Insurers are using AI to deny senior patients care
When undergoing medical care, patients expect that their providers decide upon the course in treatment based on medical standards and the providers’ clinical expertise. Providers are the ones who recommend to patients how many therapy hours they need, when they should leave a hospital or nursing home, and which specialists to see.
However, a third party is often in play, gatekeeping access to care: the patient’s health insurance.
Most clinicians’ understanding of this hurdle is that insurance companies—and their human employees—make the decisions for when care is and isn’t authorized for coverage. Increasingly, that job is being handed over to artificial intelligence
A STAT investigation has found that Medicare Advantage plans in the U.S. are using unregulated predictive algorithms to cut off patient care with surgical precision—even when their providers disagree.
Medicare Advantage meets AI
The insurance AI businesses, hired by Medicare Advantage plans to help manage seniors’ claims, claim to use their algorithms to more scientifically pinpoint when care can be cut off—under the guise of personalization.
However, providers and health systems are sounding the alarm that Medicare Advantage plans are doing the opposite: overwhelmingly cutting off life-saving care, with no adjustment for patients’ individual circumstances.
“While the firms say [the algorithm] is suggestive, it ends up being a hard-and-fast rule that the plan or the care management firms really try to follow,” said David Lipschutz of the Center for Medicare Advocacy, a nonprofit group reviewing Medicare Advantage denials. “There’s no deviation from it, no accounting for changes in condition, no accounting for situations in which a person could use more care.”
The patients, who have spent their lives paying into Medicare, are now being forced into years-long appeal processes to access care they need much sooner.
Chris Comfort, COO of New York hospice facility Calvary Hospital, said of Medicare Advantage: “We take patients who are going to die of their diseases within a three-month period of time, and we force them into a denial and appeals process that lasts up to 2.5 years. So what happens is the appeal outlasts the beneficiary.”
How insurers control access to medical care in the U.S.
The American health insurance prior authorization system was initially designed to prevent doctors from recommending expensive, un-indicated care—to promote cost efficiency across the healthcare system.
Now, insurers commonly require prior authorization for a wide array of medical encounters, from basic imaging to prescription refills. It’s become such a common hurdle that 40% of physicians report having dedicated staff just for managing prior authorizations.
Various studies have demonstrated that the additional delays in care the insurance authorization system creates hurts patients—including cancer care delays in Pennsylvania, growing hospitalizations among sick children in Colorado, and the inability of low-income patients to access opioid addiction care.
But the system helps insurers maintain (and increase) profit margins—especially when it comes to care for seniors, a population with high healthcare utilization.
With Medicare Advantage plans, insurers have been making growing profits as more seniors opt for plans with lower premiums and prescription drug coverage. That’s where predicting how much care to authorize for high utilizers becomes an effective cost-limiting measure.
The predictions for which care is necessary—and how costs can be lowered for insurers—have become so profitable, that many insurers have acquired the companies behind the most widely-used insurance AI tools. One of the biggest (and most controversial) companies in this space, NaviHealth, is owned by UnitedHealth Group.
In December, the federal government proposed changes to Medicaid, Medicare Advantage, and federal Affordable Care Act marketplace plans that would reign in some of insurers’ power to dictate care—particularly when it comes to prior authorization.
The jury is still out on how this move will impact the insurers’ reliance on AI in authorization processes—if at all.
Medtech must lead the way on building trust in medical AI
There is a gap in trust between medical AI innovators and the overall health system and government.
This news about the liberal use of insurance AI is happening against the backdrop of the brewing tension between the medical AI industry and the federal government—specifically, the FDA, which announced that it plans to regulate medical AI tools as devices.
Revelations like this—about how medical AI may get in the way of their care—is also likely to make more patients skeptical of AI overall, an issue which we discussed in our last edition. It’s our responsibility to show them—and their providers—that the technology can be trusted.
We must start by calling for our industry to prioritize using medical AI to improve patient care and access to it, not to optimize corporate profits at patients’ expense. We look forward to seeing the medtech industry step up to lead on this topic, advocating for a healthier world through technology.