While there is general consensus that existing value assessment methods are imperfect solutions for providing credible and relevant information on value specific to cancer care decisions, there is much less consensus on the appropriate methodologic advancements needed.
At the ISPOR Annual Meeting (May 18-20, 2020), the Innovation and Value Initiative (IVI) hosted a panel discussion centered on identifying gaps and practical solutions for such gaps in value assessment methods. The panel was comprised of a diverse group of individuals and perspectives in order to consider various stakeholder viewpoints.
Journal of Clinical Pathways sat down with Jennifer Bright, MPA, executive director of IVI and moderator of the panel, and Lou Garrison, PhD, Professor Emeritus and health economist from the University of Washington and discussion panelist, on the touch points of the panel and how value assessment and clinical pathways should evolve together to meet the needs of value-based cancer care.
During the ISPOR panel on the gaps in value assessment methods, each panelist was tasked with providing an overview on the gaps from the perspective of their respective stakeholder viewpoint. Can you briefly summarize the presented gaps from the patient perspective, the provider perspective, the payer perspective, and the health economics perspectives?
Mrs Bright: This panel was born first out of the methods summit that IVI hosted on February 27, 2020. It was a day-long conversation that brought together patients, methodologists, payers, employers, and other stakeholders to ponder existing areas of consensus and priority gaps for improving the practice of value assessment.
Leah McCormick Howard, JD, chief operating officer of the National Psoriasis Foundation, was on a panel during the summit, and subsequently part of the panel dialogue at ISPOR. She emphasized the unique and vital input of patient perspectives and why it is important to improve data inputs—namely, the clinical and real-world evidence about patients and patient outcomes that help us understand value in the context of heterogenous patients. We know that the quality and types of data that the patient perspective brings forward is not always “pluggable” into current value assessment models. This lack of compatibility between available data and value assessment models was a common theme that arose from the ISPOR panel.
Eric Stanek, PharmD, principal scientist at HealthCore, Inc (Wilmington, DE), also attended the February summit and participated in the ISPOR panel. He reflected on the challenges from a payer perspective, specifically, how to use all the data inputs in the context of decision challenges faced by payers. One of the ongoing challenges is the mismatch between decision context and usable information. Eric reflected on the importance of payer engagement – of coming to the table and embracing the notion that some modified or augmented methods need to be explored and that a learning culture needs to exist to truly advance the methods and practice of value assessment.
Dr Garrison: I contributed the health economist perspective to the discussion reflecting in part my experience as co-chair of an ISPOR Special Task Force on US Value Frameworks. That task force was faced with sorting out the plethora of frameworks that have emerged in the last 7 or 8 years. Ultimately, one way we sorted them out was based on the “decision context” they are being applied in, and they can be seen as inter-related in a sequence or cascade.
The cascade can be ordered as follows. First, the FDA is tasked with assessing the clinical value of a therapy or service. In other words, is the benefit risk worth it? Then once it is on the market, the health plans have to determine whether and how to incorporate it into their formulary or plan benefits. The third decision context is in the hands of the providers and plan managers through the use of clinical pathways, guidelines, and utilization management. How are these therapies or services going to be managed? The last level in the cascade—some would argue the most important—is the clinical shared decision-making context of the physician patient interaction. Certainly, patient-centeredness is critical for the physician to understand the individual patient's preferences regarding benefits and risks.
Organizations such the Institute for Clinical and Economic Review (ICER) are using cost per quality-adjusted life-year (QALY) as the core metric of value. Our task force and ICER have recognized that the quality-of-life component of the QALY has limitations and that it may not capture the patient-centered experience fully. Certainly, following our task force recommendations, many of us are looking for ways to improve the use of the cost-per-QALY metric. It is a good starting point, but it is not perfect, and in particular, it does not deal adequately with the impact of disability and is often not sufficiently patient-centered.
A major issue for value assessments moving forward will be how they are utilized in developing clinical pathways. Are pathways to be based on expert judgment about appropriateness—which they have been historically—or will they place more weight on cost-effectiveness and value? From my perspective, it is time to ask if clinical pathways should begin looking at economic value and not just clinical value. Pathway developers need to grapple with this question.
It is fair to say that the direct costs of care (eg, price of a test, price of a therapy, or cost saved by avoiding ED visits or hospitalizations) is the main consideration of value assessment. How are indirect benefits (eg, lost productivity, caregiver burden, or diminished parental quality of life) currently considered in value assessment?
Dr Garrison: Historically, economists have done parallel analyses from a health sector perspective and a societal perspective. The societal perspective—in theory and certainly in practice—tries to take into account all of these indirect effects.
In 2018, the US Second Panel on Cost-Effectiveness proposed a comprehensive “Impact Inventory” that aims—in the societal perspective—to consider a wide range of indirect effects throughout the economy. The recommendations of the ISPOR Special Task Force would imply that this inventory should deal more explicitly deal with impact of uncertainty about health outcomes and costs on patient health and well-being.
When, for example, a patient undergoes a genomic or other diagnostic test, he or she may enter that process without previous knowledge of the extent of the disease. Afterwards, much more information may become available, thereby reducing the patient’s uncertainty, which most people generally value. The value of this reduction in uncertainty needs to be recognized and reflected in value assessments.
Mrs Bright: The patient community in particular is leaning in on indirect benefit inclusion in value assessments. In particular, they are asking for more vigorous attempts to bring the value equations and data inputs that go into these assessments into the open. They want transparency on all of the components that go into a value assessment.
That being said, doing so is very complicated. There is a lot of discussion about caregiver burden and productivity, for example. How do we calculate these factors in a way that is methodologically sound and reinforcing of good decision-making? What are the component data points that help us measure these two aspects?
IVI has identified major depressive disorder as a target for method and model development where we can explore such factors as part of the examination of value. The prevalence and impact of treatment in depression is significant and patients, clinicians, and employers could benefit from greater understanding of comparative value of different treatment pathways that account for the impact of an intervention on ability to work, for example.
We all understand that these indirect costs are relevant. We just do not have a good way of knowing how they are incorporated into our decision-making and how we balance those offsets in a way that is transparent.
Is “valuing” indirect benefits in the assessments a problem of insufficient data or lack of patient-focused measures?
Mrs Bright: We lack both clear, prioritized, patient-defined measures that align with patient perspectives on value, as well as sufficient access to data that meaningfully illustrate these indirect benefits. Certainly, there is vast data available through clinical, outcomes research, and real-world sources ranging from electronic health records to registries, but often the problem is a clear definition of what is important to the patient and that has bearing on outcomes that indicate health and recovery.
Dr Garrison: I think it is a mix of both as well as some methodological issues. Certainly, our health system generates more data on costs, largely for administrative reasons. There is a wealth of information on health outcomes in our electronic health records that we are only beginning to harvest.
Indicators of impact on work productivity, caregivers, and family members are more difficult to come by. There are survey instruments for eliciting and measuring these in the context of clinical trials and patient registries that are widely used. They provide useful information on the contribution of these indirect effects to economic value. The larger concern may be that our measures of direct impact on the health and well-being of patients are not sufficiently centered on what really matters to patients.
Can you speak to whether the increase in use of genomic testing in cancer care presents any unique challenges to value assessment design and methods?
Dr Garrison: The advent of “personalized medicine” has taken longer than many predicted following the sequencing of the human genome early in this century. Yet, in the past decade, we have finally witnessed significant scientific advances, for example, in genomic testing, immunotherapy, and gene therapy. We are now applying genomic profiling to assess tumor mutational status, and screening tests of blood—referred to as “liquid biopsies”— are being tested. These technologies raise unique issues for value assessment. What is the value of reducing uncertainty about outcomes and costs with tools? Will they be cost-effective in screening for cancers? There are still many questions to be answered about the cost-effectiveness of the genomic and pharmacogenetic testing interventions.
Certainly, a large component of the potential value of genomic testing is the same as for all diagnostics: it can reduce the uncertainty about the effectiveness, cost-effectiveness, and appropriateness of a given intervention. However, value assessments of these technologies have sometimes overlooked these elements, and perhaps more importantly, our reward system of payments has been more “cost-based” than “value-based.” But if we do not reward the full value of testing, then we will not generate necessary patient-level data to demonstrate that value. We reward value in new medicines, but not so much for diagnostic testing, with the exception of a handful of examples of value-based payments for diagnostics. Our health care system has been biased against rewarding innovation in testing.
Mrs Bright: The most concerning issue we hear from patients is the time to diagnosis; many patients, particularly with chronic and co-occurring diseases, first confront the wasted time, uncertainty, and inappropriate treatment pathways that accompany the absence or poor utilization of diagnostic tools. One result is that by the time diagnosis is achieved, they may have missed opportunities for adequate treatment or positive outcomes for first-line therapies. Personalized medicine and the advances of genomic and other diagnostic testing can accelerate our ability to see which treatment pathway and for whom it may have highest potential positive outcome and highest value, but we do not have methods for how to factor such advances into our calculation of value yet.
Flexibility is an important part of the evolution of value assessment modeling. To Mr Garrison’s point, clinical pathways need to evolve in parallel; they need to go beyond the clinical data and expertise to incorporate perspectives derived from patients and that reflect patient sub-populations’ needs and treatment response.
Value assessment can be a parallel resource to clinical judgment in patient treatment. We need to think holistically about where value assessment can contribute to other decision-making, or where bringing in the economic and societal perspective could be of value.
What specific inputs should be flexible? Should the parameters for value assessment be flexible or amenable based on patient populations and characteristics?
Mrs Bright: The ability for value assessment models to examine subpopulations within a disease state is one aspect of flexibility in their design that IVI thinks is important. Patient heterogeneity is a real-world challenge to clinical decision-making and we should be working collectively to develop assessment tools that can present multiple scenarios for comparison, so that decision makers can better understand for which subpopulations there might be a different sequence of treatment that is meets both clinical and cost-effectiveness. Again, there is important work to be done to identify the factors that matter most to patient populations – especially to the extent we can define common factors across disease states. Moreover, we must define the decision contexts that require such flexible inputs (eg, for clinical pathways or benefit design/utilization management structures).
Dr Garrison: Mrs Bright is absolutely correct that these are models that must often project impacts based on immature or trial data and hence need to be flexible and updated over time as we generate more real-world data on their performance. Model parameters such as the sensitivity and specificity of the test as well as the linkage from biomarkers and surrogate outcomes to long-term mortality or cost need to be modifiable, including for different types of patients. IVI is doing us a service by promoting this open and flexible science.