Prior Filing Digest — Our Own Provisional "Systems and Methods for Medical Support"¶
Ingested: 2026-06-04 · Owner: Gene (Pharmacy Ops) · Source: _Prior_Filing/Provisional_216679-010400_Medical_Support_filed_2025-06-26.pdf (81 pp.)
Why this matters: this is Rx360's own prior patent filing. It already discloses most of Smart Scheduling and the Balance Meter in detail — which changes how we frame both patent conversations. Read §4 (the delta) and §5 (actions) before the 6/04 meeting.
⚖ Confidential — attorney work product. Stored locally for reference; do not circulate outside the working folder.
1 · What it is (the IP family)¶
| Title | "Systems and Methods for Medical Support" |
| Type | U.S. Provisional patent application |
| Attorney docket | 216679-010400/PRO |
| Filed | June 26, 2025 (electronically) |
| Parent / related | U.S. Patent Application No. 18/989,851, "Systems and Methods for Digital Prescription Management," filed Dec 20, 2024 — incorporated by reference; covers multi-provider prescription import [0001], [0082] |
| Key deadline | A non-provisional (or PCT) must be filed by June 26, 2026 to claim the provisional's priority date. |
The IP family so far: 18/989,851 (Dec-2024, Rx import) → this provisional (Jun-2025, the full medical-support system: scheduling + fall + AI agents + EMS/provider comms + tap-to-share) → non-provisional TBD (deadline 6/26/2026).
2 · What it DISCLOSES (priority as of 2025-06-26)¶
The specification is broad and reads on almost everything we're building. A provisional's disclosure establishes a priority date and is enabling support for later claims. Core building block: the LMSS (Local Medical Support System = wearable + mobile + smart pill bottle), a server, and AI agents — an agentic AI pharmacist (160), an AI triage agent (157), and a signal engine (158).
Smart-Scheduling content already disclosed (this is the big one): - Medication-schedule generation by the AI pharmacist: dose-event aggregation (merge meds at compatible times into daily "dosage events"), separating drugs with troublesome interactions, separating visually-similar pills to reduce error [0085]–[0088]. - Scheduling within the prescribed boundaries — the explicit "tolerance" idea: "2x daily" has a wide compliant window; "take at bedtime" is constrained; move a med to post-breakfast to test for better tolerance if the patient reports an upset stomach [0086]–[0088], [0012]. - Learned rhythm / adaptive adherence logic: modifies the schedule from observed behavior (sleep/eat/activity/GPS); proposes 8am from a "morning" guideline; shifts dosage events on missed/late patterns; learning + model training with success-signal feedback into personalized and population models; demographics; DST adaptation staged over 2–3 days [0089]–[0091]. - Food/timing context: empty-stomach, grapefruit-counterindication dietary guidance [0080], [0086], [0097]. - Confirmation / adherence detection: pill-bottle proximity/NFC-tag/scan + user-confirm; "incorrect medication" alerts (duplicate / non-prescribed / wrong-time) [0092]–[0095], [0153]–[0156]. - Caregiver/Circle + provider alerts, gamify/incentives [0093]–[0095]. - Approval workflow: propose an adapted schedule → send to patient / support circle / provider / AI for comment/approve/decline before replacing [0096]–[0097]. - Refill anticipation & refill-pattern logic (the exact thing we discussed adding): auto-refill within allowed refills; pause refill on interaction/adverse reaction; shift refill date earlier on supply-chain shortage (LLM news-headline analysis of drug names); shift refill date on missed-dose/adherence patterns; predict leftover medication; trip/vacation advance-refill; insurance max-advance policy [0098]–[0104]. - OCR scan of the printed Rx label (3D-rotation OCR + object detection + text recognition + contextual extraction) [0072], [0156] — this is our pilot "scan the printed Rx summary" mechanism.
Balance-Meter content already disclosed: - Fall detection: IMU/accelerometer + neural-network pattern recognition before classifying a "fall"; abort-tap UI ("We believe you may have fallen. Tap here to cancel the report.") [0045]–[0046], [0059]. - Per-individual baselines + deviation-from-self: rolling mean/EWMA/IQR/z-score ±1σ/±2σ, PCA, Bayesian online change-point detection, RF/SVM classifiers [0128]–[0129]. - Gait analysis via IMU as a triage trigger [0058], [0109]; risk states normal / moderate / high [0110], [0144]–[0150]. - Sensitivity escalation: in a high/moderate-risk state, increase sampling cadence / activate additional sensors [0148]. - Habit-personalized data-absence threshold (alert sooner/later based on how long the member usually removes the wearable) [0124]–[0126]; offline mode [0142]. - Medication context conditions fall risk — recent orthostatic-risk med starts feed the triage (consistent with our med-change watch-window).
Other disclosed: AI triage routing (call center / pharmacist / Circle / EMS); speech biomarkers (mania/dementia onset, cough/wheeze) [0121]; provider/EMS selection + send-data-ahead; LLM use (Claude/ChatGPT/Llama/etc.) [0105]; deep-links / tap-to-share; smart pill bottles (RFID/NFC/BLE/AirTag).
3 · What it CLAIMS (three claim sets)¶
The claims are narrower than the disclosure — they focus on communication initiation and data sharing, not the scheduling-tolerance or fall-scoring algorithms:
- Claim Set 1 (button-push → provider/EMS): detect interface interaction → initiate real-time comms to a healthcare provider/EMS selected based on the member's health data (the Summary §[0014]–[0017] core).
- Claim Set 2 (proximity → authenticate → select → transmit): a component brought into proximity to a computing device → authenticate (identity) → select a portion of health data by identity → transmit (§[0018]–[0023]).
- Claim Set 3 (tap-to-share, claims 1–24, pp. 76–81): the proximity/authenticate/select/transmit method + dependents: location → healthcare-facility/pharmacy → provider/pharmacist; portion includes medication-adherence / BP / fall / activity data; claim 8: status dashboard linking the medication list + scheduled medical events + adherence; claim 9: show a schedule notification when proximity occurs within a threshold of a scheduled medical event; encode health data in URL payloads → dynamic/ephemeral interface → NFC; CRM + system claim variants.
NOT claimed (disclosed but open): clinically-curated PK/PD drug-class forgiveness/tolerance windows; the fall-risk-scoring "barometer" (deviation-from-self → 3-band signal); a cross-engine "medication-change" event that simultaneously tightens fall-sensitivity and scheduling-attention; provenance-tiered reminder precision.
4 · The strategic delta — what's genuinely NEW vs our own provisional¶
Our Smart Scheduling deck's six "what to protect" elements, re-scored against the provisional:
| Deck element | In the provisional? | Verdict |
|---|---|---|
| 1 · Clinically-curated PK/PD drug-forgiveness tolerance | Disclosure has "within prescribed boundaries" + "adjust if reactions" + food/timing — but no curated PK+PD drug-class forgiveness model driving reminder behavior (weekly biologic ±1–2 d vs short-t½) | STRONGEST NEW DELTA — lead the non-provisional here |
| 2 · Dose-time context payload | Partially disclosed (empty-stomach, grapefruit) [0080],[0086] | Weak delta — refine/narrow |
| 3 · Cross-engine medication-change event | Unified signal engine disclosed; the specific shared event coupling fall-sensitivity ↔ scheduling-attention is not claimed | NEW DELTA (as a claimed mechanism) |
| 4 · Provenance-tiered reminder precision | Approval workflows disclosed; tiered-confidence → precision not articulated | Partial delta |
| 5 · Confirmation-first + adaptive de-escalation | Adaptive adjustment + missed-dose shifts disclosed [0089]–[0091]; "remove nags as routine proves out" specifically not framed | Weak delta — refine |
| 6 · Wellness-lane framing | Design posture, not a claim | n/a |
Bottom line: the genuinely new, claimable material vs our own disclosure = (1) clinically-curated PK/PD forgiveness + (3) the explicit cross-engine event + (4) provenance-tiered precision — plus, for the Balance Meter, the deviation-from-self fall-risk scoring/banding (the spec discloses the statistics but doesn't claim a fall-risk barometer score). Prediction, learned rhythm, refill logic, dose aggregation, food/timing, caregiver alerts, OCR scan, tap-to-share are already ours by disclosure — good for priority, not new claims.
4½ · Coverage cross-walk — did our this-week work cover everything it discloses?¶
Core engine: yes (~75%). Completely: no. Our Smart Scheduling deck / brief / claim-architecture converged on the engine but did not pick up ~6 features the provisional already discloses. Nothing is lost (priority held since 6/26/2025) — but they are disclosed-but-unclaimed and absent from our scoping, so the non-provisional must carry them forward.
A · Covered — provisional ∩ our work¶
| Provisional concept | In our work |
|---|---|
| Dose-event aggregation / cadence schedule [0085–0088] | ✅ deck S7 |
| Scheduling within prescribed boundaries (tolerance) [0086],[0012] | ✅ S7 (we go deeper) |
| Learned rhythm / adaptive adjustment [0089–0091] | ✅ S8 (table-stakes) |
| Food/timing context [0080],[0086] | ✅ element 2 |
| Confirmation / adherence detection [0092–0095] | ✅ S9 |
| Caregiver/Circle + provider alerts [0093–0095] | ✅ table-stakes |
| OCR scan of the Rx label [0072],[0156] | ✅ S12 / roadmap |
| Med context conditions fall risk [0041–0045] | ✅ element 3 |
| Provider-network / provenance [0073–0076] | ✅ element 4 |
| Generic refill anticipation + pattern [0098],[0101] | ◑ roadmap S18 (generic only) |
B · ⚠ Disclosed but our work MISSED — CARRY INTO THE NON-PROVISIONAL¶
(priority held, but unclaimed and absent from our deck/brief — don't leave them on the table)
| Disclosed feature | ¶ | Worth claiming? |
|---|---|---|
| Separate interacting drugs in time + separate visually-similar pills (assembly-error reduction) | [0088] | ✅ yes |
| "Wrong / duplicate medication" alert at the moment of taking | [0092] | ✅ yes |
| Pause refill on interaction / adverse reaction | [0099] | ✅ yes |
| Shift refill earlier on supply-chain shortage (LLM news-headline scan of drug names) | [0100] | ✅ distinctive |
| Trip/vacation advance-refill + insurance max-advance policy | [0104] | ✅ yes |
| Predict leftover medication / late pickup | [0102–0103] | ◑ refinement |
| Smart pill bottles (RFID / NFC / BLE / AirTag) | [0071],[0155] | ◑ hardware embodiment |
| Approval workflow (propose schedule → patient/circle/provider/AI approve) | [0096–0097] | ◑ partly in our provenance |
| Tap-to-share (proximity→authenticate→select→transmit) | [0067–0070] / Claim Set 3 | ✗ already claimed; not scheduling |
C · Our genuine delta — the NEW claim surface (our work → theirs)¶
| Our element | Beyond the provisional? |
|---|---|
| Clinically-curated PK/PD drug-forgiveness model → reminder behavior | ✅ strongest — provisional only says "within boundaries / adjust if reactions" |
| Explicit cross-engine "medication-change" event (fall-sensitivity ↔ scheduling, bounded windows) | ✅ unified signal engine disclosed; this coupling not claimed |
| Provenance-tiered reminder precision (self/imported/verified → discrete precision tiers) | ✅ approval workflows disclosed; tiered-confidence→precision not |
| Fall-risk barometer SCORE / banding (Balance Meter) | ✅ statistics disclosed; a fall-risk score not claimed |
| Confirmation-first de-escalation · dose-context as a distinct claimed element | ◑ refinements vs the provisional's adaptive shifting |
One line: we hold priority on the engine; the non-provisional should (C) claim our four deltas and (B) sweep up the disclosed-but-unclaimed features we'd otherwise leave behind.
5 · What this changes / actions¶
- Reframe the 6/04 Smart Scheduling meeting. It's not "scope a brand-new invention." It's: "our 6/25 provisional already discloses the scheduling engine; here's the genuine delta (clinically-curated PK/PD forgiveness + cross-engine event + provenance tiers) to claim in the non-provisional, due 6/26/2026." The deck's "what to protect" still holds — but now anchored against our own disclosure, not the open market.
- This answers the "disclosure timing" open question on both the Smart Scheduling and Balance Meter decks: the clock is the provisional's 12-month bar — non-provisional/PCT by June 26, 2026. Everything we're scoping should feed that filing (or continuations off it).
- Portfolio is real and coordinated:
18/989,851(Rx import) + this provisional (medical support) → one non-provisional family, or continuations. The Balance Meter and Smart Scheduling are both disclosed here → strong argument for a coordinated non-provisional that claims both, exactly the "file together?" question on both decks. - FTO note: this is our own disclosure — not prior art against us. But our novelty framing should be measured against it: lead claims on the un-claimed deltas (§4), present the already-disclosed pieces as priority we hold.
- Counsel hand-off: give counsel (a) this digest, (b) the claim-architecture draft (
Smart_Scheduling_Claim_Architecture_2026-06-04) — and ask them to map our proposed independent/dependent claims against Claim Sets 1–3 and the spec, and to confirm the 6/26/2026 deadline + whether to claim Smart Scheduling + Balance Meter in the non-provisional or as continuations. - Sweep up the disclosed-but-unclaimed features (§4½-B) in the non-provisional so they aren't left on the table: drug-interaction time-separation + visually-similar-pill separation, the "wrong/duplicate-med" alert, and the refill suite (pause-on-interaction, supply-shortage shift, trip/vacation advance, leftover prediction). They're already ours by disclosure — but unclaimed, and our this-week scoping missed them.
- Verify the parent: pull
18/989,851(Digital Prescription Management) to see what the Dec-2024 non-provisional already claims (it owns the multi-provider Rx import — relevant to our med-matching/onboarding scan work too).
6 · Cross-references¶
- Smart Scheduling: deck
Smart_Scheduling_Deck_2026-06-04(S13 claim spine, S17 roadmap),Smart_Scheduling_Claim_Architecture_2026-06-04,Smart_Scheduling_Invention_Brief_2026-06-04,KB/03_Precedents_Tools_FTO. - Balance Meter:
_VP_Priorities/Fall_Risk/(the fall-risk barometer is disclosed here too — see §2). This digest is portfolio-wide; a pointer belongs in Fall_Risk as well. - Source PDF:
_Prior_Filing/Provisional_216679-010400_Medical_Support_filed_2025-06-26.pdf.