Artificial intelligence for cancer prevention has gained momentum.
However, most of these new programs are not covered by Medicare or private health insurance, creating headwinds for companies looking to increase adoption and for patients who could benefit from the new technology.
“Traditionally, with medical devices, it takes up to seven years after a product is approved by the FDA to get reimbursed, so that's quite a challenge,” said Brittany Berry-Pusey, CEO of AI screening startup Avenda Health.
As AI capabilities continue to grow, the Food and Drug Administration has approved 882 AI and machine learning-enabled devices and programs. Nearly 600 of these are approved AI applications for radiology in the past five years. Most do not yet have billing codes that would enable reimbursement and prevent patients from having to pay out of pocket.
While some tools have shown promise in improving diagnosis and treatment of cancer patients, more data may be needed to determine whether they are more effective than traditional screening before major insurers are willing to cover their costs.
A medical robot from the French start-up SquareMind, which is designed to facilitate cancer screening using artificial intelligence, will be presented on May 22, 2024 at the Vivatech technology start-up and innovation fair in the Porte de Versailles exhibition center in Paris.
Julien De Rosa | Afp | Getty Images
One of Avenda’s products illustrates the complex process that must take place before insurers cover AI tools.
The company's Unfold AI platform for prostate cancer can help urologists find more cancer cells than traditional MRI screenings and can help determine the best treatment to reduce the risk of side effects such as incontinence and impotence from prostate cancer surgery.
The FDA approved the medical decision support program last year. Just as importantly, the American Medical Association has assigned it an interim billing code—which most AI radiology products have not yet received.
Avenda is now trying to get Medicare and insurers to cover the costs, which in many cases can take years.
“If there is no payment, it means patients have to pay out of pocket, which can be challenging … especially for our patients. This is an older patient population,” Berry-Pusey said.
Hurdles to reimbursement
The American Medical Association, the professional medical organization that assigns the Current Procedural Terminology codes that enable reimbursement, issued guidelines for setting AI-CPT codes last fall. The group said different medical specialties should help set the standards for use in their fields.
The lack of reimbursement is hindering the rollout of new AI programs for cancer screening, especially in smaller hospitals and doctor's offices, said Dr. William Thorwarth, CEO of the American College of Radiology, which represents thousands of professionals in the field. But in a letter to a congressional committee studying the use of AI in health care, he cautioned against moving too quickly.
Thorwarth wrote that reimbursement for AI costs is complex and establishing billing codes for each approved AI tool is “problematic.” He added that it is “unclear” whether the AI platforms currently covered “provide value to patients or the healthcare system.”
Medicare and private health insurers have expressed similar caution. A spokesperson for the Centers for Medicare & Medicaid Services told CNBC that the agency considers CPT codes when making reimbursements and “continually evaluates opportunities to safely and responsibly deploy new, innovative strategies and technologies, including artificial intelligence.”
Some of this caution may stem from previous experience with computer-assisted mammography in the late 1990s. Doctors have since said it led to false positives and unnecessary biopsies.
Dr. Rodrigo Cerda, chief medical officer of Independence Blue Cross, said the effectiveness of the latest programs has not yet been determined.
““The evidence is not yet quite sufficient to say that it clearly provides a positive benefit to our members and does not pose other risks that could lead to false-positive results or exacerbate the false-negative results,” Cerda said.
Patients have to pay out of their own pocket
Without reimbursement by insurance, radiology providers RadNet has started charging patients for its proprietary AI screening tool, Enhanced Breast Cancer Detection, which launched in 2022. RadNet has released data suggesting the tool helps improve cancer detection.
The company recently reduced the price of the test from $59 to $40. It said its revenue from AI in digital healthcare more than doubled year-on-year in the first quarter and patient adoption AI screening increased from around 25% to 39% of mammography patients.
RadNet executives compare the AI screening process to the radiology industry's experience with digital breast tomosynthesis, known as 3D mammography. The FDA approved the procedure in 2011 and women were initially offered screening for a co-payment. However, by the end of the decade, it was largely covered by health insurance.
“The question is, can we finally [insurers] to advocate for it? And I think the acceptance and value creation in finding more cancers will ultimately convince them,” said Dr. Greg Sorensen, RadNet's chief scientific officer.
Sorensen said RadNet has recruited an employer in New Jersey that will now cover the cost of breast cancer screenings for its employees.
The company will also will soon launch an AI-powered prostate MRI screening for $250. But that price could pose a major hurdle to adoption — and access for patients who can't afford it.
Concerns about access
Josh Trachtenberg, a neurology professor at UCLA, was willing to pay for AI prostate cancer screening because he believed it would make a huge difference to his own treatment.
Trachtenburg says when he was diagnosed with prostate cancer Last year, several doctors told him that his prostate would need to be removed, a procedure that would have caused incontinence and impotence problems.
He turned to a urologist at UCLA School of Medicine who used Avenda Health's AI program Unfold. The program was able to more accurately measure the extent of his tumor, allowing doctors to get to the cancer cells during surgery while preserving healthy tissue.
Trachtenberg fears that patients who cannot afford the additional costs of certain AI tools will pay the price with worse outcomes.
“I think most men who are not teaching in medical school are just put through the meat grinder because it's covered by insurance and that's the standard procedure,” he said.
Berry-Pusey of Avenda Health fears that patients may be denied access to new technologies altogether because uncertainty about reimbursement could hinder funding for innovation.
“As a startup, we are always looking for investors, so making sure there is a clear path to more revenue is important for our survival,” she said.
Despite the payment hurdles, investors are backing developers of AI in healthcare. Alex Morgan, partner at Khosla Ventures, is optimistic about the sector and recently participated in a large funding round for an AI company in radiology.
“If you just have a human do a set of tasks and then turn on AI, you're not going to get efficiency gains,” Morgan said, adding that the key to getting paid is “delivering differentiated, high-performance results.”
He said that in the end, technology that improves the quality of care and patient outcomes will prevail.
