Personalized Medicine: The Role of Technology and Diagnostics in Driving Change

March 3, 2020
Laura Joszt , Mary Caffrey
Volume 26, Issue 3

Rebecca Kaul, MBA, vice president and chief innovation officer at The University of Texas MD Anderson Cancer Center, said the best technology improves the patient experience.

Technology Remains a Small Part of Innovating Cancer Care Delivery

Contrary to popular belief, innovation in healthcare does not necessarily refer to technology, although technology can assist. Innovation is about improving experiences, explained Rebecca Kaul, MBA, vice president and chief innovation offi cer at The University of Texas MD Anderson Cancer Center.

The greatest new technology is not actually about the technical specs, she said. When Apple released the first iPod, the interest it generated, and its ultimate success, was not necessarily about the technical details, but instead that the iPod was something small and convenient that delivered an experience people wanted.

“For me, success is not completion of a project plan,” Kaul said. “Success is delivery of value, delivery of experience.” If new inventions cannot be converted into something people can use to change their lives, “have you really driven change? Have you really innovated?” she asked.

When comparing the healthcare sector to the consumer environment, the industry falls short, she said.

“I would say that our experience in healthcare is, at most, functional, not necessarily desirable,” Kaul said, and as a result, the industry is seeing big companies, like Apple and Amazon, trying to enter the industry to disrupt it. However, these companies are not necessarily bringing novel technology, they are trying to innovate how healthcare is delivered to bring an experience people need and want.

Delivering better care and a better experience means going beyond metrics and truly understanding the person you are delivering care to. There is a lot of buzz about data in healthcare and how to aggregate it, create interoperability, and utilize artifi cial intelligence to make meaning of the data. “We have to be careful with data,” Kaul said, “because correlation doesn’t necessarily mean causality.”

For instance, she provided the example that the divorce rate in Alabama correlates with per capita consumption of high fructose corn syrup. However, that doesn’t mean divorce rates in Alabama are caused by consumption of high fructose corn syrup. So, data are important, but it’s also important to know what to do with the data and how to dig deeper and understand what the data are saying.

“You look for those themes and patterns in data,” Kaul said. “You make creative leaps and inferences you can test. But then, the most important part is using qualitative design research to verify [or] debunk. And that’s really the point here: You can’t just take what you’re seeing in the data at face value.”

As someone who has worked in the field of innovation for a decade, Kaul has learned that 99% of eff orts should be spent finding solutions to human problems and only 1% on solutions to technology problems. She ran a technology development center, where the employees built the technology they thought would be most valuable, but it’s crucial to fi rst understand the problem being solved and the experience that should be delivered.

Kaul provided the example of an employee at MD Anderson who showed her a 30-day predictive algorithm that had been developed, but the person who created it had never thought about how the doctor would use it in practice. What should the doctor do after receiving the prediction? What intervention should there be?

“So, the question becomes: How useful is a prediction if we don’t know how we’re going to action it—if we’re not actually going to translate that prediction into something that actually impacts our patients impacts our experience?” she asked.

In another example, Kaul explained the challenge of ill-defined problems. At MD Anderson, she was told early on that there needed to be more chairs in the infusion center. This didn’t seem like a problem in her area, but she decided to take a look and found that during the course of a day, there was never a time when a chair was not available, but there was still a waiting room full of people. She then dug deeper to understand why people weren’t getting past the waiting room faster and found a myriad of issues including orders weren’t signed, labs weren’t ready, drugs weren’t mixed, and even nurses weren’t available.

The underlying problem was how to make an appointment meaningful and optimize patient flow. This was an instance when technology was useful, because the cancer center was able to utilize machine learning to understand the best time to make an appointment because everything that had to happen to meet that commitment could actually be done in time.

However, there were times when technology was not the answer. MD Anderson has been expanding its footprint with new buildings in the community, so people have a convenient place to receive care closer to home. The people in operations and technology wanted to make these new buildings more effi cient so that patients could check in at a kiosk, get a wristband, and do everything else they need to do, such as pay the bill, at the computer. Next, they took the time to understand the experience this setup would deliver. Patients would be greeted with a computer screen and as a medical record number. So, then the employees in operations and technology went to talk to the patients.

“What patients wanted was a human experience,” Kaul said. “They’re scared. They just got diagnosed with cancer.”

Usually, patients are coming in for multiple procedures, such as examinations, imaging, labs, and potentially chemotherapy. “They’re usually coming in for appointment after appointment, having to address something that is really scary, and what they want is someone to greet them, someone to give them the feeling that it’s going to be okay, and that they are being taken care of now,” she explained.

Yes, patients want effi ciency, but they want a human experience, too. This fi nding resulted in designing a better experience even before the patient arrives at the center by reviewing schedules, providing directions to get around the facility, and filling out paperwork before they arrive.

MD Anderson also created a new role for someone to greet patients as they arrived. Instead of sitting behind the desk, this staff member would be out on the floor, wearing an identifi able shirt, available to help answer questions and make patients feel comfortable when they arrived.

And in the end, they did incorporate technology in the form of biometrics to gather patient information and maintain patient records. However, “technology ends up being that last mile,” Kaul said.

Technology alone isn’t the solution, she said. Everything that goes around technology makes up the solution to a problem. “We can’t just follow shiny objects,” Kaul added. “We can’t implement technology for technology’s sake; we have to understand the problem space.”

Personalized Medicine Brings Diagnostics to the Value Equation

Can oncology care follow pathways—ensuring patients eceive evidence-based care—and at the same time meet individual needs? That was the question for the panel, “Personalized Medicine and Value-Based Care,” at Patient-Centered Oncology Care®. Moderator and co-chairman Joseph Alvarnas, MD, led the discussion with:

BRYAN LOY, MD, MBA,

physician lead for Oncology, Laboratory, and Personalized Medicine at Humana;

EDWARD ABRAHAMS, PHD, president of the Personalized Medicine Coalition, which represents more than 200 innovators, scientists, providers, and payers; and

JAMES ALMAS, MD

, vice president and national medical director of Clinical Eff ectiveness, LabCorp.

Loy said that as a pathologist, one view of the term “personalized medicine” is targeting therapies to the right patient based on abnormalities. But it is more than that, he continued. “At least at my payer, it’s what can we learn about our members—or your patients if you’re a doctor—and how can we create a better experience?” not help if patients go home to an empty refrigerator or have no transportation for follow-up care. Thus, the broader meaning of personalized medicine is to identify the “highest-quality, least toxic, most eff ective, and convenient care for our members.”

Abrahams agreed, adding that personalize d medicine means “marrying diagnostics and learning everything we can about the patient in order to prescribe the right therapies,” while accounting for patient values and experiences. This makes the term different from precision medicine, which he said is preferred by scientists.

Almas said his view of the term is informed by his experience working on the joint CMS and FDA parallel review1 for Foundation Medicine’s CDx product, which was later reopened for a National Coverage Determination.2 From there, Almas served as a medical director for the Molecular Diagnostics Services (MolDX) Program

at Palmetto GBA, the Medicare administrative contractor that developed early expertise in approving diagnostics. Based on these experiences, he still sees some barriers. In clinical trials, Almas said, “we still run into resistance with some commercial barriers—not so much with Medicare.”

Reimbursement has often been the rub, Alvarnas said. Value-based care is supposed to be about arriving at care that is high quality and sustainable, not a race to low-cost care. How can that be achieved, he asked.

At times, Loy said, “value-based care overemphasizes the dollars when, in fact, this should be about the patient’s values, which in my mind leads into shared decision making and real-world evidence.”

Abrahams agreed, saying, “Value-based care, in my estimation, is not the least expensive, it’s the most eff ective.”

Both Alvarnas and Almas commented that it is not always that simple, and Almas noted that is why the development of real-world evidence is important. Testing companies are often asked if they have a randomized clinical trial to prove the clinical utility of their products, but designing such trials is difficult. Partnering with payers to develop evidence is more sensible, he said.

Designing the right data models is also challenging, Alvarnas pointed out. The numbers must be large enough to ensure that researchers “get it right,” but turnaround is important, too. Still, the fi eld has moved a long way from registry studies that did not provide answers for 5 to 10 years. “The rate of innovation is outpacing our old models of gathering data,” Alvarnas said.

Key considerations, according to the panelists, include:

• the quality of the test or diagnostic being used;

• addressing knowledge gaps that lead to physicians not ordering the right tests;

• dealing with situations when the test is ordered, but the results are ignored, so the patient gets no benefit; and

• ensuring that reimbursement discussions address not only tests for genomic alterations but also screening, such as increasing the population tested.

Alvarnas said all these problems call for changes in physician behavior. “This is a real challenge, and it doesn’t happen infrequently,” he said. “How do you keep this from happening?”

Loy noted that current reimbursement models also do not reward physicians for the most important test of all: does the patient want a $300,000 treatment, or a drug that costs $10,000 a month? “I don’t think we can afford to run the test without first asking the question, ‘Do you want to be treated?’”

Payment models must be flexible, he said, because not all practices are at the same point in transitioning from fee-for-service to value-based care.

Alvarnas asked whether commercial payers or Medicare are more likely to drive change, and Loy credited the Center for Medicare and Medicaid Innovation for “cracking the ice” with the Oncology Care Model. There are “refi nements” that commercial payers can offer, especially making sure the voice of the employer is heard.

“Employers are in a unique position to experiment creatively around the next steps,” said Alvarnas, whose portfolio includes employer strategy at City of Hope, where he is a hematologist/oncologist.

Rewarding physicians for taking time to understand what patients want is essential, Almas said. “You’re going to want to talk to patients to truly have shared decision making about whether they want to pursue targeted treatment,” he said. If patients don’t want to do the test, “we don’t want to do a test,” because it is wasteful when the results are not used. “That doesn’t help anybody,” Almas said.

References

1. Ramamurthy L, Maxwell K, Sawchyn B, Anhorn R. Perspective: FDA/CMS parallel review advances coverage for cancer comprehensive genomic profiling. Am J Manag Care 2018; 24(SP6):SP193-SP196.

2. Inserro A, Caffrey M. CMS agrees to cover NGS for Medicare patients with breast, ovarian, other cancers. The American Journal of Managed Care® website. ajmc.com/newsroom/cmsagrees-to-cover-ngs-for-medicare-patients-with-breast-ovarian-other-cancers. Published January 27, 2020. Accessed February 4, 2020.