New Partnerships Bring New Opportunities, Challenges With Data in HEOR

Laura Joszt, MA
Laura Joszt, MA

Laura is the editorial director of The American Journal of Managed Care® (AJMC®) and all its brands, including The American Journal of Accountable Care®, Evidence-Based Oncology™, and The Center for Biosimilars®. She has been working on AJMC® since 2014 and has been with AJMC®'s parent company, MJH Life Sciences, since 2011. She has an MA in business and economic reporting from New York University.

Technological innovation is helping to reinvent the traditional health economics and outcomes research (HEOR) field as traditional and new players partner to leverage data sources.

Technological innovation is helping to reinvent the traditional health economics and outcomes research (HEOR) field as traditional and new players partner to leverage data sources. During a plenary session at Virtual ISPOR 2021, thought leaders from government, payer, technology, and academia discussed what is working and what isn’t.

Amy Abernethy, MD, PhD, fresh off a stint with the government as the former principal deputy commissioner and acting chief information officer at FDA, noted that new partnerships and capabilities have expanded HEOR from focusing activity at the end of the development of a drug or medical product to be applicable across the development life cycle.

“We also see HEOR—real-world data, real-world evidence—applicable for a new set of tasks like being able to describe the natural history of the pandemic, surveillance of real-world performance of COVID-19 vaccines and diagnostic tests, and management of our drug supply chains,” she said.

Health technology and data companies are advancing the space by building software and algorithms that can clean, link, prepare, and analyze data sets using information from various sources. Now, there is the ability to link data from a national registry, electronic health records, and obituary and grave site information from the internet to create high-quality mortality data sets.

However, these new partnerships mean there need to be new ways of working together. HEOR researchers will need new capabilities so they can manage and analyze larger amounts of data, as well as data that keep changing.

“There’s an incredible tension between traditional hypothesis-drive research and new data science approaches, such as AI [artificial intelligence] and ML [machine learning] that tend to focus on letting the data speak for themselves,” Abernethy said. “And we need to find ways to really figure out how to work through that tension.”

Recently, the digital health space has received an “unprecedented” amount of venture-backed funding, said Carolyn Bradner Jasik, MD, who heads up the clinical and research teams at Omada Health. As a result of this funding, there’s going to be a “tsunami” of new health care interventions, she said.

The industry is looking for technology to provide care in nontraditional spaces. Instead of episodic visits, data will be collected continuously, which will make the idea of encounter-based data obsolete.

Bradner Jasik also referenced the tension between the traditional and the new.

“The 3- or 4-year gestation of a randomized controlled trial, a university-sponsored study is really not practical,” she said. “Many of us have to show market traction within 2 years. We can't be doing the old methods. We have to be able to get our data and our research out more quickly.”

She cautioned about the fast pace of change, pointing out that it is a largely unregulated environment leading to “a bit of a Wild West.” The hunger for solutions may mean companies adopt solutions that haven’t been thoroughly studied.

“And though many things should be adapted, like doing virtual trials…the rigor and care with which that research is done does not need to be sacrificed just because of the pace that we're moving at,” Bradner Jasik said.

The pandemic has revealed that many health systems are financially at capacity and unsustainable under pressure, and the value chains need to be rebuilt, said Dipak Kalra, MBBS, PhD, president of the European Institute for Innovation through Health Data. Focusing on good health outcomes and value-based care models will be important for that redesign.

He highlighted 3 challenges about what is meant by outcomes in value-based care:

  • The meaning of outcomes and value. Multiple stakeholders need to be involved to reconsider which outcomes matter most, which should be prioritized, which are the most pragmatic to collect, and which are the best to use for performance and quality.
  • Many clinicians don’t document outcomes as part of their electronic health records. “We are left in a position where real-world evidence, real-world data sources of outcomes are weakened, patchy, and of variable quality,” he said.
  • Collecting outcomes directly from patients. This is done in a piecemeal way in Europe using standardized tools that are expensive and cumbersome. Instead, patients need to be engaged differently to collect the outcomes that matter most to them and they need to be equipped with practical tools to collect outcomes on a near-continuous basis.

Ultimately, the HEOR community is results based and wants to know what has been achieved for people, which is connected to value, said Harlan Krumholz, MD, director of the Yale New Haven Hospital Center for Outcomes Research and Evaluation and the Harold H. Hines Jr Professor of Medicine at Yale University.

“HEOR team, this is our time,” he said. “There’s never been a greater interest in value or in understanding the end result of all that's done within health care, and in appreciating its economic consequences, as the cost of care continues to escalate.”

Rather bluntly, Krumholz said he doesn’t think there is enough that is going well in the space right now. During the pandemic, what went well were the vaccines, but the implementation application, which is the jurisdiction of HEOR, is where there was trouble.

“My main concern is that in the future for our field, to make a big impact will require something different than where we're going and where we've been,” he said. “And yet, like so many fields, many of us are comfortable with what's always been done.”

He highlighted a few areas where he thinks there needs to be focus:

  1. Reproducibility and validity—the field can’t just be interested in a particular result over the truth.
  2. Data quality over quantity—“We're addicted to the data and the output, but we're not paying enough attention to where it comes from.”
  3. Analysis and design—the old designs cannot be retrofitted to these new “fuel sources” like ML and AI.
  4. Implementation—studies need to be built in a way that ensures the work is actionable and researchers need to do a better job thinking about the translation. Part of the implementation is keeping in mind access and equity so no one is left out.
  5. Patients—how do they get involved in prospective studies? They have can make sure researchers focus on the outcomes that matter to them. “Only with people as true teammates, can we increase the efficiency and effectiveness in our studies. We should do this not just because it's the right thing to do to include people as teammates and partners and participants, but because it's a competitive advantage.”

The success of the HEOR field depends on the ability to innovate and embrace new approaches. “This field should look very different going forward than we have in the past,” he said.

Robert Califf, MD, head of clinical policy and strategy for Verily and Google Health and former FDA commissioner, as moderator of the session asked the panelists about the tension between the need for long-term follow-up and the desire to see change now.

At least in the US health care system, employers act as a check because they are interested in long-term outcomes, Bradner Jasik said. Employers “own that heart attack for the next 20 years,” she said. In comparison, payers have a much higher churn and they want to see a return on investment within a year.

“Flashy models that people are putting out will attract the attention of payers and lead to solutions that don't have durable long-term outcomes being adopted…and that is a very big concern in the industry,” she said.

Primary prevention is the so-called holy grail, but getting investments in people who are not in the middle of a health system payment cycle, people who are healthy citizens and not yet labeled as patients, is challenging, Kalra said.

Abernethy added that with cell and gene therapies coming to the market, there is a need to think about long-term outcomes.

“We really don't know the implications of these products across time,” she said. “So, I think this issue of long-term outcomes and balancing what we know in the short term vs what we know in the long term is going to be a need, not just from the perspective of who's paying for health care, but also from the perspective of how the products work.”

Abernethy also talked about the role of data aggregators, noting that there is going to be a “ton of maturation.” As it stands now, these aggregators have both done a great service—showing it is possible to aggregate data and start to use solutions like tokenization—and a terrible disservice—they don’t care as much about core elements of research like provenance.

“It's challenge, because from what I can tell…we haven't actually really held data aggregators to task in needing to do the kind of transparent high-level work that needs to be done,” Abernethy said.

Krumholz added that the topic of data aggregators is complex. He admitted that he has benefitted from these aggregators, “and I think I have produced…useful contributions.” However, he didn’t know where the data were from and he just had to trust the companies producing them.

At the end of the discussion, Krumholz struck a tone of optimism about the future.

“You know, we talked about a lot of these challenges, but I'm really optimistic that the future is going to be better than the past,” he said. “That these are problems that can be solved. We can all be part of those solutions.”