Aug 7, 1995 - ment policy in patients with asymptomatic left ventricular ... Key Words: Captopril, heart failure, prevention, ..... uncertainty margin of ± 10%.
European Heart Journal (1996) 17, 731-740
Economic aspects of treatment with captopril for patients with asymptomatic left ventricular dysfunction in The Netherlands B. C. Michel*, M. J. Al*, W. J. Remmef, J. H. Kingma*, J. A. Kragten||, R. van NieuwenhuizenU and B. A. van Hout* * Institute for Medical Technology Assessment, Erasmus University, Rotterdam, |STIC'ARES Cardiovascular Research Foundation, Rotterdam, %St Antonious Hospital, Nieuwegein, \\de Wever Hospital, Heerlen, ^Rijnstate Hospital DH, Arnhem, The Netherlands
Objective To estimate the costs and effects of preventive treatment with captopril compared with the current treatment policy in patients with asymptomatic left ventricular dysfunction after a myocardial infarction. Methods Estimates of effects are based on the results of the SAVE trial. Costs are estimated on the basis of current treatment patterns in four Dutch hospitals. All knowledge is incorporated in a mathematical model extrapolating the SAVE results to 20 years.
costs per life-year gained can be estimated at DF1 15 799. From univariate sensitivity analysis it appears that the results are highly sensitive for the costs of treatment with captopril and the occurrence and prevention of clinical heart failure. Varying all estimates randomly between upper and lower limits — in 5000 simulations — an estimate of costs per life-year gained of DF1 15 729 is made for 20 years of treatment, with 95% of all estimates between DF1 0 and DF1 50 000. On a national level, undiscounted costs are expected to increase up to approximately DF1 42 million annually during the first 40 years after introduction of the preventative strategy. (Eur Heart J 1996; 17: 731-740)
Results and conclusions Captopril treatment is expected to increase survival at certain costs. The average additional costs per patient are estimated at DF1 2 491 in 4 years and at DF1 8 723 in 20 years of treatment. Costs per additional Key Words: Captopril, heart failure, prevention, survivor after 4 years are estimated at DF1 69 126. After angiotensin-converting enzyme inhibitors, cost-effectiveness, extrapolation of the results of the SAVE trial to 20 years, simulation, model.
Introduction Heart failure is a relatively common disease in the Netherlands. With a prevalence of 13-5 per 1000 men and 8 per 1000 women between 35 and 85 years of age, it is estimated that heart failure is present in more than 80 000 persons'1'21. In the near future this number will probably increase due to the expected increase in the number of elderly where heart failure is relatively more prevalent'1'31. Next to this demographic factor, a further reduction in early mortality after myocardial infarction is likely to cause an additional increase in the prevalence of heart failure" ". Revision submitted for publication on 20 July 1995, and accepted 7 August 1995. The study was supported by an unrestricted grant from Bristol Myers Squibb. Correspondence: Bowine C. Michel MD, PhD, Institute for Medical Technology Assessment, Erasmus University, Rotterdam, PO Box 1738, 3000 DR Rotterdam, The Netherlands. 0195-668X/96/050731 + 10 $18.00/0
Due to the high prevalence of the disease in the Dutch population, successful strategies aimed at a reduction of the progression, the incidence or the prevalence of symptomatic heart failure will have important consequences. Detection and treatment of risk factors — such as hypertension, coronary artery disease, diabetes, obesity, serum cholesterol and smoking — are examples of strategies that may delay or even prohibit the progression to symptomatic heart failure. Here, changes in life style, screening of high-risk individuals and optimal treatment of risk factors will probably have their effect. Additional benefits may be expected from new medical strategies. In this respect, the introduction of angiotensin-converting enzyme-inhibitors in the 1980s have already had a major effect on the prevalence and costs related to heart failure. It has been estimated that treatment with angiotensin-converting enzymeinhibitors for patients with symptomatic heart failure improves survival and decreases costs by postponing the 1996 The European Society of Cardiology
732 B. C. Michel et al.
progression to the more advanced — and more costly — stages of the disease'2'. Recently, the SAVE trial and the SOLVD Prevention Trial have shown that treatment with angiotensin-converting enzymeinhibitors can also delay or prevent the occurrence of symptomatic heart failure'4"51. Moreover, the SAVE trial, which investigated patients with diminished left ventricular function after a myocardial infarction, also showed a significantly lower death rate by cardiovascular causes'41. In the Netherlands, preventive therapy with angiotensin-converting enzyme-inhibitors is not yet common. In view of the economic aspects of this form of therapy several questions can be raised. Preventive treatment may be life-long and it may involve large numbers of patients. Hence, costs may be considerable. On the other hand, there also is a potential to generate savings. Under the current treatment policy, almost 429 million Dutch Guilders (about 200 million ECU) are spent annually on the treatment of symptomatic heart failure; this equals 11% of the total costs of Dutch healthcare'61. In this article the results of the SAVE trial are used to estimate the balance between costs and effects of preventive treatment with captopril, compared to the current strategy, for patients with asymptomatic left ventricular dysfunction after a myocardial infarction. Also, the macro-economic consequences of an initiation of preventive treatment on a nationwide level are estimated.
Methods The SAVE trial evaluated the effects of a preventive strategy with captopril in two groups of about 1100 patients with asymptomatic left ventricular dysfunction after myocardial infarction. The follow-up of the trial was 42 ± 10 months, 82% of the patients were males, and the average age of the patients was 60'4). This trial showed that long-term administration of captopril is associated with an improvement in survival, a decrease in the incidence of symptomatic heart failure and a decrease in the number of re-infarctions'71. In the present study two baseline scenarios were defined for the comparison of costs and effects between preventive treatment with captopril and the current strategy. The first scenario models the current treatment policy. In this scenario the probabilities concerning effects are obtained from the results reported for the placebo arm of the SAVE trial. The second scenario models the situation in which captopril is prescribed. Here, the parameters concerning effects are taken from the results in the captopril arm of the SAVE trial. To compare the costs and effects of these two scenarios a mathematical model was built. There were two reasons why this modelling technique was chosen. First, data on costs were not made available in publications about the SAVE trial. Therefore, different sources of data had to be brought together, and, for such a purpose, a model is a suitable instrument. Eur Heart J. Vol. 17. May 1996
The second reason is that preventive treatment can be seen as an investment for the future and that, consequently, the balance between costs and effects needs to be assessed over a period longer than the observations made in the trial. Again, a mathematical model is a suitable instrument to make such an extrapolation. The model that was used is a so-called Markov chain model that incorporates data about the current strategy, the effects of captopril, the costs of treatment and the epidemiology related to the patient population studied. With regard to the effectiveness of the therapy the model is based mainly on the data from the SAVE trial'4'8'9'. With respect to the current treatment policy, the costs and the epidemiology, priority was given to data specific for the Dutch situation. Two separate analyses were made. In the first analysis the results after 4 years were evaluated and compared to those of the SAVE trial using the cumulative event rates that were presented by Pfeffer et a/.[4'. In the second analysis the results were extrapolated to a period of 20 years.
The model Figure 1 presents an outline of the model. The model evaluates the diagnostic and therapeutic events that occur in 100 patients with the characteristics of the patients in the SAVE trial during 4 (the first analysis) or 20 (the second analysis) consecutive years. For the first 4 years of treatment the model copies the SAVE results as closely as possibly. After that, the model extrapolates the results with the assumption that the average effectiveness and treatment policy over the first 4 years will continue in the later years of therapy. The model describes events and treatments for 100 patients who experienced a recent myocardial infarction but who have no symptoms of heart failure. The layout of the model is identical for both scenarios with respect to the events that may occur and the treatments that are given. It is assumed that additional echocardiography (and sometimes radioventriculography) will have to be performed in patients who did not receive echocardiography as a routine follow-up of their infarction. The costs for this additional assessment are included in the captopril scenario. During the first year, patients with asymptomatic heart failure experience normal follow-up of their index infarction, including analysis of cardiac ischaemia with exercise tests or thallium scans, angiography if one of these tests is positive, and cardiac intervention if the angiography indicates that this might be necessary, in both scenarios. Because the policy on the timing of interventions differs for separate hospitals and in separate periods, depending on the waiting list and on the urgency of the indication, no distinction is made as to whether the analysis for cardiac ischaemia and subsequent interventions were performed during the initial hospitalization phase for the index infarction or during the follow-up phase,
Economic aspects of treatment with captopril
733
clinical assessment and follow-up primary infarction
1 CVA/RI/UAP
'r
follow-up stage 1
/
\
NYHAI r
CVA/RI/UAP /
death
•
