Assessment — idea-002

Single-use canister-free negative pressure wound therapy engineered for canine and equine distal-limb wounds, sold cash-pay into specialty/referral and equine practice.

This file carries no confidence tags. It cites them: [claim N: tag] reads claim N of the candidate file with the tag it carries there, and check_assessment.py refuses a citation that disagrees. It moves no score, no stage: and no status:.

How these gradings were made, disclosed rather than smoothed over. Seven of the eight propositions below were written on 2026-09-14, after the round table, nine impact reviews and three Verifier passes, and their graded: dates therefore postdate the tags they read. That was reconstruction, and it stays disclosed rather than re-dated. B7 A single-use disposable pump can hold the… was the first genuinely prospective pillar in this file, and it has now been tested. It was graded reshaping on 2026-09-24 against claims 65 and 66 while both carried unverified. The Verifier pass of 2026-09-29 returned both unconfirmed, and IR-17 claims 65 and 66 came back unconfirmed routed them to commission. B7 A single-use disposable pump can hold the…'s class did not move when the answer arrived. That is the one entry in this file where a reader can check the separation of the two axes against a date, not take it on trust.

What changed under this derivation (2026-09-24 → 2026-09-29). pivot: is still P0 the original shape and the bet was not restated; the standing was re-derived against the tags as they now stand.

• The candidate went from 147 claims to 153. It now carries 15 verified, 132 unconfirmed, 6 unverified and 0 refuted.
• The Verifier moved claim 99verified up from unconfirmed, and attempted twenty unverified lines. Claims 65-70 and 134-147 all came back unconfirmed, every one for want of an instrument or on a split reading, and none on contrary evidence.
• IR-17 claims 65 and 66 came back unconfirmed, IR-18 claim 67 came back unconfirmed and IR-19 claims 68 and 69 came back unconfirmed were recorded against register rows A39 The mean pump ON-time duty cycle for a…-A43 IEC 60601-1 is within scope for a…. All three were commission, raising REQ-0148 Extending manifest row… to REQ-0151 Quote verbatim. There was no pivot.
• The deck rebuild appended claims 148-153, all unverified.

Six claims entered rests on:, and each one did because the pillar's proposition cannot survive without it:

• 140 and 152 under B1 The wear system holds a working seal at…
• 141, 148 and 149 under B2 Veterinary surgeons switch from the human…
• 138 under B4 No live enforceable patent reads on…
• 139 under B6 It is a business

No verified claim was added anywhere, so no standing improved by addition. Every new citation is unconfirmed or unverified, and none of them changes a band. They are there so that when they come back, the pillar they bear on is the one that moves. No consequence class changed, and ## Regrades has no new row.

The bet

A dog degloves a hock, or a horse opens a distal limb, and the wound heals badly, slowly and expensively under a conventional bandage. Human medicine solved a version of this with negative pressure wound therapy, and veterinary practices already improvise with human single-use units off-label. The bet is that a device engineered for the actual problem — fur, a limb that moves and gets chewed, an owner managing it at home after discharge — is enough better than the improvisation that specialty and equine practices will buy it, and that an owner already paying for the surgery will pay a few hundred dollars more for the adjunct. The structural attraction is that a veterinary-only device has no FDA (US Food and Drug Administration) premarket pathway at all claim 2verified, claim 48verified, claim 49verified — no 510(k) (premarket notification), no De Novo (De Novo classification request), no IDE — so the gap between a design and first revenue is tooling and qualification rather than a submission queue. The structural danger is the same sentence read backwards: the absence of a premarket gate protects nobody, including us claim 17unconfirmed. Whatever moat exists has to be in the drape chemistry, the interface or the supply chain, because it cannot be in the regulatory calendar.

The bet was restated once already, around Kamus et al. 2019 (PMID (PubMed identifier) 30841889). That study is the single published equine trial of this candidate's own canister-free form factor, and it could not be completed because "difficulties with adhesion and seal prevented the completion of the experimental wound study despite the use of many adjunctive adhesives" claim 75unconfirmed. The bet is "the drape is the hard part, the closest published attempt at it failed hard enough to stop a study, and solving it is the moat." That restatement now has its own claim. Whether the Kamus failure belongs to the drape and adhesive system rather than to the canister-free architecture is on the ledger as claim 140unconfirmed. The Verifier's note says it plainly: even a full-text read leaves that attribution "an inference beyond what the paper reports". It sits under B1 The wear system holds a working seal at…, because if the architecture is the cause, a better drape does not rescue the design.

The shapes this bet is not, and which pillar each one survives. Writing these down is what justifies the reshaping grades below. Each is a business that remains when a particular pillar fails.

• An in-clinic-only veterinary NPWT unit. It survives the home-dispensing half of B2 Veterinary surgeons switch from the human… failing. The product keeps the fur-compatible drape and loses the post-discharge rationale that the decks now lean on.
• A drape and limb-interface module licensed to an existing NPWT maker or rental vendor claim 42unconfirmed. It survives B6 It is a business's channel leg and B7 A single-use disposable pump can hold the… failing, because someone else's pump carries the dose.
• A rental pump with a disposable dressing, which is the model the one direct-to-practice vendor already runs claim 119unconfirmed, claim 43unconfirmed. It survives B7 A single-use disposable pump can hold the… failing. That is why B7 A single-use disposable pump can hold the… is reshaping, not fatal.
• A canine-only launch claim 44unconfirmed. It survives the equine half of B1 The wear system holds a working seal at… failing, and that half is where Kamus and the equine distal-limb model study claim 20unconfirmed both land.
• The device sold as a substitute for hospital days, not as an adjunct on top of them. The rebuilt decks now carry this as a mechanism: owners choose discharge with the device claim 149unverified, and the course costs them less in total than the hospital days it displaces claim 150unverified. This does not replace the bet. It changes what B3 Owners pay several hundred dollars out of… would have to prove. See ## What this does not say for why B3 Owners pay several hundred dollars out of… was not regraded on it.

Pillars

Eight, the checker's cap, which forced a choice this derivation (see the last bullet of ## What this does not say). Each is a separate way this fails; consequence is graded against the business, not against a factor score. The distributor-listing proposition that stood alone as B7 A single-use disposable pump can hold the… on 2026-09-14 remains a leg of B6 It is a business (see ## Regrades), and B7 A single-use disposable pump can hold the… names the disposable-pump proposition.

• B1 The wear system holds a working seal at… The wear system holds a working seal at approximately -80 mmHg on a furred, weight-bearing, chewable distal limb for the treatment period, applied by an ordinary technician — consequence: fatal — graded: 2026-09-14 — rests on: claim 9unconfirmed, claim 23unconfirmed, claim 28unconfirmed, claim 29unconfirmed, claim 30unconfirmed, claim 46unconfirmed, claim 51unconfirmed, claim 59unconfirmed, claim 62unconfirmed, claim 140unconfirmed, claim 152unverified — settled by: a limb-model rig with instrumented pressure logging, a motion cycle and a calibrated exudate load, run first against a commercial PICO dressing to reproduce the Kamus adhesion/seal failure and only then against the candidate's drape — before, not with, the fifteen-dog/fifteen-horse study written down as claim 46unconfirmed — cost: project — in plain words: The dressing has to hold an airtight seal and steady suction on a furry, weight-bearing, chewable lower leg for the whole treatment, applied by an ordinary veterinary technician. If the seal fails, the therapy does not work and there is no product.
• B2 Veterinary surgeons switch from the human… Veterinary surgeons switch from the human single-use unit they already use off-label to a purpose-built veterinary one, and practices dispense it for post-discharge home management — consequence: fatal — graded: 2026-09-14 — rests on: claim 19unconfirmed, claim 24unconfirmed, claim 25unconfirmed, claim 47unconfirmed, claim 141unconfirmed, claim 148unverified, claim 149unverified — settled by: twenty structured calls with ACVS-affiliated small-animal and equine surgeons, asking first whether they use NPWT on degloving, shearing and graft cases at all, then what unit, at what cost, and what would make them change and send the animal home with it — cost: cheap — in plain words: Veterinary surgeons have to switch from the human single-use pump they already use off-label to a version built for animals, and send the animal home with it. If they see no reason to change, there is no customer.
• B3 Owners pay several hundred dollars out of… Owners pay several hundred dollars out of pocket on top of the surgical bill, because the owner is the payer and no third party is obliged to be — consequence: fatal — graded: 2026-09-14 — rests on: claim 12unconfirmed, claim 21unconfirmed, claim 53unconfirmed, claim 55unconfirmed, claim 56unconfirmed — settled by: structured conversations with owners at discharge from two or three referral practices, putting the choice and price as they would actually be offered, plus a human reading two pet insurers' full policy documents rather than their marketing pages and asking one insurer directly whether an invoiced NPWT line has ever been paid — cost: cheap — in plain words: The pet's owner is the payer, and has to be willing to pay several hundred dollars on top of the surgery bill, because no insurer is obliged to. If owners decline at the counter, there is no revenue.
• B4 No live enforceable patent reads on… No live enforceable patent reads on canister-free superabsorbent NPWT or on the dressing-interface chemistry — consequence: fatal — graded: 2026-09-14 — rests on: claim 5verified, claim 6verified, claim 7unconfirmed, claim 64unconfirmed, claim 138unconfirmed — settled by: counsel pulls the claim sets for the six named live US patents from USPTO full text — no connector here reaches claims text — and charts them against the candidate's drape, interface and connector; claims 64 and 138 between them name five of the six — cost: project — in plain words: No patent still in force covers a canister-free, absorbent negative-pressure wound dressing or the chemistry where the dressing meets the wound. If one does, the product cannot be sold without a licence.
• B5 Something stops a fast-follower who faces… Something stops a fast-follower who faces no regulatory delay whatsoever — consequence: reshaping — graded: 2026-09-14 — rests on: claim 16unconfirmed, claim 17unconfirmed — settled by: a prior-art scan of veterinary drape and limb-interface filings, read for whether the space is already claimed or merely crowded, then provisional filings drafted against the six named live patents — cost: cheap — in plain words: Because no regulator stands in the way of a veterinary device, a competitor can copy it quickly; something, a patent or a head start the market rewards, has to stop them. If nothing does, this becomes a different, lower-margin business.
• B6 It is a business It is a business: enough qualifying cases a year, reachable through a channel, at a price that survives dressing cost of goods and the distributor's cut — consequence: fatal — graded: 2026-09-14 — rests on: claim 13unconfirmed, claim 14unconfirmed, claim 15unconfirmed, claim 26unconfirmed, claim 27unconfirmed, claim 40unconfirmed, claim 139unconfirmed — settled by: a caseload count built from ACVS practice directories and referral volumes rather than from a prevalence estimate, priced against what a practice pays the direct-to-practice vendor today, plus two conversations with distributor category managers about terms for a first-year veterinary SKU — cost: project — in plain words: There have to be enough suitable wound cases a year, reachable through a distributor, at a price that still leaves money after the dressing's cost and the distributor's cut. If the numbers do not add up, there is no business.
• B7 A single-use disposable pump can hold the… A single-use disposable pump can hold the therapy setpoint at the dressing, for the labelled course, inside the energy, size and cost class a disposable permits — consequence: reshaping — graded: 2026-09-24 — rests on: claim 65unconfirmed, claim 66unconfirmed — settled by: instrumented bench measurement of pressure at the dressing port — not at the pump — and of pump ON-time against a leak-driven duty cycle on a limb model, with fault injection against the hardware over-pressure limit, against the PS-1 design inputs as written (IR-17 claims 65 and 66 came back unconfirmed; REQ-0148 Extending manifest row… bounds claim 65unconfirmed and does not settle it) — cost: project — in plain words: A throwaway, single-use pump has to hold the right suction at the wound, for the whole labelled course, within the battery, size and cost a disposable allows. If it cannot, the product becomes a reusable pump, which is a different design and a different business.
• B8 The no-premarket-pathway advantage is real… The no-premarket-pathway advantage is real money and is not eaten by tooling, sterile-packaging validation, battery and transport qualification, and process qualification — consequence: costly — graded: 2026-09-14 — rests on: claim 2verified, claim 18unconfirmed, claim 31unconfirmed, claim 41unconfirmed, claim 48verified, claim 49verified, claim 50unconfirmed, claim 52unconfirmed, claim 54unconfirmed, claim 57unconfirmed, claim 58unconfirmed, claim 60unconfirmed, claim 61unconfirmed, claim 63unconfirmed, claim 67unconfirmed, claim 68unconfirmed, claim 69unconfirmed — settled by: a quote from one converter and one sterile-packaging house against the PS-1 pump specification and a dressing spec — REQ-0066 The pump subassembly now has a named…, open, which per IR-17 claims 65 and 66 came back unconfirmed to IR-19 claims 68 and 69 came back unconfirmed should wait for REQ-0150 Quote verbatim's standard scopes and should quote the pack at both the 7% and the 20% duty sizes, with FCC Part 15 as a separable line — cost: cheap — in plain words: Skipping regulatory review is only an advantage if the saving is not eaten by tooling, sterile packaging, battery shipping rules and production qualification. If it is, the business is the same but thinner.

Standing

Rendered, never written: python3 engine/scripts/check_assessment.py --candidate idea-002 --render

Decision queue

Same render. Worst exposure first; inside a band, cheapest first.

Regrades

One, from 2026-09-24, and it is structural rather than an argument about severity. None was added on 2026-09-29.

No consequence class was argued up or down on the evidence. Three Verifier passes have landed since the classes were set: twenty tags moved on 2026-09-17, seventy-eight on 2026-09-24, and twenty-one on 2026-09-29. Together they moved no judgement in this file, which is the separation this section exists to keep visible.

This derivation added claims to five pillars' rests on: and regraded none of them. Adding a citation is not a regrade. It records which pillar a claim bears on, and the class was set for the proposition, not for the claims under it. Had any of the six additions been verified, it would have been padding, and that is why each one is argued in the prose rather than just listed.

B8 The no-premarket-pathway advantage is real… graded costly remains the judgement here most worth arguing with, and IR-18 claim 67 came back unconfirmed sharpened it in B8 The no-premarket-pathway advantage is real…'s favour. claim 67unconfirmed was split by the eCFR text. The lithium half, UN 38.3 under 49 CFR 173.185, matched. The "no FCC exemption available" half ran into 47 CFR 15.101's Supplier's Declaration route and into 15.103(e). So the live direction on that line is money spent qualifying against a regime that never applied, not a regime appearing late. IR-19 claims 68 and 69 came back unconfirmed leaves claim 68unconfirmed and claim 69unconfirmed exactly where the chair's reading put them, because the standards are paywalled. Every one of these is cost and schedule, not product: the device, the customer and the channel are unchanged if the number doubles. costly stands on that reasoning. If the reasoning is wrong, the fix is a row in this table, not a quiet edit.

What this does not say

• Nothing here is blocking, and that is still not good news. A further twenty-one claims were worked on 2026-09-29, and none came back refuted. Read the reasons, not the count. Ten of them are adoption or market lines, and the adoption and market connectors return unverifiable by design. claim 138unconfirmed resolved both of its patents by kind code and in-force date and nothing more, because the connector never reads claims text. Claims 65, 66 and 140 have no connector at all. The IEC standards are paywalled. And claim 141unconfirmed went unchecked because the connector crashed on a missing _cffi_backend import. That is a tooling failure, and it says nothing about the IFU (instructions for use). An absence of refutations here is an absence of a check.
• claim 99verified was deliberately left out of every rests on:. It says the 21 CFR (Title 21 of the Code of Federal Regulations) 878.4783 special controls for human NPWT, including usability testing and prescription-only use, are a human-use standard that does not reach an animal-labelled device. It sits right beside B8 The no-premarket-pathway advantage is real…'s regulatory legs. Adding it would take B8 The no-premarket-pathway advantage is real… from three verified claims to four without changing its band, and it is exactly the adjacent-verified-fact padding the spec warns against. B8 The no-premarket-pathway advantage is real…'s proposition is whether the saving is real money. It would not die if claim 99verified were false, because 807.81 and 814.1 claim 48verified, claim 49verified already carry the "no premarket gate" leg. claim 99verified is context this file reads around, not something it rests on.
• B3 Owners pay several hundred dollars out of… could fall on a technicality, and the table would show it as a disaster. B3 Owners pay several hundred dollars out of… cites claim 12unconfirmed, register row A4 Pet insurance penetration in the US is…, which says US dog pet-insurance penetration is "under 5%". Claims 111 and 112 on the same ledger record NAPHIA figures of 5.99% and 5.46% claim 111unconfirmed, claim 112unconfirmed. If a human attaches the NAPHIA report and claim 12unconfirmed is tagged refuted on the 5% bound, B3 Owners pay several hundred dollars out of… turns holed, the pillar turns blocking, and the posture becomes repair-pending. Almost nothing that B3 Owners pay several hundred dollars out of… turns on would have moved. At six percent, insurance is still not what pays for this. The impact review that followed would most likely be an absorb that names B3 Owners pay several hundred dollars out of…'s other dependants, claim 21unconfirmed above all. That is stated here in advance so that nobody reads the band flip as news when it comes. This is a consequence statement about a possible tag, not a guess about which tag it gets.
• **B3 Owners pay several hundred dollars out of… was not regraded to reshaping on the substitution mechanism, and that is the grading here I most expect to be argued with.** The case for regrading goes like this. If owners will not pay on top of the surgical bill, the device can still be sold as a replacement for hospital days claim 149unverified, claim 150unverified. That is a business in a different shape, which is the definition of reshaping. It was not regraded for three reasons.
  • Both claims are unverified deck assumptions. Softening a fatal class on the strength of an unchecked hopeful line is the rationalisation the graded: field exists to expose, even when no tag has arrived.
  • The owner is the payer in both shapes. Under substitution the owner still pays the device price in cash, so what fails is the phrase "on top of", not the pillar.
  • Substitution moves a cost onto the practice, which loses the hospitalisation revenue the device displaces. That lands on B2 Veterinary surgeons switch from the human…'s home-dispensing half claim 24unconfirmed, not on B3 Owners pay several hundred dollars out of…. If a human disagrees, the fix is a row in ## Regrades.
• **Several claims cited in rests on: point the wrong way, and the standing arithmetic cannot read direction.**
  • claim 62unconfirmed under B1 The wear system holds a working seal at… asserts that the unfavourable equine distal-limb result generalises to the canister-free form factor.
  • claim 9unconfirmed under B1 The wear system holds a working seal at… records that the canine pilot's irritation hampered attainment of negative pressure. If either came back verified, B1 The wear system holds a working seal at…'s standing would improve and its prospects would worsen. claim 140unconfirmed runs the other way: it is the claim that would have to be true for the drape to be a repair rather than a hope. The instrument cannot tell these directions apart. Where a band and the prose disagree, the prose is the correction and the table is left alone, because patching a rendering would make it a memory.
• Claims 71 and 72 verifying does not narrow B1 The wear system holds a working seal at…, and must not be read as if it did. The attribution of the unfavourable equine study to a canistered V.A.C. ATS rests on FDA records of that device claim 71verified, claim 72verified. The attribution itself is single-source with three recorded transcription errors claim 74unconfirmed. The study's own Methods text was never read claim 73unconfirmed. And a second review records its device field as data not presented claim 76unconfirmed. What this buys is a narrowing of an adverse fact, not support for anything.
• Silence from an insurer is not coverage. claim 92unconfirmed and claim 93unconfirmed record that California's pet-insurance statute names no coding system. claim 101unconfirmed through claim 106unconfirmed record insurers adjudicating off the itemized invoice and naming neither NPWT nor wound-care devices as exclusions. That undercuts the reasoning "no HCPCS (Healthcare Common Procedure Coding System) code, therefore no insurance leg". It does not establish that any insurer pays, which is exactly what claim 132unconfirmed is on record to prevent anyone assuming. claim 144unconfirmed, that insured share is higher at specialty hospitals, is also read around and not rested on, for the reason given in the B3 Owners pay several hundred dollars out of… bullet above.
• **The post-discharge home phase has no pillar of its own, and that is the cap talking, not a judgement that it does not matter.** Seven claims are about whether the home phase works:
  • claim 22unconfirmed: the dog tolerates the assembly
  • claim 134unconfirmed: owners keep it on
  • claim 135unconfirmed: owners act on a loss-of-seal alarm
  • claim 136unconfirmed: no recheck visit is needed
  • claim 141unconfirmed: a veterinary label can go where the human predicate's IFU claim 94verified does not
  • claim 149unverified: owners choose discharge
  • claim 151unverified: one unit per course Two of those, 141 and 149, now sit under B2 Veterinary surgeons switch from the human…, because B2 Veterinary surgeons switch from the human…'s own proposition names post-discharge dispensing. The rest are read around. At nine pillars the natural split would be a reshaping home-phase pillar, with the in-clinic-only shape as its fallback. Doing that inside the cap would mean merging two pillars of different consequence class, which is worse. A reader who thinks the home phase is the real bet should know it is currently carried as the second half of a fatal adoption pillar. claim 151unverified also has tension on file: the pump is sized for a 7-day course claim 65unconfirmed, and the mean equine application was 11.5 days claim 36unconfirmed.
• **The top of the queue is cheap because of how its tests are worded, and the wording is arguable.** Until this derivation, B2 Veterinary surgeons switch from the human… and B3 Owners pay several hundred dollars out of… named a "structured survey" and a "willingness-to-pay study" at cheap, and the spec's band table puts a survey at project. Both settled by: lines now name the cheapest artifact that would move the pillar: structured calls with surgeons, and conversations with owners at discharge. That matches the spec's own cheap example and the desirability deck's "can be asked this quarter without a regulator, a payer or a prototype". The owner conversations need a referral practice willing to host them. A reader who does not think that takes an hour of organising should read B3 Owners pay several hundred dollars out of… as project, which puts it level with B1 The wear system holds a working seal at… and B6 It is a business. B2 Veterinary surgeons switch from the human… stays first either way.
• critical on B1 The wear system holds a working seal at…, B2 Veterinary surgeons switch from the human…, B3 Owners pay several hundred dollars out of… and B6 It is a business is the ordinary condition of an unbuilt device, not a verdict. Two of the four can be answered with conversations. The other two need a rig or a directory-built count before the question can even be asked.
• **Exposure is not truth, resolved is not a pass, and these bands do not compare to idea-001's.** Nothing in this file says whether idea-002 is worth funding; that is Tier 3.

Every claim in the candidate file, with the pillars that rest on it and the materiality it inherits from them. 56 of 153 are read by a pillar; the rest are context — true, possibly checked, and carrying nothing. That is not an insult and it is not an error: it is the only view in this repo that can tell you where the verification effort went.