Study protocol of the SuDHO study: hyperbaric oxygen treatment added to standard care for acute idiopathic sudden sensorineural hearing loss – a multicentre randomised controlled trial

Abstract

Introduction

Idiopathic sudden sensorineural hearing loss (ISSNHL) can lead to permanent sequelae with functional impairment of hearing. Hyperbaric oxygen treatment (HBOT) may improve outcomes but conclusive evidence for its efficacy is currently lacking.

Methods and analysis

This is a multicentre, 1:1 randomised, controlled, assessor-blind trial comparing standard care alone versus HBOT plus standard care. Adult patients with ISSNHL, at least 50 dB hearing level (HL) in pure tone average (PTA4) on the affected ear and who were eligible for commencement of HBOT within 7 days from onset of hearing loss are included. Those with other causes of hearing loss and contraindications for HBOT are excluded. The primary outcome is absolute PTA4 after 14 weeks, while secondary outcomes include other measures of audiometric hearing improvement including word recognition score and patient-reported outcomes related to hearing, tinnitus, balance and quality of life. The power analyses indicated a group size of 151 participants to detect a difference in mean PTA4 of at least 10 dB HL at follow-up with 90% power based on an assumed SD of 34 dB HL with a 0.05 significance level.

Ethics and dissemination

The trial has been approved by the Regional committees for medical and health research ethics (268140), by the Norwegian Medical Products Agency and is registered in the EU Clinical Trials Information System (CTIS, 2024-510798-84-00). User representatives have been involved in the development of the study. The results will be reported according to the Consolidated Standards of Reporting Trials (CONSORT) 2025 guidelines in an international peer-reviewed medical journal as soon as possible after the last visit of the last patient.

Trial registration number

2021-002284-22 (primary registration 4 May 2021; initially registered in EU Clinical Trials Register (EudraCT)). From 2024 replaced by EU Clinical Trials Information System (CTIS): 2024-510798-84-00 (primary registration 2 October 2024).

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• Sudden Deafness treated with Hyperbaric Oxygen (SuDHO) is a large, randomised controlled multicentre trial on the effect of hyperbaric oxygen treatment (HBOT) in idiopathic sudden sensorineural hearing loss (ISSNHL) conducted at 12 ear–nose–throat (ENT) and two HBOT centres.

• The study compares established standard care versus HBOT given in addition to standard care, using pure tone average (PTA4) as primary outcome measure, while secondary outcomes include word recognition and patient-reported outcomes.

• Standardised HBOT protocol with 15 sessions of 90 min breathing 100% oxygen at 2.4 atmospheres once daily, while oral corticosteroids are recommended standard care.

• Patients aged 18–80 years with moderate or more severe hearing loss (PTA4 ≥50 dB HL) are included and HBOT starts within 7 days from onset of hearing loss.

• The study does not include a placebo arm with sham HBOT, but outcome assessors are blinded.

In this publication we have included SPIRIT checklist item numbers referring to the content of the paragraphs, numbers are given in brackets.

Introduction

Background and rationale (9a)

Hearing impairment is one of the most frequent global causes of disability. In most individuals, the hearing loss is sensorineural with gradual development related to ageing, genetics, noise or drug exposures. However, a significant proportion of cases experience a sudden loss of hearing, typically affecting only one ear, with a severity ranging from mild impairment to total deafness. When identifiable causes have been ruled out, the condition is named idiopathic sudden sensorineural hearing loss (ISSNHL), often also called sudden deafness. Individuals of all ages are affected, including many young adults. There is an increasing incidence with advancing age and only a rare occurrence in children. ISSNHL is the most frequent ‘cause’ of sudden hearing loss, and modern studies suggest an overall incidence of 18–27 per 100 000 per year,^{1 2} although both higher and lower numbers have been reported.^{3 4} According to the commonly used definition, the hearing loss develops within 3 days, and an audiogram shows a sensorineural hearing loss of 30 dB hearing level (HL) or more affecting at least three consecutive frequencies.⁵

Without treatment, about 30%–40% of patients experience spontaneous improvement or normalisation of hearing, with a better prognosis in patients with less severe hearing loss.^6–8 However, 40%–60% of cases are left with persistent, moderate to severe hearing problems including reduced overall hearing, speech discrimination and directional hearing, often accompanied by tinnitus and less frequently by vertigo or balance problems.^8–13 Depending on the extension of symptoms, non-recovery from ISSNHL can result in reduced quality of life, hampered social function and impaired work participation.^14–17

The systemic administration of steroids in the acute phase is widely regarded as standard treatment for ISSNHL, although robust evidence of improved outcome is lacking.¹⁸ Steroids counteract inflammation, which is suspected to play a role in ISSNHL pathogenesis, including viral or immunological processes in the cochlea and/or adjacent structures.^{5 19} Another key issue is hypoxia, possibly due to vascular occlusive or inflammatory processes. This may cause damage to the highly oxygen-dependent and sparsely vascularised cochlea.^{5 20}

Hyperbaric oxygen treatment (HBOT) improves tissue oxygenation including the cochlea and reduces oedema and is, therefore, assumed to have a beneficial effect in ISSNHL.²¹ Patients are placed in hyperbaric chambers with increased ambient pressure of 2.0–2.5 absolute atmospheres (ATA) and breathe pure oxygen for 80–100 min daily for 10 days or longer.

A considerable body of evidence, including systematic reviews, suggests that HBOT may improve hearing outcome in ISSNHL, indicating a higher mean hearing gain with HBOT compared with without and a higher proportion of individuals with significant improvement.^22–31 However, all available randomised controlled trials (RCTs) are hampered by small size or other, sometimes severe methodological shortcomings. Only non-randomised and mostly retrospective studies include large numbers of patients, and the results are heterogeneous.^32–34 For treatment studies of ISSNHL, sample sizes are critical due to the relatively high rate of spontaneous improvement and large variability of the recommended main outcome measure, pure tone hearing.¹⁸ The true benefit from HBOT in ISSNHL is, therefore, still uncertain and higher quality evidence from methodologically sound RCTs has been repeatedly called for.^{19 35 36}

The aim of this study is to provide reliable evidence on the efficacy of HBOT as an early treatment for ISSNHL when given in addition to standard care.

Objectives (10)

The primary objective of the study is to assess if the addition of HBOT to standard care leads to better HLs 3 months after treatment compared with standard care only.

Secondary objectives are:

• To compare the efficacy of both treatment regimens 3 months after treatment based on different hearing measures.

• To compare quality of life in treatment groups.

• To identify subgroups associated with better hearing outcome 3 months after baseline.

• To assess associations between improvement in pure tone hearing and word recognition with communication problems and quality of life.

• To assess the safety of HBOT in patients with ISSNHL.

Exploratory objectives are:

• To identify prognostic and predictive factors on hearing.

• To assess the trajectory of hearing function depending on treatment.

Methods

Trial design (3c, 12)

This is a multicentre, randomised controlled, phase II and III, assessor-blinded, two-arm, parallel group, superiority study. Patients are randomised 1:1 to Standard care or Standard care+HBOT with a total study duration of 14 weeks from randomisation.

Trial summary

Administrative information and trial summary are presented in table 1.

Table 1. Trial summary.

Title	Hyperbaric oxygen treatment added to standard care for acute idiopathic sudden sensorineural hearing loss—a multicentre randomised controlled trial
Acronym	SuDHO—Sudden Deafness treated with Hyperbaric Oxygen
Health condition studied	Idiopathic sudden sensorineural hearing loss (ISSNHL)
Countries of recruitment	Norway
Study design	Study type: interventional trial Allocation: randomised
Trial registry identifying numbers and dates of registration	Initially registered in EU Clinical Trials Register (EudraCT): 2021-002284-22 (primary registration 4 May 2021) From 2024 replaced by EU Clinical Trials Information System (CTIS): 2024-510798-84-00 (primary registration 2 October 2024) Regional committees for medical and health research ethics (REK): 268140 (primary registration 27 April 2021) Open Science Framework (OSF), registered 25 November 2025, registration DOI 10.17605/OSF.IO/R59MU
Source(s) of monetary or material support	Helse Vest (F-13100) Stiftelsen Dam (SDAM-FOR462254) The Bergesen Foundation (project support grant number 17661) The external sponsors had no impact on design, conduct or plans for analyses. No sponsor will have influence on the content of publications. The Dam Foundation demands open access publication of the results.
Protocol version	By August 2026: V.2.6/10.08.2026
Primary sponsor	Helse Bergen HF, Haukeland University Hospital. Contact person: Alf Henrik Andreassen, Department of Occupational Medicine
Secondary sponsor(s)	None
Coordinating centre	Haukeland University Hospital, Bergen
Study principal investigator	Bernd Mueller, MD PhD Haukeland University Hospital Jonas Lies Vei 65, 5021 Bergen +47 55973875 bernd.mueller@helse-bergen.no
Corresponding author	Kristina Sture Larsen, MD Haukeland University Hospital Jonas Lies Vei 65, 5021 Bergen +47 41701837 kristina.sture.larsen@helse-bergen.no
Leading study team	Frederik Kragerud Goplen, MD PhD Jan Erik Berge, MD PhD Jeanette Hess-Erga, MD Kathrin Skorpa Nilsen, MD PhD Kristina Sture Larsen, MD
Date of first enrolment	31 January 2022
Recruitment status	Recruiting
Ethics review	Approved, date: 25 June 2021 REK sør-øst C ID268140
IPD sharing statement	Plan to share IPD: a limited set of anonymised data for analyses of the main outcome measure can be shared on reasonable request.

IPD, Individual Patient Data.

Patient and public involvement (11)

Two patient representatives with experience from chronic hearing loss or ISSNHL and HBOT have been involved in the development of the study and contributed to ethical aspects, design, outcome measures and dissemination.

Study setting (13)

The study is carried out at ear–nose–throat (ENT) departments at university hospitals and larger community hospitals. Presently, they cover the three largest healthcare regions in Norway, which are served by two hyperbaric treatment centres at Haukeland University Hospital in Bergen (for Western Norway and Central Norway) and Oslo University Hospital, Ullevål (for Southern and Eastern Norway). Patients are recruited and included at 12 ENT study centres, and those randomised to HBOT receive the treatment at the responsible centre depending on their place of residence.

Eligibility criteria (14a)

Inclusion criteria

1. Age 18 to 80 years.

2. Onset of acute hearing loss within the past 7 days.

3. HBOT can be started within 7 days from symptom onset.

4. Pure tone audiometry shows sensorineural hearing loss with at least 30 dB HL loss in three consecutive frequencies.

5. Idiopathic sensorineural hearing loss is the most probable diagnosis according to clinical judgement.

6. The hearing threshold is 50 dB HL or more in PTA4 (based on 0,5 kHz, 1 kHz, 2 kHz, 4 kHz).

7. Informed consent given.

Exclusion criteria

1. Other disease, injury or condition is most likely the cause of the current hearing loss.

2. Cognitive or language barriers that hamper adherence to the study protocol.

3. Major alcohol abuse.

4. Any contraindication to treatment with hyperbaric oxygen:

   • Chronic Obstructive Pulmonary Disease (COPD), asthma or other lung disease with severely reduced lung function corresponding to Global Initiative for Chronic Obstructive Lung Disease (GOLD) III (spirometry shows Forced Expiratory Volume first second (FEV1)<50% of expected).

   • Asthma not well regulated on oral basic medication and/or with attacks within the last year.

   • Current pneumonia or less than 4 weeks from complete recovery.

   • X-ray showing emphysema-bulla or non-static atelectasis.

   • Severe cardiac failure, unstable angina, myocardial infarction within the past 6 months.

   • Current treatment with Disulfiram (Antabus).

   • Prior Bleomycin treatment (any time).

   • Prior stapes surgery.

   • Pregnancy.

Sub-study PTA4 <50 dB HL

In addition to the main study with a requirement of PTA4 ≥50 dB HL, some centres may include patients with PTA4 <50 dB HL in a substudy. With the exception of the PTA4 level, the same exclusion and inclusion criteria apply, but patients included in the substudy are not included in the sample size calculated for the main study, are randomised separately and not included in the analyses of the main study cohort. If relevant inclusion numbers are reached, analyses including substudy participants will be reported independent and clearly separated from the main study. Due to insufficient data available for a calculation of sample size, we assumed a smaller SD of 24 in PTA4 and estimated that recruitment of 58 patients per group will give a power of 80%. However, current recruitment numbers are far below this. This substudy is described in the protocol of the main study and thus registered under the same identification number in EU Clinical Trials Information System (EU CTIS). Any further details on the substudy will not be addressed in this publication.

Choice of comparators (9b)

HBOT is given in addition to standard care in the intervention group, while the control group receives standard care only. Notably, the type of standard care can be chosen freely by the responsible ENT physicians, but the study recommends a specific type and dosage of systemic steroids to increase homogeneity within and between the study groups and to avoid underdosing of an established standard treatment.¹⁹

Interventions (15)

Intervention description (15a)

HBOT: a standard protocol is used for HBOT with pure oxygen breathed for 90 min one time per day at a pressure of 2, 4 ATA for 15 consecutive days.³⁷

Standard care: the study recommends oral prednisolone 1 mg/kg or maximum 60 mg one time per day for 10 days plus tapering of −10 mg per day for 5 days. Variations in treatment are permitted but must be documented. Concomitant standard care will be summarised descriptively and compared between treatment groups based on the type of treatment received, including recommended prednisolone therapy, alternative systemic steroid regimens, intratympanic steroid treatment and other therapeutic interventions.

Modifications (15b)

There are no hearing-related stopping criteria for the study treatment. HBOT is stopped in case of pregnancy, occurrence of other contraindications of HBOT or adverse events (AEs) not compatible with continuation of HBOT.

A modification of the study dose is not intended with respect to the applied pressure, the total number or the duration of each HBOT session given. The following deviations for HBOT will be accepted for inclusion in the per protocol analysis: maximum 2 days from randomisation to start of HBOT if within 7 days after start of symptoms; 15±1 sessions within 18 calendar days from start of HBOT; first three treatments on consecutive days; maximum one treatment-free day within the first seven treatments; duration of each treatment 90±5 min; maximum two sessions with less than 85 but at least 60 min oxygen; treatment pressure 2,4±0,1 ATA.

Concomitant care (15d)

HBOT administered by other healthcare providers is prohibited for participants during the trial. No other treatment is prohibited or specifically permitted during the trial.

Assessments (25a)

Clinical examination

ENT physicians perform a standard clinical examination using standard clinical facilities including otomicroscopy to exclude known, detectable causes of hearing loss.

Hearing assessments

Audiometry uses regularly calibrated equipment in audiometric booths to assess hearing thresholds for 125, 250 and 500 Hz as well as 1, 2, 3, 4, 6 and 8 kHz. PTA4 is calculated as mean hearing threshold of the four frequencies 500 Hz, 1 kHz, 2 kHz and 4 kHz. When hearing thresholds are over the equipment maximum, the highest possible values for the equipment, usually 120 dB HL, are registered as ‘dummy code’.^{38 39} Established standard procedures are used for masking of the contralateral (better) ears. Bone conduction is assessed at baseline for the affected ear to verify the sensorineural nature of the hearing loss.

Word recognition is tested at each ear with 50 monosyllabic words, and the word recognition score (WRS) is the percentage of correctly identified words. The loudness used is either the optimal sound pressure determined in a separate, preceding speech test or set to PTA4+25 dB HL. The maximum for speech testing equipment is usually about 85 dB HL, but the actual applied loudness is recorded in the electronic case report form (eCRF).

Vestibular tests

Patients recruited at hospitals with the necessary facilities undergo vestibular testing as long as practically feasible. Testing includes cervical vestibular evoked myogenic potential (cVEMP), ocular vestibular evoked myogenic potential (oVEMP) and video head impulse test (vHIT). These three tests provide information about the function of different parts of the vestibular apparatus: cVEMP reflects the saccule and the inferior part of the vestibular nerve, oVEMP the utricle and the superior part of the vestibular nerve, and vHIT evaluates the semicircular canals, assessing both the superior and the inferior parts of the vestibular nerve.⁴⁰

Patient-reported outcome measures

All participants are asked to answer a collection of standardised and validated questionnaires at baseline and week 14. The questionnaires include the Abbreviated Profile of Hearing Aid Benefit (APHAB), the Tinnitus Handicap Inventory (THI), the Dizziness Handicap Inventory (DHI) and the RAND-12 to assess health-related quality of life. The Norwegian translations of the questionnaires are validated.^41–44

In addition, a Visual Analogue Scale (VAS) score is used to assess the subjective degree of hearing loss, tinnitus, dizziness, pain and hyperacusis at baseline, day 15 and week 14. Given the open label design of the study with a potential placebo effect and risk of bias, patient-reported outcome measures (PROMs) will not be regarded as independent measures of effect and will not alter the conclusions drawn from analyses of the primary outcome.

Outcomes (16)

Primary endpoint

• Pure tone average (PTA4 (dB HL)), absolute value at week 14.

Secondary endpoints

• Change in PTA4 (dB HL) from baseline to week 14.

• Proportion of patients with improvement of at least 10 dB HL PTA4 from baseline to week 14.

• WRS in % at week 14 and change from baseline.

• Categories of hearing recovery according to Siegel’s criteria at week 14.

• Categories of hearing severity according to the American Academy of Otolaryngology, Head- and Neck-Surgery (AAO-HNS) hearing classification at baseline and week 14.

• APHAB score at week 14 and changes from baseline.

• RAND-12 at week 14 and changes from baseline.

• Number of AEs, serious AEs (SAEs) and known side effects of HBOT.

• Subgroup analyses for gender, age, severity of initial hearing loss in PTA4, hearing curve categories, medical history, time from onset to treatment, bilateral hearing function before onset, vertigo and tinnitus at baseline, smoking.

Participant timeline

The timeline of enrolment, interventions, assessments and visits for participants is presented in table 2.

Table 2. Participant timeline (18).

Procedure	Intervention period (visit/day or week/visit window)
	1 Day 1 Screening+BL	2 Day 10 −1/+4 days*	3 Day 15 −1/+5 days* (PROM −1/+20 days)	4 Week 8 ±5 days	5 Week 14 ±5 days (PROM −5/+20 days)
ENT examination and diagnosis	X		X		X
Audiometry pure tone	X	X	X	X	X
Audiometry word recognition	X		X		X
Inclusion and exclusion criteria	X
Randomisation	X
HBOT-related examinations, including X-ray thorax if necessary	X		X
MRI cerebrum (send request)	X
VEMP and vHIT (selected centres)	X				X
Study intervention (HBOT)	<==========================>
Registration of adverse events	<============================>
PROMs
Quality of life (RAND-12)	X				X
Communication (APHAB)	X				X
Tinnitus (THI)	X				X
Dizziness (DHI)	X				X
VAS score (hearing, tinnitus, dizziness, pain, hyperacusis)	X		X		X

^*Visit 2 when 9–11 HBOT, visit 3 when 14–15 HBOT performed.

APHAB, Abbreviated Profile of Hearing Aid Benefit; DHI, Dizziness Handicap Inventory; ENT, ear–nose–throat; HBOT, hyperbaric oxygen treatment; PROMs, patient-reported outcome measures; THI, Tinnitus Handicap Inventory; VAS, Visual Analogue Scale; VEMP, vestibular evoked myogenic potential; vHIT, video head impulse test.

Sample size calculation (19)

We assumed a difference of 10 dB HL at follow-up, a correlation between baseline and follow-up of 0.5 (control) and 0.7 (HBOT) and different values for the SD based on own historic data and previous studies.^{25 45 46} For an SD of 34, a significance level of 0.05 and a power of 90%, we will need 151 participants per group, totally n=302. Considering an experience-based dropout rate of 10%, this will lead to a required sample size of about 340. Even if the SD should turn out to be much larger than expected and close to 40, a power of 80% will be reached. The calculations were based on the primary outcome and performed as computed estimation of the analysis of covariance (ANCOVA) for simulated data in several scenarios with 10 000 repetitions (figure 1).

Figure 1. Sample size calculation with different SD. ANCOVA, analysis of covariance.

Recruitment (20)

Patients will be recruited at 12 hospital-based ENT departments in Norway. Patients may be referred to recruiting centres from general practitioners, emergency departments, non-recruiting ENT departments and private ENT centres or any other doctor, within 7 days from onset of sudden hearing loss. Patients who comply with the study criteria will be invited to join the study and receive detailed information. Those who fulfil all inclusion criteria, have no exclusion criteria and give written, informed consent, are included. The participant consent form is available in online supplemental material (Norwegian text only). Inclusion rates were low for the first 2 years but have increased to about 60 per year from 2024. Regular and repetitive information of the public and relevant healthcare actors, including newspaper articles, TV broadcast and presentations at professional conferences, contributes to stable recruitment. By mid-September 2026, 231 patients were randomised, of which 27 have screening failure identified postrandomisation, withdrawn consent or were lost to follow-up, leaving 204 patients available for analyses. With continuing 60±10 analysable patients per year, we anticipate inclusion of the last patients by spring or summer 2028. Despite the delayed schedule, we do not plan extension of the study to several national or international centres to limit the risk of heterogeneity in patient management.

Assignment of intervention (21 a,b, 22, 23)

After completion of baseline assessments, the study intervention will be assigned within the eCRF (Viedoc (R)) by an automated two-armed 1:1 block-randomisation process, stratified by severity of initial hearing loss (PTA4 ≤90 vs >90 dB HL) and smoking status (active smoking vs not). Assignment is shown in the eCRF to all study personnel except for audiometric technicians. The randomisation sequence, including block sizes, was generated by a statistician, incorporated in the eCRF and is not visible nor accessible for investigators and other study personnel.

Blinding (masking) (24 a,b,c)

Audiometric technicians perform audiometry and assess the primary and the most important secondary outcomes (PTA4 and WRS). They will be masked for the applied treatment during assessments after baseline, but most importantly at the final study visit.

To ensure blinding, audiometric technicians shall not access the patient’s electronic patient file after assignment, and patients and other staff are instructed not to reveal the treatment group of any study participant to the audiometric technician. The eCRF is modified to only show strictly necessary information to the audiometric technicians and no information which directly or indirectly discloses allocation.

Unblinding is not necessary and, therefore, not permissible under any circumstance, but patients included in the study will not be excluded in case of unblinding. For all study visits after baseline, the audiometric technician must state in the eCRF whether information that disclosed a participants’ treatment allocation was revealed by any means. The statistician performing the primary analysis will be blinded to allocation labels.

Data management (26)

All trial data are registered directly into the eCRF (Viedoc (R)) by study physicians or audiometric technicians. All study participants will be given a unique identification number. The database will not contain any personal identification.

Statistical analysis (27 a,b,c,d)

The following populations are defined for analysis:

Enrolled: all participants who have signed the ICF and have been included in the study database.

Intention to treat (ITT): all participants randomly assigned to a treatment group.

Modified ITT (mITT): all participants randomly assigned to a treatment group who have not been rediagnosed with a non-idiopathic cause of the acute hearing loss.

Per protocol (PP): all participants randomly assigned to a treatment group who have followed the study protocol without major deviations.

Safety: all participants randomly assigned to a treatment group.

Subgroup vestibular function: all participants who have performed the specific tests for vestibular function (vHIT, VEMP).

For the primary outcome, a superiority test based on the following hypotheses will be performed:

• Null hypothesis:

There is no difference between the control group (standard care) and intervention group (standard care+HBOT) in mean pure tone audiometric HL (PTA4) measured 3 months after treatment

• Alternative hypothesis:

There is a difference between the control group (standard care) and intervention group (standard care+HBOT) in mean pure tone audiometric HL (PTA4) measured 3 months after treatment.

Efficacy (primary and secondary endpoints) and safety analyses will include data from the ITT population as main outcome measure. Analyses of the mITT and PP populations will be presented as additional information. Exploratory endpoints will be analysed for the PP population. The relatively large number of secondary endpoints to compare hearing in the two treatment groups shall ensure comparability with earlier studies, as a large variety of outcome measures have been used in available publications.

Demographics and baseline characteristics, efficacy and safety variables will be summarised using descriptive statistics. All efficacy analyses will be presented with the results from the hypothesis testing (by p value) in addition to estimates and 95% confidence limits of the treatment effect. We will check the structure of missing data and then use multiple imputations where it is possible and meaningful.

The primary outcome PTA4 value at week 14 will be analysed by ANCOVA with baseline PTA4 as covariate and with severity group (PTA4 ≤90 dB HL vs over) and smoking status as random effects:

PTA4(follow-up)~treatment + PTA4(baseline)+random(stratification)

The primary effect size will be the adjusted between-group mean difference in PTA4 and presented with 95% CI and p value. This analysis will be supplemented by a sensitivity analysis adjusted for study site and cotreatment, ie, adherence to the recommended prednisolone regimen and use of salvage intratympanic steroids. To account for ceiling and floor effects, ANCOVA of the audiometric outcomes PTA4 and WRS will be checked by secondary sensitivity analyses using a censored model (eg, Tobit). The model will use 115 dB HL as upper censoring limit for PTA4 and 85 dB for WRS. If assumptions are violated, a robust model will be used. As the large number of secondary endpoints and subgroups will be interpreted as hypothesis-generating, no formal multiplicity adjustment will be applied to their analyses.

A separate statistical analysis plan will prespecify all statistical analyses and will be finalised prior to database lock.

Data safety monitoring (28 a,b)

A data safety monitoring board (DSMB) has been established and has conducted a safety analysis after inclusion of approximately 150 participants. The board was composed of two physicians and one nurse covering clinical ENT, neurology and research experience. All were independent from the sponsor-institution. The DSMB-charter states that stopping of the study should be recommended when more than 5% of patients experience an adverse reaction of grade 4 or higher (life-threatening or urgent intervention required) or in case of other safety concerns. The board did not find any of this and recommended continuation of the study.

An interim analysis of efficacy parameters is not planned.

Harms (17)

All AEs are registered from day 1–22, while later medical events only are registered as AE if they are likely related to undergone HBOT. Serious AEs are immediately reported to the sponsor. Special attention is given to the registration of barotrauma and tinnitus. Severe barotrauma to the ears may worsen hearing and audiologic measures in the short term, while potential long-term consequences for hearing outcome in ISSNHL are unclear. Tinnitus may occur as additional symptom of ISSNHL but can in rare cases worsen or be induced by HBOT. Furthermore, vision changes, anxiety, fatigue and oxygen seizures are registered on designated schemes.

Auditing (29)

Monitoring is provided by the Norwegian Clinical Research Infrastructures Network (NorCRIN), a national network for research support between the six university hospitals in Norway. The lead monitor is employed at the Section for Research and Innovation at Haukeland University Hospital. Monitoring visits at each study site are performed at least once a year.

Ethics and dissemination

Research ethics approval (30)

The trial has been approved in June 2021 by the regional committees for medical and health research ethics (REK), identification number 268140, and by the Norwegian Medical Products Agency with EudraCT number 2021–0 02 284-22. Transition to the EU Clinical Trial Information System (CTIS) has been approved in May 2024, number 2024-510798-84-00. Patients are protected by the legislations regulating treatment of patients in Norwegian hospitals. The study is performed based on the Declaration of Helsinki⁴⁷ and principles of Good Clinical Practice (GCP). All study staff have received training in study procedures and GCP.

Protocol amendments (31)

Important protocol modifications have been and will continue to be submitted to medical and ethical authorities for approval, as appropriate. All analysis specifications introduced up to version 2.6 were made while investigators and analysts remained blind to accumulating outcome data and before database lock. All protocol amendments are listed in table 3, more details on amendments are listed in the published protocol on the website of the EU CTIS.

Table 3. Protocol amendments.

Document	Date
Amendment 7, Protocol V.2.6	10 August 2026
Amendment 6, Protocol V.2.5	4 March 2025
Amendment 5, Protocol V.2.4	9 September 2024
Amendment 4, Protocol V.2.3	15 February 2024
Amendment 3, Protocol V.2.2	16 November 2022
Amendment 2, Protocol V.2.1	1 January 2022
Amendment 1, Protocol V.2.0	2 December 2021
Original Protocol	27 April 2021

Consent procedure (32 a)

Potential participants will be informed about the study by an investigator or his/her representative, receive the informed consent form (ICF) and can ask questions to a study physician. When ready, the participants sign the ICF and receive a copy which includes contact information to the relevant study centres. The ICF and consent procedure are in line with local regulations and ICH guidelines.

Confidentiality (33)

Personal information about potential and enrolled participants is registered in the electronic patient journal system and an eCRF satisfying all legal and technical requirements for data-safety.

Ancillary and post-trial care (34)

Participants will be followed by hearing healthcare professionals according to the severity of remaining hearing loss and the local routines, which include further investigation, education or hearing aids as needed.

Dissemination policy (8)

Trial methods and results will be reported according to the Consolidated Standards of Reporting Trials (CONSORT) 2025 guidelines. The results from primary, key secondary and several secondary outcomes including data on AEs will be submitted for publication in relevant peer-reviewed journals within 1 year after the last visit of the last patient. Results from other parameters will be published consecutively as analyses and writing processes are being conducted. Results will also be presented at national and international medical conferences within the fields of Otorhinolaryngology and Hyperbaric medicine. There are no restrictions preventing the disclosure and publication of the results from this research project.

Discussion

ISSNHL is a condition with few treatment options supported by robust evidence, and research on the efficacy of HBOT for ISSNHL is likewise limited due to methodological shortcomings of available randomised and non-randomised trials.^{19 36} Significant hearing sequelae are present after ISSNHL in a large proportion of patients, justifying efforts to improve treatment and outcome. Despite the uncertain benefit, it seems that HBOT has been implemented in the treatment portfolio for sudden hearing loss in several countries and institutions. However, the highly specialised staff and technical equipment, the limited availability and costs, including travel and accommodation, underline the need for reliable evidence on the achievable benefit from HBOT in treatment of ISSNHL.

To our knowledge, this is the first RCT in the field to combine a robust methodological design with a sample size powered according to the recommended minimal clinically significant difference of 10 dB HL.¹⁸ Although HBOT is an established treatment modality for many conditions, this is a phase II–III trial as it combines the investigation of a possible effect in ISSNHL (phase II) with the comparison to standard care for this condition (phase III). The study methodology is chosen to avoid overselection and to include a population and clinical practice close to ‘real life’ in as many aspects as possible. Therefore, the eligibility criteria exclude primarily those with non-idiopathic causes of ISSNHL and those with contraindications to HBOT. The diagnosis is based on clinical evaluation by ENT physicians and patients will be included in the study prior to MRI in order not to delay treatment. Some participants may thus be excluded subsequently due to identification of vestibular schwannomas or other intracranial lesions, but this is expected to be rare and is also in line with current clinical practice. The lower age limit of 18 years has been set to include adult patients only, while the upper limit of 80 years aims to limit heterogeneity related to the increased risk of comorbidities associated with hearing loss.

Only patients with PTA4 ≥50 dB HL are included in the main study, as the risk of permanent hearing loss increases with greater initial severity. This patient group is, therefore, most likely to benefit the most from improved treatment. Moreover, some reviews indicate that HBOT has predominantly been shown to be effective in more severe hearing loss.^{33 48}

In most of the available research, recommendations and current practice, HBOT is applied as a supplement to systemic steroids, while intratympanic steroids are mainly used as salvage treatment.^{19 49} The treatment groups comparing HBOT plus standard care versus standard care only reflect the clinically most relevant scenario without challenging the ethics related to a withdrawal of an established standard of care.¹⁹ With the recommendation to use oral Prednisolone for 10 days plus tapering the study intends to homogenise the treatment given in both groups and to avoid underdosing in the control group, thus strengthening the results if they show a positive effect of HBOT. A third treatment arm receiving neither steroids nor HBOT was discussed in the development phase of the study. Although of clinical and scientific relevance, it was discarded as being out of the scope of this study, patients expected to be more sceptical to participate and larger numbers needed to include, what would reduce the feasibility of the study. Similarly, the randomised inclusion of intratympanic steroids as salvage treatment was discussed but discarded favouring a clear methodology and statistical strength.

The Undersea & Hyperbaric Medical Society (UHMS) recommends 10–20 sessions HBOT for ISSNHL, but most studies report 10–15 sessions, a minor number has applied 20 sessions or more.⁵⁰ One review suggests efficacy for ≥1200 min total duration of HBOT, corresponding to 13–14 sessions of 90 min.⁴⁸ Although a flexible treatment length depending on the response has also been reported, a fixed treatment length of 15 sessions was deemed as potentially effective, valid for clinical practice, solid for statistics and feasible in this study setting.

AEs are registered day 1–22. Patients in the HBO group have daily contact with healthcare professionals during HBOT, while the control group has no other scheduled events than an optional day 10 audiometry. AE ascertainment will, therefore, be biased and favour the control group, a limitation which will be taken into consideration when results are interpreted.

A double-blind design is generally regarded as ideal to compare treatment groups and to minimise bias. However, while placebo treatment with HBOT is possible and has been performed, the practical and economic consequences made it not feasible for the participating hyperbaric centres. Instead, and to make this study possible, an open-label design with assessor-blinding of those who assess the primary and most important secondary outcomes by audiometry and word-recognition testing was chosen. Assessor-blinding for the primary outcome is also categorised as low risk of bias in the Cochrane system.⁵¹ The test-nature of audiometry makes it quasi-objective and audiometric technicians will usually identify manipulation-related inconsistencies. Furthermore, wilful falsification of audiometry by patients can be considered very unlikely in the setting of this study.

Results from studies on sudden hearing loss display a remarkable heterogeneity of audiometry-based outcome measures including means of variable frequencies, single frequency-analyses, different types of dichotomised or categorised scores, comparisons of improvement or final hearing and of follow-up time. While the primary outcome in this study reflects consensus-recommendation and statistical expertise, a relatively large number of secondary outcome measures have been implemented to facilitate comparison with results from earlier studies. A follow-up time of 14 weeks corresponds to the time needed for hearing to be reliably stable.

While the focus often is on the actual hearing aspects of sudden hearing loss, the condition often results in problems with communication and is accompanied by additional symptoms such as tinnitus, hyperacusis and balance problems. These aspects are assessed by testing of word recognition and PROMS using established standard questionnaires and VAS developed for this study.

Multiple studies have tried to identify prognostic factors for outcome in SSNHL with or without HBOT or other treatment measures. This is important for treatment decisions, selection and appropriate patient information. Higher age, more severe hearing loss and down sloping audiogram shape have been related to poorer prognosis,^{6 25 52–56} while the presence of cardiovascular disease may improve the prognosis.⁶ It is not clear whether these or other factors affect the efficacy of HBOT in ISSNHL. Prior studies indicate that the effect of HBOT may be better when commenced within 14 days from onset of hearing loss.^{55 57} Evidence to start earlier is extremely weak,²³ but to reduce the heterogeneity resulting from a relatively large timing window for this acute condition, the inclusion is limited to 7 days from start of hearing loss. However, as the response to HBOT may still be heterogeneous within the relatively short time span of 7 days, subgroup analyses on time from onset to treatment will be performed.

Analyses of relevant subgroups are, therefore, planned with respect to prognosis and the effect of HBOT, and the results may contribute to improve selection of patients with expectedly more benefit from HBOT, if any.

The cochlea and the vestibular organs are anatomically distinct but closely related structures within the inner ear, and pathological processes affecting the labyrinth may therefore result in overlapping auditory and vestibular symptoms, as seen in sudden sensorineural hearing loss and labyrinthitis. To improve knowledge on the coaffection of the vestibular system in sudden deafness, a proportion of patients will undergo comprehensive vestibular testing at one or two study centres with the necessary equipment and experience.

Trial status

Recruitment started in January 2022, by mid-September 2026, 231 patients have been randomised in the main study, 204 of these are deemed likely available for analysis (including 27 patients who have not yet reached their final study visit).