Turning Reality Health Games engagement into a standards-based, de-identified signal
that the Pennsylvania HIE, payers, and public health can actually use, price, and act on.
Illustrative prototype. Every number in the charts below is synthetic — placed only to show
what the tool would display and to frame the hypotheses. No real patient, member, or FFH user data is
represented. Any correlation shown here is a design mockup, not a finding.
The idea in one paragraph
CTO + CFO hats.
Right now, Healthy Coins are an internal reward — they buy gift cards and drive engagement inside
the Academy. The opportunity is to give them a second life as an interoperable health signal. If a
coin balance, a badge, or a "Certified Patient" credential is expressed in the same standard language an
HIE already speaks (HL7 FHIR), then engagement stops being a marketing metric and becomes a
measurable, exchangeable unit that sits next to claims and clinical data. The PA HIE's core job is
moving standardized health data between organizations. We give them a brand-new, prevention-side data
stream they don't have today: proof that a person (or a whole household) is actively learning and
changing behavior — and a de-identified way to test whether that engagement tracks with fewer ER visits
and lower cost.
1. Why coins become a "currency"
A currency needs a common unit and a shared ledger. FHIR is that shared ledger for health.
The move that makes coins valuable outside FFH is translating each engagement artifact into a FHIR
resource. Once it's a FHIR resource, any HIE, EHR, or payer system can ingest it without custom
integration — that's what turns an internal point into a tradeable signal.
FFH artifact
FHIR resource
What it asserts
Healthy Coin balance / streak
Observation
Quantified engagement (e.g., "health-education-minutes", "engagement-index") with a date and value
Badge (Learn It / Live It / Share It)
Observation + Coding
A completed, verifiable competency at a defined level
Certified Patient credential (e.g., Asthma)
DiagnosticReport / Observation
Health-literacy + adherence score reportable back to a care team
Educational "prescription" from a provider
ServiceRequest / CarePlan
The order that sent the patient to FFH in the first place
Redemption / reward event
Provenance
Auditable ledger entry — who earned what, when, and how it was validated
The loop: Provider order → FFH engagement (coins/badges)
→ FHIR Observation/DiagnosticReport back to the HIE
→ HIE routes the prevention signal to the payer / care team
→ value flows back to fund the coin pool.
2. The de-identified engagement-to-outcomes tool
This is the part that answers "is there actually a correlation?" — safely.
What it measures
Engagement tier — none / light / active / power user, from coins + streaks
Credential depth — badges earned, Certified Patient status
Household effect — 1 vs 2+ engaged members under one roof
Outcome proxies — ER visits, avoidable admissions, total cost of care, appointment adherence
How it stays de-identified
HIPAA Safe Harbor: strip all 18 identifiers before analysis
Match on a one-way tokenized ID (privacy-preserving record linkage) — the HIE holds the key, FFH never sees claims
Report only cohorts, never individuals; small-cell suppression (n<11 hidden)
An expert-determination review before anything is shared externally
Architecture: FFH engagement (Supabase) → nightly FHIR export
→ tokenize + de-identify → HIE links to
claims/utilization on the token → aggregate cohort stats returned to a
dashboard. FFH sees engagement + outcome rates; it never sees a member's claims.
3. What the dashboard would show
Illustrative view — synthetic data, for shape and hypothesis only.
–31%
Illustrative ER visits: power users vs non-engaged
2.4×
Illustrative appointment adherence lift with a Certified Patient badge
$540
Illustrative avg annual cost-of-care gap per engaged member
ER visits per 1,000 members, by engagement tier synthetic
Hypothesis: higher sustained engagement associates with fewer avoidable ER visits.
Not engaged420
Light350
Active285
Power user240
Illustrative outcomes by household engagement synthetic
Hypothesis: a second engaged member in the home amplifies the effect (shared habits).
0 engaged in homeBase
1 engaged–18%
2+ engaged–37%
Higher utilizationLower utilization
Credential vs adherence synthetic
Hypothesis: specific badges/credentials predict specific behaviors better than raw coin totals.
No badge31%
Learn It44%
Live It59%
Certified Patient75%
4. The study design (so a finding would hold up)
To make any correlation credible to a payer or public-health reviewer, the tool is built around a real
study frame, not just a chart:
Cohorts: engaged vs matched non-engaged members, matched on age band, sex, geography, and baseline risk.
Exposure: engagement tier and credential status over a defined window (e.g., 12 months).
Outcomes: ER visits, avoidable admissions, adherence, total cost of care — from HIE/claims side.
Guardrail: report association, not causation, until a prospective pilot supports more.
Unit of analysis: individual and household, so the "whole family" effect is testable.
5. What it's worth — and to whom
Buyers of this signal
PA HIE — a new prevention/SDOH data stream to route to members