From Prototype to Product.

01
The gap

• 24 million Americans want cognitive screening and cannot access it.¹
• 10 million adults over 65 have mild cognitive impairment.²
• 7 million people are living with undiagnosed dementia.³

Meanwhile, two FDA-approved amyloid-targeting therapies, lecanemab (Leqembi) and donanemab (Kisunla), are only indicated for patients with mild cognitive impairment or mild-stage Alzheimer's disease.⁴ Once someone progresses past that window, they are ineligible. The clinical value of early detection is no longer theoretical.

The structural barrier is primary care.

The average primary care visit runs 15 to 18 minutes. Cognitive assessment is not part of most annual wellness visit workflows despite Medicare covering a cognitive evaluation under the Annual Wellness Visit (G0438/G0439).⁵ Reimbursement for dedicated neuropsychological testing (CPT 96116, 96146) requires a licensed practitioner, formal test administration, and documentation that most primary care offices are not staffed to produce.⁶

The result: screening happens in memory clinics after a family has already spent months to years noticing decline and working up the courage to seek help. By then, the treatment window may have closed.

First Signs Health addresses the gap upstream of the clinical encounter. It is a browser-based, three-gate screening protocol that a family caregiver administers at home, using validated instruments, and generates a clinician handoff packet that gives a doctor something concrete to act on at the first appointment. It does not diagnose. It does not replace clinical evaluation.

It fills the dead space between "I'm worried about my mother's memory" and a formal workup.

02
Why these instruments

The three-gate architecture moves from fast triage to domain-level assessment to longitudinal monitoring.

Each gate uses instruments selected against four constraints: psychometric validity, public availability (no licensing fees or access restrictions), appropriateness for lay administration, and complementary coverage across the screening-to-monitoring arc.

Gate 1 · SLUMS (Saint Louis University Mental Status)

The SLUMS was chosen over the Mini-Cog for triage. Both are brief, but the SLUMS has greater sensitivity to mild cognitive impairment. Tariq et al. (2006) reported sensitivity of 0.92 and specificity of 0.81 for MCI detection in the original validation study.⁷

The Mini-Cog is faster (3 minutes vs. 7), but its primary design target is dementia-level impairment, not MCI, which makes it less useful for early detection.⁸ The SLUMS is fully public domain from Saint Louis University, with no licensing restrictions. The Mini-Cog requires registration and restricts commercial use.⁹

Gate 1 · Short IQCODE

The Short IQCODE (16 items) is paired with the SLUMS to add an informant-report dimension. Direct cognitive testing can miss decline in high-functioning individuals who compensate well on structured tasks. The IQCODE captures functional change as observed by someone who knows the person.

Jorm (1994) established the psychometric properties of the short form, which correlates at 0.98 with the full 26-item version while cutting administration time roughly in half.¹⁰ The 16-item version was chosen over the full 26-item version to minimize caregiver burden in a home screening context where attention and patience are limited resources.

Gate 2 · ACE-III

When Gate 1 flags a concern, the ACE-III provides domain-level granularity across five cognitive areas: attention/orientation, memory, verbal fluency, language, and visuospatial function. Hsieh et al. (2013) reported sensitivity of 1.00 and specificity of 0.96 at a cutoff of 88/100 for distinguishing frontotemporal dementia and Alzheimer's disease from healthy controls.¹¹

The ACE-III was chosen over the MoCA for two reasons: it produces domain subscores that map cleanly onto clinical referral questions ("memory-predominant or language-predominant?"), which the MoCA does less precisely; and the MoCA charges a licensing fee for digital or commercial use, while the ACE-III is freely available with registration from the University of Sydney.¹²

Gate 3 · SAGE

SAGE is the only validated instrument in this set designed for self-administration without a proctor. Scharre et al. (2010) demonstrated sensitivity of 0.79 and specificity of 0.95 for detecting MCI.¹³ Four parallel forms allow repeated testing at six-month intervals without practice effects confounding the results. SAGE is publicly available from Ohio State University.¹⁴

03
The handoff packet

The product's most important output is not a score. It is the clinician handoff packet: a structured document that a caregiver prints or downloads and brings to a doctor's appointment.

The packet includes:

• Patient demographics and caregiver relationship.
• SLUMS total score with education-adjusted category and item-level detail. Short IQCODE mean score with individual item ratings.
• ACE-III total score, domain subscores (attention/orientation /18, memory /26, fluency /14, language /26, visuospatial /16), and item-level responses.
• SAGE longitudinal history with a trajectory chart showing scores across sessions.
• Administration quality metadata: caregiver's self-assessed confidence on subjective scoring, whether the timer was used, whether cueing may have occurred.
• A boxed disclaimer stating that the instruments were administered by a lay caregiver in an unvalidated administration modality.
• Validation citations for every instrument used.

The packet is designed for the receiving clinician, not the patient. It answers the three questions a primary care physician or memory specialist needs answered at intake:

1. Is there objective evidence of cognitive concern?
2. Which domains are affected?
3. How reliable is this evidence?

Most consumer screening tools give users a score and a suggestion to see a doctor. This one gives the doctor a working document.

04
Competitive positioning

Every funded competitor in this market sells to clinicians, health systems, or pharmaceutical companies. Their pricing models, regulatory postures, and product architectures are built for institutional buyers.

None of them address the person sitting at a kitchen table who suspects something is wrong with their parent's memory and has no clinical pathway yet. That person is the user First Signs Health is built for.

05
Why the market gap persists

Three structural factors keep funded competitors out of community pre-screening. These are rational business decisions. They are also the reason 24 million people remain unscreened.

1 · Enterprise pricing

Health system software contracts run $50K–500K+ annually. That model requires institutional buyers, procurement cycles, and demonstrated ROI. A free community tool generates none of those revenue signals.

2 · FDA focus

Linus Health and BrainCheck have invested years and millions in regulatory clearance. Their products are designed to support clinical decision-making within a regulated framework. A community tool that explicitly disclaims clinical authority operates outside that framework.

3 · Clinician-administered workflows

Sensitivity and specificity data for these instruments assume administration by trained clinicians. Lay-caregiver administration is an unvalidated modality. No funded competitor wants to associate its brand with results from an uncontrolled home environment, even if those results are useful as screening indicators.

06
Business model

If every funded player makes money by selling to institutions, how does a free, community-facing screening tool sustain itself?

The current prototype is free and should stay free at the point of use. Charging anxious caregivers to find out if their parent might have dementia is not a viable positioning. The revenue has to come from the value the screening generates downstream, not from the screening itself.

1 · Telehealth patient acquisition

Dementia-focused telehealth services have providers but no public-facing screening funnel. First Signs Health fills that gap: caregivers complete the protocol, get results warranting evaluation, and route directly to a telehealth partner. The partner pays a per-referral fee for pre-qualified leads who've already completed cognitive triage.

2 · Health plan partnerships

Medicare Advantage plans have a financial incentive to identify cognitive impairment early. Members with undiagnosed MCI generate higher emergency department utilization and worse medication adherence.²⁶ A pre-screening tool that compresses time-to-diagnosis has quantifiable value to a plan that bears downstream risk. Per-member-per-month or per-completed-screening, paid by the plan, free to the member.

3 · Employer wellness programs

Large self-insured employers are beginning to add cognitive screening to benefits portfolios.²⁷ A caregiver-administered tool fits naturally into an EAP or caregiver benefits package.

4 · Clinical referral network integration

Memory clinics have wait times in months.²⁸ A pre-screening tool that triages patients before arrival, with a structured handoff packet, eliminates the first 30 minutes of intake. Per-referral fee or practice subscription.

07
Architecture for production

The current prototype proves that the clinical logic works. A production version has to answer three questions:

1. Where does the data live?
2. How does it get to the clinician?
3. What can we learn from it at scale?

Data layer

Store screening results in encrypted cloud storage with user-controlled access. The consent model: explicit, granular, time-limited access tokens. HIPAA applicability depends on whether the product transmits PHI to a covered entity.²⁹

EHR integration

The handoff packet should be transmittable as a FHIR document — Observation, DiagnosticReport, QuestionnaireResponse.³⁰ Distribution: patient-portal upload is the pragmatic starting point.

Population analytics

Aggregated, de-identified screening data would be valuable for public health surveillance and instrument validation research. Downstream priority — but the data model should support it from day one.

08
Regulatory posture

The FDA's "Policy for Device Software Functions and Mobile Medical Applications" (revised September 2022, updated January 2026) defines which software functions FDA intends to actively regulate.³¹ The January 2026 update broadened enforcement discretion for certain CDS and general wellness products.³²

First Signs Health presents three arguments for remaining outside FDA scope:

1 · No diagnostic output

It provides raw instrument scores with published reference ranges and a recommendation to consult a physician. Under the 21st Century Cures Act's four-criteria CDS test, software that displays information so a clinician can independently review is excluded from the device definition.³³

2 · Non-proprietary instruments

The instruments themselves are published, publicly available cognitive screening tools with established psychometric properties. The software is administering and scoring them, not generating novel clinical interpretations from opaque models.³⁴

3 · Caregiver-directed

The product is directed at caregivers, not health care professionals. The September 2022 CDS Final Guidance removed enforcement discretion for patient-directed CDS, but the January 2026 update revisited this position, providing broader enforcement discretion for low-risk software.³²

What would trigger reclassification

If the product begins generating proprietary risk scores, algorithmic diagnostic predictions, or treatment recommendations a user cannot independently verify against published reference data, it crosses into device territory. Staying on the right side means maintaining transparency of scoring logic, presenting published cutoffs rather than proprietary thresholds, and never claiming diagnostic capability.

09
Future Vision

I built this tool becuase I see a gap in the market, and I beleive there is a market in the gap. Today I seek advice, ideas, and funding to better define the gap and develop the market for this tool in three areas.

Please email me now if you (or someone you know) can help with one of these activities.

1 · Screening awareness

Targeted campaigns reaching worried caregivers where they already search for answers, through paid search, caregiver communities, and content that surfaces during late-night "is this dementia" queries.

2 · Credibility partnerships

Collaborations with organizations caregivers already trust, like Alzheimer's Association chapters, AARP, senior care networks, and faith-based organizations serving aging populations.

3 · Post-screening pathways

High-risk results should route directly into telehealth dementia services, like Isaac Health, Synapticure, and OSF TeleDementia, where pateints (and pharamcutical companies) can benefit from earlier screening and diagnosis.