idea-019 · viability 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: the investor, and the operator who would have to get a powered abdominal garment coded, covered and paid for out of the Medicare Part B durable-medical-equipment benefit.
How to read this: every slide is a condition, not a conclusion. A bracketed claim reference points into knowledge-base/candidates/idea-019.md — the number is the claim's position in its ## Claims list, and the tag is copied from that claim as it stands right now, never adjusted here. A [no claim] marker means nothing in the file speaks to the condition at all. No number appears on this deck that is not in a claim, and there is no ask slide. See docs/deck-spec.md.
The shape of the money evidence, said first. There is no dollar figure anywhere in this candidate file — not a price, not an allowable, not a volume. The one verified reimbursement fact in the file is an obstacle: NCD 280.6 exists, covers lymphedema and venous stasis ulcers, describes a limb garment, and does not reach orthostatic hypotension [claim 9: verified]. Three of the four reimbursement lookups that would have moved this came back unconfirmed for a tooling reason — missing local caches — rather than because CMS is silent. That distinction is the most important thing on this deck.
Would have to be true: CMS would have to treat a home-use, patient-owned, prescription abdominal compression device as falling inside the Part B durable-medical-equipment benefit category and as separately billable in its own right — and the patients prescribed it would have to be Medicare beneficiaries, not a population sitting mostly under 65.
Where it stands: The benefit-category half is the Generator's own design intent, correctly labelled as an assumption, and the file names why it cannot be looked up: benefit-category determination is a CMS judgment, not a database query [claim 32: unconfirmed]. The nearest precedent is real and verified — NCD 280.6 establishes that pneumatic compression devices are covered in the home setting under the DME benefit for other indications [claim 9: verified] — and the exposure is real too: separate payability is an explicit eligibility criterion in the RAPID notice [claim 13: verified]. The age half has nothing behind it and is now on record [claim 44: unverified]; the only anchor in the file is mildly cautionary, since the two registered binder studies enrolled ages 40-80 and 40-90, ranges that straddle 65 rather than sitting above it [claim 33: unconfirmed].
What would settle it: A benefit-category determination request to CMS, or at minimum a DME MAC written inquiry — human-run and multi-month, with no connector substitute. For the age question, a CCW or registry pull on the age distribution of diagnosed primary autonomic failure, or the posted baseline characteristics of NCT01971008 and NCT03482297.
If it's false: If the device is packaged into another payment rather than separately billable, Factor 2 loses its base case entirely, and the RAPID upside case loses one of the three eligibility criteria the file claims it is on the right side of. If the paying population is not predominantly Medicare, every reimbursement claim in this file describes the wrong payer and the commercial payer landscape has not been examined at all.
Would have to be true: A code would have to exist that this device can be billed under; that code would have to pay an amount that supports a device business; and a coverage policy would have to reach this indication. All three are separate, and all three are open.
Where it stands:
hcpcs connector needs
data/hcpcs_level_ii.csv and the file is absent here, so E0650, E0651, E0652, E0656 and E0657
were not looked up at all, and no device-specific code for posture-triggered abdominal
compression was identified [claim 10: unconfirmed]. That CMS would assign one, or accept this
device under an existing family code, is now on record as an assumption [claim 40: unverified].
What would settle it: Land data/hcpcs_level_ii.csv and data/hcpcs_payment_rates.csv and re-run hcpcs and payment against E0650-E0652 and the trunk/chest appliance codes — the file names both missing files explicitly. Then read any DME MAC LCD or policy article reaching compression devices. The Scores table's Reimbursement row names exactly this and adds the part no lookup answers: benefit-category and separate-payability status is a CMS determination.
If it's false: The file's Stage 2 draft table already scores Reimbursement on the base case with the missing code and the non-reaching NCD counted as obstacles. If no code can be obtained and no coverage argument lands, there is no Part B revenue at all, and the only remaining channel is cash-pay — which no claim in this file has examined.
Would have to be true: Population × rate × price × reachable share would all have to be knowable, and the product of them large enough to matter. Stated separately, because the file carries exactly one of the four.
Where it stands:
What would settle it: CMS "Medicare Durable Medical Equipment, Devices & Supplies — by Geography and Service" from data.cms.gov, which is uncached and named in the backlog; the published DMEPOS fee schedule for the same three codes; and CCW Parkinson prevalence for the denominator. Reachable share needs human-attached primary research and will not resolve from any source in this repo.
If it's false: The file's Stage 2 draft table scores TAM on population plausibility alone, with the file's own verdict that there is "no dollar TAM at all". If the population base does not hold, or if the allowable turns out to be small against the cost of a powered servo garment, the row has nothing left — and TAM below the rubric's floor is one of the three named Stage 5 kill conditions, which this deck names rather than applies.
Would have to be true: Three gates would have to be passed, in series, before anyone can be billed: a De Novo authorization for a device class whose regulation does not yet exist, a HCPCS code assignment, and a coverage determination reaching an indication no current policy names — and the calendar would have to start before a prototype exists.
Where it stands: No claim in this file carries a cost figure, a headcount or a calendar [no claim]; this deck does not estimate one. What is on record: no prototype and no bench, animal or human data exist [claim 8: unconfirmed]; the closest classification analog is confirmed only to exist and is a limb regulation by its own text [claim 1: verified] [claim 2: verified]; and no cleared predicate or De Novo grant for this indication was identified, with the De Novo half not checked at all because data/fda_denovo.csv is absent [claim 3: unconfirmed]. That a clinical investigation under an IDE will be required rather than bench data alone is the feasibility deck's Slide 3 [claim 47: unverified], and that the three gates cannot be substantially parallelised is now on record here [claim 42: unverified]. No policy acceleration is available today: the RAPID notice is a notice with comment closing 2026-10-13 and nothing in it was in force [claim 15: verified]; TCET is paused for new candidates [claim 16: verified] and was in any case capacity-limited to roughly five candidates a year [claim 17: verified]; and the one instrument that was finalized before these was repealed the very day it would have operated [claim 18: verified]. The 510(k) footnote into RAPID requires a De Novo predicate authorized within 18 months, which claim 3 records as not identified [claim 19: verified].
What would settle it: FDA De Novo decision summaries for comparable Class II devices, read for review clock and evidence demanded, plus the published CMS HCPCS code-application cycle — both public and low cost, and both named in the Scores table. A costed development plan is Stage 7 work and is not attempted here.
If it's false: Capital intensity & time-to-revenue is already the lowest row in the file's Stage 2 draft table. If the gates are genuinely serial and the pivotal study is an IDE study, time-to-first-dollar is measured in years before any revenue, which is an investor's question long before it is an engineer's.
Would have to be true: Defensibility would have to rest on something mechanical — an exclusive licence to the on-mechanism patent family, or De Novo-plus-clinical-data first-mover status. The bill of materials will not do it: a bladder, a pump and an accelerometer are not a barrier.
Where it stands: The mechanism has been public since 2016 and is under active academic trial [claim 5: verified] [claim 6: verified] [claim 7: verified], so there is no secret to keep. One named family sits squarely on the mechanism — US2015313608, "Compression device, system, and method for decreasing abdominal venous pooling", filed 2013-11-04, naive 20-year term expiry 2033-11-04, with grant status, claim scope, current assignee and the Vanderbilt CTTC availability listing all undetermined [claim 26: verified]. That it is available for exclusive licence to one developer is an assumption, not a finding, and is now on record [claim 43: unverified]. Beyond that single number no patent search was performed at all [claim 27: unconfirmed].
What would settle it: A patent-office file-history pull on US20150313608A1 — grant status, issued claims, current assignee, term — free and a few hours; plus a call to Vanderbilt's technology transfer office about licensing and exclusivity status. Both are named in the Scores table's FTO and Moat rows.
If it's false: If the family is granted and held or licensed elsewhere, the moat inverts into a block — and the same listing that is a route in for this candidate is a route in for any entrant, which claim 34's note says in terms. Moat / defensibility carries very low confidence in the file's own draft table; this is the check that moves it in either direction.
Would have to be true: The occupied field would have to be drugs rather than devices, and no intermittent-pneumatic-compression incumbent — a company already cleared under the limb-sleeve regulation and one product-line extension away — would have to be building this already.
Where it stands: The drug comparators are named but unconfirmed for a tooling reason: this repo has no drug connector at all, so midodrine's and droxidopa's approvals, indication wording and supine-hypertension labelling cannot be checked here [claim 11: unconfirmed] — the only independent corroboration is incidental, NCT03482297 using 10 mg midodrine t.i.d. as its standard-of-care arm [claim 6: verified]. Commodity elastic binders set the effort and price comparator and are likewise unconfirmed [claim 4: unconfirmed]. On device entrants, the file is explicit that its null result is a search absence and not market absence: no competitor company was identified, no connector here reads company, financing or FDA registration-and-listing data, and the De Novo half could not be checked at all [claim 34: unconfirmed]. That IPC incumbents exist and are already cleared under 870.5800 / product code JOW is verified [claim 1: verified] — and no assignee search was ever run against them [claim 27: unconfirmed].
What would settle it: Company and financing databases plus FDA registration-and-listing for abdominal-compression entrants, and assignee searches against the IPC incumbents named in claim 1 — a licensed-database pull a human would have to attach, as the Scores table says.
If it's false: If an IPC incumbent is already building this, the fast-follower question on Slide 5 is not hypothetical and the first-mover half of the moat is gone. The file scores Competitive intensity as a moderately contested indication precisely because absence of a record is not absence of a competitor.
Would have to be true: The smallest piece of evidence that moves the money question would have to be reachable without building anything — and here it is, because three of the blocking lookups fail on missing local files rather than on missing public data.
Where it stands: data/hcpcs_level_ii.csv blocks the code question [claim 10: unconfirmed], data/hcpcs_payment_rates.csv blocks the price question [claim 23: unconfirmed], and the DMEPOS utilization release blocks the volume question [claim 25: unconfirmed]. All three are public CMS products. Alongside them, the patent file-history pull on the named family is a few hours of work against an identifier already in the file [claim 26: verified].
What would settle it: Land the two HCPCS caches and the CMS DMEPOS utilization release (docs/backlog.md 1.13), re-run hcpcs, payment and procedures against E0650-E0652, and pull the US20150313608A1 file history. That combination converts the price, volume and IP rows from "not looked up" to a number or a fact, and it costs no fieldwork.
If it's false: If the allowable for the adjacent code family turns out to be far below what a powered servo device costs to make, the base case has no unit economics, and neither the coverage argument nor a new-code application is worth starting. That is a cheap thing to learn first.
Every condition above with nothing verified behind it. Read this slide first. Note the shape: the verified claims this deck does cite are mostly obstacles (claim 9) or background (claims 12-19 about a pathway not in force), not support.
data/hcpcs_level_ii.csv, then the CMS code-application cycle.
If only one thing from this chair could be checked: whether Medicare will pay for this device at all — that is, whether CMS places it inside the DME benefit category as separately billable [claim 32: unconfirmed] and whether any coverage policy can be made to reach orthostatic hypotension [claim 41: unverified], given that the one national policy on pneumatic compression covers lymphedema and venous stasis ulcers and describes a limb garment [claim 9: verified].
Why this one rather than the others. The TAM rows are empty but cheap to fill, and the IP question is a few hours of file history. The benefit-category and coverage question is the only one that is both expensive and unresolvable by any lookup — the file says in terms that it is a CMS determination, not a database query — and every dollar in the base case flows through it. A price with no coverage is worth nothing; coverage with an unknown price is still a business.
Naming it is not a recommendation, a gate, or a kill.