Oasis Or Mirage? Digital Dementia Screening In The Neurology Desert
"Digital Paper" ≠ Innovation
In my neurotech travels the most common companies I encounter are those building digital dementia screening tools. These are a class of brain health assessments to see whether a person might have, or be at risk of developing, cognitive impairments that could be signs of neurodegenerative disease.
The number of people with Alzheimer’s disease is on pace to triple by 2060, as per CDC estimates. How best to treat them remains an open question, one that hinges on a more immediate question: Who are we treating?
Improving our ability to find people living with cognitive impairment entails identifying what disease they have, earlier in the disease’s progression. But doing this in a cost-effective, patient-centered, physician-friendly manner represents a key priority for the likes of Bill Gates. And he’s not alone.
As I noted at this year’s JP Morgan conference: “a neurotechnology home run would be a reliable diagnostic test (like a set of biomarkers) for earlier detection of mild cognitive impairment and/or Alzheimer's disease.”
Putting trial recruitment and digital endpoints aside for a moment, let’s consider the challenge of finding cognitive impairment in outpatient care. This problem represents an opportunity that could save hundreds of billions of dollars over the coming decades, while fundamentally changing the course of this disease and the search for a cure.
It’s a complex and rapidly changing market. Not all dementia is Alzheimer’s. Not all cognitive impairment is dementia. Not all screening is diagnostic. Payments are small, volumes are big. And new tools are coming, but old ones aren’t leaving: Subjective legacy assessments dominate thinking from medical school all the way through businesses’ quarterly statements.
Early detection can change a person’s life. Yet underreporting and stigma persist. Training and resources to adopt new tools aren’t always available.
For the healthcare system facing the “silver tsunami,” the need is clear: Better screening to find warning signs, predict disease onset, and help prioritize who gets a closer examination.
Inside a Messy Market
Today’s standard of care to evaluate cognitive health involves an acronym salad of written or verbally administered cognitive assessments: MMSE, Mini-Cog, GPCOG, MoCA, SLUMS. These can be run by various clinical staff and reimbursed through standard billing codes. Additional proprietary tests abound.
Most technology-based brain health tests today feature some interactive, proprietary screening process. Unfortunately, due to billing limitations, many “digital tests” are just electronic versions of the above assessments, though more creative tools are emerging.
Some startups emphasize quickness and affordability of use. Others emphasize objectivity through integrated biometrics. Virtually all incorporate AI for scoring analysis and standard reporting features for clinical teams. No two tools are the same, but many are targeting the same problem.
Beyond an aging population and growing prevalence of disease, several business pressures define the growing market of dementia screening.
Alzheimer’s Disease and Related Dementias Neurology Desert Index (ANDI), 2021
Doctor Shortage: Over 20 states are classified as a "dementia neurology desert" by the Society of Actuaries. There are fewer than 10,000 practicing neurologists in the US as per the BLS, or roughly 1 per 33,000 people on average. HRSA predicts a 19% shortfall by next year. But this is about primary care, too: 85% of dementia patients are first diagnosed by non-specialists, but many primary care physicians lack training and time to proactively assess cognitive health of their patients. Rural areas have even greater shortages of both primary and specialty care providers, underscoring the opportunity for scalable, easy-to-use tests.
Regulatory Yahtzee: FDA has sanctioned numerous devices, apps, and algorithms to help identify - but not diagnose - dementia. The agency’s gatekeeping, while understandable, is unfortunately detached from clinical practice. Their fragmented listing and lack of upkeep or public guidance about these clearances and approvals means that bricky, “computerized” tools cleared over a decade ago may hold more market power than today’s cutting edge tech. What’s more: Poor coordination between federal agencies like the NIA, NIH, CDC, and FDA on this single issue - dementia screening tools (p 59) - may be the country’s biggest missed opportunity to help millions of patients and families.
Reimbursement: A $268 payment for cognitive assessment visit, along with a handful of additional CPT codes and modifiers tied to annual wellness visits, pre-and post-test evaluation and interpretation, collectively create the clinical market for digital dementia screening. However, CMS loves legacy assessments, like the CDR scale or the FAST Screening, which are still considered reliable and standardized primarily because they’ve been around for forty years. Take a look below and ask yourself why only an estimated 8% of older Americans living with mild cognitive impairment receive a diagnosis.
Quality of Diagnosis: Biology beats Digital
Given how outdated clinical practice remains when it comes to assessing dementia, biomedical science has thankfully brought in greater degrees of accuracy through pathological testing to identify biological presence of disease. Digital dementia screening fits into this broader landscape of diagnostics.
The gold standard for definitive diagnosis of Alzheimer’s disease is the PET scan, which tags brain plaques with radioactive molecules (see this excellent history at Endpoints News.) But these are time-intensive at two hours and up, expensive, priced at $7-10k and higher per test, with only about 2,500 PET scanners available in the US.
Despite these less-than ideal conditions, CMS removed reimbursement restrictions on PET scans last October, as part of a market readiness push for new treatments. After this change, the number of claims for PET scans rapidly increased by approximately 5x, as per Eisai’s latest briefing. But this isn’t a silver bullet: The pharma still missed their patient enrollment target by 80 percent.
Cerebrospinal Fluid (CSF) testing, involving lumbar puncture, also offers high accuracy, but carries considerable expense, risk, and pain.
If there is a silver bullet, it may be a blood test.
“I think there’s a very good expectation that the blood test could replace the PET scan,” says Howard Fillit, cofounder and chief science officer at the Alzheimer’s Drug Discovery Foundation (ADDF).
To date, ADDF’s Diagnostics Accelerator has invested more than $60 million in 58 projects, mostly blood tests, and about one-fifth on digital approaches. Neurotech companies, take note: they’re currently open for proposals.
So while routine blood tests to detect Alzheimer’s disease appear to be imminent, in the US and globally, in the absence of these tests, and until their widespread rollout for early detection and soon, monitoring of side effects from anti-amyloid drugs, frontline clinical teams will still require assistance.
Quantity of Screening: Digital Beats Biology
Public Health is a numbers game: When you’re talking about 55 million people around the world, blood tests versus digital tests is the wrong debate. It comes down to “diagnostic workflows” to triage risk and direct next steps.
Digital dementia screening tools are an early step in a process that might also include a routine physical exam, aforementioned questionnaires, and “historian” (caregiver or family member) interviews, before escalating to pathological tests to pin down a definitive diagnosis (e.g. CSF, or soon, blood tests).
These companies cannot use the word “diagnostic,” reverting instead to words like “assessment” or “assistive” or “adjunctive.” In medical jargon, they’re clinical decision support tools that can be administered by medical assistants or ancillary staff without extensive training.
But these tools don’t exist in a void. AI-powered innovation is coming from all sides:
Games: Next gen tools use clock-drawing, card-playing, shape-navigating, Augmented Reality-questing, and various other interactive approaches to score brain health. Evidence is growing that, at least compared to yesterday’s questionnaires, these approaches are effective.
EHR data: In Medicare Advantage, and CMS’ GUIDE dementia care model, startups are building care management platforms that run on advanced analytics, which researchers are validating can be used to predict which patients develop Alzheimer’s disease.
MRI scans: The FDA just approved the first MRI-based, AI-enabled prognostic test to predict the future of “amnestic mild cognitive impairment” and provide the likelihood of progression to Alzheimer's disease within 5 years. AI-based MRI are changing clinical care for stroke, cancer, and brain injury; can they improve dementia care, too?
EEG: Companies like Cognision and VoxNeuro are bringing EEG-based assessments of cognitive function into the clinical setting. If biomarker-based approaches cross over from research interest to practice-based tool, opportunities to integrate clinical trial recruitment into clinical practice more tightly could be a win-win for public health and life sciences.
On the horizon, eye testing, vocal biomarkers, and other neurotechnologies are in research and development. But even if the science and technology behind all of these innovations are valid, it won’t matter much until they reach the outpatient market as a usable and cost-effective part of clinical practice.
RIDDLE Me This: A Long-Term Solution?
How can a time and cash-strapped neurology or primary care practice decide what screening tools are best, or in what order to use them? EHR analytics before wellness visits? Games in the waiting room? MRI before care plan? Or, just wait until blood tests become available?
The Alzheimer's Association, perhaps the most influential body for practicing physicians, is working on answers. Dr. Rebecca Edelmayer, senior director of scientific engagement, recently said, “The field is moving so fast...Even the recommendations we put out this year—likely they’ll need to be updated again within a year.”
What’s more: Beyond being able to stratify and reliably diagnose patients, the economics of what happens next are another quagmire. The handoff between outpatient clinicians and clinical trials recruitment requires additional confirmation testing, which varies by each combination of clinical research organization, pharma sponsor, and insurer.
What about Pharma’s newly approved and soon-coming IV-based drugs targeting amyloids? Changing prescriber behavior remains a thorny path, involving heavy yet unknown administrative lifts and overhead expenses to coordinate visits, imaging, reporting, as well as rigorous clinical monitoring and compliance.
Again: Eisai, with an $11B market cap and the only real drug on the market, couldn’t meet the needs of 80 percent of the people eager for treatment. We’re not there yet.
One neurotech chief executive I asked about those enrollment woes was blunt: “There’s just not a sufficient incentive for physicians to diagnose those patients. It’s not just about getting paid transactionally - if you’re a physician running a practice, you’re worried about what’s next: more questions, more liability, more compliance. It’s a straight up negative incentive structure.”
His short-term idea was “lining up serious payment to get doctors to do this.”
For a longer-term approach, I’d eschew the US federal government’s byzantine National Plan - wherein a multimillion dollar digital screening toolkit built by a research arm is not endorsed for official clinical use by either the regulatory or the payment arm - and turn instead to Europe.
The Real-World Implementation, Deployment, and Validation of Early Detection Tools and Lifestyle Enhancement (AD-RIDDLE) is a 24-member collaborative testing out the real world economics of an Alzheimer’s Disease toolbox, using a shortlist of government-cleared (CE mark) tools in sync with lab tests, clinical workflows, longitudinal tracking, and real-time evidence generation.
Whether or not it works, the coordinated approach is the notable innovation.
AD-RIDDLE “envisions the continuum of AD screening, diagnosis, and treatment as a pipeline, and right now, that pipeline is complex. The project aims to take apart the pipes, identify the obstacles and their location, and put them back together again so that patient and provider can flow more smoothly, making a difficult process more efficient, effective, and patient-focused.”
Innovating On Shifting Sands
So, what does this all mean for neurotech startups, investors, and partners angling for opportunities to improve dementia screening?
Short-term, R&R (regulation & reimbursement) is tricky, but don’t overlook the value of denting the number of undiagnosed patients and getting a foothold in the clinical market.
Mid-term, diversify revenue. Life sciences is an obvious market, with less regulation in exchange for what one CEO called “lumpy” (read: sporadic) contracts. Life insurers, research groups, and overseas markets are all buying. Direct pay clinics, other neurotech companies, or even consumers might become markets, too
Long-term, don’t underestimate the value of engaging directly with patient and caregiver communities as you map the next generation of your product. Their expert knowledge of living with dementia hardly factors into the above challenges, making them the largest underserved market in the world - one with untold untapped opportunities to improve efficiency and impact of clinical care, drug trials, and even federal policy.
Originally published in Forbes, April 26, 2024.
👋🏽 Collaborators and Sponsors: Build with me. LinkedIn | Twitter/X | Email
❤️ Please click the heart and share with anyone who’s into the future of brain health!