A View from the Bridge: Health Economic Evaluation - A Value-Based Framework?

Abstract

In the 25 years since Health Economics first came to press, health economic evaluation has come of age. The first issue of Health Economics was published in 1992, the same year that the Pharmaceutical Benefits Advisory Committee in Australia launched its first formal submission process for manufacturers to gain listing on its pharmaceutical benefit schedule, which included the requirement to present an economic assessment of the drug's cost-effectiveness. Within the next few years, Canada had followed with its Common Drug Review and the UK introduced the National Institute for Clinical Excellence (NICE) in 1999. These early attempts to implement formal health technology assessment (HTA) processes into national decision-making have all stood the test of time, and many other countries have followed suit, with processes in Germany, Sweden and the Netherlands all employing a formal assessment of value (O'Donnell et al., 2009). The proliferation of national processes is gaining traction in other parts of the world too, with many countries in South America and Asia developing their own guidelines and submission processes. Even Japan has entered the fray with a new HTA programme that began on a trial basis from April of this year. Although the US was an early pioneer of HTA with its Office of Technology Assessment (OTA), the US remains one part of the world where there is relatively little formal use of health economic evaluation. This gap can in part be explained by the disparate group of payers that makes up the US system. Without a ‘single-payer’ model, the impetus for a national programme of HTA is lacking. Indeed, the OTA was disbanded in the US in 1995 just as other HTA systems were finding their mark (O'Donnell et al., 2009). Even where national bodies still exist, their ability to consider cost-effectiveness information is restricted. For example, the US Preventive Services Task Force (USPSTF) is not allowed to consider cost when issuing its guidance — a bizarre situation that leads investigators to use healthcare resource use as a proxy for cost when examining efficiency in screening programmes. The recent USPSTF guidance on colorectal cancer screening used ‘colonoscopy’ procedures as its measure of resource cost in order to estimate an efficiency frontier for different screening options (Knudsen et al., 2016). Likewise, in order to get the Affordable Care Act through congress, an amendment required that the Patient-Centered Outcomes Research Institute (PCORI) be prevented from using cost per quality-adjusted life year (QALY) thresholds for prioritizing healthcare delivery — effectively outlawing cost-effectiveness analysis (Neumann and Weinstein, 2010). PCORI's own frequently asked questions page even emphasizes that proposals that include direct comparisons of the costs of care between two or more alternative interventions will not be considered for funding. Although the processes in Australia, Canada and the UK have largely survived unchanged over the past 15–20 years, there have been some challenges. In particular, NICE saw political interference in its process from the UK Department of Health when it was asked to introduce ‘end-of-life criteria’ that would enable it to approve sunitinib for patients with advanced kidney cancer. The apparent arbitrariness of some of the criteria imposed on NICE (and now assimilated into its guidance) was confirmed in empirical work by NICE's own Decision Support Unit that cast doubt on whether the general public did put extra value on health outcomes achieved at the end of life (Shah et al., 2015). Nevertheless, the principle that there could be other factors beyond cost-effectiveness that should influence decision-making began to be seriously explored. To be fair, NICE (and many other reimbursement agencies) have long emphasized that cost-effectiveness results are only one input into their (deliberative) decision-making process. But recent research interest has focused on whether it is possible to quantify additional criteria in such a way that they can be embedded formally into the HTA decision-making process. In particular, a technique known as multi-criteria decision analysis (MCDA) has been advocated as an approach to do this. Originally designed as a technique to help committees reach decisions (Goetghebeur et al., 2008), increasingly, a number of commentators have suggested that MCDA could be used for HTA by developing weights for attributes that could sit alongside cost and health outcomes (Goldman et al., 2010; Devlin and Sussex, 2011). In the US, the situation is even more confusing. In the absence of formal HTA processes of the type that have emerged within the single-payer systems that predominate in the developed economies outside of the US, costs have inevitably risen across the system. It is well known that the US spends more on its healthcare as a percentage of per-capita GDP than any other nation. In particular, pharmaceutical spending has come under the spotlight with an increasing awareness that the prices of drugs in the US far exceed those that are paid by other economies with single-payer systems and HTA processes. Oncology has become a particularly acute battleground, with some analyses suggesting exponential growth in the cost of cancer drugs over the past four decades (Bach, 2009). Against this backdrop, it was inevitable that something would have to give, and commentators in the US are now beginning to talk seriously about the need to assess the value of new products. For some insurers and other buyers in the US system, value has been defined in relatively traditional cost-effectiveness terms. Formulary submissions to the Academy of Managed Care Pharmacy invite manufacturers to present optional cost-effectiveness information alongside obligatory budget impact modelling, and cost-effectiveness has been used by at least one large US insurer to explicitly link value to patient copayments in a way that has been shown to reduce overall medical expenditures (Yeung et al., 2016). For others, talk of ‘value frameworks’ has become the latest buzzword, and a new initiative spearheaded by a collaboration between the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) and the American Society for Health Economics (ASHEcon) plans to put out a white paper on US value frameworks (ISPOR, 2016). Existing value frameworks that have emerged in the US (Neumann and Cohen, 2015) can be broadly split into two types. The first are those that have, to a greater or lesser extent, followed the cost–utility analysis type of framework for assessing value in healthcare. This approach includes the framework proposed by the Institute for Clinical and Economic Review and the American College of Cardiology and the American Heart Association guidelines (Anderson et al., 2014) on value. The second type of framework has emerged in the oncology field and appears to take a broader notion of value. The American Society for Clinical Oncology (ASCO; Schnipper et al., 2015) and the National Comprehensive Cancer Network (NCCN)'s frameworks and the Memorial Sloan Kettering Cancer Center (MSKCC)'s Drug Abacus all include cost (affordability) alongside traditional survival metrics while allowing for other potential attributes. Only the ASCO framework has attempted to develop a weighting system to their framework, at least on the benefit side (it stopped short of a formal comparison with costs). By contrast, the NCCN simply presents five dimensions in an ‘evidence block’ with no attempt to weight each dimension, while the Drug Abacus allows users to define their own weights to come up with a value-based price for many contemporary cancer drugs. Commercial organizations are also getting in on the act — with products designed to help US payers assess the ‘true value’ of specialty drugs using disease-specific, and sometimes class-specific, MCDA-like analyses. What should we, as health economists, make of these recent efforts to define value? First of all, it should be noted that a worrying amount of development of these frameworks, both in terms of MCDA and the US value frameworks that are essentially a form of MCDA, appear to have developed with little input from economists. Indeed, the lack of rigorous economic thinking is apparent — most notably in the absence of any notion of sacrifice when it comes to defining weights for different attributes within any given framework. The ASCO weighting system, for example, has points allocated to each of its attributes in order to come up with a ‘net-health benefit’, yet it is not clear that any notion of choice or sacrifice was employed by the committee when deriving the weights, which renders the results meaningless for priority setting. Second, it is clear that universal agreement does not exist on the additional attributes that should be considered. Even within the oncology area, the three multiple criterion frameworks of ASCO, NCCN and MSKCC have a wide disparity in the additional criteria they apply. Of course, health economists have much to say on value, and it is promising that ASHEcon is directly involved in critiquing the emerging frameworks in the US. But perhaps, as health economists, we need to be bolder and broader in our suggestion that assessment of value is necessarily in our purview and that, furthermore, our discipline has the tools to address the problems of value assessment that have been honed over a number of decades. It is with some irony that the US Panel on Cost-Effectiveness has just released an update to its original 1996 guidance on good practice (Weinstein et al., 1996; Sanders et al., 2016), yet few commentators have recognized that this guidance is the original value framework for the US audience. This lack of recognition may reflect the desire to include attributes in the new frameworks that go beyond length and quality of life, but if that were the case, then the methods of cost–benefit analysis (CBA) represent the original MCDA — where all measures of welfare, not just health, can be captured in a single analysis, which brings the focus squarely back to the welfarist versus extra-welfarist approaches to efficiency that health economists have been debating for many years (Hurley, 2014). Of course, the traditional welfare economics framework of CBA has not developed much traction in the health field due to the methodological challenges associated with measuring health in monetary terms. However, discrete choice experiments (DCEs) have proven to be promising in generating monetary values for health outcomes (de Bekker-Grob et al., 2012). Indeed, DCEs are a form of MCDA, being based on attributes and levels of those attributes that are important to a consumer's utility function. But crucially, the design of a DCE is consistent with the economist's notion of sacrifice such that the preference data from the experiment typically conform to random utility theory. Furthermore, DCEs provide a framework by which it may be possible to combine the traditional QALY health outcome measures with other value attributes that go beyond health (Wildman et al., 2016). The notion of value in healthcare is fundamental to our discipline. The emergence of MCDA and associated value frameworks as a way to define value is only a threat if we continue to allow the development of these frameworks in isolation from the methods of health economic evaluation – welfarist and extra-welfarist – that have been developed and refined over the past decades. A call to arms is required for health economists to engage with our non-economist colleagues to help them understand the importance of rigorous economic approaches to defining value.