Odds ratios (ORs) are calculated using the random-effects Mantel-Haenszel method. A, Hearing improvement was measured between the groups with sudden sensorineural hearing loss with and without vertigo. B, Hearing recovery rate was measured in a subgroup analysis of the grading system using Siegel criteria. C, Hearing recovery rate was measured in the subgroup receiving systemic corticosteroids. D, Hearing recovery rate was measured in the subgroup receiving intratympanic corticosteroids. Different marker sizes indicate weight.
Odds ratios (ORs) are calculated using the random-effects Mantel-Haenszel method. A, Studies included in the meta-analysis of the hazard to hearing improvement between the groups with sudden sensorineural hearing loss with and without vertigo. B, Studies included in the meta-analysis of the hazard to hearing improvement in subgroup analysis of grading system using the Siegel criteria. C, Studies included in the meta-analysis of the hazard to hearing improvement in the subgroup receiving systemic corticosteroids. D, Studies included in the meta-analysis of the hazard to hearing improvement in the subgroup receiving intratympanic corticosteroids. The dashed blue lines indicate pooled odds ratios.
Meta-analysis estimates are calculated from all studies except each omitted source in turn. The dashed blue lines indicate pooled odds ratio and corresponding 95% CIs.
eFigure. Flowchart of the Searching and Selecting Process
Customize your JAMA Network experience by selecting one or more topics from the list below.
Yu H, Li H. Association of Vertigo With Hearing Outcomes in Patients With Sudden Sensorineural Hearing Loss: A Systematic Review and Meta-analysis. JAMA Otolaryngol Head Neck Surg. 2018;144(8):677–683. doi:10.1001/jamaoto.2018.0648
What is the association of vertigo with hearing outcomes of sudden sensorineural hearing loss?
In this systematic review and meta-analysis of 4814 patients from 10 studies, sudden sensorineural hearing loss accompanying vertigo was associated with worse recovery of hearing compared with sudden sensorineural hearing loss without vertigo. However, no obvious association of vertigo with the prognosis of sudden sensorineural hearing loss was observed when corticosteroids were injected in the middle ear.
Vertigo may be a risk factor for sudden sensorineural hearing loss; future studies are needed to examine whether treatment of vertigo might contribute to the recovery of hearing.
Sudden sensorineural hearing loss (SSHL) accompanied by vertigo may portend a negative prognosis in the hearing outcome.
To investigate the association of vertigo with prognosis of hearing variables in SSHL.
A literature search of eligible studies was performed in PubMed, Web of Science, and Embase from September 26, 1973, through September 26, 2017. Studies published in English were retrieved with no restrictions on the date of publication. References were identified by screening the proceedings of relevant reviews, and annual meeting and other correlative papers were scanned manually for enrollment.
All original research studies and retrospective or prospective studies focusing on the role of vertigo in the prognosis for the hearing outcome of SSHL were systematically retrieved. Studies that did not include data regarding the association between the rate of hearing recovery and vertigo were excluded, as were reviews, comments, case reports, editorials, letters, and practice guidelines.
Data Extraction and Synthesis
Data were extracted and evaluated by 2 researchers. Data extracted included research type, number of participants with or without vertigo, treatment regime, definition of pure-tone average, criteria for hearing improvement, and length of follow-up. The quality of included studies was evaluated using the Newcastle-Ottawa Scale (scores range from 0-9, with a score of ≥6 indicating a high-quality study). The data were synthesized in Mantel-Haenszel models; the aggregate results were estimated in forest plots.
Main Outcomes and Measures
Association of vertigo with the prognosis for the hearing outcome of SSHL.
Of the 4814 unique patients identified in 10 studies, 1709 were included in the SSHL group with vertigo and 3105 were included in SSHL group without vertigo. The Newcastle-Ottawa Scale score of each study selected was greater than 7. The recovery rate of hearing was 42.13% in the group with vertigo, compared with 60.29% in the group without vertigo. Vertigo was significantly associated with a worse hearing recovery (odds ratio, 2.22; 95% CI, 1.54-3.20; I2 = 74%). Similar results were obtained in subgroup analyses of the grading system using the Siegel criteria and systemic corticosteroid therapy. However, no association of vertigo with the prognosis of SSHL was observed within the subgroup receiving intratympanic corticosteroids (odds ratio, 1.78; 95% CI, 0.64-4.94; I2 = 70%).
Conclusions and Relevance
Current evidence revealed that vertigo may be negatively associated with hearing recovery in patients with SSHL, except in a subgroup that received intratympanic corticosteroids. Corticosteroid injection may be more effective for treatment of SSHL accompanied by vertigo; future studies are needed to determine whether treatment of vertigo might contribute to the recovery of SSHL.
Quiz Ref IDSudden sensorineural hearing loss (SSHL) is clinically characterized by rapid-onset sensorineural hearing loss of more than 30 dB in at least 3 contiguous audiometric frequencies within 3 days.1-3 The incidence of SSHL is approximately 5 to 20 per 100 000 persons per year.4 At present, the pathogenesis, clinical manifestations, optimal treatments, and prognostic factors of SSHL are not clear, with spontaneous recovery rates ranging from 32% to 70%.5,6 Various factors may affect hearing recovery after SSHL, such as age, degree of hearing loss, type of hearing loss, interval from onset of symptoms to treatment, and the presence of vertigo, tinnitus, type 2 diabetes, and cardiovascular disease.7-9
Vertigo is a symptom of vestibular dysfunction and has been described as a sensation of motion, most commonly rotational motion.10 Differentiation of vertiginous symptoms from other forms of dizziness, such as lightheadedness, which is most often associated with presyncope, is important.10 Vestibular involvement in SSHL was first reported in 1949, and its incidence in previous studies was 30% to 40%.11,12 Causes may be associated with the anatomical proximity of the cochlea and vestibule. At present, systemic and intratympanic corticosteroids are considered to be the most effective and commonly accepted treatments for SSHL.13 However, little is known about the mechanisms underlying the treatment effect of corticosteroids against sudden hearing loss. Because oxidative stress is a key determinant in endothelial dysfunction, a complex interaction of corticosteroids is believed to contribute to the inhibition of the process of hearing loss and restoration of hearing.14 Some studies15,16 reported vertigo as a poor indicator in the clinical manifestations and prognosis of SSHL. Research by Cho and Choi17 and Wen et al6 did not confirm the prior hypothesis, and they found no difference in hearing recovery between groups with and without vertigo among patients with profound SSHL. Therefore, the clinical value of vertigo in the clinical manifestation and prognosis of SSHL has not yet achieved consensus. To this end, we performed a systematic review and meta-analysis of published studies to explore the association of vertigo with prognosis of SSHL.
Based on PubMed, Web of Science, and Embase, we performed a literature review of studies that assessed the association between vertigo and prognosis as well as clinical manifestations of SSHL, from September 26, 1973, through September 26, 2017. We used a combination of keywords. The first search string included sudden hearing loss OR deafness, sudden OR sudden deafness OR hearing loss, sudden; the second, vertigo OR vertigos OR spinning sensation OR sensation, spinning OR sensations, spinning OR spinning sensations; and the third, both strings.
All pertinent articles or abstracts that were published in English were retrieved with no restrictions on the date of publication. In addition, references were identified by screening the proceedings of relevant reviews, and annual meeting and other correlative papers were scanned manually for enrollment.
Selection criteria for articles included in this analysis consisted of (1) articles or abstracts of original research; (2) retrospective or prospective studies of hearing outcomes in patients with SSHL; (3) a homogeneous treatment regime (systemic or intratympanic corticosteroid therapy) in each group; and (4) populations of patients with and without vestibular symptoms. However, studies were excluded from the present analysis if we failed to retrieve the data regarding the association between rate of hearing recovery and vertigo. Investigations that included reviews, comments, case report, editorials, letters, or practice guidelines were also excluded.
Studies and data were independently extracted and evaluated by 2 of us (H.Y. and H.L.) for quantitative analysis. The characteristics of each study included research type, number of participants with or without vertigo, treatment regime, definition of pure-tone average, criteria of hearing improvement, and length of follow-up. Hearing recovery was defined using the Siegel criteria (≥15 dB) and the criteria of the Ministry of Health and Welfare of Japan (≥10 dB). The retrieved data were compared and revised in mutual agreement by the 2 researchers.
We used the Newcastle-Ottawa Scale to assess the quality of included studies. The Newcastle-Ottawa Scale checklist included the following 3 criteria: study selection (maximum, 4 stars), comparability of cohort (maximum, 2 stars), and assessment of outcome (maximum, 3 stars).18,19 A score of 6 or more stars is indicative of a high-quality study.
A sensitivity analysis and meta-regression were further performed to investigate the heterogeneity of our data and estimate the influence of each study on pooled outcomes, based on the rule of omission. We also performed detailed sensitivity analyses to explore whether final results were significantly affected by a certain study. The study was recognized as a source of high heterogeneity if the I2 value decreased greatly with its exclusion.
Meta-analysis was performed using Review Manager, version 5.3 (The Nordic Cochrane Centre) and Stata, version 12.0 (StataCorp) software. The prognosis outcomes were displayed in forest plots using the corresponding 95% CIs. The pooled association between hearing recovery rate and incidence of vertigo was analyzed as dichotomous data using the Mantel-Haenszel method, with a combined estimate of odds ratio.20 All data included in this analysis were combined quantitatively across eligible studies. The statistical heterogeneity of the studies was assessed using a P value (P ≤ .05 indicating significant heterogeneity) and an I2 statistic as a derivative of the Cochran Q statistic (<25% indicates slight heterogeneity; 25% to <50%, moderate heterogeneity; and ≥50%, high heterogeneity).21,22 A fixed-effects model was used when I2 < 50% and/or P ≤ .05; otherwise, a random-effects model was used. The potential asymmetry and publication bias were visually evaluated by funnel plot, which has insufficient power when fewer than 10 studies were involved.
Based on the search criteria, 1685 studies were initially identified from our database search. After checking the titles or abstracts, 1590 studies were first excluded. Of the remaining 95 studies, 77 were then removed because of the absence of hearing recovery in SSHL with or without vertigo. Another 8 articles were excluded from the final analysis owing to (1) duplicated data probably containing the same participants or (2) insufficient data reported. A total of 10 unique articles with 4814 unique participants were eligible for the final meta-analysis6,23-31 (eFigure in the Supplement).
The main clinical characteristics of included studies are summarized in Table 1. Among the 4814 patients, 1709 were included in the SSHL group with vertigo and 3105 were included in the SSHL group without vertigo. The duration of follow-up since the onset of deafness ranged from 10 days to 5.7 years. The level of hearing improvement was assessed according to the Siegel criteria32 in 5 studies.6,24-26,29 Patients in 7 studies were treated with systemic corticosteroids24-30; patients in 2 studies, with systemic corticosteroids and/or intratympanic corticosteroids6,31; and patients in 1 study, with intratympanic corticosteroid injection23 (Table 1).
We used the Newcastle-Ottawa Scale to assess the quality of selected studies (Table 2). Five studies24,25,27,30,31 were judged to have an inadequate follow-up period (no data or <2 months) and only achieved 2 stars in outcome. All studies received overall scores of greater than 7.
All 10 studies described the odds of poor hearing results (negative) in a dichotomous pattern (groups with vs without vertigo). We performed a meta-analysis of these data (Figure 1A), and the data showed a high statistical heterogeneity with I2 = 74%. The random-effects model was used for analysis. Quiz Ref IDOur meta-analysis estimated the aggregate mean odds of poor hearing outcomes with vertigo was 2.22 times that without vertigo (95% CI, 1.54-3.20). The recovery rate of hearing was only 42.13% in the SSHL group with vertigo, compared with 60.29% in the SSHL group without vertigo. Vertigo had a strong negative association with hearing recovery in SSHL.
Considering the bias due to different criteria of hearing recovery (threshold of pure-tone average, 10-30 dB) in the 10 studies, we performed a meta-analysis of the 5 studies6,24-26,29 in which the Siegel criteria were used to assess the treatment outcome. The I2 value of these data was 77%. The weighted mean odds ratio of worse hearing outcomes of SSHL with vertigo was nearly double (1.99) that of SSHL without vertigo in the random-effects model (95% CI, 1.05-3.77) (Figure 1B). We observed a poor association within the prognosis of SSHL.
Furthermore, this characteristic was assessed in a subgroup receiving systemic corticosteroids. Eight studies6,24-30 described hearing improvement in this subgroup. We adopted a random-effects model with a high heterogeneity (I2 = 77%) to evaluate the hearing recovery in this subgroup. The weighted mean odds ratio was 2.56 (with vertigo vs without vertigo; 95% CI, 1.47-4.45) (Figure 1C). Vertigo was associated with a worse prognosis for SSHL treated with systemic corticosteroids.
Only 2 studies6,23 were assessed in the intratympanic corticosteroid subgroup, with a combined odds ratio of 1.78 (95% CI, 0.64-4.94) in a random-effects model (I2 = 70%) (Figure 1D). We found no obvious difference in hearing status between groups with and without vertigo, which indicated that the presence of vertigo is not always negatively associated with prognosis in the context of intratympanic corticosteroid therapy.
Asymmetry was observed in the pool of data from our 10 investigations, indicating obvious publication bias, as demonstrated by a funnel plot (Figure 2). A sensitivity analysis was further performed to assess and correct the publication bias, based on the rule of omission (Figure 3). Quiz Ref IDWhen each study was sequentially excluded to assess the stability of the final results, we found that no investigation could affect the pooled risk estimate. In addition, we found that the heterogeneity was significantly reduced if the investigation by Wen et al6 was omitted from our study. In that case, funnel plot was symmetrical in general, and the heterogeneity test changed from I2 = 74% to I2 = 30%. The meta-regression by treatment regime and criteria of hearing improvement showed that these factors did not significantly affect the results (eTable in the Supplement).
Thus far, the pathogenesis, clinical manifestations, optimal treatments, and prognostic factors of SSHL are not clear. Many hypotheses have been advanced to explain its cause, including vascular diseases, allergic reaction, viral inflammation, rupture of intralabyrinthine membranes, and autoimmune diseases.33 Meanwhile, the clinical value of forecasting the efficacy of treatment of SSHL is important. Since the first report of SSNL with balance impairment in 1949,12 the role of vertigo had been extensively researched in SSHL. Many researchers6,15-17,34 have reported an association between the presence of vertigo and clinical features, as well as prognosis of SSHL. However, some findings about the exact role of vertigo in SSHL were contradictious.
To evaluate whether the presence of vertigo is associated with a negative prognosis of SSHL, we quantified the occurrence of hearing improvement in our meta-analysis. The dichotomous data in our review showed that the odds ratio of recovery in the group with vertigo was worse than that in the group without vertigo. Liu et al35 found that using a vestibular function examination, the cochlear and vestibular impairment could be more serious in patients with than without vertigo, which may result in a worse hearing outcome.
Furthermore, the association between the vertigo and poor hearing recovery remained significant in subgroup analyses of the grading system using the Siegel criteria and systemic corticosteroid therapy, which made the potential conclusion more convincing. However, in the intratympanic corticosteroid subgroup, the pooled rate of hearing improvement in the vertigo group also disfavored the group without vertigo computationally. Thus, our study suggested that the presence of vertigo was not a critical variable in determining the extent of hearing improvement in the intratympanic corticosteroid subgroup. For the inconsistent result, Newman-Toker et al36 explained that vertigo was not a certain disease, but rather a symptom with various causes. The study by Wen et al6 was focused on the feature of profound SSHL, which has a worse hearing outcome regardless of the presence of vertigo. A prospective, multicenter randomized clinical trial13 concluded that the addition of intratympanic corticosteroids to systemic corticosteroid therapy may provide a potentially more effective therapeutic option in patients with mild to severe SSHL. However, the mechanism of action of corticosteroids in the inner ear is not completely understood, which may be associated with drug concentrations in the local region.
Our study is, to our knowledge, the first meta-analysis to explore the prognostic role of vertigo in SSHL. The results were large enough to be valid, including 4814 patients in 10 studies. We also use the rule of omission and meta-regression to assess and correct the heterogeneity of our data. If the investigation by Wen et al6 was omitted from our study, the heterogeneity was significantly reduced, and a similar hearing outcome was gained. Besides, subgroup analyses of the grading system using the Siegel criteria and systemic corticosteroid therapy were found to show similar results.
Quiz Ref IDSeveral limitations must be considered in our study. First, all the included studies were generally retrospective, which brought some confounding factors, especially publication bias, to our analysis. Second, the clinical heterogeneity remains significant, although the meta-regression and subgroup analyses were calculated, and a random-effects model was applied to reduce the deviation to a certain extent. In addition, details on the cause of SSHL (eg, viral, stroke) and the audiogram shape (flat, upward rising, or downward sloping) were not included in the study. Finally, we could not include hearing level at onset into our analyses owing to the lack of corresponding data, and it might reduce reliability and induce greater heterogeneity because the type of corticosteroids, the degree of hearing loss at the initial presentation, administration route, and dosage were different in each study.6,24,25 For example, higher heterogeneity was found in the research by Wen et al,6 which focused on profound hearing loss, not all types of hearing loss. With omission of this study, the heterogeneity of the meta-analysis decreased to I2 = 30%, and the conclusion remained basically unchanged. However, the importance of the results should be interpreted with caution.
Our overall results demonstrated that the presence of vertigo may have a negative association with hearing recovery in SSHL, even when the subgroup analyses of grading system using the Siegel criteria and systemic corticosteroid therapy were taken into account. Meanwhile, no significant association between vertigo and prognosis of SSHL was observed in the intratympanic corticosteroid subgroup. Corticosteroid injection in the middle ear may be more effective in SSHL accompanied by vertigo. However, because of some confounding factors, especially publication bias, in our analysis, additional unbiased studies with prospective, larger sample sizes are needed for a more precise estimate of the association of vertigo with the clinical manifestation and prognosis of SSHL.
Accepted for Publication: March 22, 2018.
Corresponding Author: Huawei Li, MD, PhD, ENT Institute, Otorhinolaryngology Department, Eye and ENT Hospital of Fudan University, 83 Fenyang Rd, Shanghai 200031, China (firstname.lastname@example.org).
Published Online: June 21, 2018. doi:10.1001/jamaoto.2018.0648
Author Contributions: Drs Yu and Li had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Yu, Li.
Acquisition, analysis, or interpretation of data: Yu, Li.
Drafting of the manuscript: Li.
Critical revision of the manuscript for important intellectual content: Yu, Li.
Statistical analysis: Yu, Li.
Obtained funding: Li.
Administrative, technical, or material support: Yu.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Funding/Support: This study was supported by grants 2017YFA0103900, 2017YFA0103904, and 2016YFC0905200 from the National Key Technologies Research and Development Program of China and grants 81300825, 81620108005, 81230019, 81470687, ZX2017ZR-ZYZ17, and 18695840700 from the National Natural Science Foundation of China.
Role of the Funder/Sponsor: The funder/sponsor had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Create a personal account or sign in to: