Families of people with seizure disorders and complex medical needs face an impossible choice: give up privacy for safety, or accept risk for dignity. This playbook is how Stark County is changing that — through privacy-first remote sensing technology, formal data governance, and honest measurement — for individuals with co-occurring developmental disabilities and seizure disorders in Stark County.

Table of Contents

This playbook contains 10 sections. Click any title to jump directly. Total: ~51 min read

Quick Reference

How to Use This Playbook

Program Partners

The Problem This Solves

Families of people with seizure disorders and complex medical needs face an impossible choice: give up privacy for safety, or accept risk for dignity.

Approximately 25% of the intellectual and developmental disability population — over 1,250 people in Stark County alone — also live with a seizure disorder.^[1] These individuals are among the most medically vulnerable in the DD community. Their families and caregivers are often afraid to allow unsupervised time or privacy, because a seizure can happen without warning and the consequences can be fatal. The result is isolation, restricted independence, and well-intentioned but limiting safety measures that shrink a person’s life.

This playbook exists because that tradeoff should not be permanent.

Who This Program Serves

This program serves individuals with co-occurring developmental disabilities and seizure disorders in Stark County. It is specifically designed for people with complex medical needs where nighttime safety is a critical concern.

The program has three goals:

• Implement customized safety technology — seizure-related safety monitoring (post-seizure fall detection, movement alerts, and prolonged immobility notifications) and assistive devices that advance independence at home, school, day programs, and in the community
• Enhance long-term safety and enable smarter support planning — near-real-time alerting that helps caregivers respond faster and make data-informed decisions about the right level of support
• Enable people with seizure disorders to live life to the fullest — increase opportunities for self-determination and community inclusion by balancing safety with freedom

What Is Privacy-First Design?

Privacy-first is not just a technology choice — it’s a design philosophy with governance behind it. It means every layer of the system is built to collect the minimum data needed and protect it with the maximum safeguards available.

At the sensor level: thermal cameras capture heat signatures only — silhouettes, not photographs. No faces, no identifying features — though body shape and activity are visible. Privacy is inherent because heat patterns cannot reveal identity. Software filters further limit what the human viewer sees. Additional software filters can further reduce visual discernibility when needed.

Original Radiometric Thermal Every pixel carries real temperature data — not a photograph. The sensor maps heat intensity to color, showing warm and cool areas across the scene. No faces. No features. Just thermal readings that map heat, not appearance.

Grayscale Heat intensity as brightness. Used for clinical review where color is not needed.

Negative Inverted contrast — cooler areas bright, warmer areas dark. Useful for detecting edges and boundaries.

Gamma Corrected (Shadow-Enhanced) Shadow detail enhanced — reveals room context and posture without increasing identifiability.

In every rendering, the person remains unidentifiable. Privacy is inherent to the sensor — not a filter applied after the fact.

At the access level: who can view what, when, and why is governed by strict rules:

• Shift-based access windows — staff can only view live thermal during assigned shifts, and only when a specific trigger occurs (e.g., an out-of-bed alert)
• Emergency access protocol (break-glass) — emergency viewing requires a documented reason, supervisor approval, a 30-minute maximum window, and a post-event review within 1 business day
• Privacy Mode — outside scheduled monitoring hours, the system stops all recording and blocks all live viewing — no data is stored or transmitted
• Manual controls — you can cover the lens, mask regions of the room (e.g., bathroom, closet), switch to a less intrusive sensor, or pause monitoring entirely at any time with no explanation needed

All privacy and data protections are formalized in the Data Governance & Security Policy (DGSP) — a comprehensive governance document available for download in Section 10: Downloads & Resources. Key provisions are covered in the FAQ (Section 9).

An Honest Starting Point

We want to be transparent from the first page: thermal sensing detects what happens after a seizure — not the seizure itself. It sees post-seizure falls, prolonged immobility, and unusual position changes. It does not detect electrical brain activity. If a person has a seizure in bed without significant movement, thermal alone may not detect it.

How Monitoring Improves Over Time

At launch, trained staff are watching. The thermal sensor shows heat patterns on a screen — shapes and warmth, not faces or photos. Staff are the detector. The sensor is their tool. The sensor reads body heat — the person’s warmth and everything it touches. It catches what eyes at a doorway would miss: restlessness, subtle position changes, the small things that matter most.

What staff can see from day 1: bed occupancy, person on the floor, room presence, general movement. These are things a human recognizes instantly on the thermal feed.

What the system learns over time (30–90 days): while staff watch, the system quietly builds a picture of what is normal for this person — when they settle into bed, how often they move, how long they stay still, what a restless night looks like versus a calm one.

How automated capabilities improve: as the baseline builds, the system starts to flag things that would be hard for a human to notice — subtle shifts in restlessness, a longer-than-usual period of stillness after movement. Staff are still watching. The system just adds a second set of eyes that never gets tired.

Notifications always go to awake, on-duty staff. Never to someone who is sleeping. No existing supports are removed until the team has real data showing the system is reliable for this specific person.

A Note on Seizure Monitoring

Thermal sensing does not detect seizure electrical activity — only an EEG can do that. What thermal can detect are the movement and positional changes that may result from a seizure: a fall from the bed, rapid repetitive movement, or prolonged immobility after an event. We consider this a moderate confidence proxy detection. It should always be part of a broader seizure safety plan, never the only safeguard.

The System Learns Over Time

As the system gathers baseline data about a person’s movement patterns and routines, its detection confidence improves. This typically requires 30 to 90 days. During this period, existing supports should remain in place. The system does not start at full accuracy — it builds a picture of what “normal” looks like for each individual, then flags deviations from that baseline.

This is why the program uses a layered approach — thermal as the privacy-first baseline, with the option to add wearables, smart mats, or other modalities based on individual need. The CtLC (Charting the LifeCourse) scoring system in Section 3 helps determine the right combination for each person.

Adding Privacy Layers The sensor already cannot capture faces or features. Software filters let us blur the image further — as much as your family wants.

Light Privacy

More Privacy

Maximum Privacy

Every version shows the same thing: someone is there, they are moving or still — and no one can tell who they are.

^[1] Estimate based on Stark County Board of DD enrollment data and CDC epilepsy prevalence rates for intellectual and developmental disability populations.

A thermal (infrared) sensor captures heat patterns only. It does not see visible light, color, faces, fine details, or any identifying information. Below is what thermal looks like in practice.

Thermal vs. Camera: Side by Side

The comparison in the header above shows the difference.

What Thermal Sees

• Movement: Person's body heat moving in and out of the frame
• Location: Which areas of a room or bed are occupied
• Posture changes: Lying down, sitting up, standing, falling
• Absence: When a monitored area becomes empty (e.g., person left the room)

What Thermal Does NOT See

• Faces, expressions, or identifying features
• Gender, age, or appearance details
• Room details (furniture labels, book titles, personal item identification)
• Clothing color or style
• Identity — the sensor sees multiple heat signatures but cannot tell one person from another
• Fine motor movements (e.g., if you're reading or writing)

Technical Specifications

How Data Flows

The system records: time of alert, which room, what kind of movement (e.g., "person got out of bed"), and staff response time. This log helps the support team and healthcare providers see patterns over weeks or months. For full details on data practices, see the Data Governance reference in Downloads & Resources.

Real Example: A Night in Bed

Standard camera: Would show your child sleeping, waking, moving, possibly in a private or intimate moment.

Thermal sensor: Shows body heat, posture, and movement — but no face and no way to identify who the person is. Staff see a privacy-filtered view, not the full thermal data. The system can never reveal identity. If the person gets up at 2 AM, the system alerts staff and they can check in.

Families and participants retain complete dignity and privacy—staff only know if something changed, not what it looked like.

The sensor watches for heat patterns in your loved one’s room — nothing more. Think of it like a weather map: warm areas show up in bright colors, cool areas stay dark. Here is what the system actually sees in practice, and the kind of notification each situation triggers.

High Confidence

In Bed — Normal Sleep

The sensor sees a warm shape in the bed area, staying mostly still. This is normal sleep. No alert is sent. The system simply confirms your loved one is in bed and resting comfortably.

High Confidence

Fall Detected — Person on Floor

The sensor sees a warm shape suddenly appear at floor level, outside the bed. The system sends an immediate notification to the Remote Support team, who can then follow the escalation plan your team set up.

High Confidence

Out of Bed

The bed still shows leftover warmth, but the person’s heat signature has moved to another part of the room — or left the sensor’s view entirely. The system sends a notification that your loved one has left the bed. If they are up longer than expected based on their usual nighttime pattern, a follow-up alert is sent.

Moderate–High Confidence

Prolonged Immobility

The sensor sees a warm shape that has not moved for longer than your loved one’s normal baseline. This could mean deep sleep — or it could mean something needs attention. The system sends a notification so the Remote Support team can check in. This detection gets more accurate over time as the system learns what is typical.

Moderate–High Confidence

Restless Sleep

The sensor notices the warm shape in bed is shifting position more often than usual. The system logs the increased movement and sends a notification if the restlessness exceeds your loved one’s normal pattern. This information is also shared in monthly reports so the support team can look for trends.

A Note on Confidence

Not every detection is equally reliable, and we believe you deserve to know the difference.

Falls and bed exits are the easiest for the sensor to spot — the change in position is large and clear. We call these high confidence detections. Prolonged immobility and restless sleep take more time to get right because the system needs to learn what “normal” looks like for your loved one first; we call these moderate confidence, and they improve steadily over the first 30 to 90 days.

For seizure-related concerns specifically: the sensor does not detect seizures themselves. It detects what may happen after a seizure — a fall to the floor, unusual stillness, or a sudden position change. We consider this a moderate confidence detection and it should always be part of a broader safety plan, never the only safeguard.

The CtLC Scoring System

Every proposed technology configuration is scored across 6 dimensions using the Charting the LifeCourse (CtLC) system. The score determines which Tier the configuration falls into, which drives whether the team should proceed, modify, or redesign.

6 Dimensions

Each dimension is rated 1–5. The weighted total produces a score from 0–100.

Tier Classifications

The Safety-Critical Modality Gate

This is the most important rule for seizure scenarios:

Most thermal-only seizure scenarios will score Tier C. This is expected and honest. A Tier C score with documented limitations is far better than a Tier A score built on false assumptions. The Gate is protection — for the individual, the team, and the program.

To address the Gate:

• Accept Tier C — document limitations, set narrow claim scope (valid choice)
• Add event-triggered audio — partially addresses Gate, moves Technology Fit to 3
• Add wearable seizure device — substantially addresses Gate, moves Technology Fit to 3–4

Privacy First

Many families reject monitoring technology because cameras feel invasive. Thermal is different. It sees only heat and movement — not your child's body, face, or any identifying detail. The system sends a notification when it detects certain changes — like someone getting out of bed or being still for too long. A real person then decides what to do.

A traditional camera captures your loved one's face, their expression, the details of their room. Anyone who sees that footage sees them — personally and clearly.

A thermal sensor sees heat. Body shape is visible — posture, movement, whether someone is lying down or sitting up, how they shift during the night. At close range, facial movement is visible. What the sensor cannot do is tell you who that person is. No recognizable face. No identity. Just warmth and movement.

That is the privacy case — not that thermal sees little, but that thermal cannot identify. And unlike consumer camera systems, every moment of access to the thermal feed is controlled, logged, and tied to a specific staff member on a specific shift.

Common Family Concerns & Answers

What to Ask Your Organization

If your child's home, school, or day program is considering thermal:

• "What problem are you trying to solve?" (Must be clear: wandering? falls? seizures?)
• "Who responds to alerts and how fast?" (Expect answer: staff responds within 1–2 minutes.)
• "Can I see the data?" (You should get summaries, incident reports, or pattern analysis.)
• "How is my child's privacy protected?" (Expect: "Thermal only — no visual identity. Strict rules control who can see the data, when they can see it, and what they can do with it.")
• "What happens to the data?" (Expect: "Protected with the same security used by banks and hospitals. Deleted on a schedule set in your care plan — typically 30 to 90 days for routine monitoring.")
• "Can I opt out?" (Should be optional, with alternative monitoring plans.)

Your Role in Success

• Stay informed. Ask to see pattern reports monthly. Collaborate with doctors and staff on what the data means.
• Share context. Tell staff about your child's behavior patterns: "She usually gets up to use the bathroom around 2 AM but sometimes has a seizure at 3 AM. Here's what to look for."
• Troubleshoot together. If alerts seem inaccurate, let staff know. (Maybe the sensor placement needs adjustment.)
• Use the data for health. Share incident logs with your child's doctor. Use patterns to adjust medications, therapies, or routines.

Who Needs Thermal Monitoring?

Prioritize people with:

• High seizure frequency (3+ per week) + nighttime occurrence — thermal detects post-seizure movement (falls, thrashing, prolonged immobility) as a proxy; it does NOT detect seizure electrical activity itself. Over time, as the system learns this person’s patterns, it can help staff notice subtle changes they might otherwise miss
• Documented wandering/elopement risk (history of incidents, staff reports)
• Frequent falls (especially at night when observation is reduced)
• Severe sleep disturbance that affects daytime safety or behavior
• History of injury during unobserved periods (thermal alerts could have enabled faster staff response)

Don't prioritize thermal for: People with mild behavioral concerns, daytime needs (other monitoring approaches guided by the privacy tier framework may be more appropriate), or situations where issues are predictable and staff are present.

Implementation Checklist

Phase 1: Assessment (1–2 weeks)

• ☐ Identify 2–3 initial participants (clear safety need, family buy-in, manageable homes)
• ☐ Meet with families and participants. Explain privacy, get consent, address concerns
• ☐ Map residential spaces (bedrooms, hallways, bathrooms). Note dimensions, HVAC vent locations and heat sources that may affect sensor readings
• ☐ Assign staff champion(s) who will lead training and monitoring
• ☐ Request a pricing estimate from Medforall — see Equipment & Pricing for standard rates

Phase 2: Setup (1 week)

• ☐ Install sensors and connect to Medforall platform
• ☐ Test alerts, confirm staff receiving notifications
• ☐ Create response protocols: "Alert received → staff checks person within 2 minutes → logs outcome"
• ☐ Confirm Data Governance & Security Plan (DGSP) is in place, including data privacy agreements and access controls

Phase 3: Training (4 hours + ongoing)

• ☐ Train all overnight staff on: sensor capabilities, false alarm handling, response time standards, data logging
• ☐ Practice mock alerts (sensor sends fake alert; staff respond as if real)
• ☐ Distribute protocol sheets (laminated cards for staff rooms)
• ☐ Assign backup person in case primary staff member is unavailable

Phase 4: Live Monitoring (2–4 weeks)

• ☐ Monitor all incidents: time, type, staff response time, outcome
• ☐ Weekly check-ins with staff about problems or adjustments needed
• ☐ Weekly communication with families about any incidents or patterns

Phase 5: Evaluation (8 weeks)

• ☐ Analyze data: incident rates pre/post, response times, staff feedback, family satisfaction
• ☐ Gather family input: "Is your family member safer? Do they have more privacy and dignity?"
• ☐ Gather staff input: "Did this help? Any false alarms? Training gaps?"
• ☐ Decide: expand to more homes, adjust protocols, or pause program

Common SSA Concerns

ISP & Compliance Integration

Thermal monitoring must be formally documented within the Individual Service Plan (ISP) and aligned with Ohio DD regulatory requirements.

• Assessed need: Document the specific safety need thermal addresses (e.g., “nighttime seizure-related falls”)
• Service description: Include thermal sensing as Assistive Technology (AT) under the ISP. Describe monitoring scope, hours, and rooms.
• Desired outcome: Tie the technology to a measurable ISP outcome
• Privacy tier: Document which tier (T0–T5) is authorized. Default is T1 (thermal only).
• HRC review: Human Rights Committee must review and approve before implementation
• Billing: Remote Support billed through IO waiver, code AMU, at $2.37 per 15-minute unit ($9.48/hr)
• DGSP: Confirm Data Governance & Security Plan is in place (AES-256, TLS, shift-bound access, MFA, immutable audit)
• MUI: System event logs support MUI documentation. Data retained 7 years per Ohio requirements.

Consent Process

Consent is modality-specific, informed, and revocable at any time.

• Who consents: The individual (if capacity exists) and/or legal guardian. Even when a guardian consents, the individual should be informed in accessible language.
• What’s documented: Which rooms, which hours, which privacy tier, who has access, retention period, and what happens if consent is withdrawn.
• Revocability: Consent can be withdrawn at any time — verbally or in writing. System deactivated within 24 hours. Data deleted upon request.
• HRC prerequisite: HRC approval must be obtained before consent is sought.
• Renewal: Consent reviewed at each annual ISP review or sooner if monitoring scope changes.

Transition Protocol

• 30–90 day baseline: Sensor active but supplementing — not replacing — existing supports. All current staffing levels remain.
• Parallel operation: Staff continue scheduled physical checks. Thermal alerts are additional information, not a replacement for direct observation.
• Criteria for adjustment: Only after 30+ days of reliable data, consistent detection demonstrated, acceptable false positive rates, staff response times meeting standards, individual/guardian agreement, ISP formally updated, and HRC approval of any support reduction.

Detection Confidence by Use Case

Ongoing Reporting

Medforall and the Remote Support provider commit to structured reporting for every participant:

• Monthly: Alerts generated, response times, notable events, missed events, false positive rate
• Quarterly: Detection threshold review and adjustment, baseline recalibration, before/after documentation
• Confidence reporting: Updated assessment of detection confidence for each ISP-documented concern. If confidence is LOW, the report says so and recommends alternatives.

Data You Should Collect

Create a simple log for each monitored person:

After 8 weeks, you'll have 56+ data points that show whether thermal is reducing injury, improving safety documentation, and supporting your participants.

Equipment Package

The following are standard rates for the Medforall privacy-first thermal sensing system.

The Medforall system consists of three core components plus cloud service and optional remote support:

Ongoing Services

Purchase vs. Lease

Typical Deployment Costs

Single Participant (1 bedroom, 1 sensor)

One-time total: ~$2,120 | Monthly: $25 cloud + RS hours as needed

Two-Sensor Coverage (bedroom + hallway)

One-time total: ~$2,870 | Monthly: $25 cloud + RS hours as needed

Cost-Benefit Considerations

The financial case rests on three factors:

• Reduced emergency incidents — faster post-seizure response can avoid ER visits and hospitalizations. A single ER visit in Ohio typically runs several thousand dollars at minimum. A single avoided hospitalization can offset the entire first year of equipment and subscription costs.
• More efficient support allocation — remote support at $2.37/unit is substantially less than on-site overnight staffing. Even partial shifts from in-person to remote support produce meaningful savings.
• Better data for care planning — objective movement and incident data improves ISP quality, supports medication adjustments, and strengthens regulatory compliance documentation.

Thermal sensing does not replace clinical care, trained staff, or emergency protocols. It is a detection and alert tool that works alongside medical oversight, behavioral support, staff training, environmental design, and individualized care plans.

Below you will find answers to the questions we hear most often — about safety, privacy, your rights, and what to expect.

Safety & Detection

Privacy & Data

Your Rights

Ethical Commitments

Transparency & Accountability

All of these resources are available from your SSA or by contacting Stark County Board of DD directly.

For All Audiences

For Families

For SSAs

For Leadership

Contact Information