The era of big data might still be in its infancy, but payers are already dealing with a deluge of bits and bytes that threatens to overwhelm their ability to keep up with recent developments in medical science. Clynt Taylor, CEO of Intervention Insights Inc., of Burlington, Mass., told BioWorld that payers are starting to see the light when it comes to reacting to data that suggest a shift in coverage policy, but that over time, payers will replace those fixed policies with a data-driven approach that adapts to patients' needs in real time.
Taylor said the company opened its doors for business roughly eight years ago, adding that "we started by providing information that helps molecular testing labs provide an enhanced report" that spells out the association between a gene and a disease. The company's vision was expanded in 2017 to aid health care professionals and payers trying to work through the complexities of precision medicine in oncology, with an eye toward making clinicians comfortable using molecular medicine and ensuring payers that they aren't covering something that isn't appropriate for that patient.
'A race' to data-driven coverage in the offing
While mounds of data are difficult to sort through, payers are increasingly faced with just that prospect, a predicament that will only worsen over time. Taylor said the competition for client employers will press payers to update their approaches, however. "It is going to be a race to use data and data-driven approaches," he said, adding that adopting the tools that allow that will give the forward-thinking payers a competitive edge. The questions that payers' customers will pose at contract time will be "which of you can use real-time data and which have policies that can't keep up?"
"Payers are slow to respond in many ways because they're conservative by nature," Taylor commented, but they are data-driven because they want to see results. Not all payers will migrate to a data-driven paradigm readily, as most payers will sit back and watch to see whether the first to adopt such an approach experiences a rather costly meltdown. As soon as the returns suggest that such a change didn't cost more than expected, the payers who did not want to be the first on the dance floor will enjoy the music much more, Taylor said.
The transition from a policy-driven to a data-driven coverage paradigm will take time, but Taylor said some plans are exploring that approach already. "The answer is not going to be whether the new data-drive approach will work, it's just a matter of when and how quickly," he said. Taylor said he anticipates that provider relationships will have a large effect on adoption. If the payers' provider network does not have cancer centers that are using precision medicine, that might be a lower priority.
Conversely, all the genomic and proteomic data coming out of research centers will prompt payers doing business in regions with a number of cancer centers to do what they can to keep up, something that will become a necessity as oncologists in the network persistently argue for faster turnaround of coverage requests.
AI not ready for coverage decisions
That transition will likely happen with the help of human eyes and gray matter, however, despite the grand hopes surrounding artificial intelligence (AI). "Most people would say that given the way literature is published on these topics, AI is not ready for that, and most payers would say they don't want to depend on that," Taylor said, although AI and machine learning will work their way into that function.
"This new age of precision medicine is requiring a different way of thinking about the whole world of medicine," Taylor observed, predicting that the advent of a data-oriented coverage paradigm will influence the relationships between the stakeholders involved. One example is prior authorization, which as currently understood will become less commonplace, in part because the evidence accumulates so quickly, which increasingly shines a difficult light on the traditional prior authorization process.
"We are going to see more technologies and processes that replace prior authorization, but not the function it performs," Taylor said, noting that an American Medical Association poll of members suggested that the practice is having a serious impact on patient care. Prior authorization might not be very economical, either, despite being one of the hoped-for benefits. Taylor said he has discussed that with executives at some of the major payers, who told him that the economic benefit of prior authorization, "compared to the percentage of denials, is so small that we're paying more to regulate than we get in savings."
The National Comprehensive Cancer Network is an oft-cited source of evidence for coverage applications, but Taylor said NCCN guidelines might evolve too slowly at times. At present, a payer might say "just because it's so complicated and there are too many ways to look at this, we just default to NCCN," he said. In a fast-moving clinical area, however, a payer may decide to adopt a more streamlined approach whereby "if a certain bit of information comes out and it meets certain criteria, [the service] automatically gets approved." Taylor acknowledged, however, that the shift to that approach would benefit tremendously by a greater degree of interoperability between electronic health records and payers' IT systems.
Appropriate use criteria formed the radiology clinic's response to calls for pre-authorization for the pricier imaging procedures, but that process, too, could prove too cumbersome to deploy with many of the diagnostics used in oncology. One reason for that is that some tests, particularly those performed in clinical labs, will be sufficiently inexpensive to become commoditized, thus draining the cost incentive for developing those criteria. There will always be the matter of adding a marker to an existing gene panel, however, thus seeming to resurrect the code-stacking controversy, something clinicians and labs will have to address carefully.
"I think payers will want to know early in the process that patients weren't undertested or overtested," Taylor said, adding that clinicians and payers might still be able to make use of a standardized test panel if they are willing to search for a lab that has a panel that most closely matches the individual tests the clinician and payer have agreed on.