Pharmacoeconomics of Beta-Blockers: Effective and Cost-Effective Therapy in Chronic Heart Failure

Chronic heart failure (CHF) is a major public health problem with high attendant morbidity and mortality. It is also a very common condition. It has been estimated that 4.8 million Americans have the disease with approximately 400 000 new cases diagnosed each year 1. In Australia, a recent survey in general practice found that 13.2% of patients over the age of 60 were diagnosed as having either systolic or diastolic heart failure 2. Disease prevalence appears to be on the increase and is expected to reach 10 million cases in the US alone by 2007 1. CHF is also a major cause of hospitalisation, estimated to account for 2.8% of all hospital discharges and 22% of all discharges for cardiovascular disease 1. Based on the above considerations, it is hardly surprising that CHF imposes a major economic burden on the health care system of most societies. Estimates of the percentage of total health care expenditure attributable to management of CHF range from 1.0% (Netherlands) 3 to 1.9% (France) 4. This analysis includes in-patient care (the major component of health care costs for CHF), drug treatment, surgery, nursing home costs and outpatient visits. However, these analyses do not include indirect costs, such as loss of income and social security costs. Furthermore, these figures are approximately a decade out of date and may in fact be extremely conservative. With the ageing of the population, improved survival following myocardial infarction as well as a wider range and increased costs of treatments available, health expenditure for CHF will continue to rise. As a result of this, policy makers will increasingly need to justify decisions on healthcare spending. This is of particular relevance in Australia and certain European countries where a large proportion of healthcare, including pharmaceuticals, are government subsidised. Current regulations in these countries require pharmaceutical companies to demonstrate cost-effectiveness of their products before being eligible for government subsidy. Therefore, not only must the efficacy and safety of pharmaceuticals be evaluated, the cost-effectiveness of their use in the community must also be determined. Of the classes of pharmacological agents used in the management of CHF, angiotensin converting enzyme (ACE) inhibitors have been the most comprehensively analysed from a pharmacoeconomic standpoint 5–8. In the Studies of Left Ventricular Dysfunction 9 (SOLVD) trial of patients with mild to moderate heart failure, there was a significant reduction in all-cause hospitalisation, as well as a prolongation of survival, with enalapril compared to placebo. In the second Veteran's Heart Failure Trial 10 (Ve-HeFT II), enalapril was also superior to the active comparator hydralazine-dinitrate on survival, but no difference was observed in hospitalisation rates between the two groups. Based on the above data, analyses of SOLVD, VeHeFT II and other studies of ACE inhibitors suggest either a net cost saving or a modest cost per life year gained (LYG) with these agents 5–8. Even taking the most conservative scenario of a USD $2500 cost per LYG, this compares favourably to other well-accepted therapeutic strategies in cardiovascular medicine, such as aspirin for secondary prevention (USD $3100, over 2 years), statins in primary (USD $30 000) and secondary prevention (USD $40 000) and coronary artery bypass grafting for multi-vessel disease (USD $50 000) 11. Other, less conservative estimates suggest that if the entire eligible CHF population in the US were to receive ACE inhibitors, there would be a 3-year cost saving of $10 billion to the US healthcare system 1. More recently, beta-adrenergic blocking agents have been studied in patients with CHF, when added to background ACE inhibitor therapy. Given the impressive cost-effectiveness of ACE inhibitors, as described above, could adding a beta-blocker still be economically rational? The consistent findings across a series of large-scale trials of beta-blockade have been those of substantial clinical benefit. For mortality, a relative risk reduction (RRR) of 65% was observed with carvedilol (a non-selective beta-blocking agent with vasodilating properties) compared to placebo in the US carvedilol trial 12. A RRR of 35% was observed in both the second Coronary Insufficiency Bisoprolol Study (CIBIS-II) 13 and the Metoprolol CR/XL Randomized Intervention Trial (MERIT-HF) 14 with the β1-selective agents, bisoprolol and extended-release metoprolol, respectively. An earlier study of bisoprolol (CIBIS-I) 15 found that bisoprolol resulted in a 20% RRR in mortality compared to placebo; this difference did not reach statistical significance due to the relatively small number of patients studied (641). The RRR in hospitalisation in the above beta-blocker trials was 32–38% for CHF-related and 13–29% for all-cause hospitalisation 12–16. These last observations suggest a great potential for beta-blockade to be considered as cost-effective therapy, even when added to ACE inhibitors, as in the above studies. Varney 17, in the previous issue of the European Journal of Heart Failure, describes a cost-effectiveness analysis of both CIBIS trials. Factored into this pharmacoeconomic model were (i) whether the benefits with bisoprolol were limited to the duration of the study (mean follow up period of 1.3 years) or extend beyond (to 5 years); (ii) whether management was community-based (practice nurse) or shared with the hospital; (iii) different rates and costs of hospitalisation; and (iv) different costs of the drug. The worst case scenario was a UK£2761 cost per LYG over 5 years, based on community care, CIBIS-II hospitalisation rates and benefits limited to the duration of the study. The best-case scenario was a UK£771 saving per LYG, with a limited benefits scenario, shared care approach to management and projected 20% rise in patient costs. These data compare favourably with the pharmacoeconomic data on ACE inhibitors 5–8. Despite the positive results of this analysis, there remain a number of caveats to the clinical utility of this data. Firstly, patients in clinical trials are not those treated in everyday clinical practice. Patients in clinical trials tend to be younger, more compliant with therapy, more likely to reach target dose of drug and do so more quickly than in everyday practice. In everyday clinical practice, achievement of a maintenance dose of beta-blockade is likely to be both more complicated and more protracted than in the clinical trial setting. This is an important consideration in beta-blocker therapy, where considerable resource utilisation occurs early, during the period of initiation and up-titration of drug therapy. Secondly, and as stated by the author, the cost-effectiveness of bisoprolol was not calculated according to lifetime benefits and costs. Increased survival implies increased cost in terms of increasing the period that they have the potential to be re-hospitalised. It has been estimated that 67–75% of the cost of heart failure management relates to hospitalisation 18. Therefore, when analysing cost-effectiveness over a lifetime, treatment may merely defer or even increase costs rather than reduce them. Thirdly, no cost utility assessments have been made, principally because of limited data on quality of life being available from the clinical trials. The years of life saved by treatments for CHF interventions are still associated with significant disability from full health. This is important when comparing the cost-effectiveness of treatments for CHF with interventions in other disease states, in particular those that prevent the transition from full health to disease, e.g. primary prevention with statins. Nevertheless, beta-blockers have generally been shown to improve patient symptom status (particularly in those patients with advanced disease) and this may translate into favourable cost per disability adjusted life year (DALY) or quality adjusted life year (QALY) outcomes. Finally, as with many analyses in this area, indirect costs such as time off work and imposition on social security resources have not been factored into the analysis. This omission seriously under-estimates the potential cost-effectiveness of this (and other) therapies for CHF. A contentious and as yet unresolved issue is whether the beta-blockers approved for CHF differ amongst themselves in their cost-effectiveness. It is not simply a question of asking which drug is the cheapest. As illustrated, differences in therapeutic efficacy, symptomatic benefit and adverse effects are some of the important determining factors in comparing the cost-effectiveness of beta-blockers 17,19. Carvedilol is generally more expensive than metoprolol CR/XL or bisoprolol in most markets. However, when compared to placebo, therapy with carvedilol appears to result in an at least similar, if not greater, reduction in all-cause hospitalisation than bisoprolol or metoprolol CR/XL. RRR for all-cause hospitalisation with beta-blocker compared to placebo was 29% in US carvedilol 20, 23% in the Australia/New Zealand carvedilol trial 21, 20% in CIBIS-II 13 and 13% in MERIT-HF 14. Furthermore, carvedilol may be better tolerated during the early stages of initiation and up-titration. A recent meta-analysis of direct head-to-head trials of carvedilol compared to immediate-release metoprolol in CHF patients would suggest that the risk of worsening CHF during initiation with carvedilol is approximately half that of metoprolol (6.0 vs. 12.9%) 22. This observation has been further supported by the absence of an excess of early discontinuation of active therapy with carvedilol in the Carvedilol Prospective Cumulative Survival 23 (COPERNICUS) Study, in contrast to that observed in MERIT-HF 15. Thus, at present it is unclear as to whether carvedilol is more or less cost-effective than bisoprolol or metoprolol in patients with CHF. A head-to-head comparison of metoprolol and carvedilol on mortality and hospitalisation, the Carvedilol or Metoprolol Evaluation Trial (COMET), is ongoing and should provide definitive answers to these questions. As with ACE inhibitors, there is now substantial data, despite its limitations, to support beta blockers as being not only of substantial clinical benefit (symptomatic improvement, survival prolongation, hospitalisation reduction) but being economically rational as well. Despite this evidence, in practice, very few patients eligible to receive beta-blockers are actually being prescribed them. Why is this so? Concerns in the medical and scientific community regarding efficacy of these agents were resolved some time ago. Concerns regarding tolerability remain; however, data from the COPERNICUS study 23 in patients with severe CHF should help assuage much of the doubt regarding tolerability of these agents. Physicians are not generally concerned about the cost-effectiveness of therapy, but they should be. On the other hand, funders of health care resources are very concerned, as would be expected. The message to all stakeholders remains the same, however: beta-blockers are effective therapy for the CHF patient and economically rational therapy for the health care system.