Rising health care costs and pharmaceutical prices in particular are among the main factors that have prompted a steady flow of scholarly and lay press articles about moving from paying for volume to paying for value. Proposals from groups such as the Center for American Progress advocate drug pricing based on assessments of comparative effectiveness.
If implemented, value-based assessments of drugs and other health care services would influence payer, provider, and patient decision making, and likely patient outcomes as well. Each of these approaches assumes that some calculus of value could reliably be measured, would apply to most patients and other stakeholders, and could inform health care decisions. As prior Health Affairs Blog posts have discussed, a variety of value assessment frameworks have been developed to compare the health and economic impacts of drugs and other interventions. Are these frameworks ready for the task? If not, what is needed?
To begin to answer this question, several organizations have developed guidance on how value should be assessed and the processes by which organizations might make those assessments. These are not value-assessment frameworks themselves, but rather guidance for how to develop such frameworks; they include the Guiding Practices for Patient-Centered Value Assessment developed by the National Pharmaceutical Council (NPC); the Principles for Value Assessment Frameworks by the Pharmaceutical Research and Manufacturers of America (PhRMA); the BIO Principles on the Value of Biopharmaceuticals by the Biotechnology Innovation Organization (BIO); and the Patient-Centered Value Model Rubric by the National Health Council. The International Society for Pharmacoeconomics and Outcomes Research and Avalere/FasterCures have related efforts underway.
We applied the National Pharmaceutical Council's 35 Guiding Practices to frameworks developed by the American College of Cardiology/American Heart Association (ACC-AHA), American Society of Clinical Oncology (ASCO), Institute for Clinical and Economic Review (ICER), Memorial Sloan-Kettering Cancer Center (DrugAbacus), and National Comprehensive Cancer Network (NCCN). Based on that comparison and the resulting public discussion at a recent conference (Assessing Value: Promise and Pitfalls, Washington DC, 9/29/2016), we contend that:
- each framework has strengths and limitations,
- all have clear opportunities for improvements, and
- none is fully matured and ready to support health care decision making.
Despite this lack of full maturity, some of these frameworks are operating, de facto, to influence health care decision making, and the demand that has fueled their development is unlikely to abate in an increasingly value-conscious environment. It is essential, therefore, that stakeholders also demand the necessary improvements to these frameworks, and that their developers respond promptly and accordingly.
The frameworks are in flux. Developers of each framework have recently released or have announced the future release of improved versions. For example, ASCO published its expanded 2.0 version and is developing a 3.0 software tool that would allow patients to enter weights to reflect their preferences for particular value parameters. DrugAbacus updated its approach by adding two components of value, and ICER requested comments on proposed revisions early this year.
If paying for value is critical to the evolving health care system, and assessing value underlies that evolution, but each framework has limitations, how can the field of value assessment progress to truly discerning and valid decision support? Since these frameworks are in flux and many of their limitations are broadly shared, we have identified the following key considerations that might help to advance the field:
Key Considerations for Improvement of Value Frameworks
Assessments of value should be separate from assessments of budget impact and affordability
Value can be defined loosely as the net benefits that a drug or other health care service provides relative to its net cost (including cost and any relevant cost offsets). Importantly, that assessment should be based on the net benefits and net costs at an individual patient level. By contrast, budget impact is a population-level measure that reflects the product of the number of patients who might receive a service and the net cost per patient. Budget impact is a dollar amount pertaining to a particular payer or other economic perspective (e.g., $10 billion per year for the United States or $100 million for an employer or health plan), but it does not indicate whether that expenditure is a good or poor value. Affordability reflects whether a given payer or society at large has the means and willingness to sustain the budget impact of a drug or other service.
The ICER framework assesses value at the individual patient level, which is appropriate, but also uses budget impact to assess value at a US population level, which is not. Budget impact and affordability are critical considerations, but value assessment frameworks should not use them as part of their value assessments. Budget impact should be separate from measures of value.
Value assessments should incorporate what is important to patients, even if the end user for a framework is the payer
Given that patients are the recipients of health care services and ultimately patients and consumers pay for that care through taxes, insurance premiums, and forgone employee compensation, frameworks should incorporate components of value that are important to patients. Not including the patient perspective may lead to assessments that over- or under-estimate a therapy's value.
Patients should be actively involved both in the development and refinement of frameworks and in the assessment of the value of individual health care services. Although greater outreach is occurring, to date, none of today's frameworks have comprehensively and effectively included patients in this fashion.
Value assessments should adopt broad system perspectives in what they assess and how they assess it
Each of the major value frameworks incorporates benefits from drugs and health care services but does so with limitations. ASCO, NCCN, and DrugAbacus do not adjust for potential cost offsets of reduced hospitalizations, imaging, or surgery, while ICER and the ACC-AHA do incorporate these direct medical expenditures.
A practical accounting of value should incorporate not only direct medical costs, but also indirect costs, including the patient's ability to work (productivity) and caregiver burden. None of these value assessment frameworks incorporates these indirect costs quantitatively, although ICER has proposed an approach to do so. The recent report of the Second Panel on Cost Effectiveness in Health and Medicine recommended that all assessments consider not only the health care system perspective but also the broader societal perspective.
Value is dynamic and needs to be considered and captured as such
An assessment of value reflects what is known about net costs and benefits at a particular moment. But for many interventions, evidence continues to accrue and indications evolve, even as an assessment report is developed and published. Thus, the value of an intervention is likely to change over time, and a value assessment that does not account for the "moving target" of evidence and innovation may render outdated or otherwise misleading findings. Yet most of these frameworks do not have explicit provisions for updates or reassessment, although the guidelines in which the NCCN framework is incorporated are updated at least annually.
Value assessments should be transparent and reproducible
The assessments produced by value frameworks vary in the extent of information shared about the evidence used, deliberations about that evidence, and the underlying economic models. Transparency is necessary for the credibility and ongoing improvement of each framework and for the correction of findings where necessary.
Transparency should also extend to and enable reproducibility. A key element of science, reproducibility entails providing enough information about data and methods to enable others to duplicate results. Most of today's framework developers do not share sufficient information to enable reproducibility. For example, the DrugAbacus equation used to calculate an Abacus price of a drug is provided on the DrugAbacus website, but the programmed values for each domain, based on clinical trial data and certain market information, are not provided. Although ICER's methodology and evidence sources are well described, the executable models and associated computer code of its economic models are not publicly available. The methods used by NCCN are publicly available, but the results of its Evidence Block scores are not directly reproducible, as they represent the average of panel members' individual subjective scores.
A diversity of value assessment approaches that reflect the differing needs of stakeholders should remain; value assessments should reflect user preferences
Stakeholders need assessments of value that address their perspectives and decision-making needs. Since framework assessments will likely influence such diverse areas as payer decisions about what to include in health care benefits and doctor-patient dialogues about treatment choices, one framework will not be adequate for all uses. A diversity of frameworks reflecting the differing preferences of patients and practical needs of payers, rather than only one "best" framework, is necessary.
A regional payer may place higher value on aiding patients with particular disease profiles, risk factors, or sociodemographic attributes, reflecting the patients that they serve contrasted with those served by other payers. At the individual patient level, value should reflect the specific needs or preferences of those patients. For example, one cancer patient may prefer a therapy that minimizes side effects or burdens on a spouse or other caregiver, whereas another patient may accept such side effects or burdens if they accompany a higher likelihood of extended survival. No single framework with a predefined set of weighted variables can reflect the needs of disparate stakeholder groups or even address the inevitable heterogeneity within a particular group.
Insurance reform is needed along with improvements in value assessment frameworks so that patients can gain access to high-value services (and low-value services are discouraged)
As public- and private-sector payers experiment with and convert more of their health benefit offerings to value-based payment models, they need credible, evidence-based assessments of health care interventions to inform policies that encourage more use of high-value services and discourage use of low-value ones. For example, value assessment frameworks that reflect patient input and preferences can directly inform value-based insurance design (VBID), in which services designated as high-value are associated with low or zero patient financial responsibility; low-value services carry greater costs for patients under VBID. A mix of value assessment frameworks can inform the design of various value-based payment models that account for the perspectives of patients, clinicians, payers, and policymakers.
Summing Up
Value assessment frameworks offer important potential for helping to make purchases of drugs and health care services more cost-effective. However, these frameworks still require improvement, and critical policy choices should not yet be based upon them. It is essential that framework developers and stakeholders strive to make the needed improvements, particularly since framework findings are in fact already exerting increasing influence on clinical decisions, payment policies, and the broader national dialogue on pricing and access to health care.
Health care access mediated by value-based payment sends signals to innovation, and those signals need to accurately reflect what patients value and society desires. It is critical that the means of value assessment be up to the challenge. Pursuing the points of improvement presented above could help to meet that challenge.
from Health Affairs BlogHealth Affairs Blog http://ift.tt/2mPcTNg
No comments:
Post a Comment