idea-005 · 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.
This is an unattended, multi-channel home sleep recorder — nasal pressure/thermistor airflow, pediatric-sized effort belts, pulse oximetry, body position — scored against pediatric rather than adult AASM criteria, for children aged 5-17 with suspected obstructive sleep apnea who do not carry a comorbidity (neuromuscular disease, craniofacial syndrome, Down syndrome, sickle cell disease, obesity hypoventilation, significant cardiopulmonary disease) that would push them toward attended in-lab testing. The regulatory picture cleared up materially in the last Verifier pass: the launch product code is identified (MNR, Class II, 21 CFR 868.2375), and the candidate's own founding premise — that no home sleep test carries a pediatric indication — was refuted by a cleared device whose Indications for Use state verbatim that "both pediatric and adult patients may be tested" [claim 3: refuted]. That refutation cuts three ways at once: it strengthens the regulatory pathway, and it weakens both the moat and the competitive-intensity story, since the whitespace the candidate was built on turns out to already be occupied. Underneath the regulatory story sits a harder one: the AASM's 2017 position paper does not recommend home testing for pediatric OSA diagnosis at all, for a reason that names exactly the channel set (no CO2, no arousal/EEG detection) this candidate proposes [claim 28: verified]. Since that pass, five more claims were appended directly to the file — none clinical or regulatory, all of them explicit, unverified, never-checked commercial assumptions about addressable share, gross margin, payer-campaign affordability, De Novo cost, and licensing value [claim 34: unverified] through [claim 38: unverified] — so the file's evidence remains exactly as commercially bare as it was: every one of the seven verified claims is clinical or regulatory, and not one is commercial.
In-lab polysomnography is the diagnostic reference standard for pediatric OSA by professional-society consensus [claim 29: verified], and this candidate's whole premise is that the burden of getting one — access, wait, an overnight hospital stay — is worth trading against a home alternative. That burden is real where it has been measured, but the measurement is foreign: a Québec quality-improvement study of 82 consecutive pediatric adenotonsillectomy patients recorded a mean referral-to-consultation delay of 9.0±7.0 months and a total referral-to-surgery delay of 15.2±8.6 months, substantially exceeding the region's own access targets [claim 32: verified] — a real, verified number, from a single-payer Canadian system, not the US market this candidate launches into. A US-specific PSG wait-time, no-show, or cancellation figure was searched for directly and not found, as distinct from not searched [dossier: pediatric-obstructive-sleep-apnea-testing] — the US demand premise still rests on that gap.
The regulatory story moved the most of any factor to date, and it moved by refutation rather than confirmation. predicate_or_analog and the Generator's own narrative both assumed no HSAT device carried a pediatric OSA indication; that assumption is now refuted by primary source — SNAP Diagnostics' SAM Model 9-10000 (K240100, cleared 2024-06-04, product code MNR) states verbatim that "both pediatric and adult patients may be tested," and the WatchPAT family is separately cleared to 12 years and older [claim 3: refuted]. The same pass independently verified the launch product code and its regulation — MNR tied to 21 CFR 868.2375, though the codified eCFR text at that section actually describes a "breathing (ventilatory) frequency monitor," a respiratory-rate alarm device, not a "ventilatory effort recorder" — a documented mismatch between the classification database's regulation field and the CFR text itself [claim 26: verified]. One quality signal also came back clean of drama: a firmware recall against a marketed MNR recorder, Nox Medical's Nox T3s (Z-0031-2022, terminated, 2021-09-07), is a resolved event about a different incumbent, not a live safety signal [claim 27: verified].
That refutation is double-edged, and the file is explicit that it was scored in both directions rather than averaged. It helps the regulatory pathway — a pediatric-permitting predicate demonstrably exists. It damages the moat, because the "cleared pediatric indication itself" was named as one of three defensible assets and is now verifiably not exclusive [claim 10: unconfirmed], and it damages the competitive-intensity read, because the incumbents are described as not "one validation study away" from this indication when one has already arrived [claim 13: unconfirmed]. That claim's tag covers a compound statement: its first half checks out against primary FDA records for ResMed, Itamar/ZOLL and Philips, but its second half — whether incumbents will bother — is permanently unverifiable, and the file carries one tag for both halves.
Underneath the "is it legally reachable" question sits the harder "should it exist" question, and that is where the clinical evidence lands hardest. The AASM's 2017 Position Paper states plainly that "use of a home sleep apnea test is not recommended for the diagnosis of obstructive sleep apnea in children," citing the field's inability to find HSAT literature on devices that monitor CO2 or detect arousals — measurements the task force viewed as critical in pediatric populations [claim 28: verified]. Arousal detection requires EEG, and an EEG-free montage is exactly this candidate's mechanism:. Weighed against that: a 2026 systematic review found that in one of the largest simultaneous HSAT-versus-PSG comparisons, mean AHI was 2.5±3 by home test against 2.7±2.9 in lab, p>0.05 — close agreement, at values near the pediatric treatment threshold — but the same review found only 3 of 9 included studies used simultaneous testing at all [claim 30: verified]. The best home-testing accuracy figure in the file points the same way: an Australian cohort of 47 children aged 5-16 obtained an acceptable recording on every home study attempted, with a 6.6% false-positive and 3% false-negative rate for OSA diagnosis [claim 31: verified] — but that study used home Type-2 polysomnography, a montage that includes EEG, so it validates home application by families rather than this device's specific no-EEG mechanism. The candidate's own dominant clinical risk — false negatives on an EEG-free montage, since pediatric OSA presents at lower AHI with more arousal-based events than the adult disease this hardware class was built around — remains entirely unconfirmed [claim 11: unconfirmed], and it now sits directly downstream of a published guideline naming that exact gap.
Money is where the file has moved the least. The reimbursement claim — that commercial and Medicaid policies generally exclude pediatric HSAT from coverage — is unconfirmed on a missing data cache, and the conditional the whole business case turns on is unresolved: whether that exclusion, if real, is worded as an age restriction or a device-clearance restriction [claim 8: unconfirmed], [claim 9: unconfirmed]. A secondary, unverified lead names three CMS G-codes (G0398/G0399/G0400) sometimes cited for home sleep testing, sourced only to a coding-vendor page rather than a CMS document, and explicitly not established as a finding [dossier: pediatric-obstructive-sleep-apnea-testing]. Every TAM component is unconfirmed — volume [claim 5: unconfirmed], price [claim 6: unconfirmed], and addressable share [claim 7: unconfirmed] — and the standard Medicare-procedure-volume check would be a wrong-denominator artifact for this population regardless, since children aged 5-17 are largely absent from Medicare FFS. FTO has no claim in the file at all — not even a gap-acknowledging one [claim 25: unconfirmed] — in a product code with 38 clearances over six years.
Since the last version of this narrative, five more claims were appended to the file, dated as a business-case assumption pass on 2026-09-03, and every one of them is unverified and explicitly recorded as "never checked": whether the device would win a material share of the addressable pediatric PSG referral pool by year 3 of commercial availability [claim 34: unverified]; whether the assumed $150-300 net price yields a positive gross margin after cost of goods, cloud scoring, and customer support [claim 35: unverified]; whether a post-clearance payer-policy-change campaign is affordable and completes on a calendar a single company can finance [claim 36: unverified]; whether the De Novo branch's cost and calendar materially exceed the mid-single-digit-millions, 2-4-year estimate the file states for the 510(k) branch [claim 37: unverified]; and whether an incumbent adult HSAT manufacturer would pay enough for a licensed pediatric scoring algorithm and validation dataset to sustain a standalone licensing business [claim 38: unverified]. None of these five touches the clinical or regulatory ground the seven verified claims stand on — they sit exactly where the reimbursement and TAM claims already left the file thin, and they widen rather than close the gap between what is evidenced about whether the device could work and what is evidenced about whether anyone would pay for it.
One numbering wrinkle is worth naming plainly, because the new claims change how misleading it reads. Claim 33 names "claim 38" as its nearest revealed-behaviour twin for a parent-preference question, but the study it actually describes — a wait-time cohort — is [claim 32: verified], not whatever sits at position 38 [claim 33: unconfirmed]. Before this pass, that cross-reference was simply stale, since the file held only 33 claims and no claim 38 existed at all. It no longer is merely stale: the file now has 38 claims, and position 38 is a real claim — the incumbent-licensing assumption [claim 38: unverified] — that has nothing to do with parent preference or the Québec study. The mismatch in substance is unchanged (the intended twin is still [claim 32: verified]), but its surface reading is now actively misleading rather than obviously broken. This is not something for a narrative to fix — claims are append-only and positional, and correcting the cross-reference is the Generator's job, not this artifact's.
refuted: a cleared device already states pediatric use is permitted, on a roster of 38 MNR
clearances [claim 3: refuted]
All three decks carry source_updated: 2026-09-03, matching the candidate's current updated: field, and python3 engine/scripts/check_decks.py --candidate idea-005 reports all three current, none STALE. Each deck's own text was written on 2026-08-30 and opens with the same disclosure: the prior build (2026-08-24) said this candidate had no verified claims at all; it now has seven, and none of them is commercial. The five business-case claims [claim 34: unverified] through [claim 38: unverified] were appended after that deck text was written and are not discussed in any of the three decks — they sharpen the same commercial-evidence gap each deck already names, but no deck's own language reflects them directly.
The desirability deck turns on whether anyone downstream of clearance will actually act on this device's result. Clearance was never the blocking constraint — a pediatric-permitting predicate already exists [claim 3: refuted] — so the deck reads the AASM's 2017 position as the first candidate explanation for why the practice hasn't moved anyway.
"whether pediatric sleep physicians and ENTs will act on an unattended study in an otherwise-healthy child" — the load-bearing condition of the desirability deck (built 2026-08-30).
The viability deck turns on a single sentence of payer-policy wording that no connector in this repo can reach. The verified evidence doesn't touch it directly — none of the seven verified claims is commercial — but it makes the deck's own reading of the coin flip less even: a professional-society position recommending against the modality is exactly the kind of document a medical-policy committee cites when it writes an exclusion.
"how the pediatric home-sleep-testing exclusion is worded in real payer policies — as an age restriction or as a device-clearance restriction" — the load-bearing condition of the viability deck (built 2026-08-30).
The feasibility deck turns on the mechanism question underneath the regulatory and guideline questions: not whether a predicate exists, but whether removing the EEG channel this device omits costs enough diagnostic agreement to matter. The deck reads the verified evidence as bracketing that question rather than answering it — favorable concordance data that is three studies deep, and the best home-testing accuracy figure obtained with a montage that keeps the very channel this device removes.
"whether an EEG-free montage can agree with concurrent PSG at pediatric AHI thresholds well enough for a treatment decision" — the load-bearing condition of the feasibility deck (built 2026-08-30).
The candidate's regulatory whitespace closed by refutation — a pediatric-permitting predicate demonstrably exists and the practice has not moved onto it regardless [claim 3: refuted] — and a professional society now supplies a documented, on-point reason why: the absence of exactly the channel (arousal/EEG detection) this device also omits [claim 28: verified]. If a retrospective rescoring of archived paired pediatric PSG recordings, with the EEG channels stripped out, shows that removing that channel costs little diagnostic agreement at the AHI values where treatment decisions are actually made, does that reopen a market that clearance alone could not — or does the deeper reason nobody has walked this legally open path turn out to be commercial, sitting in payer-policy wording no connector in this repo can read? And even a favourable answer to both would still leave five newly recorded, entirely unchecked commercial assumptions — addressable share, gross margin, payer-campaign affordability, De Novo cost, and licensing value [claim 34: unverified] through [claim 38: unverified] — standing exactly where they are now: named, and not yet examined.