idea-017 · 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: the user. The adult with chronic respiratory failure consequent to COPD who would sleep under a mask at high inspiratory pressures and wear a sensor on the ear or chest; the spouse or caregiver who has to make that happen on a Tuesday night; the pulmonologist or sleep physician who would have to prescribe it and then set the ventilation from a recording made in a bedroom; and the DME supplier's respiratory therapist who would deliver it.
How to read this: every slide is a condition, not a conclusion. A bracketed claim reference points into knowledge-base/candidates/idea-017.md — the number is the claim's position in its ## Claims list and the tag is copied from it, never adjusted here. A no-claim marker means nothing in the file speaks to the condition at all. See docs/deck-spec.md.
First build. This candidate is at stage: verified with eight verified claims, and every one of them is a supply-side fact: a CMS instrument, a CFR section, a product code, a clearance record, a recall record, a trial result. Not one of them says a user wants this. Three of this deck's conditions had nothing in the file behind them before this build and are now write-backs (claims 31-33).
Would have to be true: A reachable US population of adults would have to be in the specific window this device is built for — chronic respiratory failure consequent to COPD with persistent hypercapnia in the weeks after a hospitalised exacerbation — and their trajectory without home NIV would have to be bad enough that they would accept a nightly mask to change it.
Where it stands: The clinical stakes for exactly this window are the strongest thing in the file: in HOT-HMV, adding home NIV to home oxygen in patients with persistent hypercapnia after a life-threatening COPD exacerbation extended median admission-free survival to 4.3 months versus 1.4 months, with 12-month mortality not significantly different (28.1% versus 32.2%) [claim 15: verified] — with the rider on that same claim that the trial was UK-conducted and carries no US regulatory or reimbursement weight. How many US patients are in the window is not established: the population base of approximately 16 million US adults with diagnosed COPD is recollection with no primary table attached [claim 12: unconfirmed], and the share of them meeting NCD 240.9's criteria is an explicit unsourced Generator estimate [claim 11: unconfirmed], whose note warns that HOT-HMV's 2,021-screened-for-124-eligible ratio is a trial-recruitment figure and must not be read as prevalence.
What would settle it: A human-attached CDC/BRFSS or NHIS prevalence table for diagnosed COPD, plus a published estimate of chronic hypercapnic respiratory failure within that population — neither is reachable by any connector in this repo (the market connector is unverifiable by design), so this is a human retrieval of a licensed or primary source.
If it's false: The TAM factor loses its population leg and the viability deck's Slide 3 fails on the same unresolved number; more immediately, a device aimed at a window too narrow to find patients in has no user at all.
Would have to be true: The patient would have to sleep, nightly, under a mask at high-intensity bi-level settings with a fixed back-up rate above their own respiratory rate, and on prescribed nights additionally apply a transcutaneous CO2 sensor at the ear or chest and keep it attached through the night — sustaining that through the adherence bar Medicare audits.
Where it stands: Both halves are on the record as assumptions and neither is answered. The file states plainly that nothing establishes adherence in ordinary practice as opposed to under trial supervision, and that nothing anywhere establishes at-home sensor application by patients or caregivers [claim 29: unconfirmed]. That same claim's note records the coverage consequence: NCD 240.9 imposes an auditable adherence bar for continued payment — use "at least 4 hours per 24-hour period, on at least 70% of days in a 30-day period", re-established at 6 months and again at 7-12 months. On the pressure the patient would have to tolerate, the file's own figure is wrong and has been corrected downward: the claim that the coverage envelope requires IPAP above 20 cm H2O is [claim 5: refuted] — the NCD's own sentence reads "a minimum IPAP >= 15 cm H2O and backup respiratory rate of at least 14 breaths per minute", and "20 cm H2O" appears nowhere in the section. The 20-plus figure is trial practice, not coverage: HOT-HMV's median setting was IPAP 24 cm H2O (IQR 22-26), EPAP 4, back-up rate 14 [claim 15: verified]. Whether a layperson can actually place the sensor is now on record as an assumption of this deck's own making [claim 31: unverified] — see Slide 6, where it is stated as the observation to make first.
What would settle it: A prospective adherence observation in ordinary practice, not a trial: enrolled patients tracked against the NCD's own 4-hour/70%-of-days threshold over a 30-day period, with sensor-nights logged separately from ventilator-nights. No connector reaches this; it is primary research and a human has to commission it.
If it's false: The therapy stops being paid before the capped-rental period ends, which makes it simultaneously a desirability failure and the revenue failure on the viability deck's Slide 7. Note the one piece of good news buried in the refutation: the tolerability question can be reframed at IPAP >= 15 rather than above 20, which is a materially easier bar.
Would have to be true: A treating pulmonologist or sleep physician would have to be willing to set and document a patient's NIV prescription from an overnight transcutaneous CO2 recording made in the patient's bedroom, in place of an attended in-laboratory titration — and then to prescribe a new entrant's device over the ResMed or Philips platform already in their clinic.
Where it stands: Both consents are unanswered assumptions, and the file flags the first hardest. On the substitution: practitioner acceptance of a home overnight tcCO2 recording as an adequate basis for setting and documenting NIV settings is explicitly written as an assumption with no primary research behind it [claim 30: unconfirmed] — and the coverage half of that claim was checked this pass with an uncomfortable result: NCD 240.9 contains the word "transcutaneous" nowhere, and every CO2 measurement it specifies is arterial (entry at "PaCO2 >= 52 mmHg by arterial blood gas during awake hours"). The same note records the other edge of it: the NCD does not require an attended in-laboratory titration and waives formal sleep testing where the clinician judges sleep apnea is not the predominant cause — so the facility event this device displaces may not be a covered study in the first place. On the brand switch: whether prescribers would choose a new entrant over the incumbent devices they already use is written as an assumption with no connector behind it, and the adoption connector returned unverifiable [claim 27: unconfirmed]; what is on the record is the entrenchment, with both incumbents holding current clearances (K161492, K102465). The third consent is the DME supplier, who dispenses and bills the rental and would have to stock and service an extra consumable sensor [claim 28: unconfirmed]. The fourth is Medicare itself, which has already said yes to the therapy category: NCD 240.9 exists, effective 2025-06-09 [claim 1: verified].
What would settle it: Structured interviews with twenty US prescribing pulmonologists and sleep physicians who currently order home NIV — asking specifically whether they would sign a settings change off a home overnight tcCO2 trace, and what documentation they would want behind it — plus a human retrieval of LCD L33800 and policy article A52517, which is where the documentation requirements would actually live and which no connector in this repo can read [claim 6: unconfirmed].
If it's false: The whole site-of-care shift collapses into a device that records a number nobody acts on, and the viability deck's reimbursement-documentation condition (Slide 2) fails with it. This is this deck's load-bearing condition — see Slide 8.
Would have to be true: The patient would have to accept the Medicare Part B cost share on a capped-rental item plus any non-covered charge for the sensor and its consumables, and the non-monetary costs — a mask and fixed back-up rate that override their own breathing, a heated sensor taped to the ear or chest overnight, and a spouse sharing the bedroom with a ventilator.
Where it stands: The monetary half cannot be stated at all, and the blank is the point: no claim in this file carries a dollar amount for E0471. The DMEPOS fee schedule allowable was not retrieved because the payment cache is absent in this environment [claim 10: unconfirmed], and the capped-rental classification itself is unconfirmed because the 2006 Federal Register rule uses a legacy document number the policy connector rejects outright [claim 8: unconfirmed]. Whether patients would accept whatever that cost share turns out to be is this deck's own write-back [claim 32: unverified]. Whether the sensor is separately payable at all is a viability question and is stated there rather than here — see the viability deck's Slide 2 [claim 34: unverified]. The non-monetary half has no entry in the file in either direction [no claim]; the nearest thing is the tolerability assumption on Slide 2 [claim 29: unconfirmed], which is about adherence rather than about what the patient gives up.
What would settle it: A free download of the CMS DMEPOS fee schedule for E0471 gives the allowable and therefore the 20% coinsurance exposure — that is the cheapest single number in this whole file. The qualitative cost needs patient interviews in the post-exacerbation population, which is primary research.
If it's false: Patients abandon therapy at the point of first invoice rather than at the point of first mask, and the adherence bar on Slide 2 is never reached — which is a revenue failure before it is a clinical one.
Would have to be true: The evidence the prescriber actually demands — a US specialty-society guideline or a colleague's experience, not a regulatory clearance — would have to point at home NIV for chronic hypercapnic COPD, and specifically at CO2-guided titration as the way to do it.
Where it stands: The trial evidence is real and verified, and the guideline evidence is absent from this file entirely. HOT-HMV's admission-free-survival result is confirmed against the PubMed record [claim 15: verified]. But the mechanism the device copies — that the trial titrated NIV to overnight transcutaneous CO2 — is not anchored, and the file is explicit that this is a genuine source silence rather than a tooling gap: the full structured abstract was retrieved and read, it describes the design, population, interventions and median settings, and it says nothing about transcutaneous CO2 or about what the titration target was [claim 16: unconfirmed]. No claim in the file names a US clinical practice guideline of any kind; that condition is this deck's write-back [claim 33: unverified].
What would settle it: For the mechanism, the JAMA paper's methods section for PMID 28528348 — free to read for a human, paywalled to every connector here, and named in the candidate's own Scores table as "free and fast". For the guideline, a search of ATS/ERS/AASM published recommendations on long-term home NIV in hypercapnic COPD, attached by a human.
If it's false: The sales conversation has no third-party endorsement to stand on, and the prescriber consent on Slide 3 gets materially harder — a new entrant asking a physician to change practice on the strength of one UK trial whose titration protocol nobody has read.
Would have to be true: In the first handful of real patients, the patient or caregiver would have to apply the sensor unaided on a prescribed night and produce a trace the practitioner judges interpretable — and the practitioner would have to actually change a setting off it. The falsifying observation is specific: traces that are unreadable, sensors found detached in the morning, or practitioners who look at a good trace and order an attended study anyway.
Where it stands: Nothing in the file establishes at-home sensor application by patients or caregivers; the claim that covers it says so in those terms [claim 29: unconfirmed], and the proposition is now on the record in its own right [claim 31: unverified]. There is one adjacent fact worth having: the SenTec Digital Monitoring System with the V-Sign sensor carries a cleared indication whose own wording includes "if under clinical supervision — home environments" [claim 24: unconfirmed] — which is a cleared indication for home use under supervision, not evidence that an unsupervised layperson can place the sensor.
What would settle it: A ten-patient usability study run by a human, in real homes, on patients in the target population — first-night success rate for sensor placement, fraction of traces the practitioner calls interpretable, and fraction that produce an approved settings change. It is human-factors work a 510(k) would want anyway (see the feasibility deck's Slide 6), so it is not spend that only serves this question.
If it's false: The recording channel — the entire novel half of the device — produces no usable data outside a clinic, and the product reduces to a bi-level ventilator competing on price with two entrenched incumbents.
Every condition above with nothing verified behind it — the no-claim markers, plus the unverified claims this deck itself put on record. Read this slide first.
adoption connector returns
unverifiable.
Note the shape of this list, not only its length. The rubric scores no factor for whether a user wants the thing, so a desirability deck resting entirely on unverified, unconfirmed and one refuted reference — with not a single verified claim standing behind a user-side condition — is the expected result and is information about the pipeline as much as about this idea. The one verified claim this deck leans on (claim 15) is about what happened to 124 patients in the UK under trial supervision, not about what a US patient would choose.
If only one thing from this chair could be checked: whether a treating pulmonologist or sleep physician would be willing to set and document a patient's NIV prescription from a home overnight transcutaneous CO2 recording, in place of an attended in-laboratory titration [claim 30: unconfirmed].
This one rather than the others because every other user-side condition is downstream of it. If the prescriber will not act on a home trace, patient adherence does not matter, sensor usability does not matter, and the device is a bi-level ventilator with an expensive accessory. It is also the condition where the file's own evidence has started to move against the comfortable reading — NCD 240.9 specifies arterial PaCO2 throughout and says "transcutaneous" nowhere — while the document that would actually settle the documentation question, LCD L33800, has never been read by anything in this repo.
Naming it is not a recommendation, a gate, or a kill. It is the answer to "check what first."