Consensus Statements on Tinnitus Treatment: A Delphi Study by the Korean Tinnitus Study Group

The study addresses the need for consensus-based guidance on tinnitus treatment due to its high global prevalence and significant impact on sleep, attention, work, social communication, and mental health. Existing evidence from randomized clinical trials is limited, and prior guidelines (e.g., AAO-HNSF 2014) offered few strong recommendations across commonly used therapies (sound therapy, TRT, CBT, medications, supplements, acupuncture, neuromodulation). To support clinicians, the Korean Tinnitus Study Group conducted a Delphi survey to synthesize expert opinion into consensus statements on patient candidacy and treatment modalities for tinnitus.

Prior to the Delphi survey, a preliminary systematic review was conducted to inform the statements. Reviewers examined 2 Cochrane reviews (post-2010), 16 narrative/review articles, and 4 randomized controlled trials identified via PubMed. The discussion further summarizes evidence: limited or mixed support for medications (e.g., steroids in sudden sensorineural hearing loss–associated acute tinnitus; insufficient evidence for Ginkgo biloba and acamprosate), psychotherapy (TRT shows benefit despite low-quality evidence; CBT improves tinnitus-related distress, mood, and quality of life), auditory rehabilitation (hearing aids effective with hearing loss; possible benefits of cochlear and middle ear implants; variable evidence for BAHA and CROS), neuromodulation (overall inconclusive in guidelines; mixed recent reports), and other modalities (limited support for hyperbaric oxygen and acupuncture; emerging support for internet/smartphone-based CBT and self-management; physical therapy effective in somatic tinnitus linked to cervical spine/TMJ disorders).

Design: Two-round modified Delphi survey of expert otologists to develop clinical consensus statements on tinnitus treatment. Timeline and recruitment: Panelists recruited in 2021; survey conducted January–March 2022. Thirty-five experts were invited; 28 completed Round 1 and qualified for Round 2. All participants were fellowship-trained in otology with ≥10 years of experience; 2 worked in primary clinics and 26 in secondary/tertiary institutions. Survey development: Initial pool of 38 statements across 6 categories (candidates for treatment; medication; TRT/CBT; auditory rehabilitation; neuromodulation; miscellaneous). After Round 1 scoring and an online discussion meeting, similar items were combined and duplicates/unnecessary items removed, yielding 31 statements for Round 2. Administration: Surveys conducted via email; responses anonymous. Scoring scale: 1–9 based on GRADE guidance (1–3 high disagreement, 4–6 important but not critical, 7–9 high agreement). Consensus criteria: Inclusion (positive consensus) required ≥75% scoring 7–9 and ≤15% scoring 1–3; Exclusion (negative consensus) required ≥75% scoring 1–3 and ≤15% scoring 7–9. Reliability measures: Content validity ratio (CVR) threshold ≥0.37 (P < 0.05) to indicate significant agreement; Kendall’s coefficient of concordance (W) computed overall and by category, where values closer to 1 indicate greater convergence (interpretation: 0.9 unusually strong; 0.7 strong; 0.5 moderate; 0.3 weak; 0.1 very weak). Consensus meeting: After analysis, an expert meeting finalized inclusion/exclusion of statements.

• Overall consensus: 19 of 31 statements met consensus criteria.
• Candidates for treatment: All 3 statements reached consensus (e.g., tinnitus ≥5 minutes on most days; unilateral tinnitus with ipsilesional hearing loss; tinnitus accompanied by dizziness).
• Medication: 3 of 8 statements reached consensus, including: • Steroids effective for acute tinnitus with sudden hearing loss (100% agreement; CVR 1.000). • Antidepressants and anxiolytics effective for managing comorbid depression/anxiety in subjective chronic tinnitus (100% agreement; CVR 1.000; effect pertains to mood, not tinnitus loudness). • Carbamazepine effective for typewriter’s tinnitus (100% agreement; 96.4% agreement excluding authors of the reference; CVR 1.000).
• TRT/CBT: Both statements reached consensus: • TRT effective for subjective chronic tinnitus (92.9% agreement; CVR 0.857). • CBT effective for subjective chronic tinnitus (89.3% agreement; CVR 0.786).
• Auditory rehabilitation: 5 of 7 statements reached consensus: • Hearing aids effective for subjective chronic tinnitus with hearing loss (100% agreement; CVR 1.000). • Sound generator therapy effective for subjective chronic tinnitus (85.7% agreement; CVR 0.714). • Cochlear implantation effective for subjective chronic tinnitus in severe-to-profound hearing loss (96.4% agreement; CVR 0.929). • Middle ear implantation effective for moderate-to-moderately severe hearing loss (78.6% agreement; 75.0% excluding authors; CVR 0.571). • Sound generator therapy effective for tinnitus with hyperacusis (75.0% agreement; CVR 0.500). • BAHA and CROS did not reach consensus (agreement ≤39.3% and 35.7%, respectively).
• Neuromodulation: None of 5 statements reached consensus (e.g., rTMS, TENS, tVNS, tDCS, tRNS).
• Miscellaneous treatments: All 6 statements reached consensus, mostly indicating ineffectiveness of certain modalities: • Hyperbaric oxygen therapy ineffective for chronic tinnitus (85.7% agreement; CVR 0.714). • Acupuncture ineffective (96.4% agreement; CVR 0.929). • Sound conditioning devices ineffective (85.7% agreement; CVR 0.714). • Ear moxibustion ineffective (96.4% agreement; CVR 0.929). • Smartphone/internet-based platforms for sound therapy, counseling, education effective for chronic tinnitus (82.1% agreement; CVR 0.643). • Physical therapy for cervical spine/TMJ disorders effective for somatic tinnitus (89.3% agreement; CVR 0.786).
• Agreement levels (Kendall’s W): Overall W = 0.476 (weak-to-moderate). By category: Medication W = 0.635 (strongest); Auditory rehabilitation W = 0.534; Miscellaneous W = 0.373; Candidates W = 0.107; TRT/CBT W = 0.107; Neuromodulation W = 0.094 (lowest).

The consensus provides actionable guidance where high-quality evidence has been limited. Experts strongly endorsed patient candidacy criteria (duration, unilateral tinnitus with ipsilesional hearing loss, associated dizziness), aligning with clinical vigilance for underlying pathology. They favored hearing aids for tinnitus with hearing loss and sound therapy for chronic tinnitus, consistent with pragmatic clinical practice and AAO-HNSF options. Psychotherapeutic approaches (TRT, CBT) were positively endorsed, reflecting benefits in distress reduction despite heterogeneous trial quality. Pharmacologic consensus was focused: steroids for sudden hearing loss–associated acute tinnitus; psychotropics for comorbid mood/anxiety rather than tinnitus per se; carbamazepine for typewriter’s tinnitus. Neuromodulation did not achieve consensus, mirroring guideline caution and mixed research findings. Negative consensus for hyperbaric oxygen, acupuncture, sound conditioning devices, and ear moxibustion underscores limited evidence and supports avoidance in routine care. Conversely, digital platforms (internet/smartphone) and targeted physical therapy for somatic tinnitus were viewed favorably, reflecting evolving modalities with practicality and patient accessibility. These findings address the need for clinical direction and can inform guideline updates and individualized management strategies.

A two-round modified Delphi process among Korean otology experts yielded 19 consensus statements on tinnitus treatment, with high agreement for patient selection criteria, CBT/TRT, and auditory rehabilitation, and negative consensus for several alternative modalities. The statements offer practical, consensus-driven guidance to clinicians managing tinnitus. Future research should prioritize high-quality randomized controlled trials to evaluate neuromodulation techniques, clarify the roles of BAHA/CROS and other auditory devices, refine psychotherapeutic and sound therapy protocols, and assess the effectiveness of digital interventions across diverse populations.

Consensus was derived from a single national expert group, potentially limiting generalizability to other healthcare systems and practice environments. Overall agreement was weak-to-moderate (Kendall’s W = 0.476), with very low convergence in some categories (e.g., neuromodulation W = 0.094). The Delphi method reflects expert opinion rather than definitive evidence, and several areas (e.g., BAHA/CROS, neuromodulation, complementary therapies) suffer from limited or low-quality trials. Survey timing (2022) may precede emerging evidence. Statements may reflect feasibility and confidence in local clinical practice rather than globally established efficacy.