The Slow, Steady Battle To Fix Cancer Care

In In The News by Barbara Jacoby

By: Carmen Nobel


After a cancer patient’s first 30 days in the American medical system, the bills start stacking up—right next to the pile of paperwork explaining benefits. “It’s very hard for patients to match these things up,” says Dr. Thomas W. Feeley, who recently joined the Harvard Business School faculty as a senior fellow to study health-care reform.

Feeley, the Helen Shafer Fly Distinguished Professor of Anesthesiology at The University of Texas MD Anderson Cancer Center, is on a quest to make the payment process easier and, hopefully, less expensive. Last December, MD Anderson teamed up with insurance provider UnitedHealthcare to test a fixed-payment structure for a select group of head and neck cancer patients. Rather than receiving separate charges for every test, treatment, and appointment, patients in the three-year pilot will pay for their cancer care in a single lump sum, based on a holistic treatment plan comprising one of eight “bundles.”

“We’re hoping that this is something that at least takes that stress of being treated for cancer out of the patient experience,” says Feeley, who still serves as head of MD Anderson’s Institute for Cancer Care Innovation.

The bundles represent the treatments clinicians select for patients and take into account those patients with other chronic conditions. While the payment model undoubtedly simplifies billing matters for the patients, the main impetus is to lower health care costs while simultaneously improving quality of care.

In the traditional fee-for-service model, hospitals secure permission from the patient’s insurance company before performing costly tests such as CT scans. In the bundled payment model, the onus is on the doctors to choose the most cost-effective care bundle.

“The model lets doctors take the risk,” Feeley says. “The onus of responsibility is on us to decide on the most effective, least costly cycle of care. The theory is if we take on the responsibility, we’ll work to control our own costs. Health care is 17 percent of our gross domestic product. If we can drive down costs while improving outcomes, that will be much more favorable to our economy and to our competitiveness as a nation, especially if we can spend more on things like education. ”

The bundled payment pilot test is small—only 13 patients so far, possibly scaling to 150 by the end of the three-year trial. But it has made big waves in the health care industry, not only flying in the face of the common fee-for-service model, but also involving the biggest medical insurer in the United States; one of the top-ranked hospitals in the world; and a multiyear research project at HBS that aims to repair the American health-care system.

The bundled payment idea is part of a larger reform proposal called Value-Based Health Care Delivery, based on research by Michael E. Porter, the Bishop William Lawrence University Professor and founder of the Institute for Strategy and Competitiveness at HBS. “Value” in this case is defined as patient outcomes per dollar spent; the goal is to achieve the best possible outcome at the lowest possible cost. In contrast, today’s American health care system tends to emphasize billing for individual procedures, regardless of how the patient fares.

The Value-Based Agenda

Developed in 2006, the Value-Based Health Care Delivery strategy includes a seven-part agenda to restructure health care. Part one calls for creating “integrated practice units,” each organized around a particular medical condition. As it happens, MD Anderson has been organized that way since 1996. “Unlike traditional health care organizations, we’re organized around the type of cancer that the patient has,” says Feeley. “So if you come to our thoracic center with lung cancer, you can see a medical oncologist, a surgical oncologist, a radiation oncologist all in the same place—and all specialists who are focused on just lung cancer.”

MD Anderson’s integrated practice units became the subject of a 2008 business case, co-authored by Porter and Sachin H. Jain. Thus began a long-standing collaboration. After Porter wrote the case, “he came to the center and said, ‘OK, now we have to start collecting some data,’” Feeley says.

And so for the past seven years, Feeley has been working with HBS researchers to bring the value-based care agenda to life at MD Anderson, with hopes that it eventually will be standard operating procedure in American health care. Since joining the HBS faculty part-time, Feeley has commuted between Boston and Houston every other week, serving as a liaison between academia and practice. “It allows me to stay grounded in both places,” he says.

Along with integrated practice units and payment bundling, the value-based health-care plan includes measuring health outcomes of every patient, determining the actual costs of patient care, integrating medical systems, expanding the geographic reach of providers, and building an information technology platform that supports value-based care.

None of it comes easy, due to deep-seated, convoluted systems already in place.

Take measuring costs, for instance. Feeley can state confidently that cancer is the most expensive disease to treat. Cancer patients and survivors account for about 0.5 percent of the US population, and about 6 percent of the nation’s health care expenditures. He’s aware that 46 percent of uninsured patients–and 22 percent of insured patients—spend most of their life savings toward trying to save their lives. But he can’t tell you the true cost of individual patient care throughout most of his hospital’s history.

“We don’t know what it really costs to care for any individual patient,” Feeley says.  ”Most of the costing systems are based on the charges, and in health care there’s been so much cost-shifting. People end up paying $30 for a single aspirin to cover the costs of things that aren’t reimbursed.”

Cost-measuring progress is slow but steady. In addition to the bundled payment pilot, MD Anderson is one of many hospitals testing a process called time-driven activity-based costing (TDABC), which is meant to marry cost-measuring with value-based care. TDABC was pioneered by Robert S. Kaplan, the Marvin Bower Professor of Leadership Development, Emeritus, at HBS. It essentially involves two basic measures: the resources needed to perform an activity, and the time it takes to perform it. Feeley, Kaplan, and a team at UnitedHealthcare are using TDABC to track the costs for the bundled payment pilot.

“Bob Kaplan made the point early on that if you don’t know what it costs to care for a patient, how are you supposed to negotiate a price with the insurers?” Feeley says. “All these things are phenomenally intertwined.”

Measuring Outcomes

Michael Porter’s Institute for Strategy and Competitiveness co-founded the International Consortium for Health Outcomes Measurement (ICHOM). The group assembles international teams of clinicians and patients to develop outcome measure sets that can be used to drive improvement in health care performance, for public reporting, and in bundled pricing initiatives. ICHOM is also helping clinicians collect and report the results.

Measuring patient outcomes poses particular challenges, in part because “outcome” is a surprisingly nebulous concept. The Value-Based Health Care model divides outcome into a tiered hierarchy: health status achieved (e.g., survival), process of recovery (e.g., how long it takes patients to get back to work), and sustainability of health (e.g., whether an illness reoccurs). In the model, survival is the obvious priority. But subsequent research shows that patients don’t necessarily feel that way.

Over the past several years, MD Anderson formed focus groups of cancer patients and asked them an important open-ended question: What does a positive outcome mean to you? “We were getting all these different answers, and very few of them mentioned survival,” Feeley says. “We came to the realization—as did ICHOM in doing similar work on an international basis—that there were a number of things that mattered to patients that weren’t necessarily the kinds of things that we as clinicians thought about.  This was a new idea to us, this idea that, gee, cancer patients aren’t always primarily interested in a cure. At least, they don’t put it high up on their list early on.”

Breast cancer patients in particular voiced concerns about maintaining close relationships. Prostate cancer patients talked about performance anxiety.

As for head and neck cancer patients, “All these patients were motivated to get back to normal swallowing and speaking,” Feeley says. “But in our old electronic medical records, we had never asked questions about swallowing. You couldn’t even tell from those records when and whether a patient was eating normally, which of course is important. So all of this gets to the fact that there are things that are important to clinicians, and there are things that are important to patients. And right now in health care, we don’t measure either of them very well.”

The bundled payment pilot uses a series of outcomes that align with the Value-based Health Care model’s three tiers of outcomes. These outcomes are collected by the clinicians in their workflow as well as through patient-reported outcome surveys, which are given periodically to patients to learn, from them, how they are doing.

Information Technology Headaches

The move toward Value-Based Health Care Delivery depends on an information technology platform that can support and enable it. It’s a huge challenge and a sore point for Feeley. When he talks about the current state of health-care IT, his Midwestern calm gives way to Bostonian bluntness.

“Information technology in health isn’t as good as it is in other sectors,” Feeley says. “Here’s an observation: In 1969, there was information technology that put people on the moon. But we didn’t have computers in the intensive care unit until 1982 or 1983. We had a computer that calculated cardiac flow rates out of the heart. And that was a very crude device that depended on fitting a tube into the heart and injecting cold fluid. So health care has always been 15 years behind, and it has remained phenomenally behind. Even after this worldwide explosion of technology, everyone just continues to approach it in kind of a random motion: ‘Oh, this over here looks good. OK, now what?’”

The American Recovery and Reinvestment Act of 2009 included the Health Information Technology for Economic and Clinical Health (HITECH) Act, a $25.9 billion program that prompted many hopeful startups to flood the market. But Feeley says the industry doesn’t need new software as much as it needs a new approach to software.

Everybody wants to make money in this sector, but no one’s approaching it with a plan ,” he says. “Take the HR software vendors, for example. All those vendors did was say, ‘Oh, how can we help health care organizations get meaningful-use dollars?’ They didn’t think, ‘Well, how is health care technology actually going to help patients?’ And they’re still struggling with that. So–no offense to every entrepreneur who’s sitting out there thinking this is a wide-open field–yes, it’s a wide-open field, but it’s totally disorganized.”

Getting The Government On Board

Still, Feeley remains bullish on the value of value-based health care, and is heartened by high-level interest by government officials.

Over the past few months, Feeley, Porter, and Kaplan met with Julian Harris, associate director for health at the Office of Management and Budget, and his team at the White House. They’ve had conversations with multiple officials at the Centers for Medicare and Medicaid Services (CMS). Most recently, Feeley sat down with US Secretary of Health and Human Services Sylvia Mathews Burwell—who published an op-ed article in a March issue of the New England Journal of Medicine titled Setting Value-Based Payment Goals—HHS Efforts to Improve U.S. Health Care. At that meeting Burwell shared her belief that CMS needed goals to move away from fee-for-service payments, Feeley says.

In January, Burwell announced an initiative to move Medicare toward value-based care. The Department of Health and Human Services has set a goal of moving 30 percent of Medicare payments into “alternative payment models” by the end of 2016, and 50 percent by the end of 2018.

“It’s complicated,” Feeley says. “We know they don’t yet have the systems to get them there. But everyone seems to be moving, however slowly, toward this system that Michael Porter recognized the need for several years ago, before most of us did. I’m sure he’s very gratified, as is Bob Kaplan, that people are seeing the need for this. And when the federal government does something, as the largest provider of health care in the nation, others will follow.”