SNMMI Hosts AI and Patient Access Summits

March 28, 2022

Last week, SNMMI hosted two summits in Bethesda, MD: the Artificial Intelligence Summit and the Barriers to Patient Access Summit. These summits were the first of their kind for the society and took place over the course of three days (March 21–23). The aim of the summits was to bring together thought leaders to discuss current obstacles in the field of nuclear medicine and provide solutions. Below are some of the highlights of both summits.

AI Summit

The Artificial Intelligence (AI) Summit was organized by the SNMMI AI Task Force and brought together stakeholders including data scientists, industry representatives, end users of AI, and federal regulators and funders (Food and Drug Administration and National Institutes of Health). AI is impacting both radiology and nuclear medicine. The central theme of this summit was that AI models need to be aligned with the needs of patients. The data show us that representation does not equal fairness. In that vein, some of the following questions were posed: How do we make decisions that respect justice? How can we address biases in medicine?

The aim is to further distributive justice using AI—that is, improve the quality of care for patients instead of widening the gap among certain populations. Ways in which we achieve this aim is by ensuring privacy, security, and health equity. Many attendees favored open access algorithms with regular maintenance and post-deployment quality control and upgrades.

Nuclear medicine imaging, especially PET, is geared toward harnessing a patient-friendly experience in which patients can see their own images and understand what they mean. Nuclear medicine technologists are a vital part of the patient experience, and they likely will be the ones to set the table regarding workflow; thus, industry should include them in evaluations regarding AI. Additionally, digital twins were discussed and how the field is well positioned to create theranostic digital twins to enhance treatment planning and administration. One need underlying all AI algorithm development is the need for well-curated data. This is one area where SNMMI can begin to make an impact to advance the field.

Slide presentations from the AI Summit may be found here.

Access Summit

The Access Summit kicked off with a global scan of the field. The availability of imaging equipment is low in middle to low-income countries, and finding a well-trained workforce is a problem. These issues are also faced by nuclear medicine in the U.S. today. Though the demand for technologists is high, training programs are costly to universities and are often the first ones to be cut. Medical physicist residency programs are also inadequate. Problems also exist in relation to infrastructure and supply and coverage, coding, and reimbursement.

Richard Wahl, MD, SNMMI president, commented that the field is growing tremendously in the therapy space—so much so that nuclear medicine physicians are becoming "nuclear oncologists." To ensure patients have access to all the field has to offer, several speakers explained the importance of diversity in clinical trials leading to more meaningful results for real-world care and treatment. Diversity should be a factor not only for trial participants but for staff and care teams as well. Community involvement from all age groups is also necessary to make sure certain studies, such as the New IDEAS trial, met their representativeness goals.

Although there has been a decline in general nuclear medicine studies, there has been a huge increase in PET studies. Also, it is forecasted that in the future, most revenue in nuclear medicine in will come from radiopharmaceutical therapies. While manufacturers have been gearing up to provide a variety of doses to patients for scans by creating both centralized and decentralized distribution pathways, the rest of the nuclear medicine community is scrambling to prepare for the deluge of patients. Large academic sites are putting in second cyclotrons. SNMMI is working to aid the workforce pipeline by identifying challenges for entry and retention and assisting in recruitment.

Other barriers exist with coverage, coding, and payment. Medicare needs major reform, especially with respect to its physician fee schedule and hospital outpatient prospective payment system (HOPPS). For example, the HOPPS “packaging” or “bundling” policy has been a major impediment since 2008 for patients in re access to the latest diagnostics for diseases including Alzheimer’s disease, Parkinson’s disease, and various cancers. After a 3-year pass-through (or “grace”) period, diagnostic radiopharmaceuticals are reimbursed at a rate less than 10% of their average sales price (plus 6%). Lack of appropriate hospital reimbursement leads to fewer hospitals offering the drugs, especially those in rural areas. The Facilitating Innovative Nuclear Diagnostics Act (FIND) Act, HR 4479/S 2609, is a legislative fix to this patient access problem.

Other battles need to be won on both the coverage and payment fronts. Amyloid PET is covered by Medicare only if the patient is a participant in a CMS-approved clinical trial. These tracers are also “packaged” and therefore not reimbursed appropriately. We eagerly await CMS’s final decision on the monoclonal antibody therapies coverage for Alzheimer Disease in hopes that they will cover amyloid PET tracers broadly. Policy decisions made by Medicare are usually followed by private payers, so it is crucial to get pro-patient policies right the first time.

The summits concluded with several solutions put forward about how SNMMI and others may further expand patient access. John Sunderland, PhD, suggested that the society act as a nexus for information distribution about workforce pipelines by sharing programmatic templates for medical physicist residency and other training programs. When academic sites relinquish exclusivity of tracers, that allows both commercial and additional academic manufacturers to set up distribution channels faster. The society can also work with private payers to eliminate prior authorizations for imaging when a referring physician uses a clinical decision support system (with corresponding appropriate use criteria). Currently, the society is examining decreasing the time delay for ANDAs through priority review channels.

In summary, nuclear medicine, molecular imaging, and radiopharmaceutical therapy is a thriving, research driven field that is improving human health. AI will continue to advance in this space and will help extend quality care to those who currently may not have access.