizon for decision-making in chronic diseases. Validity of estimate of measure of benefit The estimation of health benefits (QALYs) was not explicitly described as the benefits had been taken from a published paper (Clarke et al. 2002). The authors provided a survival analysis which should facilitate comparisons with the benefits of other interventions. Discounting was performed, which would appear appropriate as the time horizon was greater than one year.
Validity of estimate of costs Given the health care perspective adopted, all the relevant cost categories and their appropriate costs were taken into consideration. The resource use data and unit costs were not reported, but supplementary material was available in the online version of the article. The price year and the sources of resource use and unit costs were adequately reported. The costs were discounted, which would appear appropriate given that the time horizon was greater than one year. Sensitivity analyses of the costs were conducted to assess the robustness of the estimates used.
Other issues The authors compared some of their findings with those from other studies and their results were in agreement. The conclusions appear to have reflected the scope of the analysis. However, the authors reported some limitations to their study. For example, co-medications other than concomitant statin use were not taken into account, although the proportion of patients on major categories of co-medication showed no statistical difference. In addition, the degree of compliance and serious adverse events, the rate of which was low and similar in both arms, were also not taken into consideration. Despite these facts the authors stated that their estimate was most likely to be a conservative one.
Implications of the study The authors recommended the use of atorvastatin in the primary prevention of CVD.
Source of funding Supported by a grant from Pfizer Inc.
Bibliographic details Raikou M, McGuire A, Colhoun H M, Betteridge D J, Durrington P N, Hitman G A, Neil H A, Livingstone S J, Charlton-Menys V, Fuller J H. Cost-effectiveness of primary prevention of cardiovascular disease with atorvastatin in type 2 diabetes: results from the Collaborative Atorvastatin Diabetes Study (CARDS) Diabetologia 2007; 50(4): 733-740 Other publications of related interest Because readers are likely to encounter and assess individual publications, NHS EED abstracts reflect the original publication as it is written, as a stand-alone paper. Where NHS EED abstractors are able to identify positively that a publication is significantly linked to or informed by other publications, these will be referenced in the text of the abstract and their bibliographic details recorded here for information.
Colhoun HM, Betteridge DJ, Durrington PN, et al. Primary prevention of cardiovascular disease with atorvastatin in type 2 diabetes in the Collaborative Atorvastatin Diabetes Study (CARDS): multicentre randomised placebo-controlled trial. Lancet 2004;364:685-96.
Clarke P, Gray A, Holman R. Estimating utility values for health states of type 2 diabetic patients using the EQ-5D (UKPDS 62). Med Decis Making 2002;22:340-9.
Indexing Status Subject indexing assigned by NLM MeSH Adult; Aged; Anticholesteremic Agents /pharmacology; Atorvastatin Calcium; Cardiovascular Diseases /complications /drug therapy; Cost-Benefit Analysis; Diabetes Complications /prevention & Diabetes Mellitus, Type 2 /drug therapy; Female; Heptanoic Acids /pharmacology; Humans; Male; Middle Aged; Models, Economic; Primary Prevention; Pyrroles /pharmacology; Quality-Adjusted Life Years; control AccessionNumber 22007000556 Date bibliographic record published 31/10/2007 Date abstract record published 31/10/2007 |
|
|
|