By Carrie Hart, Director of Market Access at Optellum

Summary

• Hospitals are seeking to adopt AI and other new technologies
• The Centers for Medicare & Medicaid Services (CMS) can pay for new, outpatient-based technologies under the New Technology Ambulatory Payment Classification system
• Optellum’s AI-enabled Lung Cancer Prediction score has been shown to be cost-effective, and may help providers and Medicare Advantage carriers with their Star ratings
• LCP is currently being reimbursed, with a focus on Medicare beneficiaries
• Continual assessment of the value that new technologies bring is crucial

In an effort to improve patient outcomes, hospitals innovate and invest in new technologies such as artificial intelligence (AI), robotics, and other ways of delivering precision medicine. But there is a lingering question around who pays for these technologies and how?

How are new technologies paid for and by whom?

In the US, medical procedures are covered in whole or in part by insurance through public or private providers.

• Medicare is a federal health insurance program primarily for individuals aged 65 and older, as well as certain younger individuals with disabilities. It consists of different elements including hospital insurance, medical insurance, approved private health care plans (Medicare Advantage)  and approved prescription drug coverage.
• Medicaid is a joint federal and state program that provides health coverage for low-income individuals and families. Eligibility and specific benefits can vary by state, but Medicaid typically covers a broad range of health services, including hospital and doctor visits, prescription drugs, and long-term care.
• There are a number of private health insurers with different polices. Some of the largest include UnitedHealth Group, Anthem, Inc, Aetna (a CVS Health company) and Cigna Corporation.

A lung biopsy on a patient being diagnosed or treated for a disease like lung cancer would typically be covered by insurance, if the insurer or payer determines that the procedure is necessary for diagnosis or treatment planning. But how do some of the newer technologies fit in? What if that patient’s diagnosis has involved the use of a new technology such as AI?

For situations like this, when a technology is deployed in the hospital outpatient site of service, the Centers for Medicare & Medicaid Services (CMS) evaluate if the technology is eligible for payment via a New Technology Ambulatory Payment Classification (APC) methodology. When new technologies or procedures are introduced in healthcare, they may impact reimbursement and payment classifications. In some cases, CMS may create or modify APCs to accommodate these innovations. New Technology APCs are reserved for comprehensive services or procedures that are truly new and meet the eligibility criteria as defined by CMS.  As adoption of the new technology increases so should the number of claims. This claims data will be analyzed, and the procedure will ultimately be assigned to an APC grouping based upon clinical and resource use. Other payers may follow suit, adjusting their coverage and payment methodologies.

One example of this is the Lung Cancer Prediction (LCP) Score developed by lung health company Optellum. In mid 2022, hospitals that bill CMS for use of the Optellum Lung Cancer Prediction for their Medicare patients became eligible for a New Technology APC reimbursement at a rate of $600-$700 under two temporary CPT codes, 0721T and +0722T, to describe quantitative CT tissue characterization performed separate or concurrent, respectively, to a computed tomography (CT) scan. The clinically validated LCP score is part of Optellum® Virtual Nodule Clinic, and is based on imaging AI. The goal behind the technology is to improve clinical decision making, with the aim of improving patient outcomes:  for malignant-nodule patients, faster times to diagnosis and for benign-nodule patients, fewer investigations, which may in turn result in fewer procedure-related complications. Crucially, the technology has been demonstrated to have the potential to improve patient outcomes[1] and to be cost-effective[2].

How is value determined?

The concept of value in healthcare is a dynamic interplay between the outcomes achieved, the patient experience, and the costs incurred, reflecting a paradigm shift from a traditional fee-for-service model to a more holistic approach focused on improving health outcomes and enhancing overall system efficiency. We think about value in a number of ways:

• When you’re submitting an insurance claim you think about value to the patient and the value to the payor, which constitute the clinical utility of an innovative diagnostic process. In other words, the costs that are removed by using the new technology such as fewer unnecessary procedures, while maintaining or improving the quality of care delivered and hence patient outcomes
• What about the value of an improved workflow for clinicians in terms of better processes? When we show the value of a new technology we can look at the steps that are removed as well as added, the time saved or the possibility of tasks being performed by nurse navigators rather than by clinicians.

As the use of AI becomes more prevalent in medicine, studies are now being conducted on market adoption and the use of insurance claims for reimbursable technologies. One such study, published in the New Engand Journal of Medicine AI[3], concludes that while commercialization of AI is still new, it is growing. Academic health centers are leading the charge with 70% of zip codes with academic centers having at least one medical AI billing, compared to 9% in zip codes without such centers: underlining a need for further investigation of barriers and incentives to ensure equitable access and wider integration of AI technologies. Patients may be incentivized to visit practices, whether academic or innovative community centers, with these state-of-the-art technologies.

However, the data is representative only of the commercially insured U.S. national population for patients under 65 years of age. It does not show the Medicare population. At least half and possibly the majority of lung-nodule patients are over 65 years old. At Optellum, we are primarily seeing reimbursement on Medicare plans in those demographics.

In evaluating and choosing Medicare Advantage plans for enrollment, consumers can assess plan design as well as CMS-developed Star Ratings[4] where customer satisfaction and health plan quality metrics are rated on a five-star scale. Therefore, coverage and reimbursement of technology that expedites time to diagnosis for malignant nodules is crucial for Medicare Advantage plans whose Star Ratings include a measure of customers’ ability to get appointments and care quickly.  

What’s next for insurable AI?

Clinical utility is imperative. When demonstrating value, we talk to our research collaborators and customers about how to demonstrate that to the payers, how did it inform and/or support clinical decisions, and what their questions are likely to be, for example: What are the FDA cleared indications for use of the technology? Can you produce documentation of why you chose to use the technology? How clinicians use the information, did this change their clinical assessment and treatment path?

We may see more of these technologies bundled together under larger systems that incorporate diagnosis, treatment and risk or recurrence. The large datasets used by AI such as Optellum’s can provide valuable insights into understanding lung cancer and propelling patients into other state-of-the-art technologies such as robotic bronchoscopy. It’s important to keep studying these improvements and innovations and their impact on value-based care and quality metrics. 

The integration of diagnostic AI tools, treatment modalities, and risk assessment within comprehensive systems is poised to reshape the landscape of patient care. While skepticism may persist, forward-thinking hospital administrators stand to usher in a new era of healthcare, marked by increased access to life-saving procedures and a profound understanding of their diverse patient populations. The journey towards enhancing healthcare value is an ongoing expedition, one where embracing innovation and scrutinizing its impact pave the way for a healthier and more resilient future.

[1] Artificial Intelligence Tool for Assessment of Indeterminate Pulmonary Nodules, Radiology 20222, https://doi.org/10.1148/radiol.212182

[2] Artificial Intelligence-assisted Versus Clinician Only Evaluation of Indeterminate Pulmonary Nodules: A Comparative Effectiveness Study, ATS Abstracts, https://www.atsjournals.org/doi/pdf/10.1164/ajrccm-conference.2023.207.1_MeetingAbstracts.A6211

[3] Characterizing the Clinical Adoption of Medical AI Devices through U.S. Insurance Claims, New England Journal of Medicine, https://onepub-media.nejmgroup-production.org/ai/media/b35da8b4-b078-492b-ae20-bf938063e91f.pdf

[4] 2024 Medicare Advantage and Part D Star Ratings | CMS