|
Cost-effectiveness of implantable cardioverter defibrillators relative to amiodarone for prevention of sudden cardiac death |
Owens D K, Sanders G D, Harris R A, McDonald K M, Heidenreich P A, Dembitzer A D, Hlatky M A |
|
|
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 Use of third generation implantable cardioverter defibrillators (ICDs) in the prevention of sudden cardiac death.
Type of intervention Secondary prevention; treatment.
Economic study type Cost-effectiveness analysis.
Study population The study populations were hypothetical cohorts of patients at either high or intermediate risk of sudden cardiac death. The base-case analysis assumed patients were 57 years old and had survived previous cardiac arrest and thus were at high risk of sudden cardiac death. Patients were at risk for ventricular tachycardia or fibrillation, nonarrhythmic cardiac death, noncardiac death, illness or death related to amiodarone, and perioperative morbidity and mortality related to ICDs.
Setting The economic study was conducted in the USA.
Dates to which data relate Resource and effectiveness data were derived from previous studies conducted between 1983 and 1996. Cost data were derived from studies conducted between 1992 and 1995, and converted to 1995 prices.
Source of effectiveness data The effectiveness data were derived from a review and synthesis of previous studies.
Modelling The authors developed a decision tree and a Markov model, using SMLTree software (version 2.9) to combine probability estimates for consequences and costs for the three regimens over time.
Outcomes assessed in the review The outcomes assessed in the review were the base-line probabilities, associated with each treatment, required for the model inputs. For ICD treatment these included the probability of perioperative death, the probability of nonarrhythmic cardiac mortality, the probability of arrhythmic mortality, and the probability of neurologic impairment, given survival, of either ventricular fibrillation (VF) or ventricular tachycardia (VT). For the amiodarone treatment the outcomes also included the probability of nonarrhythmic cardiac mortality, arrhythmic mortality, the probability of rehospitalization after an arrhythmic event and the probability of dying given toxicity as a side effect of treatment.
Study designs and other criteria for inclusion in the review The authors used the results of past and ongoing randomised and randomised controlled trials. Other criteria for inclusion were not explicitly stated.
Sources searched to identify primary studies Criteria used to ensure the validity of primary studies A modified version of a rating system, developed by the US Preventive Services Task Force, was used to evaluate the quality of evidence for all important model inputs. These were identified as RCT evidence, time-series studies, expert opinion, clinical experience, descriptive studies or case reports.
Methods used to judge relevance and validity, and for extracting data Number of primary studies included The authors reviewed approximately 35 studies to determine probabilities and data required for the modelling aspects of the study.
Methods of combining primary studies Investigation of differences between primary studies Results of the review For ICD treatment the probability of perioperative death was estimated to be 1.8% (range: 0.6 - 3.6%), the probability of nonarrhythmic cardiac mortality was 11.3% (range: 9.04 - 13.56%), the probability of arrhythmic mortality was 0.2% (range: 0.16 - 0.24%), and the probability of neurologic impairment given survival of VF was 10% (range: 5.0 - 40%); for VT the figure was 0.0% (range: 0.0 -20.0%). From the review the authors estimated that ICD implantation would reduce total mortality rate by either 20% or 40% at one year, relative to amiodarone. For the amiodarone treatment the probability of nonarrhythmic cardiac mortality was 5.0% (range: 2.7 - 15.7%), arrhythmic mortality was 8.6% (range: 3.3 - 11.4%), the probability of rehospitalization after an arrhythmic event was 7.0% (range: 5.6 - 8.4%) and the probability of dying given toxicity as a side effect of treatment was estimated to be 0.41% (range: 0.0 - 9.0%).
Measure of benefits used in the economic analysis Benefits in the economic analysis were expressed in life-years gained and quality-adjusted life years (QALYs). The authors employed the Time Trade-off (TTO) method to evaluate health states, which were based on the authors' assumptions. Health outcomes were discounted at 3% annually.
Direct costs Direct costs included the costs of medical care associated with each treatment. For the ICD implantation this included the cost of the device itself, the annualized cost of generator replacement and the cost of complications such as lead failure and infection, plus generator replacement. For amiodarone the costs included initial hospitalization, the annual treatment costs, treatment for acute mild toxicity and hospitalization costs for severe toxicity. Costs for ICD treatment were derived from published data from hospitals in northern California and analysis of Medicare claims. Costs of amiodarone treatment were based on published estimates and adjusted to reflect current practices. Discounting was applied at a base-case rate of 3%. The price year was 1995.
Indirect Costs Indirect costs were not provided.
Sensitivity analysis Sensitivity analyses were carried out on all model variables, for the high-risk population. The key parameters were: quality of life measures associated with ICD and amiodarone therapy, the relative risk reduction (RRR) in mortality caused by ICD implantation, the frequency of generator replacement, and cost of initial implantation. This was achieved by a series of one-way analyses referenced to the marginal cost-effectiveness ratio of ICD compared with amiodarone.
Estimated benefits used in the economic analysis For high-risk patients, assuming the ICD reduced the RRR by 20%, the number of QALYs gained was 4.18 for ICD and 3.68 for amiodarone. For intermediate risk patients the benefits were 6.32 QALYs and 5.81 QALYs respectively. If ICD treatment is assumed to reduce the total mortality rate to 40%, high-risk patients in the ICD group would live 1.17 QALYs more than the amiodarone group and intermediate risk patients would live 1.28 QALYs more.
Cost results For a high-risk patient, the total costs were $88,400 (range: $78,800 - 111,600) for the ICD treatment and $51,000 (range: $42,400 -60,700) for the amiodarone regimen. For an intermediate risk patient, the corresponding figures were $110,500 (range: $98,700 -138,900) and $71,400 (range: $61,400 - 84,300).
Synthesis of costs and benefits The costs and benefits were combined by means of a cost/QALY measure. Assuming a 20% mortality reduction rate due to ICD, the marginal cost-effectiveness ratio for ICD relative to amiodarone was $74,400 per QALY saved for a high-risk patient and $76,800 per QALY saved for an intermediate-risk patient. If the RRR was assumed to be 40%, the corresponding figures were $37,300 for a high-risk patient and $36,300 for an intermediate-risk patient. The sensitivity analyses were reported for high-risk patients only. If the RRR was reduced to 10%, the marginal cost-effectiveness ratio of ICD compared with amiodarone is $518,700 per QALY gained. If the RRR was more than 50%, this figure falls to $30,200 per QALY. With the RRR set at 20%, the quality of life measures associated with ICD remaining at the base-line value of 0.75, but the measure for amiodarone therapy changed to 0.65, the marginal cost-effectiveness ratio is $43,300. If the frequency of generator replacement is increased from 4 to 5 years, the ratio falls to $63,800. The cost-effectiveness results of changes in the cost of initial implantation were not reported.
Authors' conclusions The use of an ICD in the prevention of sudden cardiac death will cost more than $50,000 per QALY unless it reduces all causes of mortality by a figure of 30% or more when compared with the alternative treatment of amiodarone. Evidence available at the time of the trial does not definitively support or exclude such a benefit and its attendant cost. However, ongoing trials will have sufficient statistical power to provide more precise estimates of the cost-effectiveness of ICDs.
CRD COMMENTARY - Selection of comparators A justification was given for the comparator used, namely that amiodarone is one of the most effective and least toxic of the antiarrhythmic drugs available. You, as a user of this database, should judge if this is a frequently used drug in your own setting.
Validity of measure of benefit:
The study was very comprehensive and detailed in terms of its methods and analyses. Due to uncertainty surrounding the estimates, however, and as acknowledged by the authors, its findings will require validating later once the results of larger randomised trials are available.
Validity of estimate of cost:
Hospital costs were analysed from the perspective of the hospital. All the relevant cost categories were included and costs were discounted appropriately. However, costs and quantities were not reported separately. This makes it more difficult to transfer the cost findings to other settings.
Other issues The study examined a third strategy, namely that of amiodarone-to-ICD in the event of patients being unable to tolerate the side-effects and toxicity of drug therapy. This was not included in the abstract as the principal aim of the study was to compare ICDs with the comparator of amiodarone alone. The authors found that both ICD and the amiodarone therapies were dominant, relative to this third regimen. A good deal of variability arose due to technical aspects and life span of the ICDs themselves. The authors assumed a 4-year replacement cycle, which was one year less than at least one of the manufacturer's recommendations, and the study showed that variations around this periodicity would have had a significant impact on the cost-effectiveness ratios. The findings were particularly sensitive to the reduction in all causes of mortality associated with the use of ICDs.
Implications of the study Further study is required into the relative risk reduction, life span and reliability of ICDs in order to increase the quality and reliability of input data to trials involving these devices.
Source of funding Supported in part by the Cardiac Arrhythmia and Risk of Death Patient Outcome Research Team grant (HS 08362) to Stanford University from the Agency for Health Care Policy and Research, Rockville, Maryland. Dr Owens is supported by a Career Development Award from the Veterans Affairs Health Services Research and Development Service.
Bibliographic details Owens D K, Sanders G D, Harris R A, McDonald K M, Heidenreich P A, Dembitzer A D, Hlatky M A. Cost-effectiveness of implantable cardioverter defibrillators relative to amiodarone for prevention of sudden cardiac death. Annals of Internal Medicine 1997; 126(1): 1-12 Other publications of related interest A critical appraisal of this study can also be found in ACP Journal Club 1997;126(3):61.
Indexing Status Subject indexing assigned by NLM MeSH Amiodarone /therapeutic use; Anti-Arrhythmia Agents /therapeutic use; Cost-Benefit Analysis; Death, Sudden, Cardiac /prevention & Decision Trees; Defibrillators, Implantable /economics; Humans; Markov Chains; Middle Aged; Quality-Adjusted Life Years; Risk Factors; control AccessionNumber 21997008032 Date bibliographic record published 31/01/1999 Date abstract record published 31/01/1999 |
|
|
|