Reference products and why people stop using them — chronic subjective tinnitus neuromodulation

1. Summary

The domain dossier's Section 5 flagged one gap above all others: no predicate-specific discontinuation or dropout figure, with a stated reason, had been located at abstract depth for the founding bimodal device, Lenire. This report retrieves full text for the three US-relevant predicate studies and finds retention data in all three, though none reports a stated per-patient reason for non-return, and all three carry a disclosed financial tie to the manufacturer, Neuromod Devices Ltd. [1][2][3]. Two independent, non-manufacturer-affiliated studies of the standard-of-care comparator (hearing aids used as tinnitus sound therapy) quantify device return rates directly, with stated reasons [4][5]. Separately, the FDA (US Food and Drug Administration)'s own device classification registry shows the predicate does not compete alone: product code KLW ("Masker, Tinnitus," 21 CFR (Title 21 of the Code of Federal Regulations) 874.3400) has at least 61 cleared 510(k)s on file, including two decided in 2026, and zero recalls in the last 20 years [6][7][8][9][10]. As of this report's access date, the predicate's own newer registered protocols still carry no posted results [11][12].

2. What Changed

_Baseline (v1). No prior version; this establishes the starting point for future diffs._

3. Details

The predicate's own attrition, at full-text depth

The domain dossier had read the predicate's pivotal trials at abstract depth only and found no per-arm dropout count. Full text of the settings-optimization RCT (randomised controlled trial) (TENT-A2, n=194 randomized, industry-sponsored by Neuromod Devices Ltd.) shows three participants withdrawn before device fitting — two did not attend the enrollment visit and one was withdrawn at that visit for an undisclosed medical condition — leaving 191 who completed fitting. Of those 191, 178 (93.2%) attended the 6-week interim visit, 173 (90.6%) attended the 12-week final visit, and 165 (86.4%) attended the first post-treatment follow-up. No reason is given in the results section for attrition during the treatment period itself [1] [single-source · manufacturer]. A separate, smaller real-world US chart review (n=220, Alaska Hearing & Tinnitus Center) reports: "At FU1, only two patients were lost to follow-up, and six more patients were lost to follow-up at FU2... resulting in a high retention rate of 96.4%," and notes that five of the six lost at the second follow-up had already shown improvement at the first — i.e. the paper frames this as loss to follow-up, not treatment discontinuation, and both authors disclose a Neuromod financial tie [2] [single-source · manufacturer].

A second, independent-cohort US chart review (2026)

A third, previously unretrieved study — a single-site chart review of 140 US patients fitted with Lenire at New York Hearing Doctors, published 2026 in the American Journal of Audiology — gives the most granular disposition data found in this domain to date. Of 140 patients fitted, 24 were lost to follow-up before the 6-week visit (116 attended, 82.9%); a further 30 were lost before the 12-week visit (89 of the 140, 63.6%, had paired data at both timepoints). The authors' own interpretation, tested directly: of 21 bothersome-tinnitus patients who attended the 6-week visit but not the 12-week visit, 57.1% had already met the minimal-clinically-important-difference threshold at 6 weeks, which the paper offers as evidence that "not attending the FU2 appointment does not necessarily indicate a negative treatment outcome." No field safety report, correction, removal, or recall was triggered by this cohort. One author (Hubert H. Lim) discloses a financial interest in and consulting relationship with Neuromod Devices Ltd., and Neuromod's internal funds supported study personnel and ethics fees; patients self-paid for the device and services [3] [single-source · manufacturer].

The broader FDA-cleared "tinnitus masker" device class

The dossier's Section 5 covered only the predicate (QVN) and three non-device standard-of-care categories (TRT, CBT, sound therapy in general). A distinct, older FDA product code exists for tinnitus-specific hardware: KLW, "Masker, Tinnitus," Class II (FDA device class II), regulation 21 CFR 874.3400 [6]. A 510(k) (premarket notification) search on this code returns at least 61 cleared devices dating back to at least 1996 (the connector's own window floor), including two decided in 2026: K260413 (an FDA 510(k) clearance number), "Levo Gen 2 Tinnitus Therapy System" (Levo Medical Corporation, decided 2026-08-13) [7], and K260614, "Tinnitus Therapy" (Sivantos GmbH, decided 2026-06-10) [8], plus a 2023-era wave that includes "Peace N Quiet," "Tinearity G1," and hearing-aid-integrated tinnitus modules from GN Hearing and Starkey [9]. Zero device recalls were found for this product code in the last 20 years [10]. This is a much larger and more active reference-product landscape than the QVN classification the predicate alone occupies — QVN has had zero follower 510(k)s in 3.5 years per the domain dossier's own scan — though none of these KLW clearances is itself evidence that any specific masker device outperforms or is preferred over the predicate; the registry records existence and recency of clearance, nothing about outcomes.

Standard-of-care comparator: quantified hearing-aid discontinuation

The dossier's own gap list named the absence of any discontinuation-rate table for the hearing-aid/sound-generator comparator. Two literature searches this pass locate quantified figures. A retrospective chart review at a Korean university audiology clinic found that of 1,138 patients prescribed hearing aids, 81 (6.14%) returned them; "ineffectiveness of the device was the most frequent reason, accounting for 32.0% of returns," with financial difficulty cited more by younger adults and management difficulty more by elderly adults [4] [single-source]. A US, NIH-funded cohort study of 105 first-time hearing-aid adherents found that 21 (20%) returned their devices for credit within a one-month trial period, and that "the primary differences between these two groups of adherents were in the poorer aided outcomes obtained by those who returned their devices" [5] [single-source]. Neither study is affiliated with a hearing-aid manufacturer or with Neuromod; together they give this domain its first quantified, reasoned discontinuation figures for any comparator, though at different return-rate magnitudes (6.14% clinical-return vs. 20% trial-period-return) reflecting different study designs and populations.

Registry status, checked on this report's run date

Two of the predicate's newer registered protocols named in the domain dossier were re-checked directly against ClinicalTrials.gov on 2026-09-01: TENT-A3 (NCT05227365 (a ClinicalTrials.gov registration number), actual enrollment 112, status COMPLETED) and a University of Minnesota customized-stimulus usability study (NCT07092033, actual enrollment 19, status COMPLETED). Both still carry hasResults=false — no change from the dossier's 2026-08-27 finding — meaning neither protocol's planned outcomes (THI/TFI change; compliance and satisfaction rate) have been posted as findings [11][12] [sponsor-reported].

4. Sources

[1] Conlon B, Hamilton C, Meade E, et al. "Different bimodal neuromodulation settings reduce tinnitus symptoms in a large randomized trial" — Scientific Reports (published 2022-06-30; accessed 2026-09-01). PMID (PubMed identifier) 35773272; full text via PMC9246951 — https://pmc.ncbi.nlm.nih.gov/articles/PMC9246951/ [peer-reviewed] [2] McMahan EE, Lim HH. "Retrospective chart review demonstrating effectiveness of bimodal neuromodulation for tinnitus treatment in a clinical setting" — Communications Medicine (published 2025-04-28; accessed 2026-09-01). PMID 40295853; full text via PMC12037789 — https://pmc.ncbi.nlm.nih.gov/articles/PMC12037789/ [peer-reviewed] [3] Kasper CA, May JM, Crossland NE, Lim HH. "Bimodal Neuromodulation for Tinnitus in a Clinical Practice Setting: Clinically Significant Benefit for Patients With Moderate or Worse Symptoms" — American Journal of Audiology (published 2026, month not given in the PubMed record; accessed 2026-09-01). PMID 41528257; DOI 10.1044/2025_AJA-25-00090 — https://pubmed.ncbi.nlm.nih.gov/41528257/ [peer-reviewed] [4] "Clinical reasons for returning hearing aids" — Korean Journal of Audiology (published 2014-04-01; accessed 2026-09-01). PMID 24782945 — https://pubmed.ncbi.nlm.nih.gov/24782945/ [peer-reviewed] [5] "Differences Between Older Adults Who Do and Do Not Try Hearing Aids and Between Those Who Keep and Return the Devices" — Trends in Hearing (published 2021, month not given in the PubMed record; accessed 2026-09-01). PMID 34057370 — https://pubmed.ncbi.nlm.nih.gov/34057370/ [peer-reviewed] [6] openFDA device/classification, product_code=KLW — "Masker, Tinnitus," Class II, regulation 21 CFR 874.3400, review panel EN (current registry record; accessed 2026-09-01). 21 CFR 874.3400 — https://api.fda.gov/device/classification.json?search=product_code:KLW [federal-registry] [7] openFDA device/510k, k_number=K260413 — "Levo Gen 2 Tinnitus Therapy System" / Levo Medical Corporation, decided 2026-08-13, product_code=KLW under 21 CFR 874.3400 (accessed 2026-09-01). K260413 — https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm?ID=K260413 [federal-registry] [8] openFDA device/510k, k_number=K260614 — "Tinnitus Therapy" / Sivantos GmbH, decided 2026-06-10, product_code=KLW under 21 CFR 874.3400 (accessed 2026-09-01). K260614 — https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfpmn/pmn.cfm?ID=K260614 [federal-registry] [9] openFDA device/510k search, product_code=KLW, decision_date 1996-09-08 to 2026-09-01 — 61 total matches (this repo's connector floor, truncated at 50 shown), including K233435 "Peace N Quiet" and K223694 "Tinearity G1," under 21 CFR 874.3400 (accessed 2026-09-01). 21 CFR 874.3400 — https://api.fda.gov/device/510k.json?search=product_code:KLW [federal-registry] [10] openFDA device/recall, product_code=KLW, event_date_initiated 2006-09-06 to 2026-09-01 — 0 records, under 21 CFR 874.3400 (accessed 2026-09-01). 21 CFR 874.3400 — https://api.fda.gov/device/recall.json?search=product_code:KLW [federal-registry] [11] ClinicalTrials.gov record NCT05227365 — "Treatment Evaluation of Neuromodulation for Tinnitus - Stage A3" (TENT-A3), Neuromod Devices Ltd. (record dated 2022-12-27; status COMPLETED; hasResults=false; accessed 2026-09-01). NCT05227365 — https://clinicaltrials.gov/study/NCT05227365 [trial-registry] [12] ClinicalTrials.gov record NCT07092033 — "Assessment of Customized Bimodal Stimulation for Tinnitus," University of Minnesota (record dated 2026-08-04; status COMPLETED; hasResults=false; accessed 2026-09-01). NCT07092033 — https://clinicaltrials.gov/study/NCT07092033 [trial-registry]

5. Sourcing & Gaps

Well established: Nothing in Section 3 rests on two independent, non-affiliated primary sources for the same specific quantity — every predicate-attrition figure comes from a single study, and the three predicate studies, while independently conducted cohorts, all share a disclosed Neuromod financial tie, so they are not treated as mutually corroborating even where their retention percentages point the same direction (roughly 82–96% retained through the final follow-up, none reporting a discontinuation reason). The zero-recall finding for product code KLW rests on a single federal registry query, which is the normal evidentiary shape for a registry fact rather than a gap.

Thin: All three predicate-attrition figures [1][2][3] are single-source and manufacturer-tied. Both hearing-aid discontinuation figures [4][5] are single-source, though neither is manufacturer-affiliated. The KLW clearance count [6][7][8][9] is a registry scan, not independently cross-checked against a second database.

Rescoped from class 3: None in this report. Every proposition is class 1 (a named registry record's status or count) or class 2 (a measured retention/return figure in a cited study), not a question about what a person would do.

Out of scope: MAUDE (Manufacturer and User Facility Device Experience) adverse-event narratives for Lenire or any KLW-coded device (no connector in this repo reads MAUDE — this is the subject of a separate, already-planned manifest row for this domain's pattern, per the potassium-monitoring-ckd sibling report's own precedent). HCPCS (Healthcare Common Procedure Coding System), payment-rate, and coverage-policy material (no local CMS (Centers for Medicare & Medicaid Services) cache; already flagged at the domain level). CPT (Current Procedural Terminology) descriptors (AMA-licensed). Non-US clearance or recall records (no connector in this repo queries a non-US source).

Not searched vs. not found: A per-patient stated reason for on-treatment attrition in TENT-A2 or the two chart reviews was searched for at full-text depth this pass and not found — a stronger negative than the domain dossier's original "not searched at full-text depth" note, but still an absence in what these three specific papers report, not proof no such analysis exists anywhere (e.g. in a supplemental file not retrieved here). Outcome data for any specific KLW-coded device beyond its clearance date was not searched — the 510(k) database records clearance, not performance, and no clinical-outcomes literature search was run per named KLW product. A head-to-head trial of the predicate against a hearing aid, TRT, or CBT arm was not searched again this pass; the domain dossier already recorded none located as of 2026-08-27.

[inference] The three predicate-linked retention figures (93.2%/90.6% at TENT-A2's two on-treatment visits; 96.4% in the Alaska chart review; 82.9%/63.6% in the New York chart review) are not the same measurement — different denominators, different visit schedules, and different definitions of "attended" — so reading them as a single converging "the predicate's attrition rate is low" claim would overstate what any one paper supports; this report presents them as three separate, individually-thin figures rather than combining them into a plain-stated line, which is my own judgment about how to apply the ≥2-independent-source rule here rather than something any of the three papers states.

6. Claim Candidates