The proliferation of high-cost cancer therapies has prompted efforts to quantify their value to healthcare stakeholders and society. The inherent challenge in achieving this goal is that each stakeholder holds differing perspectives of value. Calculators that have emerged to measure the value of cancer therapies reflect the interests of their developers and vary substantially in their intention, format, and usability. Input from each stakeholder is critical to construct a tool that has applicability to individuals, populations, and the healthcare system.

In today’s healthcare delivery and reimbursement environment, the focus has shifted from volume to value. Purchasers and payers now compel providers and pharmaceutical manufacturers to demonstrate the value of services and products through metrics that can be measured. In theory, the provision of value-based care also improves the patient experience through higher quality and satisfaction.

The shift to value-based care has been rapid in high-cost specialties where clearly established care protocols can result in predictable outcomes, such as cardiology and orthopedics. In oncology, where personalized treatment approaches are more common, the value-based paradigm has been slower to take root. Scientific advances in our understanding of cancer defy population-based approaches to care, which often form the foundation of value-based reimbursement programs.

Our evolving recognition of cancer, as a complex gallery of molecularly driven diseases, has led to the development of niche drugs, often costing $100,000 or more per year of treatment.

 Escalating drug costs have prompted several individuals and organizations to develop tools or, “calculators,” intended to pinpoint the value of new drug therapies. To date, 4 value calculators have been developed for public dissemination of information, and more are likely to emerge.

 representing payers, providers, pharmaceutical companies, and patients to discuss meaningful considerations for building a value calculator. The group examined the strengths and weaknesses of existing calculators and reached consensus on additional considerations that are relevant across stakeholder groups.

The panel was moderated by Bruce A. Feinberg, DO, vice president and chief medical officer, Cardinal Health Specialty Solutions. Experts who lent their opinions to the discussion included Alan Balch, PhD, chief executive officer, Patient Advocate Foundation; James C. Gilroy, senior director, Market Access, Eli Lilly & Co; Ira Klein, MD, MBA, FACP, senior director, Quality, Strategic Customer Group, Janssen Pharmaceutical Companies of Johnson & Johnson; Michael A. Kolodziej, MD, national medical director, Oncology Solutions, Aetna; and Ted Okon, MBA, executive director, Community Oncology Alliance.

The need to define value in healthcare stems from the steady growth of healthcare consumption. Following the enactment of Medicare, Pauly predicted that a health insurance setup that reduces the price of healthcare to consumers, but not the cost of producing healthcare, would result in excess consumption and a loss of economic prosperity.

 Since then, health spending, as a share of Gross Domestic Product, has increased from 6.2% to 17.5%.

Government actions that limit healthcare demand would be politically unpalatable, so payers have moved to control the supply side. Evidence—or a lack of it&mdash;to support a healthcare intervention, validates controls on supply. The growth of the evidence base for cancer therapies has enabled the development of empirical tools for measuring their value. Despite the limitations inherent in the new value calculators, their effort to bring rigor to the assessment of novel therapies is a positive development, especially in light of the fast-growing array of choices to treat cancer.

Existing value calculators tend to be grounded in the developer’s interpretation of value. In developing a value calculator that considers the needs and objectives of all stakeholders, a useful starting point is to begin with an equation for value and examine stakeholder perspectives that define it.

By definition, value-based care incorporates cost into clinical decision making. In the context of healthcare, then, value can be expressed as an equation

This seemingly simple idea becomes more complex when one considers what constitutes 

. Porter noted that outcomes are “inherently condition-specific and…no single outcome captures the results of care.” Cost, he said, refers to the total cost for a full cycle of care, and that “to reduce cost, the best approach is often to spend more on some services and reduce the need for others.”

The relevance of outcomes also differs among stakeholders, and even within a single stakeholder group. Payers, for instance, would welcome an outcome which measures the total cost of care affected by a therapeutic regimen. Patients may prefer a continuum of outcomes ranging from cure to amelioration of disease to assurance that they will be able to resume normal activities.

Understanding which of these outcomes is attainable is critical for a patient to determine the value of the available treatment options. For a variety of reasons— fragmentation within healthcare, reimbursement that discourages providers from taking the time to explain options, or the lack of price transparency associated with first-dollar coverage, to name a few&mdash;patients historically have been shielded from intelligence they need to discern treatment options that match their preferences. As cancer treatments become more specialized, patients will need tools to help them decide which outcomes are important to them. This becomes even more important as changing benefit designs increasingly shift costs to patients.

If defining the numerator (outcomes) seems complex, then defining the denominator (cost) is equally challenging. Existing calculators, for measuring the value of a drug, consider its cost in isolation outside of the larger context of total cost of care for an episode. As such, a calculator may not necessarily be useful for determining the intervention’s true value (

For patients, costs are both absolute and relative. A patient’s benefit design, for instance, defines his absolute out-of-pocket cost for an intervention. The patient’s willingness to pay, however, depends on how that cost affects his ability to afford aspects of everyday life, such as utilities, rent, or transportation. The availability of co-payment assistance for physician visits, hospital bills, and the drug itself also factor into willingness to pay. This series of tradeoffs is unique to each patient and contributes to his judgment about the “worth” of a drug’s cost.

As perspectives on cost differ from one stakeholder to the next, so do perceptions of value. Let us consider how each stakeholder formulates value judgments.

 To an oncologist, the cost half of the value equation is simple: does the patient believe he can afford a particular treatment? Whether the patient can accept the economic burden of a particular therapy is a decision that only the patient can make. Many oncology practices have financial counselors on staff to help guide the patient, but in the end, the choice is up to the patient.

The outcomes half of the equation is more complicated. In treating patients, oncologists think about several types of outcomes, primarily:

		Survival: is there a chance for a cure?

		Endpoints: given the patient’s cancer type and stage, how important are response rates or progression-free survival (PFS)?

		Toxicity: what level of fatigue can a patient live with? How will neuropathy affect the patient’s life or livelihood?

 Payers may view value through 2 lenses: population-based care and personalized medicine.

At the population level, value-based insurance design can be implemented for some high-prevalence conditions that have,

		clear, evidence-based protocols for care

		the potential for high costs when those protocols are breached.

Making beta blockers accessible to patients who have had a heart attack, for instance, is an evidence-based, low-cost investment whose value is reflected in the potential for avoidance of readmission or a recurrent cardiac event. Similar value judgments can be made about the accessibility of treatments for people with type 2 diabetes to help them avoid complications, or to people with asthma to prevent exacerbations.

For low-incidence, high-cost diseases in which a small minority of the population incurs a disproportionate share of costs, decisions about value are more frequently made on an individual basis. Sometimes, the value judgment lies neither with the patient nor the payer, but with the provider group if the group is obligated to follow clinical pathways or has taken on financial risk.

 For brand manufacturers, value lies in true innovation. Is a drug first in its class? Does it represent a breakthrough over existing therapies? Even drugs that are not first-in-class may demonstrate value if it can be shown how some populations respond to the product relative to other drugs in its class.

Manufacturers measure the value of new products using FDA-specified endpoints for efficacy and safety. As more specialized therapies are developed, however, some of those endpoints, such as overall survival (OS), PFS, or time to progression, may not capture the true value of the drug, especially for subpopulations of so-called, super responders.

The industry’s ability to communicate value for a product is further complicated by the fact that its true value may not be fully understood at launch. Determining the value of a drug is a dynamic process that takes into account real-world experience, though the industry is limited to marketing efficacy on the basis of clinical trial data.

 Patients often perceive value in terms of what they want from their care. Is the patient, for instance, most concerned about making it to her child’s wedding? Controlling nausea? The patient’s treatment goals can lead to a discussion with her oncologist about outcomes, such as PFS or side-effect profiles, that are most meaningful to the patient.

On the cost side of the equation, a drug’s cost is part of the larger milieu of treatment-related expenses. Considering the patient’s individual treatment goals and the overall expense of treating her cancer, the cost of the drug itself may become less relevant as a focal point.

In recent years, 4 organizations have released calculators, and other tools, for assessing the value of high-cost specialty drugs. The first, launched 3 years ago, was Memorial Sloan-Kettering Cancer Center’s (MSKCC), DrugAbacus. In 2015, the American Society of Clinical Oncology (ASCO) released its Value Framework, which was followed by the rollout of the National Comprehensive Cancer Network’s (NCCN), Evidence Blocks. This year, the Institute for Clinical and Economic Review (ICER) began to assess the value of individual drugs through evidence reports.

Each of these tools has substantial differences. These tools are summarized here, along with the expert panel’s analysis of their strengths and weaknesses.

Developed by Peter Bach, MD, director of the Center for Health Policy and Outcomes, at MSKCC, DrugAbacus is an online tool that prompts users to make their own judgment about whether the cost of a drug is in line with its value.

In the top panel, each drug’s average wholesale price is represented by the gray points. Its “Abacus price” is represented by the green and red points, which rise or fall, as the user adjusts the inputs below.

The DrugAbacus begins with a fixed dollar value for efficacy, then allows the user to adjust that value up or down on the basis of 6 inputs (

). The user can adjust the importance of each input, in essence assigning a value to each of the 6 components. The resulting price, or “Abacus price,” can be compared to the drug’s average wholesale price at launch.

The Abacus’s baseline price for efficacy and the magnitude of the variable inputs are based on data supplied by the manufacturer to the FDA. Strengths and weaknesses of the model are summarized in 

Memorial Sloan Kettering Cancer Center DrugAbacus

Articles

Evidence-Based Oncology

A review of the various value calculators that have emerged to measure the value of cancer therapies&mdash;these calculators reflect the interests of their developers and vary substantially in their intention, format, and usability.

Considerations for Designing "Value Calculators" for Oncology Therapies

’s managing editor Surabhi Dangi-Garimella, PhD, recently interviewed Steven D. Pearson, MD, MSc, FRCP, founder and president of the Institute for Clinical and Economic Review (ICER), Boston. ICER has been actively involved in changing how drugs are evaluated and priced. In this interview, Pearson talks about the need for value tools, compares the various tools that have been developed, explains the need for understanding the budgetary impact of healthcare costs, and why stakeholders are interested in innovative tools such as indication-based pricing.

 I think value has always been a part of the decision-making process for patients, health systems, and insurers—both public and private&mdash;forever. I think there has always been a mix of different approaches to doing it. In the United States, we tend to assess comparative clinical effectiveness, and we are less clear and transparent about how value has been assessed or has been woven into decisions in clinical guidelines, coverage policies, etc. Ethicists, health policy experts, and insurers have, for a long time now, recognized that we need to do a better job of being transparent. And any time that people feel [that] the overall cost of healthcare seems to push the envelope of what people can afford, they come back with greater attention for the methods that can be used to help them make better judgements about value.

Can you describe the key features of ICER’s value framework? Does it distinguish itself from the frameworks developed by the American Society of Clinical Oncology (ASCO) and by the National Comprehensive Cancer Network (NCCN)?

 At ICER, we have been generating evidence reports that integrate value for about 9 years now. Our methods have evolved over time, because a large part of our goal was to figure out how to provide evidence for value in a way that would be helpful for multiple decision makers in the healthcare system. Over the past 2 years, we have worked to bring together different stakeholders to gain their input as we revised our approach, trying to make it as explicit and transparent [as was possible].

ICER’s framework is specifically designed for payers; to do that, it has to include perspectives of other stakeholders. But it’s [also] supposed to be a tool that’s more focused on the payer perspective—we look at the cost to payers and not the out-of-pocket (OOP) cost to patients. Our general approach was to discuss with various stakeholders the elements that they would want payers to think about as they made [coverage] decisions. Based on that, we generated a list of components of value. However, there was never any consensus on these elements because there was debate over whether budget impact or measure of affordability should be a part of the value assessment.

What our system develops is something called the “care value,” which is a long-term perspective that is used as an anchor to query the value that will also have a short-term impact on the budget. So ICER’s framework identifies the threshold at which cost might create a stress on affordability.

ASCO’s framework aids the shared decision-making process between an individual doctor and an individual patient. We, too, have developed tools to assist that process: it’s called Proven Best Choices. However, our value assessment framework itself is more oriented toward integrating patient perspectives into tools that are more likely to be used by payers. We develop benchmarks for the cost-effectiveness threshold—in the long term as well as the short term&mdash;that would have budget impacts that none of the other [frameworks] officially use, although some have said that the ASCO model tries to do 2 things at once by using the OOP costs and payer costs in the same model.

: Does ICER’s framework extend beyond drugs and devices, and help establish the value of a treatment pathway?

 Yes, it can be applied to evaluate the value of anything, including introducing behavioral healthcare into primary care. So anytime you look at the long-term cost effectiveness of something, you are creating a pathway—trying to determine what happens to a patient next year and the year after that. It can also be used to compare sequencing of treatments as well.

Can you comment on the societal willingness to pay for a drug or treatment? Do we, as a society, need a better understanding of this?

You do make certain assumptions in our framework. You can use different thresholds, which, in turn, are linked to various methodological anchors. There’s of course no perfect answer to what society is willing to pay across the board for different services. Clear empirical evidence does come out of England, where they believe these thresholds are too high. In their healthcare system, the true threshold at which you start to lose value by adding new things is much lower. Numerous studies have been conducted to understand society’s willingness to pay for certain things. Some studies also lump this value to the nation’s GDP [gross domestic product], so that poorer nations would have a much lower threshold. So what we end up doing, since there’s no perfect answer, is we use a range of $100,000 to $150,000, which does seem to represent a consensus, if there is one. That’s more with the long-term effectiveness.

But in the short term, we need to understand “how much is too much.” For this, we survey payers and ask them, say, in the next 2 years, how much new spending can you afford? Would it need you to raise premiums, delay services, or hold back on other areas that may cause patients more harm than good? We did get an answer when we tried that.

At the societal level, we made a very big assumption that the US society would want an alarm bell that would indicate when healthcare costs are growing much faster than our overall economy. We therefore developed a way to figure out when a single agent would potentially contribute enough to make a significant impact that would tilt healthcare spending. This would mean that everyone from patients to doctors to insurers to manufacturers may need to think hard and hopefully work together to ensure that if it has a good long-term value, we bring it into the healthcare system in a way that it doesn’t create significant short-term budget impact problems.

: Healthcare providers would like to see standardization across the value tools and frameworks. Your comments?

The United States is the only developed country that does not have a national or federal process [for healthcare]. Our healthcare system has always been decentralized, so it’s not surprising that we do not have a single approach. I think it’s healthy to have the discussion and comparison, because these tools have very different objectives. I do think that it would be better if manufacturers and payers understood the approach to assessing value, because manufacturers often do not know what the evidence standards are or how the value will be assessed by payers. This creates uncertainty and lack of efficiency.

I don’t think we’ll have 1 tool for every purpose—that wouldn’t be the right approach to start with&mdash;but at least what we are doing with drugs, where we plan to put out reports on all new drugs at the time of FDA approval, we do hope that by providing a transparent framework, it can create a paradigm so the use and assessment of value can be more consistent.

: At ICER’s recent policy summit, stakeholders held discussions on indication-specific pricing for drugs. What were some of the findings? Are payers and manufacturers interested in developing and implementing this model?

Payers and manufacturers are interested in this model. Sometimes it gets pitched as a great idea and they are definitely interested in learning more about it. The paper that was published off of the summit tried to explain the basic approaches to indication-specific pricing (ISP) and to explore the international experiences.

 I think they are interested in it, along with outcomes-based contracting and some other approaches, as tools that are certainly worth considering in specific circumstances. I think they know that any one approach will not be a silver bullet to solve all issues around affordability and value and access—but payers and manufacturers are both interested in innovation. They do feel a bit stuck in the current models of pricing and coverage and negotiation, and they are looking for ways by which they can achieve their mutual goals. I think ISP will get a lot of close looks, although it’ll be a lot harder than many people think when they first started out looking at it, and that is one of the purposes of our paper. 

		Pearson SD, Dreitlein B, Henshall C. Indication-specific pricing of pharmaceuticals in the United States health care system. Institute for Clinical and Economic Review website. 

http://icer-review.org/wp-content/uploads/2015/03/Final-Report-2015-ICER-Policy-Summit-on-Indication-specific-Pricing-March-2016_revised-icons-002.pdf

. Published March 2016. Accessed April 4, 2016.

In this interview, Steven D. Pearson, MD, MSc, FRCP, founder and president of the Institute for Clinical and Economic Review, talks about the various value tools that have been developed, the need for understanding the budgetary impact of healthcare costs, and why stakeholders are interested in innovative models such as indication-based pricing.

Q&A With ICER's Steven D. Pearson

The American Society of Clinical Oncology (ASCO)’s Value in Cancer Care Task Force has carved a conceptual framework for new cancer treatments, evaluating each one for clinical benefit, toxicity, and cost compared with available standards of care. Published in the 

 the framework will form the basis of a user-friendly tool for use by physicians as they discuss new treatments with their patients. ASCO’s CMO, Richard Schilsky, MD, FACP, FASCO, spoke with Surabhi Dangi-Garimella, PhD, managing editor of 

 on how the organization plans to utilize the feedback received to improve the Value Framework, so physicians can use it for clinical decisions at the point of care.

 Who did ASCO invite to participate on the Value in Cancer Care

 The development of the framework was quite a lengthy process. ASCO convened a multidisciplinary group to participate on the Value in Cancer Care Task Force, which included not only oncologists but also patient advocates and representatives from the pharmaceutical industry, payers, and experts in health economics. During multiple meetings, the Task Force drew up the blueprint for the framework, the initial draft of which was published last June, and we are making revisions to that version as we move forward.

 There are several different value frameworks that providers and payers can use, especially in oncology. What distinguishes ASCO’s Value Framework from the National Comprehensive Cancer Network (NCCN) Evidence Blocks and the Institute for Clinical and Economic Review (ICER) Value Assessment Framework?

 From what I know about the landscape, all of these value frameworks—including NCCN, ESMO [European Society for Medical Oncology], DrugAbacus&mdash;they are all evaluating the same information. Some of them have proposed to evaluate clinical benefit or toxicity or cost, but all the information that we used for value assessment comes from published clinical trial literature. I think that’s the big difference from the NCCN approach, which, of course, uses clinical trial data but also considers expert opinion and consensus to define their Evidence Blocks. Our framework, on the other hand, is focused around discussions between doctors and patients.

DrugAbacus aims to help payers and others to establish the best price of a drug based on what its efficacy is; but that is not at all a consideration of the ASCO framework. ICER has used components that are also related to affordability to the healthcare system. While this is a point of consideration, it’s not really the focus of the ASCO Value Framework—our focus is on providing guidance to doctors for discussions with patients. While there are many similarities [with the other frameworks], there are also some important differences.

 Does the Value Framework consider differences between acute versus chronic disease management? 

 The ASCO framework does not really characterize it in those terms, but we do have 2 different versions of the framework: there’s one for the advanced setting, for the treatment of patients with metastatic disease where cure is not expected to be the outcome, and another version for use in the adjuvant setting, where we are talking about the potential for cure. In either case, the framework can account for what the duration of treatment is, whether it’s a short-term treatment, as can occur in the adjuvant setting, or whether its chronic therapy that might need to be administered over years for the management of someone who has advanced cancer. So the framework can account for either of those cases.

 How about rare versus more common cancers. Are there different considerations under the framework? 

 Not really. Our framework is really applicable in any kind of a setting; [however,] it works best in a scenario where there has been a prospective randomized trial. Sometimes, in the very rare tumor types, it’s just not possible to conduct a randomized controlled trial because there are not a lot of patients [to be able to conduct such a trial]. However, even in that situation, we are able to use outcomes like response rate that might be reported in a similar trial, as a clinically relevant endpoint in the Value Framework.

 Have patient-reported outcomes (PROs) been considered when deriving the Net Health Benefit score for ASCO’s framework? 

 The initial version of the framework that was published in June—in the advanced disease setting, when patients are likely to be symptomatic&mdash;does provide bonus points if a new treatment produces significant reduction in cancer-related symptoms. In the revised version of the framework, where we will continue to provide bonus points for symptom palliation, we will also provide separate bonus points for improvement in quality of life. We have not specifically included other PROs in the framework at this point, largely because it’s still quite rare that those are actually captured or reported in clinical trial data.

 Can you speak to physician awareness on the existence of these value frameworks? What has been the feedback from providers and other stakeholders on the ASCO framework?

What we have so far is a “conceptual framework.” But what has been published is not really practical for use at the point of care. What we are planning to do later this year is to develop a software application that incorporates all the elements of the Value Framework into an easy-to-use software that can be used on a laptop or a tablet at the point of care, where doctors can actually use it to engage patients in a conversation [around value]. Once we have that ready to go, we will push that out quite aggressively so it gains widespread adoption.

 has promised to have a significant impact on the clinical decision-making process in oncology. Are there plans to weave the 2 tools together? 

I think these things will converge over time. One of the limitations of the Value Framework is that there is a lot of information that doctors and patients would like to know that is not actually available in the medical literature—there are significant evidence gaps&mdash;and we expect CancerLinQ to be able to fill in those evidence gaps. In particular, the limitation that we have identified with the Value Framework is that it doesn’t allow doctors to discuss regimens that have not been directly compared to each other in a clinical trial. We hope that CancerLinQ data can be used for that purpose.

 Are there ethical considerations when oncologists participate in the cost/value discussion? Should oncologists be advocates of the societal cost of care?

Oncologists always want to do what is best for their patient. There’s always been concern about oncologists having financial incentives for prescribing more expensive treatments. But a lot of that is going away, considering many of the new cancer drugs are now given orally. Secondly, the way in which physician reimbursement is moving—with the recent Medicare proposal that changes how physicians will be reimbursed for infused drugs

—there will be substantial reduction in physician reimbursement. The entire field is now placing a much greater emphasis on quality and value as a basis of reimbursement. I think tools like the Value Framework will play an important role in helping guide doctors to make valuable clinical decisions.

We have received more than 400 comments during the public comment period following the publication of our framework, and they have helped refine some of the ways in which we assess clinical benefit and toxicity in the framework. We’ll be publishing those soon. This will also contribute to the software application that we plan to build and release later on this year. Once the software is in the hands of treating physicians, it will give us a more realistic picture of how useful it turns out to be. The framework is intended to be a moving document that’ll need to be modified with time, and the Net Health Benefit score will be continuously reassessed with respect to its value, because value can change with different scenarios. We recognize that value assessment is not static, and it’s something that we will continuously monitor. 

		Dangi-Garimella S. ASCO's conceptual value framework released: tool for shared decision making. 

http://www.ajmc.com/newsroom/ascos-conceptual-value-framework-released-tool-for-shared-decision-making

. Published June 22, 2015. Accessed April 28, 2016.

		Caffrey M. ASCO announces partner for groundbreaking CancerLinQ platform. 

http://www.ajmc.com/focus-of-the-week/0115/asco-announces-partner-for-groundbreaking-cancerlinq-platform

. Published January 21, 2015. Accessed April 28, 2016.

		Dangi-Garimella S. ASCO, COA unhappy with proposed CMS Part B reimbursement model. 

http://www.ajmc.com/newsroom/asco-coa-unhappy-with-proposed-cms-part-b-reimbursement-model

. Published March 9, 2016. Accessed April 28, 2016.

The American Society of Clinical Oncology (ASCO)’s Value Framework evaluates the clinical benefit, toxicity, and cost of new interventions compared with the standard of care. ASCO’s CMO, Richard Schilsky, MD, FACP, FASCO, explains their plans to make the framework useful for clinical decisions at the point of care.

In Conversation With ASCO's Richard Schilsky

CMS and commercial insurers are increasingly paying for medical care through alternative payment models (APMs) that reward “value”.

 These APMs include accountable care organizations, bundled payments, “pathway” programs with physician risk and reward, and advanced medical homes, as well as tying fee-for-service (FFS) payments to quality metrics.

Outpatient cancer care is an important target for these efforts: the 2015 Medicare Access and CHIP Reauthorization Act of 2015 legislation made the move to APMs in outpatient oncology care inevitable, tying FFS Medicare payments to the reporting and achievement of specific quality targets. In 2015, the Center for Medicare & Medicaid Innovation (CMMI) announced its oncology care model (OCM)

 The OCM incorporates bundled payment with a shared savings program based on spending for all care provided to cancer patients upon the initiation of chemotherapy, including chemotherapy and supportive care drugs, day surgeries, diagnostic tests, emergency department visits, and inpatient stays. In March 2016, CMMI announced its intention to pilot additional APMs in outpatient oncology. The interventions include reducing physician reimbursement for the use of expensive chemotherapy and providing tools for physicians to assess cancer drug selection based on measures of safety, efficacy, effectiveness, and price.

Adoption of value-based outpatient oncology reimbursement by CMS is appealing. Under these reforms, services that used to generate revenue for physicians become costs they need to manage. In the more structured bundled or medical home models, payers reimburse defined payments for the care of each patient for a set period of time (an “episode”) and physicians have the freedom to treat patients as they choose. The payers benefit by moving away from existing policies that reward physicians for doing and billing more. In essence, APMs act to transfer some risks associated with high levels of cancer care spending from insurers to physicians or physician based organizations.

Yet, these policies also raise some challenges. In this perspective, I briefly review 2 vexing concerns with value-based outpatient cancer treatment.

Will High Drug Prices Be Tamed by Bundled or Pathway-based Payments? 

One driving rationale for APMs in outpatient oncology care is the high and rising spending on prescription drugs used to treat cancer and its symptoms.

 Stakeholders point to the fact that both the ever-growing prices of these drugs and the extent of their overuse likely account for a nontrivial amount of high spending levels and trends.

 Are bundled payments that include prescription drugs the answer to taming drug-related spending? The answer lies with whether one believes drug prices are a technical or probability risk to physicians or physician based organizations:

 This is the classic form of risk individuals buy insurance to mitigate. It is characterized by the occurrence of unpredictable events. In this context, probability risk is related to the spending required to treat a patient developing a particular cancer and seeking treatment from a specific physician. These are risks that lie outside a physician’s control.

This risk is related to the treating physician’s clinical skill and efficient management of the patient’s cancer, including diagnosis and treatment. These risks lie under the physician’s control.

Basic insurance theory suggests APMs are a means by which insurers should seek to transfer the technical risks associated with spending to physicians while retaining the management of probability risks to ensure patient access to care. What components of spending risk can physicians control? Clearly, physicians have some control over how many lines of therapy they use, which regimens they choose, and how creative they get in applying drugs to treat cancers off-label. Consequently, the “use” aspect of drug spending should be considered a technical risk. The promise of APMs is that the use—eliminated by transferring spending risk to physicians&mdash;will be “waste” and thus its elimination will not entail any harm to patients.

Identifying which aspects of use are overuse and misuse is difficult, particularly when limited treatment options are available for deadly cancers or in periods of rapid technological innovation and diffusion. One worry is that APMs that transfer substantial risk to physicians to control use may entail significant risks to patients in the form of enticing physicians to skimp on care or shun specific types of patients altogether.

 For these reasons, many stakeholders believe the use of pathways for cancers exhibiting these characteristics are preferred APM models at this time.

With respect to drug prices being a technical or probability risk, the story of ziv-aflibercept’s (Zaltrap) US market launch in August 2012 is helpful to keep in mind.

 Soon after Sanofi Pharmaceuticals Inc launched Zaltrap, physicians from Memorial Sloan-Kettering Cancer Center wrote in a 

 op-ed piece that they wouldn’t prescribe the drug because its list price ($11,407 per person per month of treatment in 2014) was twice as much as Genentech’s bevacizumab (Avastin), a drug with similar expected outcomes for colorectal cancer patients. In response, Sanofi said they would reduce the price of the drug by 50%.

Will physicians extend this bargaining power to other drugs’ prices under APMs? Yes, perhaps in selected cancers where innovation has created multiple branded and/or generic options that provide equivalent therapeutic benefits for patients. The cancers that have benefited from this type of recent treatment availability include some blood cancers, breast cancer, colorectal cancer, and lung cancer.

 Here the application of bundled payments or medical home will likely empower physicians to seek more substantial discounts and rebates from manufacturers, mitigating price growth.  

However, I have some reasons to be cautious about the net effect of this leverage, since spending risks associated with drug prices for truly novel drugs appear to entail probability risk. Despite public outcry,

 manufacturers of novel cancer drugs have not pursued moderating pricing strategies. Indeed, physicians already have some leverage over the acquisition prices of generics and drugs with competitors in a therapeutic class. Their leverage comes from 340B drug discounts limited to the entities that qualify for the program and/or the volume of drug purchasing they are pursuing in a given period of time by their group purchasing organizations.

 Only some of these discounts are passed onto insurers or to patients under current payment policies. Nothing about the use of APMs will alter these arrangements, keeping such drug revenues locked in the coffers of physician groups, hospitals and other third parties. Furthermore, physicians have control over the prices paid for novel and existing oral drugs covered under cancer patients’ pharmacy benefits. These prices are negotiated by payers or increasingly by pharmacy benefit managers.

 Therefore, taming the prices of truly novel cancer drugs, particularly oral cancer drugs, will likely require other policy solutions.

The key tool to mitigating patient and physician concerns regarding APM-related stinting challenges and moving us toward innovative pricing policies for prescription drugs to treat cancer (such as risk-sharing agreements between insurers and pharmaceutical manufacturers) are valid, reliable, comparable, and time-sensitive outcomes data. Much-touted efforts are underway by the American Society of Clinical Oncology and US Oncology, among others, to expand metrics for assessing cancer treatment quality delivered in the outpatient setting across many more cancers and also for different outcome measures for comparative purposes. This flurry of activity belies a simple fact: in cancer, there is no settled-upon methodology for measuring outcomes and comparing outcomes between institutions, care settings, or physicians.

Quality measurement in this area is currently stymied for at least 3 reasons:

		First, modeling prognostic outcome correlates requires consensus on treatment for a given cancer patient, inclusive of state and possible line of therapy. However, the pace of technological change in both cancer diagnoses and treatment options hampers this consensus building.

		Second, like other areas of medicine, patient selection and treatment intensity bias outcome measurement. Risk adjustment of these measures as they relate to outpatient cancer care requires an analyst to observe patient cancer stage, line of therapy, and the presence and outcomes of diagnostic tests from claims data. FFS Medicare claims do not currently provide this information, inhibiting progress on risk-adjusted comparative outcome development.

		Finally, the construction of comparative outcome measures presumes identifiable and unique physicians. Yet, the task of defining unique physicians is increasingly problematic, particularly in the delivery of cancer care where merger and acquisition activities between physicians, community oncology groups, hospitals, and other arrangements have substantially increased. There is also a flurry of cancer care center “affiliations” occurring nationwide. These activities can cause a dramatic shift in the patterns of cancer care in a limited time frame, making it hard to assign treatment to a given physician.

Affiliation and consolidation activities among physicians, oncology groups and/or hospitals can also compound selection and treatment biases in the construction of comparative cancer measures. The extent of the potential biases that such activities introduce to these metrics is intimately related to the underlying economic and scientific rationales driving these activities. We can presume that these arrangements provide participants with several advantages, all impacting patient selection into treatment and/or the treatment patients receive, including the syndication of treatment pathways for the care of specific cancers, the creation of closer ties between institutions for patient referrals, access to shared physician time and expertise, improved bargaining power with insurers, and more control over intra-institutional prices.

I am unaware of any ongoing efforts by CMS to systematically track and characterize these relationships for quality measurement and reimbursement under the existing APM policies. New data collection efforts are likely required to systematically track consolidation and affiliation activity, and its impact on patient flow across settings. This is an important area for future research. 

Rena M. Conti, PhD, is assistant professor of health policy and economics at The University of Chicago.

Departments of Pediatrics Hematology/Oncology and Health Studies

: The author has no relevant or material financial interests that relate to the research described herein.

		Cutler DM. Payment reform is about to become a reality. 

2015;313(16):1606-1607. doi:10.1001/jama.2015.1926.

		Press MJ, Rajkumar R, Conway PH. Medicare’s new bundled payments: design, strategy, and evolution. 

2016;315(2):131-132. doi:10.1001/jama.2015.18161.

		Newcomer LN, Gould B, Page RD, Donelan SA, Perkins M. Changing physician incentives for affordable, quality cancer care: results of an episode payment model. 

. 2014;10(5):322-326. doi:10.1200/JOP.2014.001488.

https://innovation.cms.gov/initiatives/oncology-care/

. Updated March 3, 2016. Accessed April 12, 2016.

		CMS proposes to test new Medicare Part B prescription drug models to improve quality of care and deliver better value for Medicare beneficiaries [press release]. Baltimore, MD: CMS; March 8, 2016. 

https://www.cms.gov/Newsroom/MediaReleaseDatabase/Press-releases/2016-Press-releases-items/2016-03-08.html

		Bach PB, Mirkin JN, Luke JL. Episode-based payment for cancer care: a proposed pilot for Medicare. 

. 2011;30(3):500-509. doi:10.1377/hlthaff.2010.0752.

		Bekelman JE, Kim M, Emanuel EJ. Toward accountable cancer care. 

. 2013;173(11):958-959. doi:10.1001/jamainternmed.2013.635.

		Howard DH, Bach PB, Berndt ER, Conti RM. Pricing in the market for anticancer drugs. 

. 2015;29(1,Winter):139-162. https://www.aeaweb.org/articles?id=10.1257/jep.29.1.139.

		MD Anderson, UnitedHealthcare launch new cancer care payment model [press release]. Houston, TX: The University of Texas MD Anderson Cancer Center; December 16, 2014. 

http://www.mdanderson.org/newsroom/news-releases/2014/md-anderson-unitedhealthcare-new-cancer-payment.html

		Conti RM, Berndt ER. Winners and losers of the Zaltrap price discount. 

http://healthaffairs.org/blog/2013/02/20/winners-and-losers-from-the-zaltrap-price-discount-unintended-consequences/

. Published February 20, 2013. Accessed April 12, 2016.

		Conti RM, WV Padula, RA Larson. Changing the cost of care for chronic myeloid leukemia: the availability of generic imatinib in the UAS and the EU. 

. 2015;94(Suppl 2):S249-S257. doi:10.1007/s00277-015-2319-x.

		Bach P. Part B payment for Medicare drugs. Memorial Sloan Kettering Cancer Center website. 

http://www.drugabacus.org/part-b-analysis/

		Experts in Chronic Myeloid Leukemia. The price of drugs for chronic myeloid leukemia (CML) is a reflection of the unsustainable prices of cancer drugs: From the perspective of a large group of CML experts. 

. 2013;121(22):4439-4442. doi:10.1182/blood-2013-03-490003.

		Tefferi A, Kantarjian H, Rajkumar SV, et al. In support of a patient-driven initiative and petition to lower the high price of cancer drugs. 

. 2015;90(8):996-1000. doi:10.1016/j.mayocp.2015.06.001.

		Siddiqui M, Rajkumar SV. The high cost of cancer drugs and what we can do about it. 

. 2012;87(10):935-943. doi:10.1016/j.mayocp.2012.07.007.

		Conti RM, Bach PB. The 340B drug discount program: hospitals generate profits by expanding to reach more affluent communities. 

 2014;33(10):1786-1792. doi:10.1377/hlthaff.2014.0540.

		Polite BN, Ward JC, Cox JV, et al. Payment for oncolytics in the United States: a history of buy and bill and proposals for reform. 

. 2014;10(6):357-366. doi:10.1200/JOP.2014.001958.

Measuring the quality of oncology care and associating it with reimbursement, and high drug prices remain important concerns of value-based outpatient cancer care. A healthcare economist reviews the current status and suggests a potential path forward.

Value in Cancer Care: An Economist's Perspective

Peter Bach, MD, MAPP, director, Center for Health and Policy Research, Memorial Sloan Kettering Cancer Center, has been a vocal proponent of appropriate drug pricing, particularly of anticancer agents. To this end, the Center developed a tool, DrugAbacus, which uses various measures such as efficacy, toxicity, and innovation to identify an appropriate price for an approved drug.

’s managing editor, Surabhi Dangi-Garimella, PhD, Bach spoke about DrugAbacus, the importance of using value frameworks, and using the European market as a model to recalibrate the healthcare system in the United States.

 What is the importance of the value discussion in oncology?

 I think we are having the discussion of value across all domains of healthcare. In the United States, we spend more per person on healthcare than in any other country. Even after adjusting for any measures of economic productivity, we are still outspending any other country. And, it’s impossible to convince yourself that we are getting any incremental value, period, let alone any incremental value of the excess spending. If we compare ourselves with other countries, there does not seem to be any rationale, except that we simply spend more for all units of healthcare.

If you try to ask if these prices align with what matters to patients, to science, or to the society as a whole, you find the prices are disassociated from any metric of value. It makes sense in a sector, especially on the drug side, where there is government-issued monopoly, where there is unrestrained pricing power, and there is mandatory buying across all payer sectors and across all cancer drugs. We are now in a situation where a year of therapy with cancer drugs costs 3 times the median income in this country, and so it makes sense to query the high prices and try to rationalize them.

Several organizations, such as ASCO [American Society of Clinical Oncology], NCCN [National Comprehensive Cancer Network], ICER [Institute for Clinical and Economic Review], ESMO [European Society for Medical Oncology], the American Heart Association, the American College of Cardiology, along with ours, are honing in on the question, ‘could we start linking therapies to their value?’

So, the purpose of the DrugAbacus is to ask how this would work. What would the equation be? This just moves us toward health technology assessment that is being pursued or already implemented in every other Western country. This should have been instituted in the United States as well. Absent a new framework, the options include a mandatory price control, which is unattractive, or we find a balance between how we pay for therapies and how accessible they are. The other problem we face in this country is that insurers have very limited means of managing drug spending, and the way they do it is by impeding access to treatments. What I am suggesting is a value framework for finding the appropriate prices, so that when those prices are found and enabled or accepted by manufacturers, insurers respond, in turn, by ensuring there is appropriate access, because the debate about access and coinsurance is, in reality, a debate about mismatches between value and price.

 Does your Center plan to expand beyond drugs prices and evaluate the value of a treatment pathway?

 The idea of indication-specific pricing can address regimen-level questions. DrugAbacus purports to find appropriate launch prices for drugs based on available data. Regimens collectively cost something as well, but we are trying to understand the incremental benefit that a newly approved drug provides and whether it is priced accordingly. The approval is usually based on a comparison of monotherapy versus standard of care, or standard of care with or without the new drug. The other scenario is of stacking a very expensive new agent on top of or following or preceding, what is usually a cheaper generic treatment. So, from a mathematical perspective, [computing the value of the regimen] may not be very critical.

 Improving healthcare value would need strategies to reduce spending and also reduce waste, as your recent paper in 

 emphasized. In your opinion, what approaches do we need to significantly improve the delivery of healthcare in the United States?

 A part of saving money is reducing the unit prices of items in healthcare, and most analysts will tell you that one of the reasons we spend more money on healthcare, per person, than Germany does, even though they cover all of their citizens and have outcomes that are as good or better than ours, and the answer is unit price. It’s not the volume of service. We may do a few more procedures per person than they do, but on balance, they have more days in hospital and more utilization of services than we do.

The reality is that the drivers of healthcare spending in the US are unit price, not volume. So we need to rationalize the price of services, not just for drugs—we pay far more for each day in the hospital; we pay our healthcare providers more money, and some of that can be adjusted by how we cost shift. The way we finance medical school education, for example, is through physician salaries. Europeans tend to directly fund their medical schools, so it’s free or not very expensive. Most of this is driven by market power, so physicians get paid more because they can extract monopoly rents by gaining control over their local markets and they extract higher wages from private insurers, something they cannot do as easily in European markets.

That’s why we are focused on drug prices. We are paying 2-to-6 times the European rates for many of these drugs. When we analyzed Pfizer’s portfolio—just their top 10 drugs&mdash;to measure the impact of Pfizer leaving the United States and paying taxes in Ireland instead,

 if we pay the prices that the Irish pay for Pfizer’s products, we found that it would save us somewhere around $6 billion. [The deal between Pfizer and Allergan has, however, fallen through following new inversion regulations by the US Treasury department.

So, if we want to manage healthcare spending, you can make the argument that the excess price above that, which is being paid ambiently in European countries, is actually wasted dollars. It does not make the pill any better by paying 4 times as much for it.

Can you speak about the recent grant that your Center received from the Arnold Foundation? How do you expect to use these funds?

 We will continue to work on value frameworks for drug prices, continue to do regulatory and legislative analysis, and actively engage with state and federal government programs, as well as many of the private payers and large companies, trying to advance value frameworks for drugs. The primary goal is to provide proof of concept that it is possible to demonstrate that we could actually price things more appropriately that continues to spur innovation, while providing better access to patients. I think it is well within reach.

I think this is the right time to have proof of concept and start to roll out programs that could start tying drugs to their value, and that’s what the grant is intended to support. 

		Bach PB, Conti RM, Muller RJ, Schnorr GC, Saltz LB. Overspending driven by oversized single dose vials of cancer drugs. 

		Pfizer would cut its corporate tax bill if it merges with Allergan. NPR website. 

http://www.npr.org/2015/11/02/453885663/pfizer-would-cut-its-corporate-tax-bill-if-it-merges-with-allergan

. Published November 2, 2015. Accessed April 27, 2016.

		Humer C and Banerjee A. Pfizer, Allergan scrap $160 billion deal after U.S. tax rule change. Reuters website. 

http://www.reuters.com/article/us-allergan-m-a-pfizer-idUSKCN0X3188

. Published April 6, 2016. Accessed April 27, 2016.

Peter Bach, MD, MAPP, director, Center for Health and Policy Research, Memorial Sloan Kettering Cancer Center, speaks about DrugAbacus, the importance of using value frameworks, and using the European market as a model to recalibrate the healthcare system in the United States.

Peer Exchange: Spring 2016 Oncology Stakeholders Summit

Considerations for Designing "Value Calculators" for Oncology Therapies

Q&A With ICER's Steven D. Pearson

In Conversation With ASCO's Richard Schilsky

Value in Cancer Care: An Economist's Perspective

A Q&A With Peter Bach