idea-009 · desirability deck
Every slide states a condition that would have to be true, then reports where it stands using the candidate file's own claim and its own confidence tag. A deck never upgrades a tag, invents a number, or recommends anything, and there is no ask slide.
The chair: an adult with clinician-diagnosed lower-extremity lymphedema who is in the maintenance phase of complete decongestive therapy and cannot self-don a flat-knit garment — that inability is the indication. Alongside them: the certified lymphedema fitter who usually chooses the product, the prescribing clinician the intended_use: requires, the DMEPOS supplier who stocks it, and the household member who may end up doing the wrapping.
How to read this: every slide is a condition, not a conclusion. A bracketed reference points into knowledge-base/candidates/idea-009.md — the number is the claim's position in its ## Claims list, and the tag is copied from it and never adjusted here. [no claim] means nothing in the file speaks to the condition. See docs/deck-spec.md.
One thing to hold while reading: this candidate has five verified claims and all five are regulatory. Nothing on this deck is verified, and nothing on this deck could have been — adoption is unverifiable by design and no connector in this repo will ever answer a question from this chair.
Would have to be true: There would have to be a population of adults who are prescribed maintenance-phase below-knee compression, who try a flat-knit garment, and who fail to get it on — often enough that "cannot self-don a flat-knit garment" describes a real cohort rather than a marketing segment.
Where it stands: The candidate defines its indicated population by that inability and records it as the file's own sharpest problem: the indication is defined by an inability while the product's premise is a different self-application succeeding in the same people [claim 23: unconfirmed]. How many such patients there are is unread — the Medicare claim volume for A6583 was never searched because data/medicare_procedure_volumes.csv is absent [claim 15: unconfirmed] — and the 3-5 million prevalence figure is aggregator-grade and is flagged in the file as not a TAM input and not to be multiplied by anything [claim 18: unconfirmed]. That garment failure actually happens at scale is on record only as this deck's own assumption [claim 26: unverified].
What would settle it: Chart review or structured interviews at two or three lymphedema clinics on how many maintenance-phase patients are switched off flat-knit garments and why — a human/Scout retrieval, no connector. The volume half re-runs unattended once a human places the CMS utilisation file per data/README.md.
If it's false: The indication has no cohort behind it and Factor 3 (TAM) loses its population term before price or share is even reached — the file already scores TAM at the floor for want of any number at all.
Would have to be true: A patient who could not pull on a knitted tube would have to wrap several overlapping hook-and-loop straps around their own lower leg, tension each one until its printed elongation window aligns, and repeat that at every donning — with hand arthritis, obesity, or limited hip flexion, unaided and without a fitter.
Where it stands: This is the candidate's crux and it rests on nothing. The file states it as an assumption, not a fact, and records that adoption "patients who cannot self-don a flat-knit garment can successfully self-apply a multi-strap wrap and read the tension indicators, including patients with hand arthritis, obesity or limited hip flexion" returned unverifiable — no free, authoritative, machine-queryable source exists, and the Verifier explicitly refused to substitute its own estimate of patient capability [claim 23: unconfirmed]. The Gate Check says the same thing in its own words: zero evidence on either side. The separate question of whether that success repeats daily over months, rather than once under supervision at the fitting, went on record only with this deck [claim 27: unverified].
What would settle it: Supervised donning sessions with patients recruited from the indicated population — people who have already failed flat-knit self-donning — scored on first-attempt success and on applied pressure measured against the target profile. A human act at Stage 7; no connector reaches it, now or ever.
If it's false: The product has no user. Every other slide in this deck and the whole of the viability deck describe a device nobody in the indication can put on, and the file's Factor 3 component-3 share assumption — which embeds this same switching premise [claim 17: unconfirmed] — goes with it.
Would have to be true: Four separate people would have to say yes before one wrap reaches one leg: a licensed clinician has to direct its use, a certified lymphedema fitter has to select it over four brands they already stock, a DMEPOS supplier has to carry it, and Medicare has to be the payer it is assumed to be.
Where it stands: Clinician acceptance is not a marketing nicety here — the intended_use: writes it in, requiring use "under the direction of a licensed clinician" — and it is an unevidenced assumption, unverifiable by design [claim 25: unconfirmed]. The fitter and supplier are the same problem one step further along: the file records that the fitter, not the patient, usually selects the product, and that whether fitters and suppliers would stock a fifth brand against four established ones is unverifiable without commissioned interviews [claim 24: unconfirmed]. The payer link — that lymphedema compression items are a distinct Medicare Part B benefit category for items furnished on or after 2024-01-01 — is unread against any primary source, because the benefit category is statutory and no connector in this repo reaches the US Code, CMS transmittals or MLN Matters [claim 10: unconfirmed]; and whether coverage extends to any lymphedema diagnosis rather than a subtype is likewise unread [claim 12: unconfirmed]. Where the patient cannot manage the straps alone, a fifth person — a household caregiver — has to be there, and that went on record only with this deck [claim 30: unverified].
What would settle it: Structured interviews with certified lymphedema fitters, DMEPOS compression suppliers and prescribing lymphedema clinicians on brand-switching and on what they would accept as maintenance therapy — one commissioned study answers claim 24 and claim 25 together, and the file notes it also unblocks TAM component 3. The payer half needs the DME MAC Local Coverage Determination and Policy Article read directly, L-number pulled from the CMS Medicare Coverage Database rather than guessed.
If it's false: The consent chain breaks at its narrowest link. If fitters do not stock it, Factor 5 (moat, channel half) and Factor 8 (competitive intensity) both worsen at once; if clinicians do not direct it, the indication as written cannot be satisfied at all.
Would have to be true: The patient's out-of-pocket share would have to be small enough that they take the wrap home, and the safety cost of self-applying 30-50 mmHg over a limb whose arterial status nobody checked would have to be controlled by labelling that patients actually follow.
Where it stands: The money side cannot be stated, and no number may be invented to fill it. Medicare is assumed to pay 80% of the lesser of the actual charge or the DMEPOS fee schedule amount with deductible and coinsurance applying [claim 11: unconfirmed] — but there is no dollar figure for A6583 anywhere in this file, because data/hcpcs_payment_rates.csv is absent and nothing was searched, so the 20% is a percentage of an unknown. The net selling price is unverifiable by design [claim 16: unconfirmed]. Quantity and replacement limits per limb per benefit period — which decide how often a patient pays again — are unestablished [claim 14: unconfirmed]. The safety side is stated more clearly and is worse for this product than for its comparators: the material risk is application over undiagnosed peripheral arterial disease, making an arterial-insufficiency contraindication and an ABI screening statement mandatory labelling content, and the file records that this hazard is worse here than on fitter-applied incumbents precisely because the indication is self-application without a fitter [claim 21: unconfirmed]. Whether patients would actually arrive with arterial status documented is this deck's assumption [claim 28: unverified].
What would settle it: A human places data/hcpcs_payment_rates.csv and data/medicare_coverage_policies.csv per data/README.md and the payment and coverage lines re-run unattended; the quantity/replacement language then comes from the LCD Policy Article read directly. The PAD/ABI half settles only through Tier 3 clinical and regulatory labelling review against the recognised consensus documents on compression in the presence of arterial insufficiency — no connector answers it.
If it's false: A cost the patient will not pay makes the fitter's stocking decision easier, not harder, and it is decided against this product. On the safety side, an unscreened PAD patient self-applying high compression is a real-world harm pathway, and it lands on Factor 6 (mechanism and clinical risk), which is currently blank for want of any verified claim.
Would have to be true: Within the first weeks, the patient would have to see something they attribute to the wrap — a limb that measures smaller, a shoe that fits, a night without the heaviness — because a maintenance therapy with no visible weekly reward is a therapy people stop.
Where it stands: The mechanism is the least contested thing in the file and is still not verified: sustained gradient compression is stated to be established maintenance therapy in complete decongestive therapy, so mechanism risk is low and the risk is application risk — but no Stage 3 connector answers a mechanism or clinical-risk claim, and the Stage 0 scans that look like substitutes (literature, trials, recalls) return a ScanResult carrying no confidence tag and cannot verify anything [claim 21: unconfirmed]. So the therapy's efficacy is clinical reasoning on record, not evidence. Whether this device delivers enough of it to be noticed by the wearer depends on the patient reaching the target tension at all [claim 23: unconfirmed], and the perceived-benefit condition itself went on record only with this deck [claim 29: unverified].
What would settle it: Serial limb-volume or circumference measurement in the same supervised-donning cohort as Slide 2, run long enough to see the maintenance signal; plus Tier 3 clinical review of the consensus literature on self-applied adjustable wraps for maintenance-phase lymphedema. Both are human acts.
If it's false: Adherence collapses in month two, the reorder that the DMEPOS economics assume never comes, and the fitter who recommended it does not recommend it again.
Would have to be true: A handful of supervised donning sessions — patients drawn from the indicated population, an instrumented leg or a pressure sensor under the wrap, first-attempt success and applied pressure recorded — would have to be capable of falsifying the crux before any tooling is cut.
Where it stands: The candidate says so itself, twice: the desirability crux is "testable cheaply with a handful of supervised donning sessions" and today has zero evidence on either side [claim 23: unconfirmed], and the Gate Check records separately that the nine rubric factors do not score whether a user wants the thing, so the gate would return the same result whether this premise were true or false. Nothing in the file says the sessions have been designed, costed or scheduled — the Diligence Package section is empty [no claim]. Note what this slide costs relative to the alternatives on the viability deck: it needs no CMS cache, no eCFR endpoint, no patent key and no FDA judgment.
What would settle it: Run them. Recruit through a lymphedema clinic, use prototype straps with the printed elongation windows, measure sub-bandage pressure against the target profile, and count how many patients get it right unaided on the first attempt and on the thirtieth. A human act at Stage 7, and the one on this deck that no infrastructure fix substitutes for.
If it's false: If even this cannot be observed cheaply, the crux is not testable before capital is committed, and the desirability question stays exactly where the file leaves it — unanswered in both directions, with the rubric structurally unable to notice.
Every condition above with nothing verified behind it — which here is every condition on this deck. The candidate carries five verified claims and all five are regulatory [claim 1: verified], [claim 2: verified], [claim 3: verified], [claim 4: verified], [claim 5: verified]; not one of them speaks to a user.
unverifiable by connector, human only.
intended_use: itself.
The shape of this list is the point, not only its length. docs/deck-spec.md predicts it and CLAUDE.md records it as a standing finding: the rubric scores regulatory permission, payment, market and defensibility, and no factor scores whether a user wants the thing. On this candidate the mismatch is unusually stark — the file's evidence is five verified claims about what a regulation says, sitting on top of a premise about human hands that has never been observed.
If only one thing from this chair could be checked: whether a patient who cannot self-don a flat-knit garment can self-apply this multi-strap wrap to a reproducible target tension, unaided [claim 23: unconfirmed].
This one rather than the others for three reasons the file supplies itself. It is the only condition on the deck whose falsity makes every other slide moot — a device the indicated population cannot don has no consent chain, no adherence question and no perceived benefit. It is the cheapest to check: the file calls it testable with a handful of supervised donning sessions, against a reimbursement question that needs CMS files and an FDA question that needs a 513(g). And it is invisible to every mechanism this repo has — adoption returns unverifiable by design, the Gate Check states that this premise changes nothing about the gate's output, and the composite of 6.0/100 over 1/9 factors would read identically whether the premise were true or false.
Naming it is not a recommendation, a gate, or a kill. It is the answer to "check what first."