Issue: May 25, 2016
May 25, 2016
15 min read

Value in cancer care must ‘start with patients, not dollars’

Issue: May 25, 2016
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact

The skyrocketing cost of cancer care has come under intense criticism from patients and health care providers.

Several promising treatments — including traditional biologics and more novel approaches, such as immunotherapy and gene therapy — have demonstrated remarkable efficacy.

Yet, newly approved cancer drugs cost an average of $10,000 per month, and the price for some exceeds $30,000 per month.

These unprecedented price tags have monopolized conversations about value in cancer care. However, other nonmonetary factors — particularly the outcomes component of care — hold at least as much weight in the value equation.

“There are financial outcomes and measurements, as well as clinical outcomes, that are very much a part of the conversations doctors are having with patients,” Gary H. Lyman, MD, MPH, co-director of the Hutchinson Institute for Cancer Outcomes Research at Fred Hutchinson Cancer Research Center, professor of medicine at University of Washington School of Medicine and medical oncologist in the breast cancer program at Seattle Cancer Care Alliance, told HemOnc Today. “We are continuing to try to do a better job capturing cost, but we are also trying to capture value through outcomes such as survival, quality of life and treatment experience.”

The National Comprehensive Cancer Network, ASCO, ASH, the American Society for Radiation Oncology and other medical societies have formed task forces to determine which outcomes can best be used to create a framework for high-quality, high-value medical care.

Advances in information technology infrastructure — such as electronic health records and the launch of ASCO’s CancerLinQ database — offer the potential to house outcomes data to centralize cancer treatment and establish national benchmarking.

“The only way to integrate cost and financial impact with outcomes and quality of care is to pull it all together on a large scale,” Lyman said. “It is going to take real time, but that is the direction we are headed.”

HemOnc Today spoke with oncologists, researchers and representatives from professional societies about how the quest to provide value in cancer care must extend beyond expenditures alone, the role that clinical and patient-reported outcomes of oncology trials can play in establishing a value framework, and what clinicians can do in their practices to ensure their patients receive top-quality care at a reasonable cost.

Focus on cost

Cost and outcomes are inextricably linked.

The price Americans with cancer pay to achieve better outcomes has increased dramatically over time. An analysis by Howard and colleagues, published last year in Journal of Economic Perspectives, showed the inflation-adjusted cost for each additional year lived increased from $54,000 in 1995 to $207,000 in 2013.

This trend can have devastating consequences.

A survey by Zafar and colleagues showed costs prompted more than half of patients with cancer to limit treatment adherence. Twenty percent of patients took less than their prescribed amount of medication, 19% partially filled prescriptions and 24% avoided filling prescriptions.

The financial toll also is well documented. Cancer is the leading cause of personal bankruptcy in the United States, and a study by Ramsey and colleagues — published in Health Affairs — showed patients with cancer are 2.65 times more likely to go bankrupt than matched healthy controls.

“As outcomes are improving, we have to get a handle on costs,” Thomas W. Feeley, MD, head of the Institute for Cancer Care Innovation at The University of Texas MD Anderson Cancer Center and senior fellow at the Institute for Strategy and Competitiveness at Harvard Business School, told HemOnc Today. “Value is the balance between the outcomes our patients achieve and the cost to deliver those outcomes. When the cost is too high, the value is reduced.”

Similarly, when costs increase, outcomes do not necessarily improve.

Soneji and Yang evaluated cancer expenditures and mortality rates from 1982 to 2010 in the United States and 20 Western European countries. Their results, published last year in Health Affairs, showed the United States had 729,000 excess breast, colorectal, prostate and lung cancer deaths compared with European countries, despite spending $1.6 trillion more on cancer care.


The ratios of incremental cost to quality-adjusted life-years saved — $1.97 million for prostate cancer, $402,000 for breast cancer and $110,000 for colorectal cancer — “exceed most accepted thresholds for cost-effective medical care,” Soneji and Yang wrote.

The researchers also reported 1.12 million excess lung cancer deaths in the United States, even though lung cancer expenditures in the country were $406 billion higher than in Europe. Each additional $19,000 spent on lung cancer in the United States equated to one quality-adjusted life-year lost.

An open, honest discussion between oncologists and their patients about efficacy, toxicity and costs is necessary to ensure value, according to Lowell E. Schnipper, MD, professor of medicine at Harvard Medical School, clinical director and chief of hematology and oncology at Beth Israel Deaconess Medical Center, and a HemOnc Today Editorial Board member.

Lowell E. Schnipper

Schnipper chairs ASCO’s Value in Cancer Care Task Force. The panel — previously called the Task Force on Cost in Cancer Care — was renamed to show money is not the only metric by which value is measured.

In this spirit, the task force last year released a conceptual framework to measure the value of a cancer treatment based on its “net health benefit” score, which accounts for clinical benefit and toxicity.

“It is most important to relate what our patients are getting in terms of clinical benefit to the cost of medications,” Schnipper said. “We are seeing substantial costs in terms of price per month for many new anticancer drugs, which are by no means necessarily curative. It is our obligation to find a metric agreeable to oncologists that represents the clinical impact of a new treatment or combination of treatments that can then be presented as options to the patient. That way, a patient will know what a therapy costs, but also how much or how little it is likely to benefit them.”

Patients before dollars

Conversations about value often are focused on cost — or how costs affect outcomes — but patients’ treatment experiences should play a role in the equation, experts said.

“The discussion often begins with cost, but I am not sure that is the right place to start,” Joseph Alvarnas, MD, clinical associate professor of hematology and hematopoietic cell transplantation at City of Hope, as well as chair of ASH’s committee on practice, told HemOnc Today. “Our discussion of value has to begin with patient-centered outcomes. We do a good job of describing costs, but we do a very bad job of describing outcomes. When we talk about value, we need to start with patients, not dollars.”

Maximizing and coordinating resources can help ensure value for patients, according to Joseph Alvarnas, MD. “If we really consciously think about how we deliver care, and do so in a way that maximizes alignment around the needs of patients and their families, we’ll deliver far better, more sustainable health care,” Alvarnas said.

Photo by Fred Lee.

Researchers and clinicians often view OS as the holy-grail endpoint for randomized controlled oncology trials. However, there has been a push to identify other outcomes that indicate clinical value.

A study by Kim and Prasad, published in JAMA Internal Medicine, showed 67% of cancer drugs approved by the FDA between 2008 and 2012 achieved their approval on the basis of surrogate endpoints and had not demonstrated an OS benefit at the time of approval.

Response rate served as the basis for approval in 53% of these cases. PFS or DFS were used as surrogates in 47% of cases.

“Survival is the most important endpoint for many patients, but it is not the only endpoint,” Lyman said. “We are converting many cancers into chronic conditions. Quality of life and the experience of cancer treatment — particularly in a cancer that is not going to be eradicated — are becoming more important.”


Patients with widespread disease might prefer to focus on DFS, quality of remaining life, and management of symptoms or psychosocial distress, according to James Helstrom, MD, MBA, chief medical officer and interim chair of anesthesiology at Fox Chase Cancer Center.

“The presumption is that developing measures that incorporate all the points on the decision-making compass will empower patients, who will sometimes choose quality over quantity,” Helstrom said. “These measures can help avoid the big cost drivers — that is, hospital and ICU stays — at end of life.”

Patients’ preferences may differ from what clinicians expect.

“If you talk to patients and their families about what is important to them, what they articulate first and foremost tends to be things physicians do not necessarily think about,” Alvarnas said. “Patients want to be informed. They want to participate as decision-makers. When cure is possible, they want to move toward that in the most efficient, least toxic way. When a cure isn’t possible, then patients want to be able to plan in meaningful ways and actively participate in the strategy of their care delivery.”

Patient-reported outcomes (PROs) may hold the key to optimizing value-based care for patients with cancer. These subjective reports — given directly by the patient and not interpreted by medical professionals — have become a more common component of oncology trial reporting.

“Patient-reported outcomes contribute to the value proposition because they place the patient, rather than the treatment, front and center in the discussion,” Helstrom said. “Presumably, treatment goals are established after careful discussion at the outset of the patient–physician relationship. Depending on how they are collected, patient-reported outcomes can give valuable feedback and serve as a check to make sure that everyone on the care team remains on the same page.”

Patients measure value in a number of quantifiable ways that can be captured and implemented, Alvarnas said.

“Returning to work, the timeliness of recovery and functionality over time are all immensely important to patients,” he said. “These are factors that should be reported far more regularly in clinical trials. Patient-reported outcomes are something that we have underutilized, and it would be very important to standardize tools for measuring distress, depression and pain.”

Pros and cons of PROs

Strategies for the proper implementation and reporting of PROs remain ongoing.

In 2004, the NIH initiated the Patient Reported Outcomes Measurement Information System (PROMIS), designed to allow for the easy collection of physical, mental and social patient-reported outcomes. The initiative is based on the belief that PROs can capture treatment benchmarks that are not quantifiable in clinical factors, and that a better understanding of these outcome measures will lead to high-value care.

In 2015, the Oncology Nursing Society launched the ONS Quality Improvement Registry, a data registry used to benchmark and improve patient outcomes and improve overall cancer care.

PROMIS and the ONS Quality Improvement Registry represent much-needed efforts to improve the tools used to measure PROs, Feeley said.

“In some studies we [at MD Anderson] conducted in a head and neck cancer population, we learned that some of the outcome metrics that were very important to patients — such as speaking or swallowing — were not being captured in the record,” Feeley said. “We started using tools so that our clinicians would have regular patient-reported outcomes as part of their decision-making process.”

However, barriers still exist with regard to the full implementation of PROs.

“Not every clinician is using them,” Feeley said. “Not every IT system easily collects them.”

Standardization in the reporting of PROs is necessary to best use these measures to assess drug efficacy, Schnipper said.

“The biggest limitation [of patient-reported outcomes] is that many clinical trials don’t report them in a quantifiable, scientifically organized way that makes them credible,” he said. “If you don’t have quantifiable data, you cannot make a value assessment. If we have a framework, it needs to include credible, scientifically driven data.”


This is further challenged because the collection of patient-reported data is often ongoing. Quality-of-life and PRO data frequently are published after survival and efficacy data.

“If clinical benefit persists and we learn that systems are much improved, that quality of life is better, then a subsequent analysis would show that treatment has improved the net health benefit,” Schnipper said. “When we juxtapose that against cost, the value would actually be perceived as being greater.”

Despite these potential benefits, the use of PROs has not been carried out through drug approval.

Gnanasakthy and colleagues conducted a retrospective review of new therapies approved by the FDA’s Office of Hematology and Oncology Products between 2010 and 2014 that showed of 40 approved drugs, only three received PRO–related labeling.

“If we are going to be making broad decisions based on patient-reported outcomes, we have to ensure that the outcome measures are validated very carefully in a way that maximizes the alignment of the care delivery team with the needs of patients and their families,” Alvarnas said.

Value via technology

Access to relevant data that can be used to achieve benchmarks and improve patient care could hold the key to optimal value-based health care.

The availability of macro-data allows health care facilities to function at a high level and provide competent, value-based care, Helstrom said.

“The collection, collation, consumption and analysis across an entire practice or institution is very important,” he said. “Being able to structure discrete collection from individual applications so that these data can be leveraged to address complex questions is necessary. If we want to understand 30-day readmission rates, we can pull data from electronic medical record coding, billing and case management to generate a dataset for analysis.”

Data collection matters on an individual patient level, too.

“It makes sure nothing falls through the cracks,” Helstrom said. “We tend to call these ‘tick-tock considerations,’ instances in which treatment or follow-up, like surveillance, must happen within a specific window or according to a schedule. Automating these are a huge advantage, as is the ability to perform routine oversight regarding pharmacy matters, like dosing.”

Tools that measure PROs can help track patients over time. For example, a clinician can be alerted if a patient reports dramatic changes in depressed symptomatology or suicidal ideation, Feeley said.

“However, there are a lot of barriers to doing this,” he said. “Not every clinician is comfortable using these measures, and we don’t have the IT systems that easily collect patient-reported outcomes. This is not to say it can’t be done, but you have to have a culture where clinicians are willing to use these tools. We are seeing organizations adopt these systems more broadly, and we have clinicians who find them extremely useful in their day-to-day patient care.”

Value can be generated from electronic systems if they can be applicable to real-world patients, Alvarnas said.

“In general, health care technology systems tend to add to our level of work and impede our work flow,” he said. “Part of our challenge as physicians has been not only to continue to serve our patients, but to do so while having to click more boxes and jump through more electronic hoops. It would be phenomenal if we could apply large data models to patients in the real world and abstract the data to learn more about the patients we serve.”

For instance, The University of Texas MD Anderson Cancer Center has created the Adaptive Patient-Oriented Longitudinal Learning and Optimization (APOLLO) program that houses large volumes of patient data. The goal is to use these data across disciplines to assist in diagnosis, prognosis and therapy selection.

Similarly, ASCO has introduced CancerLinQ, a health information technology platform designed to address these challenges and allow clinicians to access electronic health records in a rapid, streamlined manner.


CancerLinQ will help clinicians assess the benefits and harms associated with different treatment options, Lyman said.

“To balance benefits, harms and costs on a large scale, it needs to be integrated electronically and eventually into a way of informing clinical decisions, or the shared decisions between the doctor and the patient,” Lyman said. “If there are two or three treatment options that are appropriate based on guidelines, they are still likely to differ on the basis of longevity, toxicity and cost. This information needs to be at the doctor’s fingertips ... and this is only going to happen with an IT structure that can capture data automatically.”

CancerLinQ likely will have the most impact on community practices.

“Those who participate in clinical trials represent a highly selected population of patients who don’t generally have the other comorbid illnesses that might compromise their ability to receive full doses of the treatment, or to tolerate the side effects,” Schnipper said. “As a result, it’s quite possible that a regimen will perform extremely well as reported in The New England Journal of Medicine or Journal of Clinical Oncology, but it might not perform as well in a completely unselected group of patients. We can learn this from CancerLinQ, and we can identify whether CancerLinQ’s data are supporting a value assessment we’ve come to — or if it’s refuting it.”

The CancerLinQ rollout began late last year, when it was made available to 15 U.S. oncology practices, ranging from a six-person private firm to NCI–designated comprehensive cancer centers.

“This is the only way that cost and financial impact, integrated with outcomes and quality of care, can ever be truly pulled together on a large scale,” Lyman said.

Everyday value

How specific components of value — including the cost of drugs, what outcomes are evaluated in clinical trials, and harnessing technology to assess performance and create benchmarking — are implemented may be outside the hands of an individual oncologist.

Still, there are specific steps clinicians can take in their everyday practice to help patients achieve high-value outcomes.

One way to ensure value is to capitalize on resources, Alvarnas said.

“We are only one part of a team,” he said. “That team may include social workers, nurses, clinical care managers and nurse coordinators. Our typical mode of operation has been to work with these professionals in a Balkanized way where we are disconnected from the activities of the other team members.

“Patients benefit from an integrity in that alignment of health care professionals,” Alvarnas added. “If we really consciously think about how we deliver care, and do so in a way that maximizes alignment around the needs of patients and their families, we’ll deliver far better, more sustainable health care.”

Oncologists can take a closer look at the costs of treatments and tests, Feeley said.

“I would encourage oncologists to be a little more demanding of knowing what the costs of therapies are and what the patients can expect,” he said. “I also would suggest that oncologists start talking to patients about out-of-pocket costs — not only in the short term, but in the long term, because I’ve seen too many patients ultimately make decisions because of financial considerations that could have been addressed upfront.”

Many experts with whom HemOnc Today spoke emphasized that value begins on an individual patient level.

“You have to know the patient’s social background,” Schnipper said. “Are they employed or unemployed? Does their insurance arrangement impose a very substantial burden on them? Doctors need to know that so that they can say to a patient with recurrent breast cancer, ‘We can no longer use the hormone pill, which was an easy first step. Now we are moving into a phase where we really need to use chemotherapy.’”


Oncologists should know how treatment side effects will affect specific patients, Schnipper added.

“If one medicine causes neuropathy, a physician who knows his or her patient is a pianist or cellist may instead recommend a treatment that does not put at risk those things patients really love to do,” he said.

Patients should be at the forefront of their care discussions, Helstrom said.

“We should begin with a thorough conversation regarding treatment goals,” he said. “From there, we should constantly reassess patient preferences and have a low threshold to include supportive oncology and psychosocial support services.”

Applying guideline recommendations to individual patients can yield enormous value, Lyman said.

“Clinical judgment and the knowledge clinicians have of their patients always need to be woven in, and sometimes that is going to be in distinction to usual guideline recommendations,” Lyman said. “The platform on which to build should be the conversation with the patient, followed by the patient’s specific set of circumstances, such as their age, their comorbidities and their disease. With an integrated evidence base, we can modify the guidelines appropriately to do the best for the individual patient.” – by Cameron Kelsall

For differing views, see commentaries from:

  • Derek Raghavan, MD, PhD
  • Ethan Basch, MD
  • Deborah Watkins Bruner, RN, PhD, FAAN
  • Carolyn Gotay, PhD
  • Heidi D. Klepin, MD, MS
  • References:

    Alvarnas J, et al. JAMA Oncol. 2015;doi:10.1001/jamaoncol.2015.3449.

    Bennette CS, et al. Med Decis Making. 2016;doi:10.1177/0272989X16636847.

    Claassens L, et al. J Clin Oncol. 2011;doi:10.1200/JCO.2010.32.3686.

    Doward LC, et al. Health Qual Life Outcomes. 2010;doi:10.1186/1477-7525-8-117.

    Durkee BY, et al. J Clin Oncol. 2015;doi:10.1200/JCO.2015.62.9105.

    Efficace F, et al. Cancer. 2015;doi:10.1002/cncr.29489.

    Gnanasakthy A, et al. J Clin Oncol. 2016;doi:10.1200/JCO.2015.63.6480.

    Gross CP and Emmanuel EJ. JAMA Oncol. 2016;doi:10.1001/jamaoncol.2015.3706.

    Howard DH, et al. J Econ Perspect. 2015;doi:10.1257/jep.29.1.139.

    Jones GH, et al. Blood. 2016;doi:10.1182/blood-2015-11-680058.

    Kim C and Prasad V. JAMA Intern Med. 2015;doi:10.1001/jamainternmed.2015.5868.

    Ramsey SD, et al. Health Aff (Millwood). 2013;doi:10.1377/hlthaff.2012.1263.

    Ramsey SD, et al. J Clin Oncol. 2016;doi:10.1200/JCO.2015.64.6620.

    Schilsky RL, et al. J Clin Oncol. 2014;doi:10.1200/JCO.2014.56.2124.

    Schnipper LE, et al. J Clin Oncol. 2015;doi:10.1200/JCO.2015.61.6706.

    Soneji S and Yang J. Health Aff (Millwood). 2015;doi:10.1377/hlthaff.2014.0174.

    Tefferi A, et al. Mayo Clin Proc. 2015;doi:10.1016/j.mayocp.2015.06.001.

    Zafar SY, et al. Oncologist. 2013;doi:10.1634/theoncologist.2012-0279.

    Zagadailov E, et al. Am Health Drug Benefits. 2013 Jul;6(5):264-74.

    For more information:

    Joseph Alvarnas, MD, can be reached at

    Thomas W. Feeley, MD, can be reached at

    James Helstrom, MD, MBA, can be reached at

    Gary H. Lyman, MD, MPH, can be reached at

    Lowell E. Schnipper, MD, can be reached at

    Disclosure: Alvarnas, Feeley, Helstrom, Lyman and Schnipper report no relevant financial disclosures.