idea-004 · narrative
A synthesis across this candidate's claims, dossier and decks — every statement here traces to something that already exists elsewhere in the knowledge base. A narrative never invents a fact, upgrades a tag, or recommends anything.
How to read this. This is a synthesis, not a new source of truth: every claim it cites carries the tag the candidate file gives it right now, copied not paraphrased, and every line quoted from a deck is quoted verbatim. See docs/narrative-spec.md.
What this build read, and what it found stale. The candidate file in full (75 claim lines; updated: 2026-09-29), its one linked dossier (neonatal-hyperbilirubinemia-home-phototherapy, built and refreshed 2026-08-27), and all three decks (each built 2026-09-29 and reported ok by check_decks.py). The candidate links no seed, so nothing here draws on a macro-trend seed. Two sources are behind the candidate's own claims, and this narrative says so rather than repeating them: the dossier's reimbursement section predates the 2026-09-29 Verifier passes and still describes the CMS (Centers for Medicare & Medicaid Services) code, fee and coverage files as missing; and the candidate's Verified Composite, Gate Check and Wrap blocks were last written before those passes, so their counts ("3 of 49 claims verified", Reimbursement blank "on a missing cache") no longer describe the claim list as it stands. Neither is edited here; re-scoring is the Scorer's job, and nothing below computes a score.
This is a prescription, home-use LED phototherapy wrap for newborns of at least 35 weeks with uncomplicated jaundice whose bilirubin sits between the treatment threshold and the escalation threshold, meant to replace an inpatient admission or readmission, with an onboard light sensor that logs the dose actually delivered for the ordering clinician. Its launch route is on record as a Class II (FDA device class II) 510(k) (premarket notification) under an existing product code claim 1verified, with a crowded predicate pool that already includes two cleared wearables claim 2verified. On 2026-09-29 the Verifier moved the three Medicare reimbursement legs to verified — the billing code exists claim 3verified, its descriptor names a photometer claim 4verified, and Medicare publishes a monthly capped-rental allowable for it claim 5verified — but newborns are not Medicare beneficiaries claim 6unconfirmed, so every verified reimbursement fact belongs to a payer this population does not have. Six of 75 claim lines are verified, none is refuted, and every one of the six is about permission or price; nothing verified yet says anyone would prescribe it, that families would sustain it, or that the dose sensor can be built.
The problem. Phototherapy for neonatal jaundice works; what this candidate is about is where it is delivered. The file's premise is that a material share of in-band infants of 35+ weeks are treated today by admission or readmission rather than at home claim 17unconfirmed. The size of the moment has a US signal: 3,628,934 US births in 2024 claim 36unconfirmed, and in a Kaiser Permanente Northern California cohort of 148,162 infants of 35+ weeks who had no phototherapy in the birth stay, 1.8% crossed the threshold after discharge claim 37unconfirmed. But home treatment already exists: in a US review of 359 newborns across 9 clinics, 44% used a home biliblanket claim 38unconfirmed, and the 2022 AAP guideline "narrowly" raised treatment thresholds claim 11unconfirmed, which shrinks the treated population rather than enlarging it. The dossier adds, as context only, that at one US hospital the 2022 revision was followed by fewer diagnoses and less pre-discharge phototherapy without more readmissions [dossier: neonatal-hyperbilirubinemia-home-phototherapy], and that the AAP guideline's full text could not be retrieved by any automated fetch, so what it says about home phototherapy is unread [dossier: neonatal-hyperbilirubinemia-home-phototherapy].
Why now. No seed is linked, so the "why now" in this file is the device record itself: two wrap/wearable garments were cleared under the same product code in 2025-2026, BiliWrap and SnugLit claim 2verified, within ten LBI clearances in ten years claim 47unconfirmed. The dossier reads LED replacing fiberoptic sources across the class as the precondition for a garment form factor [dossier: neonatal-hyperbilirubinemia-home-phototherapy]. That same record is the candidate's hardest competitive fact: SnugLit's cleared indication covers hospital and home use, caregiver-administered, across 1.5-5.5 kg and 40.5-57.5 cm, with no gestational-age floor and no bilirubin-band ceiling claim 2verified — so the narrowed 35-week, in-band indication the file uses as its de-risking lever buys no separate lane. What the cleared indication does not claim is a dose or adherence record claim 2verified, and that is the one axis the candidate proposes to occupy.
The mechanism and its risk. A narrow-band LED wrap with an onboard radiometer logging irradiance times exposure time. The device-specific risk the file names is under-dosing at home, as delivered irradiance falls with distance and reduced coverage and no nurse is there to see it claim 14unconfirmed. The only observation of a wearable in use is a 12-infant Dutch pilot of a different device: 10 of 12 completed within 48 hours and 2 were switched to conventional intensive phototherapy because bilirubin kept rising claim 41unconfirmed. That is both the failure mode and the case for the dose record — the record is how a clinician would tell an infant who needs escalation from a wrap that is not delivering. The broader home-versus-inpatient evidence conflicts: a 2024 meta-analysis found readmission higher after home phototherapy (RR 4.61, very-low-quality evidence) claim 39unconfirmed, a 2026 review of 31 studies found a pooled 3.5% (re)admission after home initiation and no study evaluating cost-effectiveness claim 40unconfirmed, and the US clinic review found 4% readmitted with a home biliblanket against 13% without claim 38unconfirmed. The dossier records that almost all of this controlled literature is non-US [dossier: neonatal-hyperbilirubinemia-home-phototherapy].
The regulatory design. The launch route is treatment delivery only, and the file checks it against the stated intended use: Class II, 21 CFR (Title 21 of the Code of Federal Regulations) 880.5700, product code LBI, 510(k) — a verification that lapses the moment any bilirubin measurement claim is added claim 1verified. The measurement extension would land on in-vitro test-system types on a different review panel claim 31verified, and the transcutaneous route remains unidentified after a name sweep claim 32unconfirmed. Whether FDA (US Food and Drug Administration) would clear the narrower indication plus a dosimetry feature on bench data alone claim 30unconfirmed, and whether it would read a dose-and-adherence record sent to the clinician as a phototherapy feature rather than a measurement function claim 74unverified, are both open.
What was checked on 2026-09-29, and what that did and did not settle. Caching the CMS files let three reimbursement lines resolve. E0202 (a HCPCS code for durable medical equipment) exists as "Phototherapy (bilirubin) light with photometer" claim 3verified; that descriptor wording names a photometer, verified as wording and explicitly not as a payment rule claim 4verified; and Medicare's DMEPOS (durable medical equipment, prosthetics, orthotics and supplies) schedule carries one line for it, a capped rental billed per month, non-rural $75.86 (AZ) to $89.25 (AL), with no purchase line claim 5verified. The coverage lookup run the same day found no cached Medicare Billing & Coding article naming the code, which the claim records with its limits rather than as a denial claim 5verified. Each of these tag moves is a change in what the repo could reach, not a change in the world. All three are Medicare's, and the payers that matter here are Medicaid and commercial claim 6unconfirmed; the one Medicaid figure in the file is Louisiana's $33.71 claim 44unconfirmed, on a billing unit nobody has established claim 51unconfirmed.
What the same day's rebuilds added. The business case and the three decks rebuilt on 2026-09-29 put eight new assumptions on record, all unverified and never yet attempted by a Verifier. Four are payer-side: that the payers that matter may pay newborn home phototherapy under the per-diem service code S9098 (a HCPCS temporary national code) rather than E0202 claim 68unverified; that a monthly-rental payer pays a full month for a 2-4 day episode claim 69unverified; that E0202 has no Physician Fee Schedule row claim 70unverified; and that an avoided readmission is paid to the hospital per case, so the saving accrues to the payer claim 71unverified. The decks added four more: that families complete the scheduled outpatient serum draws without nursing claim 72unverified; that payers enforce the photometer as a payment condition, not only as wording claim 73unverified; the FDA reading of the dose record above claim 74unverified; and that a battery wrap keeps skin-contact surfaces within temperature limits for a full episode claim 75unverified.
None of these six says the device can be built, that a clinician would order it, that a family would sustain it, or what the payers this population actually has would pay.
No claim in the file is tagged refuted. Everything below is unconfirmed (attempted, not settled) or unverified (never attempted).
Who wants it
US jurisdiction the claim itself flags as wrong claim 43unconfirmed; an implementation study names lack of reimbursement and clinician-education reach as the main barriers claim 42unconfirmed.
Who pays, and how much
Who else is there
Whether it can be built and cleared
PATENTSVIEW_API_KEY is unset — on the rubric's only automatic-kill factor claim 28unconfirmed.
All three decks were rebuilt on 2026-09-29 from the candidate as it stands, and check_decks.py reports each ok; none is missing and none is STALE. That is a wiring check, not a content one. The three agree on one thing unprompted: every verified claim they cite describes permission, price or competitors, and none carries the condition the deck turns on. They disagree on which condition breaks the idea first.
The desirability deck holds its condition unchanged from its previous build, and says why: the week's verified claims settled that a code exists and what Medicare pays, which is a question about the money after someone has said yes. It adds one condition of its own, that families manage the outpatient serum draws without a nurse claim 72unverified, and it names the tension the file's own design creates — excluding bilirubin measurement to stay on an existing predicate claim 1verified leaves the clinician holding exactly the monitoring gap the regulatory strategy creates.
"whether discharging clinicians and general pediatricians will substitute home phototherapy for inpatient treatment in the eligible band without an added home bilirubin measurement" — the load-bearing condition of the desirability deck (built 2026-09-29).
The viability deck moved its condition. Its previous build turned on supplier margin under the E0202 allowable claim 24unconfirmed; the business case's S9098 assumption claim 68unverified moved it upstream, because a margin under E0202 presupposes that the payers that matter pay E0202 at all. If they pay the per-diem service code, the deck reads, the photometer in the descriptor claim 4verified decides nothing and the buyer becomes a home-health provider. It also notes that the verified Medicare legs leave the Verified Composite's Reimbursement row blank as last recomputed, before the tags moved — which matches what this build found.
"whether state Medicaid programs and commercial plans in the largest birth states pay newborn home phototherapy under the equipment code E0202 at all, rather than under the per-diem service code S9098" — the load-bearing condition of the viability deck (built 2026-09-29).
The feasibility deck holds its condition and reads the week's verified move as adding weight to it: because the code's descriptor names a photometer claim 4verified, the sensor is now a descriptor requirement as well as the only differentiator the cleared wearable leaves open claim 2verified. It also casts the sensor as the safety instrument against the failure mode seen in the Dutch pilot claim 41unconfirmed, and it puts on record how FDA would read the record claim 74unverified and whether a battery wrap stays within skin-temperature limits claim 75unverified.
"whether an onboard radiometer can measure delivered irradiance at the skin, in a garment that moves against a neonate, within a tolerance a reviewer and a payer would accept as a dose record" — the load-bearing condition of the feasibility deck (built 2026-09-29).
Where they part: the desirability condition makes the device irrelevant however well it is built; the viability condition could remove the equipment code the whole business model and the photometer argument lean on, whoever prescribes it; and the feasibility condition is the only one of the three that needs no clinician, payer or regulator to test, only a bench. The viability and feasibility conditions also meet in one place. The feasibility deck's added weight on the sensor rests on the E0202 descriptor, and the viability deck's condition asks whether E0202 is the code the relevant payers use at all claim 68unverified, and, if it is, whether they enforce the photometer claim 73unverified.
When a US discharging clinician sends an in-band newborn of 35+ weeks home on phototherapy today, what do the state Medicaid programs and commercial plans in the largest birth states actually pay for — an E0202 light rental, where a verified photometer descriptor might make a dose record matter, or a per-diem home phototherapy service under S9098, where it would not — and does that same payer rule say anything about the home bilirubin monitoring this device deliberately leaves out?
_shared/plain-language.md