STRENGTHS AND LIMITATIONS OF THIS STUDY
-
This study will be the first network meta-analysis (NMA) to comprehensively compare the efficacy of multiple acupuncture-related modalities for sudden sensorineural hearing loss.
-
The study employs rigorous methodology, including a comprehensive search strategy across nine databases and five clinical trial registries, and adherence to Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines for NMA.
-
The use of the Cochrane risk of bias tool (V.2.0) and Grading of Recommendations, Assessment, Development and Evaluation methodology will ensure a thorough assessment of study quality and strength of evidence.
-
The inclusion of both Western and Chinese databases will provide a comprehensive overview of available evidence, though this may also introduce potential publication bias due to the predominance of Chinese studies in acupuncture research.
-
Heterogeneity in acupuncture protocols and outcome measures across included studies may limit the interpretability of pooled estimates, necessitating careful subgroup and sensitivity analyses.
Introduction
Sensorineural hearing loss is a type of hearing impairment that stems from issues within the inner ear, the auditory nerve or the central auditory pathways. A specific, acute form of this condition is sudden sensorineural hearing loss (SSNHL), which is clinically defined as an idiopathic, rapidly developing hearing reduction of at least 30 dB affecting three contiguous audiometric frequencies within 72 hours.1 Patients with SSNHL can also experience tinnitus (usual), dizziness or vertigo. With a profound negative impact on hearing and quality of life, SSNHL is an otologic emergency that warrants prompt evaluation and treatment. SSNHL has an annual incidence rate ranging from 5–27 cases per 100 000 individuals. In the US alone, it is estimated that approximately 66 000 new cases of ISSNHL occur each year.1 Hearing loss imposes substantial societal and economic burdens. According to data from the WHO, disabling hearing loss in adulthood is the most prevalent form of disability globally and ranks as the third leading cause of years lived with disability.2
While the aetiology remains uncertain in the majority of SHL cases, common causes include vascular obstruction, labyrinthine membrane breaks or viral infection.3 Conventional western treatments for SHL primarily involve corticosteroids, which are aimed at reducing inflammation and oedema in the inner ear.4 Other therapies such as vasodilators, antiviral agents, rheologic agents and hyperbaric oxygen therapy have also been explored, though their efficacy remains inconclusive.1 Despite the above-mentioned treatment options, a proportion of patients experience persistent hearing deficits and other annoying symptoms (eg, tinnitus, insomnia, anxiety), underscoring the need for additional therapeutic interventions.
Acupuncture, an important part of traditional Chinese medicine, has been increasingly recognised for its potential therapeutic benefits in various disease conditions, including ear-related conditions such as SSNHL5 6 and tinnitus.7 Dating back millennia, ancient Chinese medical texts, including the Huangdi Neijing (Yellow Emperor’s Inner Canon), documented the use of specific acupoints for treating hearing loss. Acupuncture has shown benefits over other therapies by offering a minimally invasive option with a low risk of systemic side effects, which can be particularly advantageous for patients who are unresponsive to or intolerant of conventional treatments like corticosteroids. The therapeutic benefits of acupuncture are thought to arise from its ability to modulate the body’s physiological functions; for SSNHL, this includes enhancing cochlear microcirculation, modulating auditory nerve activity and reducing inflammatory responses. Two systematic reviews (SRs) and meta-analyses8 9 assessed the therapeutic effect of acupuncture for SSNHL. Although findings obtained from these two studies8 9 generally support the efficacy of acupuncture for SSNHL, of note, various acupuncture modalities are commonly employed in clinical practice for the management of SSNHL, involving electroacupuncture (EA), manual acupuncture (MA), acupoint injection and warm needling, among others. Furthermore, these acupuncture-related modalities are typically adopted either independently or in conjunction. Despite the widespread application of acupuncture for SSNHL, there persists a paucity of studies comparing the efficacy of multiple acupuncture-related modalities. However, the relative efficacy differences among these acupuncture-related modalities remain unclear, and determining which specific acupuncture modality or combination of modalities can exert maximal therapeutic effects is an important clinical question for acupuncturists. Elucidating this question would aid them in prospectively selecting the optimal regimen to achieve the best outcomes.
While there have been traditional pairwise meta-analyses on this topic, our study is the first to employ a network meta-analysis (NMA) approach to comprehensively compare the relative efficacy of various acupuncture-related therapies, which offers distinct advantages over previous work.8 9 Unlike pairwise meta-analyses limited to direct comparisons, NMA allows comprehensive comparisons across multiple acupuncture modalities simultaneously, even in the absence of direct head-to-head trials.10 By integrating both direct and indirect evidence within a network of the included trials, NMA can provide a more comprehensive assessment of the comparative efficacy of acupuncture-related modalities for SSNHL. To our knowledge, no NMA studies have been published on this topic. By applying this distinct methodology, our study addresses a key knowledge gap and differentiates our work from previous studies.8 9 However, it is important to acknowledge that while NMA can generate estimates for treatments that have not been directly compared, such indirect comparisons should be interpreted with caution. They are valuable for improving the precision of combined estimates and generating hypotheses but may require future head-to-head trials for confirmation.
Taken together, we perform the first SR and NMA study that aims to critically appraise and synthesise the available evidence on the comparative efficacy of different acupuncture-related modalities for SSNHL. Subsequent findings will contribute to identifying the optimal treatment regimen for clinicians and SSNHL patients, providing evidence-based guidance for making clinical decisions.
Methods
Study registration
This SR and NMA study will adhere to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (uploaded in the online supplemental file 1).11 12 Concurrently, the study has been registered on PROSPERO with the registration number CRD42024525763. No deviations from the registered protocol are anticipated; however, any necessary modifications during the study’s execution will be documented in PROSPERO and detailed in the final publication.
Search strategy and data sources
We will conduct a comprehensive search for eligible trials on acupuncture for SSNHL across nine databases, from their inception to December 2024. The search will cover five English databases (PubMed, Embase, Web of Science, CINAHL, Cochrane Library) and four Chinese databases (Chinese Biomedical Literature Database, China National Knowledge Infrastructure, VIP Database for Chinese Technical Periodicals and Wanfang). The search strategy will combine medical subject headings (eg, ‘hearing loss, sudden’, ‘hearing loss, sensorineural’ in PubMed) and free-text terms/acronyms across three categories: medical conditions (eg, ‘sudden hearing loss’, ‘idiopathic sensorineural hearing loss’), interventions (eg, ‘acupuncture’, ‘electroacupuncture’) and study designs (eg, ‘randomised controlled trial’). Boolean operators ‘AND’ and ‘OR’ will be used for focused and comprehensive searches, respectively. The PubMed search strategy is presented in table 1 and will be adapted for other electronic databases. Reference lists of included studies will also undergo screening.
Search strategies in PubMed
Additionally, five major clinical trial registries, namely, the National Institute of Health Clinical Registry (Clinicaltrials.gov), International Clinical Trial Registration Platform, Australian New Zealand Clinical Trials Registry, ISRCTN registry and Chinese Clinical Trial Registry will be searched for ongoing unpublished trials. Reviewers of our group will try to contact primary investigators of identified ongoing trials via email to request the latest data.
Eligibility criteria
The eligibility criteria for study inclusion in this review will be formulated using the Population, Intervention, Comparison, Outcomes and Study (PICOS) framework (ie, Participants, Intervention, Comparison, Outcomes and Study design).13
Participants
Eligible participants will comprise adults diagnosed with SSNHL. Specifically, SSNHL is defined as an abrupt ≥30 dB hearing reduction across three consecutive frequencies within ≤3 days, attributable to cochlear, vestibular nerve or higher auditory processing disruptions.14 Participants with non-idiopathic SSNHL aetiologies will be excluded, as these are usually associated with definite conditions and require management accordingly.1
Experimental interventions
Eligible interventions in the experimental group can consist of a specified acupuncture modality alone, a combination of multiple acupuncture-related modalities, or acupuncture paired with other active therapies that match the control group. Drawing from similar meta-analysis studies,15 16 our review will limit eligible acupuncture-related therapies to invasive techniques involving needle insertion, such as EA, MA, fire acupuncture, warm needling, bloodletting puncture and acupoint injection. Non-invasive modalities, such as transcutaneous electrical acupoint stimulation, laser acupuncture and acupressure, will be excluded.
Control comparators
Permissible control group comparators include: (1) Active therapies for SSNHL, such as corticosteroids (either systemically or via intratympanic/postauricular injection), vasodilators, hyperbaric oxygen, antiviral agents and rheologic agents;1 17 (2) Acupuncture therapies differing from the experimental group and (3) Placebo or sham acupuncture interventions. Notably, waitlist-control will be excluded due to their known tendency to inflate treatment effects and their inherent differences in validity compared with placebo controls, particularly for subjective outcomes.
Outcomes
With references to meta-analysis studies of similar topics,17–19 potentially eligible randomised controlled trials (RCTs) should report at least one of the following outcomes.
Primary outcome
The primary outcome is the change in average pure-tone audiometric (PTA) thresholds across multiple frequencies (ie, 0.5, 1, 2, 4 and 8 kHz).
Secondary outcomes
-
The proportion of recovered patients
The proportion of recovered patients can be evaluated based on specified criteria, such as Siegel’s Criteria or other measures that categorise patients’ recovery as complete, marked, slight or none based on hearing improvement in PTA average,17 or a ≥10 dB improvement in PTA.19
-
Tinnitus symptom severity
Tinnitus symptom severity can be measured by rating scales (eg, visual analogue scales) and/or standardised questionnaires on functional effects20 (eg, Tinnitus Handicap Inventory, Tinnitus Reaction Questionnaire, Tinnitus Functional Index or psychoacoustic tests (eg, pitch match, loudness match) where available.
-
Psychological symptom severity
Psychological symptom (eg, anxiety, depression, insomnia) severity will be assessed using standardised questionnaires, for instance, the Self-Assessment Anxiety Scale21 for anxiety, the Hamilton Depression Scale22 for depression and the Pittsburgh Sleep Quality Index (PSQI23 for insomnia.
-
Quality of Life
Patient’s quality of life will be evaluated using validated generic instruments like the 36-Item Short Form Health Survey (SF-36) or disease-specific questionnaires where available from the included studies.
-
Adverse events (AEs) associated with interventions
The number of AEs linked to the interventions will be evaluated.
Study designs
Studies included in this review will be limited to RCTs. Non-RCTs, case series, observational studies and reviews will be excluded.
Study selection
Using EndNote V.X9 software (Clarivate, USA), two researchers will import search records from all databases, removing duplicates and screening titles/abstracts against inclusion criteria. Subsequently, full texts will undergo eligibility screening. Reference lists of included studies will be examined to identify add-on qualified trials. Disagreements will be resolved by a third investigator. A PRISMA flowchart (figure 1) will illustrate the selection process, accompanied by a list of excluded studies with specified justifications.
Flow diagram of study selection.
Data extraction and management
Two researchers will independently extract data from the included studies using Microsoft Excel, which mainly covers study information, participant details, intervention specifics, methodological aspects and outcome indicators. Missing information will be sought from authors. To ensure accuracy, the two researchers will cross-check their extracted data. Any discrepancies will be resolved through discussion or consultation with a third senior researcher.
Risk of bias assessment
Two raters will independently evaluate the methodological quality of each RCT using the Cochrane’s risk of bias (ROB, V.2.0) tool across five domains: randomisation, deviations from intended interventions, missing outcome data, outcome measurement and selective reporting.24 Each domain will be rated as low, unclear or high risk of bias for each individual RCT. An overall trial rating of unclear (some concerns in ≥1 domain), low (low bias risk across domains) or high (high bias risk in ≥1 domain or concerns across multiple domains) will be assigned by the reviewers. A third party will be invited to resolve any disagreements between the two raters.
Statistical analysis
Continuous data will be analysed using mean differences (MDs with 95% CIs, while dichotomous data will use ORs with 95% CIs. Given the anticipated clinical and methodological diversity among trials (eg, variations in acupuncture protocols, participant characteristics and co-interventions), a random-effects model will be the primary approach for all analyses. This model assumes that the true effect size varies between studies and provides a more conservative estimate of the treatment effect. The issue of multiple outcomes and potential type 1 error inflation will be acknowledged, and results will be interpreted cautiously, particularly for secondary outcomes.
Assessment of heterogeneity
Heterogeneity will be assessed using Cochran’s Q test and the I² statistic, where I² values of 25%, 50% and 75% will be considered indicative of low, moderate and high heterogeneity, respectively.25 However, it is important to note that I² represents the proportion of total variance attributable to between-study variance rather than a direct measure of heterogeneity itself. Potential heterogeneity sources will be explored through subgroup and sensitivity analyses.
Network meta-analysis and network geometry
The Bayesian NMA will be conducted using a hierarchical random-effects model framework. For continuous outcomes, a normal likelihood with identity link function will be employed, while binomial likelihood with logit link function will be used for dichotomous outcomes. Vague (non-informative) prior distributions will be specified for treatment effects (normal distribution with mean 0 and variance 10 000), while between-study heterogeneity will be modelled using a uniform distribution (0, 5) for the SD parameter. Three Markov chains will be run simultaneously with different initial values, each consisting of 50 000 iterations following a 20 000-iteration burn-in period. Convergence will be assessed using the Brooks-Gelman-Rubin diagnostic, with convergence deemed satisfactory when potential scale reduction factors approach 1.0 (specifically <1.05) for all monitored parameters. Trace plots and density plots will be visually inspected to confirm adequate mixing and convergence. Model fit will be evaluated using the deviance information criterion, with lower values indicating better fit.
Graphical representations depicting interconnections between diverse interventions will be generated using WinBUGS (V.1.4.3) and R (V.3.6.2) software. The node dimensions will correspond to the quantity of trials assessing each intervention, while line thicknesses will signify the magnitude of direct comparisons. Ranking probabilities and rankograms will be produced to compare and rank the efficacy of various acupuncture modalities for treating SSNHL. The Surface under the Cumulative Ranking (SUCRA) will be computed to rank interventions based on their therapeutic effect. A higher SUCRA score indicates superior efficacy, with a score of 1 denoting the optimal intervention.26
Inconsistency assessment
The node-splitting method will be used to evaluate the consistency of NMA results. The principle of a NMA is to obtain consistent results, where direct evidence aligns with indirect evidence. To achieve this, each node of the NMA will be analysed to test for statistical differences between direct and indirect evidence. If no significant inconsistency is detected (p>0.05, indicating failure to reject the null hypothesis of consistency), a consistency model will be applied; otherwise, an inconsistency model will be employed. Factors contributing to the discrepancies will be investigated, which can involve evaluating the degree of contribution from direct and indirect comparisons, as well as the extent of heterogeneity within direct comparisons. When substantial inconsistency persists, it is appropriate to rely solely on the more reliable direct comparison results to guide decision-making.
Subgroup and sensitivity analyses
If applicable, subgroup analyses will be performed in terms of the following characteristics of the included studies: (1) Different time points for primary outcome measurement, (2) Different patient age groups, (3) Different types of acupuncture modalities, (4) Different control interventions and (5) Individualised acupuncture protocols versus standardised acupuncture protocols. To ensure the robustness of the meta-analysis results, sensitivity analyses will be conducted using the leave-one-out method to assess the impact of each individual study on the overall results.
Safety evaluation
The safety assessment of acupuncture versus control interventions will involve pooling AE counts from the original studies through a meta-analytic approach. Fainting, intolerable procedural pain, post-needling discomfort and needling site bruising are primarily associated with acupuncture-related therapies’ AEs.
Grading the quality of evidence
Two independent reviewers will strictly adhere to the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) system to assess the quality of evidence, which has become the consensus for rating the strength of recommendations.27 Five aspects will be considered: methodological quality (risk of bias) of the included studies, extrapolation of the study population, interventions and outcome measures, consistency of results across studies, precision of the pooled study results and the extent to which the included studies meet the criteria (publication bias). The quality of evidence assessment will follow the GRADE guidelines strictly and involve the following four steps: (1) Presenting direct and indirect treatment estimates for each comparison in the evidence network, (2) Rating the quality of each direct and indirect effect estimate, (3) Presenting the NMA estimates for each comparison in the evidence network and (4) Rating the quality of each NMA effect estimate. Ultimately, the quality of study evidence will be rated as high, moderate, low or very low.
Publication bias assessment
If the primary outcome is reported in 10 or more included studies, funnel plots will be used to evaluate potential publication bias. Additionally, Egger’s test will be performed to evaluate the influence of significant publication bias on the meta-analysis results, with a p value <0.05 indicating the presence of significant bias.28
Ethics and Dissemination
As a review, all data used are publicly available and derived from published clinical studies, with no individual patient data involved, ethical review and approval is not needed for our study. The findings of this study will be disseminated through publication in a peer-reviewed academic journal.
Patient and public involvement
Patients or the public were NOT involved in the design, or conduct, or reporting, or dissemination plans of our research.
Discussion
This study elucidates the rationale and outlines methodological procedures for a SR and NMA, with the objective to comprehensively evaluate the relative efficacy of various acupuncture therapies for SSNHL. This study holds clinical significance as Western medicine like corticosteroids, though moderately effective, often fails to benefit a proportion of SSNHL patients and may result in some adverse effects. Thus, there is an urgent need to explore safe and effective adjunctive therapies.
With deepening mechanistic research into traditional acupuncture, accumulating evidence suggests acupuncture may exert therapeutic effects on SSNHL through multiple pathways. First, prior studies indicate acupuncture can significantly improve cochlear blood flow, thereby providing better oxygen and nutrient supply to the damaged cochlea.29 Second, acupuncture may alleviate hair cell injury by reducing oxidative stress in the inner ear. Furthermore, some researchers propose that acupuncture could promote neuroplasticity remodelling of the central auditory pathway, aiding in the restoration of impaired auditory conduction.30 Additionally, emerging neuroimaging research suggests acupuncture can modulate the neural functional connectivity across various brain networks associated with auditory processing, thereby exerting a synergistic regulatory effect on hearing and tinnitus improvement.31
Various acupuncture-related modalities are frequently employed for SSNHL in clinical practice, including MA, EA, warm needling and acupoint injection. These modalities can be used independently or in combination. However, the relative efficacy differences among various acupuncture-related therapies in treating SSNHL remain to be elucidated. For acupuncturists, determining which single acupuncture modality or combination of multiple modalities can achieve optimal therapeutic effects is an extremely important clinical question. Only by clarifying the optimal acupuncture treatment regimen can acupuncturists prospectively formulate more standardised treatment protocols and truly leverage the full advantages and potential of acupuncture as an adjuvant therapy for SSNHL. Through the well-established approach of NMA, this review will comprehensively assess and compare the therapeutic effect among multiple acupuncture-related modalities, providing evidence-based guidance for otolaryngologists and SSNHL patients in selecting the optimal acupuncture therapy.
Of note, our study presents several innovative aspects. First, to our knowledge, no NMA studies have been published that aimed to compare the relative efficacy of multiple acupuncture-related modalities for SSNHL. As the first NMA study of this important clinical topic, our study bridges this research gap. Second, in terms of methodological design, we establish a rigorous PICOS framework for our SR and NMA study. Simultaneously, by setting comprehensive study endpoints (eg, objectively measured hearing threshold, psychological symptom scores), the multidimensional therapeutic effects of acupuncture for treating SSNHL will be evaluated. Third, we emphasised quality control by employing the Cochrane ROB V.2.0 tool to rigorously assess the risk of bias in the enrolled RCTs and grading the overall quality of evidence based on the GRADE system. These methodologies ensure the credibility of our final results.
Clinical and policy implications
By identifying the optimal acupuncture modality or combination of modalities for SSNHL, this study will provide evidence-based guidance to help clinicians make informed treatment decisions. Specifically, the ranking of different acupuncture approaches will enable acupuncturists and otolaryngologists to select the most effective interventions for individual patients, potentially improving treatment outcomes and patient satisfaction. For healthcare policymakers, our findings may inform clinical practice guidelines, reimbursement decisions and resource allocation for integrative medicine services. Additionally, by identifying gaps in the current evidence base, our study will highlight priority areas for future research, potentially guiding funding decisions and research agendas. If certain acupuncture modalities demonstrate superior efficacy with acceptable safety profiles, this could support their broader integration into comprehensive SSNHL management protocols, particularly as adjunctive therapy for patients who have incomplete response to conventional treatments or seek complementary approaches.
Methodological considerations and potential sources of bias
Several methodological considerations warrant careful attention in interpreting our planned NMA. First, potential sources of bias may arise from the characteristics of the included studies. Publication bias remains a concern, particularly given that most acupuncture studies originate from China. We will attempt to mitigate this through comprehensive searching of both published and unpublished literature, including trial registries. Nonetheless, we acknowledge this limitation may persist. Performance bias may be present in studies with inadequate blinding, particularly relevant for acupuncture trials where blinding of acupuncturists is challenging. We will carefully assess these risks using the Cochrane Risk of Bias V.2.0 tool.
Second, heterogeneity among included studies represents a significant challenge. Clinical heterogeneity may stem from variations in: (1) Patient populations (eg, age, duration of symptoms), (2) Acupuncture protocols (eg, specific acupoints used, treatment frequency and duration), (3) Co-interventions and control treatments and (4) Outcome assessment methods and timing. Substantial heterogeneity may limit the interpretability and generalisability of pooled estimates. We will address this through detailed subgroup analyses, meta-regression if sufficient studies are available, and transparent reporting of heterogeneity measures.
Third, inherent limitations of the NMA methodology must be discussed. The validity of indirect comparisons fundamentally depends on the transitivity assumption—that studies comparing different interventions are similar in their distribution of effect modifiers. In addition, the credibility of NMA findings is generally lower for indirect comparisons compared with direct head-to-head comparisons. Furthermore, for treatment comparisons based exclusively on indirect evidence without any direct trials, our findings should be considered hypothesis-generating rather than definitive, potentially indicating the need for direct comparative trials as a confirmatory step.
Limitations
Beyond the methodological considerations discussed above, several additional limitations warrant acknowledgement. First, our analysis will examine each outcome separately without accounting for potential correlations between multiple outcomes (eg, hearing improvement and tinnitus reduction) measured within the same patient population. This pragmatic simplification may overlook important relationships between outcome domains and should be considered when interpreting results. Second, the geographical distribution of evidence presents concerns for external validity. As with many previous SRs and meta-analyses in the field of acupuncture, we expect most included studies to originate from China. This geographical concentration, along with documented concerns regarding publication bias in Chinese acupuncture literature, where positive results in acupuncture trials may be more likely to be published,32 may limit the generalisability of our findings to Western healthcare settings.
Data availability statement
Data sharing not applicable as no datasets generated and/or analysed for this study. Not applicable.
Ethics statements
Patient consent for publication
Ethics approval
Not applicable.
