idea-008 · 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. The candidate file in full as it stands on 2026-09-29 (84 claims: 21 verified, 50 unconfirmed, 9 unverified, 4 refuted), the one dossier its domains: lists (internal-hemorrhoids-office-based-treatment, built 2026-08-27), and all three decks, each rebuilt 2026-09-29. check_decks.py --candidate idea-008 reports all three decks current: no lens is missing and none is STALE. The candidate lists no seeds:, so no macro-trend seed contributes context here.
A single-use bipolar radiofrequency applicator with impedance-based dosimetry and automatic cutoff, deployed through a disposable anoscope, to treat grade II and grade III internal hemorrhoids in the office without sedation, sized to what Medicare pays for the office procedure. The file is wrapped at stage verified, and 21 of its 84 claims are verified — every one a regulatory-database fact about other devices or a CMS (Centers for Medicare & Medicaid Services) figure about payment, volume and setting; none says the cutoff works, what the kit costs, whether anyone would buy it, or whether anyone else owns the approach. The money moved in the candidate's direction when its original payment line was refuted: the office is the higher-paying setting, at $249.07 rather than the ~$158 the file first assumed. The regulatory map is detailed and ends at one edge, grade III, where no cleared bipolar or RF (radiofrequency) applicator indication reaches. All three decks, rebuilt today, meet at that edge, and all three name a question no connector can answer.
The problem. Grade II-III internal hemorrhoids that have failed conservative management are treated in the office with banding or heat, or escalated to surgery. The file's sourced picture of why that matters sits at grade III and is still unconfirmed, because published findings wait for Stage 7 appraisal: in a randomized noninferiority trial at grade III, rubber band ligation was not noninferior to hemorrhoidectomy on 12-month quality-adjusted life years, with 47.5% recurrence against 6.1% claim 59unconfirmed; in a single-centre ligation cohort, grade III recurred at 32.8% against 14.6% at grade I-II claim 60unconfirmed. The surgical alternative carries the costs this device exists to avoid — across 79 randomized trials (n=9,232), closed hemorrhoidectomy had the lowest recurrence while the surgical options carried the most bleeding, urinary retention and incontinence claim 61unconfirmed. Whether recurrence is actually what drives escalation today is itself unbacked claim 24unconfirmed, and so is what office proceduralists do with grade III patients now claim 48unconfirmed. The dossier adds, as context only, that the AGA's 2026 practice update places office banding or infrared coagulation before surgery for grades 1 to 3 and offers surgical consultation to grade 3 patients who fail banding [dossier: internal-hemorrhoids-office-based-treatment].
The design choice. The Generator's own scope note in the candidate file explains grade and sedation as deliberate levers: grade II-III is the band where an office device could plausibly displace a facility procedure, and keeping the procedure sedation-free is what keeps it in the office and on the non-facility payment. That note is design rationale, not a claim, and the file says what it commits the device to: working durably enough at grade III to be chosen over referral to surgery. That durability is unbacked claim 33unconfirmed.
The mechanism. Bipolar RF to the submucosal pedicle, with impedance-based dosimetry and an automatic cutoff as the mitigation for the named failure modes — recurrence at grade III, pain, bleeding, and rare thermal injury to the sphincter or rectal wall claim 12unconfirmed. Whether impedance can predict injury depth closely enough to devascularise the pedicle and spare the sphincter is unbacked, and the Verifier records that no connector answers bench physics claim 31unconfirmed. The dossier notes that no search was run for impedance-dosimetry or automatic-cutoff precedent in any cleared electrosurgical device [dossier: internal-hemorrhoids-office-based-treatment].
Why now. No macro-trend seed is linked to this candidate, so there is no seed direction to name. What the file itself shows changing is the evidence base, not the market: CMS fee-schedule and utilisation data became checkable in this repo on 2026-09-11, and a single-use anoscope indicated through grade III was cleared in 2025 claim 55verified.
What has been checked, in order. The first Verifier pass (2026-08-23) established the regulatory envelope: 21 CFR (Title 21 of the Code of Federal Regulations) 878.4400 is Class II (FDA device class II) claim 16verified, its product code is GEI on the SU panel claim 17verified, and the frontmatter's unidentified Part 876 route resolves to named Class II regulations claim 18verified. The second pass (2026-08-25) found what the file had missed by searching device-name strings: a bipolar RF applicator cleared for internal hemorrhoids with a four-clearance US history under GEI claim 40verified, whose indications all stop at grade II claim 42verified, claim 52verified. That pass refuted three earlier readings — that the panel evidence pointed to Part 876 claim 19refuted, that no RF-specific office hemorrhoid device existed in openFDA claim 23refuted, and that no predicate had been identified claim 46refuted. The dossier (2026-08-27) located the CMS fee-schedule file and four published recurrence findings, which were copied in as claims 57-62. On 2026-09-11 the CMS caches arrived: the RVU (relative value unit) file refuted the file's original payment figures, which had the facility and non-facility amounts swapped claim 5refuted; the office is the higher-paying setting at 7.42 against 4.72 total RVUs claim 58verified; $188.99 of the office payment is practice expense claim 74verified; and Medicare fee-for-service billed 8,090 services across the three codes, a floor rather than a market claim 7verified, with 66% of 46930 billed in the office claim 29verified. The 2026-09-29 pass re-ran the newly cached coverage, HCPCS (Healthcare Common Procedure Coding System), payment and De Novo (De Novo classification request) lookups. No Medicare NCD (national coverage determination) or LCD (local coverage determination) in the cached index names 46930 — recorded in the Verifier notes on several lines, and put on record as its own proposition by today's viability deck claim 78unverified. The payment connector reads the wrong schedules for these codes, so its "no row" result says nothing either way. Every De Novo query containing "hemorrhoid" hits a spurious substring match, and the usable queries are nulls about a string, so the De Novo route is still not established either way claim 14unconfirmed. One tag moved: the 25-year hemorrhoid clearance scan is now unconfirmed claim 47unconfirmed. Today's three deck rebuilds appended claims 76-84, all unverified.
What the scoring sections say, and how current they are. The candidate's Verified Composite section is dated 2026-08-29 and its Gate Check 2026-08-25; the recorded Stage 5 result is queued-research. Both predate the 2026-09-11 verifications of claims 7, 29, 58 and 74, and the composite's own tally still reads 64 claim lines where the file now has 84. The file's ## Wrap section likewise still counts 16 of 64 verified and quotes earlier wording of the desirability deck's load-bearing condition. This narrative reports those sections as they stand and does not recompute them; bringing them up to date is not this artifact's job.
Twenty-one claims, all verified. Every one is about the regulatory envelope, other devices' cleared scope, or what CMS pays and counts.
The regulatory envelope
Other devices' cleared scope
What CMS pays and counts
Refuted — stated plainly. Four lines are refuted, and all four are corrections to earlier readings rather than to the candidate's premises.
The money.
unconfirmed for the pipeline claim 1unconfirmed, and the ligation codes' descriptors
are in the same position claim 3unconfirmed.
The people.
The device and the pathway.
The field.
A note on the dossier's age. It was built on 2026-08-27, before the CMS caches arrived. Its reimbursement section records payment, coverage and procedures as blocked on missing cache files; the candidate file's later Verifier notes supersede that, and this narrative reads the candidate file for those facts. The dossier is left as it is.
Desirability (built 2026-09-29, current). The patient and proceduralist chair. It finds the volume and setting verified and the recurrence gap sourced, and then finds nothing at all behind the human decisions: switching, patient acceptance, pain relative to banding, and a grade III patient choosing the office over surgery. Its own summary is that not one of the twenty-one verified claims says a patient or a proceduralist wants this. Where it departs from the other two decks is order: it argues the switch is decided before any cost comparison is reached, and treats the 90-day global period claim 57unconfirmed as what would give the physician a reason to pay for durability at all. Its two write-backs are claim 76unverified and claim 77unverified.
"whether a proceduralist will trade an owned, paid-off instrument for a per-case disposable, on a grade III durability promise that is unproven at launch" — the load-bearing condition of the desirability deck (built 2026-09-29).
The condition rests on claim 15unconfirmed and claim 63unconfirmed.
Viability (built 2026-09-29, current). The investor and operator chair. It agrees that the rate is checked and the office pays more, and it is the deck that puts that good news under the most pressure: a verified $188.99 practice-expense share is a gross figure, and whether any of it is free for a device is its write-back claim 80unverified. It also names coverage for the first time claim 78unverified, claim 79unverified. It disagrees with desirability on order — it treats the switch as downstream of the subtraction, not upstream of it — and agrees with feasibility that the capital case turns on grade III, since a clinical study forced before clearance would move onto the critical path. It adds a second question beside its quoted condition: "how much of the $188.99 practice-expense share is actually free to pay for it".
"what a dosimetry-bearing disposable actually costs to land" — the load-bearing condition of the viability deck (built 2026-09-29).
The condition rests on claim 32unconfirmed and claim 65unconfirmed, with claim 74verified and claim 80unverified beside it.
Feasibility (built 2026-09-29, current). The engineering chair. It agrees with the other two that FTO is unexamined and the landed cost unsourced, and it adds four conditions no other lens reaches: disclosed predicate parameters claim 81unverified, a usable anatomical thermal-spread limit claim 82unverified, one-kit clearance claim 83unverified, and sterilisation of the sensing electronics claim 84unverified. Its argument is that grade III is where the engineering, the evidence and the regulatory strategy meet, and it lays out both sides of the fork the file records — narrow to grade I-II, where the nearest predicate matches exactly on ground held since 2012, or keep grade III and argue across codes to a 2003 device — and leaves the choice to FDA and a human.
"whether this device can be evidenced and cleared at grade III, the half of its own indication for which no bipolar or RF applicator predicate exists" — the load-bearing condition of the feasibility deck (built 2026-09-29).
The condition rests on claim 42verified and claim 51unconfirmed.
Where they agree, and where they don't. All three decks agree on three things: grade III is the only ground no verified RF record occupies claim 52verified; the kit's cost has no source claim 32unconfirmed; and FTO has never been searched claim 11unconfirmed. They disagree on which question comes first. Desirability puts the proceduralist's decision first and the price second; viability puts the subtraction first and the proceduralist downstream of it; feasibility puts the grade III indication first, because both the evidence the proceduralist would need and the calendar the investor would fund depend on it. None of the three is STALE, so the disagreement is a real difference between chairs, not an artefact of build dates.
At grade III — the one band no cleared bipolar or RF applicator covers — can this kit be both cleared and paid for: will FDA accept a grade III indication on predicates cleared to grade II, and does a dosimetry-bearing disposable land at a cost the office payment actually has room for once the session's own costs are met?
Each half is a question the file has already named and no connector can answer. The regulatory half is the feasibility deck's condition claim 42verified, claim 51unconfirmed, settled only by an FDA pre-submission. The money half is the viability deck's condition, with one term unsourced claim 32unconfirmed, claim 65unconfirmed and the other's usable size open claim 80unverified. The desirability deck's question — whether a proceduralist switches claim 15unconfirmed, claim 63unconfirmed — sits behind both: it can only be asked of a kit that can be cleared for the patients it is meant for, at a price that exists. Around all of it sits an automatic-kill factor nobody has looked at once claim 11unconfirmed.
Naming this question is not a recommendation, a gate, or a kill.
_shared/plain-language.md