PART IV - COST AND COST-EFFECTIVENESS
TABLE OF CONTENTS
BOOKSHOP
Updated on April 3, 2019
PART IV

Cost and cost-effectiveness

Isabelle Durand-Zaleski

Summary

The objective of this chapter is to highlight cost issues in interventional cardiovascular medicine. In developed countries, treatment decisions for patients with coronary artery disease are based on patients’needs and seldom on economic information. Indeed, there is evidence that a nationwide revascularisation policy prolongs life expectancy by over one year at a cost of about $40,000 per added year of life. Thus, the debate has shifted to a discussion about the best and most cost-effective revascularisation strategies. For patients with acute coronary syndrome, the availability of facilities rather than the actual costs determines the choice between thrombolysis and percutaneous coronary intervention (PCI). However, when PCI is feasible, the economic discussion moves to the type of procedure and the use of stents. The high cost of drug-eluting stents (DES) has led many health authorities to develop guidelines which include the economic aspects and the likely added value of using DES. These guidelines are supported by many economic evaluations of DES conducted either with clinical trials or from registries. In the latter case, the indication bias must be dealt with. It is worth noting that both researchers and policymakers had to deal with the lack of evidence of improved survival resulting from the use of DES. This had led to the use of such endpoints as utility (quality-adjusted life years) or willingness to pay.

General scope and definitions

The objectives of this chapter are first to highlight the debates in interventional cardiovascular medicine that involve cost issues and secondly to help interventional cardiologists in their critical appraisal of journal articles that present economic evaluations. It is not to present an exhaustive discussion of the several hundred articles indexed in Medline that deal with coronary interventions and costs, nor to discuss the many reviews on that same topic.

Like the rest of the economy, interventional cardiology has been subjected to the law of diminishing marginal returns which implies that any incremental benefit, such as reduction in mortality and morbidity, costs more to achieve than previously. The 2% absolute reduction in STEMI mortality from GISSI costs an additional $320, while the additional 1% in GUSTO costs $2,750 [1, 2]. Substituting PCI for thrombolysis was proposed by the ESC in 2005 [3]. The technological change increased the treatment costs by an average $2,000 to $3,000 per patient with a possible additional absolute reduction of 2% mortality conditional on the time to revascularisation [4]. Subsequent trials to improve the benefits of revascularisation yielded only a reduction in morbidity without significant effect on mortality. Adding stents to conventional PCI increased procedural costs by approximately $2,000 per patient and reduced the need for repeat revascularisation without measurable effect on mortality [5]. Adding abciximab costs $1,122 for modest additional clinical benefits such as reduction in subacute thrombosis [6]. The advent of DES increased costs by an average $2,000 with a delayed benefit on repeat revascularisation but no significant advantage with regard to death or myocardial infarction [7].

Because cardiovascular diseases are a leading cause of mortality and make up around 10% of total health expenditure in developed countries, any change in practice has an effect on medical expenditure. This explains the scrutiny to which interventional cardiology has been subjected, first on PCI itself and later on the use of stents, which led in turn to questions about the effectiveness of this additional spending at the population level. Figure 1 and Figure 2 show the rapid uptake of coronary interventions in some countries and the differences in practice, while Figure 3 shows the disparities in revascularization procedures for STEMI.

These findings led to the first economic debate of interest to interventional cardiologists which concerns macroeconomic aspects: what is the relationship between service utilisation and mortality?

The question resulting from the ESC recommendations on the use of PCI for STEMI was the following: how efficient is PCI in reducing mortality when used in a routine setting outside of the controlled environment of a clinical trial?

FOCUS BOX 1General scope and definitions
  • The uptake of revascularisation using PCI with or without stent has been rapid in developed countries
  • There is a positive correlation between the rate of PCI and the wealth of a country
  • The relationship between the rate of PCI and cardiac mortality is less straightforward

International comparisons performed by the Organisation for Economic Cooperation and Development (OECD) have attempted to address the complex issue of the relationship between provision of care and outcome. The high utilisation of revascularisation procedures observed in some countries does not necessarily translate to improvements in outcomes. However, the OECD noted that “better access to high-quality acute care for heart attack, including timely transportation of patients, evidence-based medical interventions and high-quality specialised health facilities such as percutaneous catheter intervention-capable centres have helped to reduce 30-day case-fatality rates” ( Figure 4A and Figure 4B)[8].

Health technology assessment and interventional cardiology

Why is economic evaluation an important component of health technology assessment and of its current sequel, comparative effectiveness research? This is the second question of interest, which focuses on the microeconomic level and decisions made at the individual patient level. Economic evaluations, in addition to effectiveness measures, are necessary to ensure that resources are allocated where they produce the greatest health benefit.

ECONOMIC EVALUATION AND POLICY DECISIONS

The general purpose of an economic evaluation in the field of health care is to relate the costs of a diagnostic or therapeutic strategy to its outcomes. Thus the two components of the evaluation are a measure of effectiveness and an estimate of costs. Economic evaluation is currently both a decision-making tool and an evolving academic discipline.

The comparison of medical strategies that both use different resources and yield different outcomes requires a different approach. Cost-effectiveness analyses compare strategies, the costs of which are expressed in monetary terms, while outcomes are expressed in a single medical unit, for example lives saved, life expectancy or life years. The final result of a cost-effectiveness analysis is often expressed as a ratio of cost to life-years gained or cost to lives saved. A positive result means that the increase in costs results in a better medical outcome. The lower the cost-effectiveness ratio, the more efficient the strategy is [9, 10, 11, 12, 13].

Comparing any two strategies yields one of the following four situations:

  1. in very simple cases, one strategy scores better on both outcome and cost, i.e., yields greater effectiveness at lower cost than the other strategy, it is superior (dominant);
  2. if it yields lower effectiveness at higher cost, it is inferior (dominated);
  3. and if it yields greater effectiveness at higher cost, we need to decide whether the incremental costs are worth paying compared to the effectiveness gained;
  4. if it yields lower effectiveness at lower cost we need to decide whether the achieved cost savings are justified compared to the effectiveness lost.

In the case of situations 1) and 2) the decision is straightforward. In situations 3) and 4) a trade-off needs to be made between costs and effectiveness, and strategies to be compared based on the cost-effectiveness ratio.

Figure 5 depicts the cost-effectiveness plane in each of the four situations. This representation is very commonly used to assess innovations in interventional cardiology.

THRESHOLDS FOR THE ADOPTION OF A STRATEGY

What is the correct amount of money that a society should pay to prolong by one year the life expectancy of one person? Some countries have set an explicit threshold, others not. The World Health Organisation suggested in 2003 that this threshold be based on the per capita national GDP. An intervention that prolongs life by one year at a cost less than 1 x per capita GDP is “cheap” and a strategy that prolongs life by one year at a cost less than 3 to 5 x per capita GDP is “expensive or even too expensive”. In the UK the threshold usually mentioned is £30,000 or $50,000 [14].

Other countries, such as the USA, do not use such thresholds on the grounds that they result in discrimination between populations. However, there is a general consensus that cost information is important to inform comparative effectiveness research for policy decisions [15].

This cost effectiveness plane is widely used to study innovations in interventional cardiology, as shown in the next part of this chapter. The discussion on the threshold for the adoption of technologies has been more scientific by quantification of the uncertainty. Uncertainty concerns both the medical result, e.g., reduction of mortality or major adverse cardiac events and the resources used (and possibly their costs). It is now common for economic evaluations of coronary interventions to use the more sophisticated representation of uncertainty, which is based on bootstrap replications and show 1,000 results obtained by taking one random effectiveness result and dividing it by one random cost result.

FOCUS BOX 2Confidence intervals for cost effectiveness
  • Uncertainty of a variable within a population is conventionally estimated using statistical central limit theorem assuming a normal distribution and large sample size
  • Economic data for populations tends to have small sample sizes and a skewed distribution
  • Bootstrapping is an automated computer simulation method for the statistical inference of uncertainty (confidence intervals), and is suited to such complex datasets
  • In bootstrap methodology rather than making assumptions about the underlying population the raw data itself is used. Random samples (replicates) of the same size are drawn from the original data at least 1,000 times to create empirically a distribution and hence the confidence intervals
FOCUS BOX 3Health technology assessment and interventional cardiology
  • There is a well-accepted method to estimate the benefit of a new technology in relation to its cost: a cost-effectiveness ratio is the amount of money to be paid in order to obtain a gain in survival or a reduction in adverse events
  • The World Health Organisation suggested in 2003 that the maximum amount to be paid be based on the per capita national GDP. An intervention that prolongs life by one year at a cost less than 1 x per capita GDP is inexpensive and a strategy that prolongs life by one year at a cost less than 3 to 5 x per capita GDP is expensive or even too expensive. In the UK the threshold usually mentioned is £30,000 or $50,000

Figure 6 shows the cost-effectiveness plane, on which the difference in survival and costs of CABG vs. PCI in high-risk patients has been shown by 1,000 dots, each one representing a possible result, given the uncertainty in the trial results. On this graph, drawn from the trial results, PCI appears to provide better outcome at lower cost than in CABG [16]. Initial costs of PCI are lower than in CABG by an average $24,000 and, despite subsequent need for revascularisation, PCI remains cheaper after 5 years (by $20,000) while there is no significant survival difference (despite a trend towards improved survival with PCI). The survival benefits in favour of PCI are shown by the fact that most simulations are on the right-hand side of the plane and the cost benefit by the negative value of the difference: most simulations fall into the lower right quadrant, which is the cost reducing/quality increasing or dominant option.

Another example with a different distribution of benefits and costs is shown in Figure 7 which depicts the cost-effectiveness analysis of invasive versus conservative management of elderly patients. In that case, only 4% of the dots are in the dominant quadrant and 96% are in the cost-increasing /quality-increasing quadrant. This means that the invasive management using PCI or CABG of patients aged 75 years or older with chronic angina pectoris of Canadian Cardiac Society class PII, compared to their medical management, was more expensive but prevented the occurrence of adverse events [17].

It is possible to translate bootstrap replications into a more readable decision tool, known as the acceptability curve. To build the acceptability curve, the software counts how many dots are below a given cost-effectiveness threshold and produces a probability. An example of an acceptability curve is shown in Figure 8. In this case, there is a 100% probability that the use of a sirolimus-eluting stent prevents one repeat revascularisation for less than $10,000 and roughly an 80% probability that the cost per repeat revascularisation is below $5,000 [7].

Economic evaluations and interventional cardiology

The field of interventional cardiology has widenend to include transcatheter aortic valve implantation (TAVI) and renal denervation (RDN) in addition to coronary interventions. The clinical benefits have been studied in other chapters. Here we focus on the cost effectiveness, at the individual level, of using PCI to treat patients with acute coronary syndromes, TAVI for aortic stenosis in patients who are high risk or contraindicated for surgery and RDN for resistant hypertension.

Economic evaluation of PCI

The broadest perspective on the cost effectiveness of coronary revascularization is probably given in the study by A Cutler on a 17-year follow-up of Medicare beneficiaries in the USA [18]. The nationwide revascularisation policy was estimated to prolong life expectancy by an average 1.1 years at a lifetime cost of about $38,000, resulting in a ratio of $33,000 per added year of life.

For patients with acute coronary syndrome the availability of facilities, rather than the actual costs, determines the choice between thrombolytic and percutaneous coronary interventions. If the treatment chosen is PCI, stent use is almost systematic (95% of procedures include coronary stenting) and between 40% and 50% of stents used are DES [19].

Because of their high cost, the benefits of DES are discussed not only in medical literature but almost as much in economic literature. Between 2007 and 2009, four systematic reviews were published on drug-eluting stents. Each country produced an economic evaluation, which usually concluded that DES were good value for money in only a limited number of indications and populations. For example, the Cochrane review suggested that “The increased cost of drug-eluting stents and lack of evidence of their cost-effectiveness meant that various health funding agencies attempted to limit or regulate their use in relation to price premium” [20, 21, 22, 23, 24, 25].

A fine demonstration of the influence of indications on the cost effectiveness of DES utilisation is provided by the 18-month follow-up of the BASKET trial. In this trial, the outcomes of patients randomised to receive either BMS or DES, irrespective of indication for PCI, were compared as well as the total medical costs. When the entire population of 826, comprising 545 patients receiving drug-eluting stents and 281 patients bare-metal stents, was considered ( Figure 9) the cost-effectiveness analysis by bootstrap shows that there is a 28.5% chance that DES are cost-increasing and quality-reducing, and there is a 71% chance that they are cost-increasing/quality-increasing. Restricting the analysis to the low-risk population ( Figure 9B) produces very different results: there is a 74% chance that DES in this population are cost-increasing/quality-reducing or dominated as described in Figure 5. On the contrary, the use of DES in high-risk patients ( Figure 9C) has a 70% chance of being cost-saving and quality-increasing or dominant, which should be a strong argument in favour of this strategy. The results presented in the cost-effectiveness plane are translated into acceptability curves for each risk group ( Figure 10). In high-risk patients, there is a 90% chance that the use of DES prevents one adverse cardiac event at a cost less than 10,000 euros, while in low-risk patients there is an 80% chance that the cost-effectiveness ratio is above 200,000 euros per event prevented [26].

Few economic studies of DES vs BMS have been added over the past years, results usually favor the use of DES over BMS in subgroups at greater risk of restenosis or with multiple vessel disease [27, 28, 29].

FOCUS BOX 4Economic evaluation and interventional cardiology
  • Revascularisation procedures are accepted to be cost effective in developed countries
  • The cost effectiveness of systematically using DES instead of BMS is more controversial

Economic evaluation of TAVI

A review of published economic evaluations was performed to present the evidence of the cost effectiveness of TAVI. Predictably, the choice of the target population, which determines the reference strategy is key to the efficiency. In patients not suitable for surgery, TAVI is more costly but more effective than medical management, with an incremental cost-effectiveness ratio (ICER) of £12,900 per quality-adjusted life-year (QALY). Of note, the efficiency is much worse in patients whose life expectancy is less than five years, the cost effectiveness ratio for patients with a 1-2 year life expectancy is above £70,000/QALY. For patients suitable for surgery, TAVI is both more costly and less effective, the lower surgical risks being associated with the worst economic results [30], ( Figure 11).

For patients with intermediate surgical risk, a model-based economic evaluation based upon analysis with data derived from the PARTNER II study found that TAVI was both more effective and cheaper compared to surgery [31].

FOCUS BOX 5Economic evaluation and TAVI
  • TAVI is likely to be cost effective in patients not eligible for surgery
  • Using TAVI instead of surgery is not cost effective

Economic evaluation of renal denervation

Economic evaluations or renal denervation requires a long term extrapolation of the results on blood pressure. The assumptions that need to be made concern:

  • the sustainability of the benefit demonstrated after 6 months extended to a 10-year interval given a possible loss of adherence to treatment since the results of the Simplicity-3HTN study have not been able to support a primary favorable effect of RDN alone;
  • the linear reversibility of the cardiovascular risk: risk models are used to estimate the increased risk of cardiovascular events given increased blood pressure, but it is likely that blood pressure reduction induced by treatment produces a lesser degree of regression of cardiovascular risk than the one that can be theoretically estimated by risk scores or equations

Thus models that use input variables from short term trials and estimate benefits from risk scores or risk equations are likely to overestimate the benefits of denervation. For example, the model based on SIMPLICTY-HTN2 results predicted a relative risk reduction of 0.7-0.8 corresponding to a reduction of 32 (15; 40) mm Hg in systolic blood pressure. The estimated efficiency was very good, $3,071 per QALY or $2,715 per life-year gained. All other published economic models are based on the same Simplificy-HTN2 inputs which are much more favorable than Simplicity-3HTN results. [32]

Economic evaluation of percutaneous closure of patent foramen ovale (PFO)

The closure of PFOs found in patients who experienced a cryptogenic stroke has been found to decrease the risk of recurrent neurological events in five prospective randomized trials. The meta analysis and economic evaluation of PFO closure used a model to extrapolate the short term results of percutaneous PFO closure compared to medical treatment over 20 years and estimate a cost effectiveness ratio. With an initial procedural cost of over $20,000 compared to a yearly cost of $600 for medications, it is expected that the duration of follow up will have a major impact on the result. Indeed the efficiency increased from $180,000 per life year saved at 2 years to $20,000 per life year saved at 20 years. This is explained by the initial hypothesis of a sustained benefit of PFO closure over time compared to medication and more importantly by the amortization of the initial procedural PFO cost over a long time period. [33].

Any chapter on health economics would be incomplete if it left the reader with the impression that policy decisions result directly from the calculations described above. The framework for the analysis of how policymakers use the information provided by health professionals or health economists is provided by M Goddard et al in their article on priority setting in health [34]. The world of healthcare delivery is not limited to clinicians and economists, other stakeholders such as politicians, bureaucrats, taxpayers, interest groups and the industry. Public and professional institutions seek to maximise their power and influence in the healthcare sector and policymakers seek to maximise their political support. Consumers (patients) are utility maximising and healthcare providers (in the private sector) are profit maximising. With this in mind, the processes of organisation and delivery of health services and technologies appear more complex than application of the results of economic evaluations would suggest.

Politicians select programmes that maximise votes or that minimise their personal or corporate risk of liability. They prioritise the implementation of technologies which benefit a large part of the population (or the population whose support is most important for politicians). Patient interest groups will support the technologies which benefit their members and increase their legitimacy as patients’ advocates. This latter point implies that interest groups will also seek to expand the use of technologies to the largest possible number of people rather than to a select group, often with the support of the industry which thus increases its market share. Interest groups will also make use of policymakers’ concerns for political support and render politically dangerous the denial of services. This is possible when the population concerned is very large, e.g., patients at risk from heart disease.

Conclusions

This chapter has attempted to present some highlights from the very rich economic literature pertaining to interventional cardiology. There remain more questions than answers. This is not surprising given the pace of technological change in interventional cardiology. For clinicians who work in a strictly regulated environment and have limited control over their use of medical technologies, this chapter will hopefully shed some light on the motives for policy decisions. For clinicians who might make decisions on the resources used to treat their patients, we hope to have provided the means of looking for evidence that will allow them to make informed decisions from both clinical and economic perspectives. One critical point to understand the results of economic models that compare medical treatments to interventions is the importance of the time horizon, ie over how long are the costs and benefits estimated. Because the cost of medications is recurrent over the patients’ lifetime, while an intervention and the implementation of a device are usually a one-off cost, the longer the time horizon, the more favourable the efficiency of the interventional procedure. This is different when interventional cardiology is compared to surgery because both require an initial costly intervention.

Personal perspective - Isabelle Durand-Zaleski

The economic literature in interventional cardiology is rich and getting richer as new procedures are being developed to replace either surgical or medical treatments. Most economic evaluations now use models to extrapolate the results of trials with relatively short follow up periods (1- years) over longer time periods, up to 10-20 years or lifetime. These extrapolations allow the authors to include recurrent events such as stroke or MI and their associated costs, and also to spread the initial cost of an expensive device such as TAVI. The models rely on assumptions about the selection of patients and the sustainability of health benefits which need to be carefully considered by clinicians who plan to develop a new technique.

SHARE YOUR COMMENT