Stress urinary incontinence — conservative and device-based treatment

Stress urinary incontinence (SUI) is a common, non-life-threatening condition whose first-line therapy (pelvic floor muscle training, PFMT) is cheap, guideline-endorsed, and demonstrably hard to sustain. The domain question a device-based candidate has to answer is not whether SUI is common — it is — but why the existing conservative and device options are abandoned at the rates they are, and whether a closed-loop home device changes that mechanism or just adds a new failure mode to the same one.

Read the thinness before reading the content. This is a first-pass dossier built in a single research session. Section 5 (reference products and abandonment) and Section 4 (standard of care) carry real, sourced, quoted content — including one product's own pivotal trial dropout numbers, pulled from the raw ClinicalTrials.gov results API rather than inferred. Sections 2, 3, and 7 are effectively unsearched. Section 6 (reimbursement) is blocked in the same way it is in idea-001 itself: the CMS data files the connectors need do not exist in this checkout, and CMS's own coverage-database pages return HTTP 403 to a direct fetch, the same bot-mitigation pattern already documented for fda.gov elsewhere in this repo. One new primary source was found there anyway (Noridian PDAC coding guidance, §6).

This file carries no confidence tags. Every line below is context with a source attached. To affect a score, a line must be copied into a candidate's ## Claims as unverified via engine/scripts/append_research_claims.py and pass the Verifier or the Corroborator. See README.md in this directory.

Citation format. Sources are numbered [1], [2], ... in order of first use, and the full resolvable identifier for each number is in ## Sources at the end of this file. This is a formatting convenience only — it carries no confidence information, exactly like the rest of this file.

1. Condition & epidemiology

as_of: 2026-08-27 · sources searched: connectors literature "stress urinary incontinence prevalence United States women" --since 10y --limit 20 (470 total matches, floor) plus literature-detail on the top hit · not searched: incidence, demographic concentration by race, geographic distribution, prevalence trend broken out by parity/BMI beyond the single multivariable model below

Verbatim:

"Among 6,446 participants, the weighted prevalence of UI was 22.3% (95% CI, 20.8-23.8), significantly higher than the 17.1% reported in 2005-2010 (P<0.001). Mixed UI was most common (46.6%), followed by stress (16.2%) and urgency (9.5%)." — [1]

Read this narrowly. The 16.2% figure is the stress-predominant share of women who already have UI by the ISI≥3 definition, not "16.2% of US women have SUI" — computing a population-level SUI prevalence from these two percentages (22.3% × 16.2% ≈ 3.6% of all women) would be this dossier doing arithmetic the source paper did not do, and is flagged here rather than done. idea-001's own Claims section carries an unsourced "more than 15 million US women report SUI symptoms" line (candidate claim on prevalence, tagged unconfirmed, sourced only to "a market claim"); this NHANES figure is the closest primary replacement found this pass, but it is not a like-for-like number and should not silently retag that claim's magnitude as confirmed.

Gaps: Incidence (as opposed to point prevalence); demographic concentration by race and geography; a population-level SUI-only prevalence figure computed and reported as such by a primary source, rather than derived here from a subtype share.

2. Care pathway & clinical setting

as_of: 2026-08-27

Not searched. No connector or web query was run for this section this pass. Where a woman is typically first seen (primary care, OB/GYN, urogynecology, pelvic floor physical therapy), what triggers referral to a specialist, and what a PFMT-supervision visit actually consists of in terms of staffing and equipment are all open. The one fact bearing on this section that surfaced incidentally in §4: PFMT is most often delivered by a pelvic floor physical therapist over a course of supervised visits, and idea-001's home device is a substitute for at least part of that visit cadence — but no source below actually characterizes visit frequency, duration, or setting.

Gaps: the entire section.

3. User groups & incentives

as_of: 2026-08-27

Not searched. No source was queried this pass. Recorded so the shape is visible: prescriber (urogynecologist, urologist, or PT-referring primary care physician), operator (the patient herself, for a home device — unlike most domains in this repo, patient and operator are the same person here), purchaser (the patient or her DME benefit), payer (Medicare Part B DME or commercial), and the pelvic floor PT practice whose in-clinic 90912/90913 revenue a home device may substitute for or add to, are five parties with plausibly divergent interests. None of this is evidenced below.

Gaps: the entire section.

4. Standard of care & clinical evidence

as_of: 2026-08-27 · sources searched: connectors literature on PFMT-adherence and electrical-stimulation queries (multiple, see §5 and below), literature-detail on 4 PMIDs, one directly-fetched AUA guideline page · not searched: KDIGO-equivalent urogynecology-specific guideline documents beyond AUA/SUFU, ICS (International Continence Society) consensus statements, industry-sponsored comparative-effectiveness studies beyond what is cited here

Verbatim, for anything load-bearing:

"For subjective cure of SUI, we found moderate-quality evidence that ES is probably better than no active treatment (risk ratio (RR) 2.31, 95% CI 1.06 to 5.02)... Low-quality evidence suggests that there may be no difference in cure or improvement for ES versus PFMT (RR 0.85, 95% CI 0.70 to 1.03)... Adverse effects were rare: in total nine of the women treated with ES in the trials reported an adverse effect." — [2]

"There is high certainty that PFMT is more beneficial than control for all types of UI for outcomes of cure or improvement and quality of life." — [3]

"In patients with SUI or stress-predominant MUI, clinicians may offer the following non-surgical treatment options: (Expert Opinion)" — pessary, vaginal inserts, "Pelvic floor muscle exercises (PFME) ± biofeedback" — AUA/SUFU SUI Guideline, Statement 11, as retrieved directly from the guideline's own page — [22]

Read this carefully. The negative finding here matters as much as the positive one, per this dossier's own instructions: the Cochrane review's own authors, not a title, say the comparison this candidate's mechanism most needs to win — ES versus PFMT — has no established advantage either way, on low-quality evidence, after 56 trials. A closed-loop EMG-gated device is a mechanistic refinement on the "ES" arm of this literature, and no source found this pass tests that specific refinement against open-loop ES or against PFMT alone — consistent with idea-001's own Claims section, which already flags "closed-loop EMG-gated stimulation produces a better continence outcome... by a margin a clinician would act on" as an unverified deck-surfaced assumption (candidate claim, feasibility deck, 2026-08-24).

Gaps: No comparative trial of closed-loop EMG-gated stimulation specifically (vs. open-loop ES or vs. PFMT) was found or is likely to exist pre-market — that is a bench/pilot question, not a literature gap. ICS consensus statements, not searched. Cost-effectiveness data beyond what the Cochrane review states was not found ("we did not identify any economic evaluations").

5. Reference products — and why people stop using them

as_of: 2026-08-27 · sources searched: connectors literature (multiple queries, see below), connectors trials + trial-detail on NCT03782116, a direct clinicaltrials.gov/api/v2 fetch for that trial's posted results (not just its protocol), connectors clearance-summary on five K-numbers, connectors recalls on two product codes and by firm name · not searched: MAUDE (no connector exists anywhere in this repo — docs/backlog.md 1.9), any post-market real-world adherence data for InTone/ELITONE/InWave beyond their own pivotal studies

This domain's incumbent set has two layers: the non-device standard of care (PFMT, delivered in clinic or at home unsupervised) and the prescription home NMES devices this candidate's product code pool competes against directly. Both have real, quoted abandonment records.

Pelvic floor muscle training (PFMT / Kegel exercise), unsupervised or clinic-taught

• What it is: first-line conservative therapy, voluntary pelvic floor muscle contraction exercises, with or without clinician supervision or biofeedback.
• Price / what it realizes per unit: not established here; typically billed as PT visits (CPT 97110/97112-class codes, out of scope — AMA-licensed) rather than a device charge.
• Evidence behind it: §4 above — high-certainty benefit over control, per Cochrane [3].
• Adoption: the guideline default; no adoption rate figure found.
• Abandonment record — why people stop:

• Verbatim, for anything load-bearing:

"Long-term adherence to PFMT varied between 10% and 70%... Losses to follow-up during the long-term period ranged between 0% and 39%." — [7]

"One hundred and thirty-two (63%) women completed all questionnaires, 55 of whom (41.7%...) reported adherence to PFMT, 75 (56.8%...) discontinued training... Although relatively high rates of 5-year adherence to training were demonstrated among our patients, this adherence was not associated with superior treatment outcomes." — [5]

"forgetting to do the exercises and boredom with the exercises were the factors most strongly related to low treatment adherence, while motivation and commitment were associated with high adherence" — [6]

This is the class-2 revealed-behaviour twin docs/research-lane-spec.md §3.1 asks for. The unanswerable class-3 question in idea-001's desirability deck is "would women find an intravaginal home device acceptable enough to start and continue treatment... given a non-invasive external alternative exists" (candidate claim, deck-surfaced, tagged unverified). Nobody can answer that. What is answerable, and answered above: women already prescribed the guideline-recommended, lowest-friction, no-hardware conservative therapy discontinue it at rates from roughly 30% to 90% depending on cohort and horizon (1–39% loss to follow-up plus 56.8% self-reported discontinuation at 5 years in the most detailed single cohort), and the stated reason in the one study that asked is not device acceptability but forgetting and boredom — a habit-formation problem, not an invasiveness problem. That is evidence the underlying behaviour (sustained pelvic floor exercise) is hard to sustain regardless of modality; it is not evidence a home NMES device fixes that, and it should not be read as such.

InTone / InToneMV (InControl Medical) — this candidate's own named predicate pool

• What it is / cleared indication: intravaginal probe, electrical stimulation plus biofeedback (manometric on some clearances, per below). Product code KPI, 21 CFR 876.5320, Class II. — [16], [19]
• A finding not in idea-001's own frontmatter or Claims: InControl Medical holds at least three distinct KPI clearances under variants of the "InTone" name, not two. Searching "InToneMV" by applicant name returns K134020 (decided 2014-02-25) as the top match — a different clearance from K133826 (decided 2014-01-06), which is the number idea-001's frontmatter and Claims cite. Both are named "InToneMV." Their Indications for Use also differ materially:

"The lnToneMVV device is a nan-implanted electrical stimulator Indicated for use In the treatment of female urinary incontinence... Intended for acute and ongoing treatment of mixed urinary incontinence... The biofeedback feature can be used for muscle re-education purposes." — K133826, verbatim (OCR artifacts as returned) — [10]

"InToneMV is intended to provide electrical stimulation and/or visual biofeedback (via manometry) for the treatment of male and female urinary and fecal incontinence." — K134020, verbatim — [11]

K133826 names female-only, urinary-only, EMG-implied biofeedback ("muscle re-education"); K134020 names male and female, urinary and fecal, and explicitly manometric (pressure) biofeedback. This is a genuine finding this pass surfaced independently, not a re-derivation of anything already in idea-001: a single applicant filed at least two same-named-device 510(k)s with different indications and different feedback modalities within one month of each other, which sharpens (not resolves) idea-001's already-open question of which InControl clearance, if any, is the right predicate comparison. Not further investigated this pass — flagged for the Verifier rather than resolved here.

• Evidence behind it: company-sponsored open-label pilot, n=33 enrolled (28 completed), 12-week single-arm design, no control group. — [4]
• Adoption: not established; no real-world utilization figure found.
• Abandonment record — why people stop:

• Verbatim:

"Thirty-three women were enrolled; five patients withdrew and were excluded... Device usability was good, with a median SUS of 86.3 and a median expected use of 107% (33-140%)." — [4]

Read this narrowly. A 15.2% pre-completion withdrawal rate in a 12-week open-label pilot with no comparator arm is a real number but a thin one — it says nothing about durability beyond 12 weeks, and the reported System Usability Scale (86.3, "good") is a company-sponsored, unblinded, self-report instrument. No independent real-world adherence or discontinuation study for InTone/InToneMV was found. Recorded as "not found," not merely "not searched" — the specific query Zynex InWave neuromuscular stimulation urinary incontinence returned zero PubMed records, and no InControl-Medical-specific adherence study beyond the pilot itself surfaced across the queries run this pass.

ELITONE / ELITONE URGE (Elidah, Inc.) — adjacent market analog, not a KPI-code predicate

• What it is / cleared indication: transcutaneous (external, perineal skin electrode) stimulator. Product code QAJ, 21 CFR 876.5330, Class II. K183585 (2019, stress UI); K223884 (2023, urge UI). — [14], [15], [17]

"ELITONE is a non-implanted muscle stimulator designed to treat stress urinary incontinence in women." — K183585 IFU, verbatim — [14]

• Price / what it realizes per unit: not established; not searched for a direct-to-cash or DME price point this pass.
• Evidence behind it: sponsor-run RCT, NCT03782116, "Stress Incontinence Trial With Elitone Device," randomized/parallel/quadruple-masked, sham-controlled, sponsor Elidah, Inc. — [9]
• Adoption: not established.
• Abandonment record — why people stop (this pass's strongest single finding):

• Verbatim: this table is built directly from the resultsSection JSON returned by https://clinicaltrials.gov/api/v2/studies/NCT03782116?format=json&fields=ResultsSection (fetched 2026-08-27), not from a summary page — the field-level values are: STARTED treatment 57 / control 30; COMPLETED (intervention period) 40 / 18; NOT COMPLETED 17 / 12; drop/withdraw reasons "Lost to Follow-up" 11/7 and "Withdrawal by Subject" 6/5; second-period (analysis-set) STARTED 40/18, COMPLETED 35/13, NOT COMPLETED 5/5 for "Protocol Violation". Primary outcome measurements: treatment group value "52.8" spread "36.3", control group value "7.9" spread "55.9", n=35/13. — [9]

Read this carefully in both directions, the same discipline as the potassium dossier's capillary-sampling section. A 29.8%–40.0% attrition rate between enrollment and intervention completion, in the sponsor's own registered, sham-controlled pivotal trial of a cleared prescription home stimulation device, is the closest thing in this domain to product-level abandonment evidence for the exact device category idea-001 proposes — home, self-administered electrical stimulation for SUI. It is higher than the InTone pilot's 15.2% and sits inside the wide 0%–39% loss-to-follow-up range the PFMT systematic review [7] reports for pelvic exercise itself. What this does and does not establish: it is evidence that the home-electrical-stimulation product category, not just PFMT, loses roughly a third to two-fifths of enrolled trial participants before intervention completion, for a mix of lost-to-follow-up and self-withdrawal reasons the registry record does not further break down. It is not evidence about post-approval, non-trial, prescribed-and-paid-for use, which is the setting idea-001 actually sells into and which no source located this pass addresses for any device in this domain.

InWave (Zynex Medical) — cleared under the same KPI code, closest indication match

• What it is / cleared indication: K121820, decided 2012-08-24, KPI / 876.5320. — [13]

"The InWave is intended to provide electrical stimulation and neuromuscular re-education for the purpose of rehabilitation of weak pelvic floor muscles for the treatment of stress urge and mixed urinary incontinence in women." — verbatim IFU — [13]

• Evidence behind it, adoption, abandonment record: not found. The literature query "Zynex InWave neuromuscular stimulation urinary incontinence" returned zero PubMed records. No pivotal study, adherence figure, or complaint record for InWave specifically was located by any query run this pass. Recorded as a genuine absence, not an unsearched gap — this is the one incumbent in this domain with essentially no public evidence trail beyond its own cleared Indications for Use.

Recalls, across the whole reference-product set

• recalls --product-code KPI --since 2010-01-01 and `--product-code QAJ --since 2010-01-01`, run directly this pass: zero records, both codes. — [20]
• Independently corroborates a null already on record in knowledge-base/signals/log.md (2026-08-23 entry): five separate scans — both product codes over a 24-month window, plus firm-name searches for InControl Medical, Elidah, and Zynex — all returned status: none. — [24]
• Read this as an absence of a query result, not a safety finding. No recall exists in either code searched, over the windows searched; that is not evidence the products are safe, only that no recall record surfaced. recalls is a Stage 0 ScanResult and carries no confidence tag by construction either way.

Gaps: No MAUDE-adjacent complaint narrative for any named product — no connector reaches MAUDE anywhere in this repo. No independent (non-sponsor) real-world adherence study for any of the four named devices (InTone/InToneMV, ELITONE, ELITONE URGE, InWave). No pricing data for any device. No post-market, non-trial discontinuation figure for the home-NMES category — everything quantitative above about home-stimulation abandonment comes from a single pivotal RCT's trial-conduct data, not from prescribed real-world use.

6. Reimbursement landscape

as_of: 2026-08-27

Blocked in the same way as idea-001's own Claims section. hcpcs E0740, payment E0740, and coverage E0740 were each re-run this pass and returned the identical unconfirmed result already on record in the candidate file: data/ in this checkout holds no CMS HCPCS Level II file, no DMEPOS fee schedule, and no Medicare coverage-policy cache. — [23]

Read the PDAC finding narrowly, against idea-001's own unresolved claim 16. The candidate's Claims section already carries "per PDAC coding guidance, E0740 is intended to describe... devices with integrated monitoring capability" tagged unconfirmed, sourced only to "an unchecked web-search recollection." The Noridian JD DME page fetched this pass (med.noridianmedicare.com/web/jddme/policies/dmd-articles/pdac-code-review-e0740...) states the same substance in near-identical language, retrieved directly from what presents as the DME MAC/PDAC's own published coding-review article rather than a secondary summary. This is a stronger source than what the candidate currently cites, but it is still a web fetch of a contractor page, not a typed connector, and it has not been independently corroborated against a second primary document — it is drafted as a claim below at the strength that retrieval actually supports.

Gaps: the full NCD 230.8 text (blocked, not merely unsearched — CMS gates automated access the same way fda.gov does); DMEPOS allowed amount for E0740 (blocked on missing data/ cache); any commercial-payer coverage policy; RTM billing codes (CPT 98977/98980/

1. are AMA-licensed and out of reach here by design, same as in idea-001 itself.

7. Technology trajectory

as_of: 2026-08-27

Not searched, beyond what fell out of §5. The KPI-code clearance-rate acceleration idea-001's own signal log already surfaced (roughly 2.8/year 2014–2023 rising to 14 across 2024–2026 YTD, six of the most recent clearances including several apparent contract-manufacturer listings and one magnetic, not electrical, stimulator) is carried here by reference rather than re-derived — see knowledge-base/signals/log.md, 2026-08-23 entry. — [24] What has changed in intravaginal EMG-sensing hardware specifically, surface-EMG artifact-rejection methods, or closed-loop stimulation control over the last decade was not searched this pass.

Gaps: the entire technology-trajectory question beyond the one cross-referenced clearance count, which is itself an unfiltered ScanResult and cannot carry a confidence tag (docs/backlog.md 1.10).

8. Market players & dynamics

as_of: 2026-08-27

Gaps: funding, launch, and exit activity for any named player (no connector reaches PitchBook/Crunchbase-class data); FDA Establishment Registration & Device Listing for the full KPI/QAJ roster (no connector); whether Elvie or Perifit hold an undiscovered pelvic-floor clearance under a different device name (the same open question idea-001 already carries, unresolved here too).

9. Open questions — and what would settle them

Class-3 questions are recorded here and never answered here — see the substitutions above.

10. Provenance & staleness

What this dossier has never looked at: the entire care pathway (§2) and user-groups (§3) sections; any incidence (as opposed to prevalence) figure; MAUDE or any complaint-narrative source for any named device, because no connector in this repo reaches it; independent, non-sponsor real-world adherence or discontinuation data for any of the four named prescription devices — everything quantitative on home-device abandonment in §5 comes from each device's own sponsor-run pivotal trial, not from prescribed, paid-for, real-world use; pricing for any device or the PFMT care pathway; CPT-coded billing detail (AMA-licensed, permanently out of reach here by design); the full text of Medicare NCD 230.8 (located, not retrievable — CMS gates automated access); and any technology-trajectory content beyond a single cross-referenced clearance-rate count. Absence of scanning is not absence of change, and a later reader should treat this as a first pass, not as coverage of the domain.

Sources

1. PMID 42606983 — "Prevalence of Urinary Incontinence in the United States 2015-2020," Urogynecology (Philadelphia, Pa.), 2026 Aug 17 (NHANES 2015–March 2020 analysis). https://pubmed.ncbi.nlm.nih.gov/42606983/
2. PMID 29271482 — "Electrical stimulation with non-implanted devices for stress urinary incontinence in women," Cochrane Database of Systematic Reviews, 2017 Dec 22. https://pubmed.ncbi.nlm.nih.gov/29271482/
3. PMID 36053030 — "Conservative interventions for treating urinary incontinence in women: an Overview of Cochrane systematic reviews," Cochrane Database of Systematic Reviews, 2022 Sep 2. https://pubmed.ncbi.nlm.nih.gov/36053030/
4. PMID 25074260 — "InTone: a novel pelvic floor rehabilitation device for urinary incontinence," International Urogynecology Journal, 2015 Jan. https://pubmed.ncbi.nlm.nih.gov/25074260/
5. PMID 26397715 — "Pelvic floor muscle training for female stress urinary incontinence: Five years outcomes," Neurourology and Urodynamics, 2017. https://pubmed.ncbi.nlm.nih.gov/26397715/
6. PMID 29095511 — "Factors influencing long-term adherence to pelvic floor exercises in women with urinary incontinence," Neurourology and Urodynamics, 2018. https://pubmed.ncbi.nlm.nih.gov/29095511/
7. PMID 22847318 — "Does it work in the long term?--A systematic review on pelvic floor muscle training for female stress urinary incontinence," Neurourology and Urodynamics, 2013 Mar. https://pubmed.ncbi.nlm.nih.gov/22847318/
8. PMID 36701620 — "Digital Therapeutic Device for Urinary Incontinence: A Longitudinal Analysis at 6 and 12 Months," Obstetrics and Gynecology, 2023 (Renovia Inc.-funded RCT, NCT04508153). https://pubmed.ncbi.nlm.nih.gov/36701620/
9. NCT03782116 — "Stress Incontinence Trial With Elitone Device," ClinicalTrials.gov, sponsor Elidah, Inc., status COMPLETED, hasResults=true, last update posted 2025-01-13. Protocol: https://clinicaltrials.gov/study/NCT03782116 . Results data pulled directly from https://clinicaltrials.gov/api/v2/studies/NCT03782116?format=json&fields=ResultsSection (fetched 2026-08-27).
10. openFDA device/510k, k_number=K133826 — INTONEMV, Incontrol Medical, LLC, decided 2014-01-06; Indications for Use per FDA 510(k) summary https://www.accessdata.fda.gov/cdrh_docs/pdf13/K133826.pdf.
11. openFDA device/510k, k_number=K134020 — INTONEMV, Incontrol Medical, LLC, decided 2014-02-25; Indications for Use per FDA 510(k) summary https://www.accessdata.fda.gov/cdrh_docs/pdf13/K134020.pdf.
12. openFDA device/510k, k_number=K150180 — InTone, Incontrol Medical, LLC, decided 2015-03-16.
13. openFDA device/510k, k_number=K121820 — INWAVE, Zynex Medical, Inc., decided 2012-08-24; Indications for Use per FDA 510(k) summary https://www.accessdata.fda.gov/cdrh_docs/pdf12/K121820.pdf.
14. openFDA device/510k, k_number=K183585 — Elitone Device, Elidah, Inc., decided 2019-02-11; Indications for Use per FDA 510(k) summary https://www.accessdata.fda.gov/cdrh_docs/pdf18/K183585.pdf.
15. openFDA device/510k, k_number=K223884 — ELITONE Urge Urinary Incontinence Device, Elidah, Inc., decided 2023-02-24; Indications for Use per FDA 510(k) summary https://www.accessdata.fda.gov/cdrh_docs/pdf22/K223884.pdf.
16. 21 CFR 876.5320, "Nonimplanted electrical continence device" — eCFR versioner, title 21, issue 2026-08-19.
17. 21 CFR 876.5330, "Transcutaneous electrical continence device" — eCFR versioner, title 21, issue 2026-08-19; special controls added 86 FR 73971 (Dec. 29, 2021).
18. 21 CFR 884.5940, "Powered vaginal muscle stimulator for therapeutic use" — eCFR versioner, title 21, issue 2026-08-19.
19. openFDA device/classification, product_code=KPI — "Stimulator, Electrical, Non-Implantable, For Incontinence," device class 2, regulation 876.5320, review panel GU.
20. openFDA device/recall, product_code=KPI and product_code=QAJ, event_date_initiated window 2010-01-01 to 2026-08-27 — zero records both codes, queried directly 2026-08-27.
21. Noridian Healthcare Solutions, JD DME, "PDAC Code Review - E0740 Non-Implantable Pelvic Floor Electrical Stimulator." https://med.noridianmedicare.com/web/jddme/policies/dmd-articles/pdac-code-review-e0740-non-implantable-pelvic-floor-electrical-stimulator (fetched 2026-08-27).
22. American Urological Association / SUFU, "Stress Urinary Incontinence (SUI) Guideline," Guideline Statements 8 and 11. https://www.auanet.org/guidelines-and-quality/guidelines/stress-urinary-incontinence-(sui)-guideline (fetched 2026-08-27).
23. data/README.md and this repo's hcpcs/payment/coverage connectors, re-run 2026-08-27 against HCPCS E0740 — all unconfirmed: no local cache at data/hcpcs_level_ii.csv, data/hcpcs_payment_rates.csv, or data/medicare_coverage_policies.csv.
24. knowledge-base/signals/log.md, entries dated 2026-08-23 — KPI 510(k) clearance-rate scan and KPI/QAJ recall-null scan (both ScanResults, no confidence tag, carried here by reference).