|Year : 2017 | Volume
| Issue : 2 | Page : 50-61
Interventions in the management of blood viscosity for idiopathic sudden sensorineural hearing loss: A meta-analysis
Department of Otolaryngology; Department of Hearing and Speech Sciences, Vanderbilt University, Nashville, Tennessee, USA
|Date of Submission||23-Dec-2016|
|Date of Acceptance||19-Jan-2017|
|Date of Web Publication||15-Jun-2017|
Department of Otolaryngology, Vanderbilt University Medical Center, 602 Oxford House, 1313 21st Avenue South, Nashville, TN 37232-4480
Source of Support: None, Conflict of Interest: None
Aims and Objectives: To estimate the efficacy of interventions in the management of blood viscosity for idiopathic sudden sensorineural hearing loss (ISSNHL) using meta-analysis method. Materials and Methods: An extensive search for literature was performed to identify publications from 1966 to 2016 in Cochrane controlled clinical trials register online, Medline, EMBASE, BIOSIS Preview, Web of Science, China Knowledge Resource Integrated Database and Wanfang. Only randomized control trials (RCTs) with valid control groups were included in this study. Each trial was graded for methodological quality using a 6-point standard. Meta-analysis was performed to assess the efficacy of intervention in managing blood viscosity on hearing improvement (primary outcome), relief of tinnitus and vertigo (secondary outcomes) for ISSNHL. Tests for sensitivity, heterogeneity as well as publication bias were also performed to evaluate the validity of results. Results: Forty-nine RCTs with a total of 4978 patients were included in this meta-analysis. The overall odds ratio of all interventions was 3.12 (P < 0.01). Significant effect sizes were shown on fibrinogen reduction, plasmapheresis, and anticoagulation as well as hemodilution. The mean methodological score was 3.0. There was great homogeneity between studies. No evidence of publication bias was found. Sensitivity test showed reliability and robustness of the results. Conclusion: This meta-analysis provided evidence to support the effect of interventions that manage blood viscosity in the treatment of ISSNHL. There is a need for more high-quality RCTs in the future, especially for hemodilution, anticoagulation as well as plasmapheresis.
Keywords: Anticoagulation, blood viscosity, fibrinolysis, fibrinolytic therapy, hemodilution, idiopathic, meta-analysis, rheopheresis, sudden deafness, sudden hearing loss, systematic review
|How to cite this article:|
Li Y. Interventions in the management of blood viscosity for idiopathic sudden sensorineural hearing loss: A meta-analysis. J Health Res Rev 2017;4:50-61
|How to cite this URL:|
Li Y. Interventions in the management of blood viscosity for idiopathic sudden sensorineural hearing loss: A meta-analysis. J Health Res Rev [serial online] 2017 [cited 2017 Jun 22];4:50-61. Available from: http://www.jhrr.org/text.asp?2017/4/2/50/208116
| Introduction|| |
Idiopathic sudden sensorineural hearing loss (ISSNHL) often presents with a sudden deafness without a clear precipitating cause. ISSNHL is diagnosed based on an abrupt or rapidly progressing hearing loss of at least 30 dB in at least three contiguous frequencies over a period of no more than 3 days. Its incidence has been estimated at 8–15/100,000 persons per year., The age of the patient population is averaged between 40 and 54 years, and there is almost an equal gender distribution of the disease.
Despite recent advances in medicine, the etiology of ISSNHL remains unclear. Different theories have been propounded, including disturbance of cochlear blood flow, autoimmune disease, viral infections, or a combination of such mechanisms.,, As a result, various treatments are applied for ISSNHL. These include vasoactive substances, antiviral agents, vitamins, fibrinolytics, hyperbaric oxygen and more commonly, oral or intratympanic steroids. However, there is lack of an evidence-based approach to evaluate the effectiveness of most treatments of this condition.
Although there is unlikely a single cause to the disease, disturbance of cochlear microcirculation is now commonly considered as a leading one. It has been demonstrated in animal models that mammalian cochlear is sensitive to changes in microcirculation. Loss of function of the organ of Corti could be triggered by even trivial impairment of perfusion. In addition, the labyrinthic artery, which is a functional end artery, is the only blood supply of the inner ear. Therefore, it is not likely to compensate for the disturbance of cochlear blood flow by shunting from the periphery. Studies also show significant correlation between cardiovascular risk factors and incidence of ISSNHL, indicating vascular event in the pathogenesis of disease., The fact that most cases of ISSNHL are unilateral and of instant onset also supports the hypothesis that ISSNHL has a vascular cause.
Local blood flow is largely determined by blood viscosity. Since the blood flow velocity decreases in the microcirculation area, high blood viscosity leads to a higher chance of local sludging, which causes further slowing of the blood flow in the microcirculation area with a decrease of the shear rate, resulting in increased blood viscosity. This decreases the capacity of O2 transport in the blood and causes microthrombosis and damage to the endothelium. Therefore, interventions that reduce blood viscosity could improve microcirculation and be of therapeutic effectiveness to this disease. The major factors influencing blood viscosity include red blood cells and levels of plasma protein or fibrinogen. Therefore, any intervention that reduces the plasma level of fibrinogen or large-molecule protein, erythrocyte deformability, or inhibits hematocyte aggregation could effectively reduce blood viscosity. Rheologic treatment (e.g., hydroxyethyl starch, dextran, and pentoxifylline), in this sense, has numerous biologic actions, including hemodilutional effect, inhibition of platelet and erythrocyte aggregation. In addition to defibrinogenase (e.g., Batroxobin) or plasmapheresis (e.g., fibrinogen apheresis) that reduces plasma fibrinogen and low-density lipoprotein (LDL) level, they, in general, are believed to be effective reducing blood viscosity and leading to improved microcirculation, greater capillary flow, and a better oxygen supply of the cochlea.
As the etiology of ISSNHL remains unclear, quite a few clinicians apply a multimodal treatment strategy, assuming that one or more treatment approaches may reverse the pathophysiologic changes in the auditory system. Although this strategy ensures at least one effective treatment may be provided to the patient, it has obscured the effect of any individual treatment as well as exposed the patients to a number of side effects. Therefore, this study is aimed to identify and assess all randomized control trials (RCTs) that isolate the effectiveness of interventions that manage blood viscosity for ISSNHL, and to produce reliable estimates of the efficacy of this treatment approach using a meta-analysis method.
| Materials and Methods|| |
This study was conducted according to the methodology suggested by the Cochrane Collaboration in the Cochrane Handbook for Systematic Reviews of Interventions. An extensive search for literature was performed to identify the primary studies from the year 1966 to August 2016 in several generally accepted databases including Cochrane controlled clinical trials register online, Medline, EMBASE, BIOSIS Preview, Web of Science, China Knowledge Resource Integrated Database, and Wanfang database. Search for “Grey literature” was also performed through OpenGrey (http://www.opengrey.eu/) and Clinical trials (http://www.clinicaltrials.gov/). Search strategy is formulated as recommended by Cochrane Collaboration consisting of the search terms “sudden,” “hearing loss” OR “deafness,” “Fibrinolytic agents” OR “Hemodilution,” OR “Plasmapheresis.” Their combinations changes in spelling and additional terms (e.g., specific drug names) varied according to the searching sources used. The terms were first expressed in English and then translated to Chinese.
Studies were eligible for meta-analysis, if they met the following criteria: Use of recognized criteria to define ISSNHL; RCT design with a control group receiving placebo and/or the same therapeutic protocol as the experimental group except for the intervention (s) being studied; intervention must be targeted only or mainly at managing the blood viscosity in patients, including use of some vasoactive drugs (e.g., Pentoxifylline), defibrinogenase (e.g., Batroxobin), plasma volume expanders (e.g., Dextran, hydroxyethyl starches), or plasmapheresis (e.g., heparin-induced extracorporeal LDL precipitation, fibrinogen/LDL apheresis). Any other treatment for ISSNHL mainly through other mechanism (s) is not eligible for this study even though it might be of some effect of managing blood viscosity (e.g., Vitamin C, Ginkgo Biloba extract, and Prostaglandin I).
The author then screened the titles and abstracts of potentially eligible studies identified, examined the full articles to determine whether they met the inclusion criteria, and extracted data using data extraction forms. Any uncertainty was marked down and rechecked until a solution was made.
The methodological quality of all the included trials was evaluated to determine the validity of each study. The criteria for assessing the study results were adapted from the Users' Guides to the Medical Literature series published by the American Medical Association and have been used by other researchers., These criteria included randomization, blinding method, baseline similarity, and concealment of allocation.
The major outcome measure was hearing gain in different frequencies assessed by pure tone audiometry. For studies that only report dichotomized outcome (e.g., pure-tone average [PTA] scores reported categorically as “improvement” or “no improvement”), odds ratio (OR) was chosen as the effect size that compares therapeutic outcome between experimental groups and control groups; in studies where outcomes are reported in forms of continuous data, standardized mean differences (SMDs) were selected as the effect size. Meta-analyses were performed using RevMan 5.3 (The Cochrane Collaboration, Oxford, England) and Stata 11.1 (StataCorp, College Station, TX, USA). The Q-test of heterogeneity was conducted to evaluate the level of homogeneity of all included studies, and I2 and τ2 values were also calculated to assess the level of between-study heterogeneity. Random-effect model was selected for calculation of the overall effect sizes beforehand regardless of the result of heterogeneity test since we did not expect a single true effect size for different types of treatments. Potential publication bias was investigated using visual assessment of the funnel plot and Peters test for the effect sizes in the form of OR.
| Results|| |
Search for literature was performed for the first time in March 2012 and was updated twice in July 2014 and August 2016. A total of 2229 citations were found in the initial literature search, deduction of 463 duplicated ones yielded 1766 potentially relevant citations. The initial screen had ruled out 1676 citations by reviewing the titles and abstracts. Full-text review on the remaining 90 articles yielded 49 studies that met our inclusion criteria. A flow diagram of search and selection is shown in [Figure 1]. Most excluded studies were either retrospective-designed or observational, or they did not have a valid control group.
Of the 49 RCTs included in the meta-analysis, there were a total of 4978 patients with 5062 affected ears. Among them, 3462 patients were included in 39 studies of fibrinolytic therapy, 1022 patients in 4 studies of hemodilution, 132 patients in 1 study of plasmapheresis, and 362 patients in 5 studies of anticoagulation. The mean age of patients ranged from 42 to 56, and male to female ratio was 1.18. All patients started treatment within 45 days after disease onset, the majority of them were treated within 2 weeks after symptom (s) presented. Hearing threshold and percentage of accompanied symptoms at baseline varied between studies [Table 1] and [Table 2].
|Table 1: Summary of treatment regime of the included randomized control trials|
Click here to view
All studies reported hearing gain assessed by PTA, of them 46 studies reported dichotomized outcome (i.e., improvement versus no improvement). Pooled effect sizes are presented in [Figure 2]. The overall OR of all interventions that manage blood viscosity was 3.12 (95% confidence interval [CI], 2.71–3.59, P< 0.01). These RCTs were further divided into four subgroups based on the mechanisms of interventions being studied, although subgroup analysis did not show significant between-group difference (χ2 = 0.66, P = 0.88). The mean OR was 3.01 (95% CI, 2.55–3.55, P< 0.01) for fibrinolytic therapy, 2.82 (95% CI, 1.27–6.22, P = 0.01) for hemodilution, 3.37 (95% CI, 1.99–5.72, P< 0.01) for anticoagulation, and 3.95 (95% CI, 1.88–8.33) for plasmapheresis. Evidence was shown for the overall efficacy of all interventions that manage blood viscosity for ISSNHL as well as the efficacy of each individual type of these therapies. All the effects ranged from medium to large based on Cohen's “rules of thumb” for effect sizes.
|Figure 2: Forest plot of odds ratios for effect of interventions that manage blood viscosity in hearing improvement assessed by PTA. CI: Confidence interval, IV: Inverse variance, PTA: Pure-tone average|
Click here to view
Q-test for the above studies that reported dichotomized outcome did not reveal any significant between-study heterogeneity (P = 0.99). No evidence of heterogeneity was found in subgroups that have 2 or more RCTs. There was great homogeneity between studies both in subgroups and all together (I2 = 0%, τ2 = 0.00).
The results of hearing gain in dB were only reported in 6 RCTs with a total of 716 cases. There were 2 studies on fibrinolytic therapies, 3 studies on hemodilution, and 1 study on anticoagulation. The effect sizes for SMD in hearing gain are presented in [Figure 3]. The overall difference in hearing gain between the experimental groups and the control groups was not significant (SMD = 0.15, P = 0.22). Further, lack of significant difference in hearing gain was seen in two of the three subgroups, including hemodilution and anticoagulation. Only fibrinolytic therapy showed a significantly higher hearing gain favoring the experimental group. Moderate heterogeneity was found for studies that reported continuous outcome (I2 = 55%, τ2 = 0.05), although Q-test did not show this heterogeneity is significant (P = 0.09). The nonsignificant results, as well as the moderate heterogeneity of included studies, may partially be due to the small sample size of identified RCTs that reported continuous outcome in the form of hearing gain in dB.
|Figure 3: Forest plot of standardized mean differences for effect of hemodilution therapy in hearing improvement assessed by PTA. CI: Confidence interval, IV: Inverse variance, PTA: Pure-tone average|
Click here to view
Only a total of 7 studies had other types of outcome reported, including self-assessment of hearing capability, relief of tinnitus, and vertigo. Relief of tinnitus was reported in all 7 studies with 965 cases; its pooled effect size is shown in [Figure 4]. The overall effect size for relief of tinnitus was 2.20 (95% CI 1.58–3.06, P< 0.01), which indicates interventions in the management of blood viscosity were effective in relieving patients' accompanying tinnitus. Relief of vertigo was reported in 6 studies with 259 cases. Significance was shown in relief of vertigo by interventions in management of blood viscosity (OR = 2.87, P< 0.01) [Figure 5]. The other outcome was only available in one study with limited patient numbers, therefore, no meta-analysis was performed on these outcomes.
|Figure 4: Forest plot of odds ratios for effect of interventions that manage blood viscosity in relieving tinnitus. CI: Confidence interval, IV: Inverse variance|
Click here to view
|Figure 5: Forest plot of odds ratios for effect of interventions that manage blood viscosity in relieving vertigo|
Click here to view
Only 9 studies reported adverse events, two of them were hemodilution, and 7 were fibrinolytic therapies. Adverse events of hemodilution were mostly seen in skin, nervous system, and gastrointestinal system, whereas the most common side effects for fibrinogen reduction therapy were elevated GPT, injection site bleeding, and gastrointestinal reaction.
Visual scanning of the funnel plot did not find obvious evidence for publication bias since studies were distributed quite symmetrically on both sides of the mean effect size [Figure 6]. Peters test did not show any evidence for small study effect (P > 0.05). Both results indicated there is probably not a publication bias that may affect the results of this meta-analysis.
|Figure 6: Funnel plot of included studies for assessment of publication bias|
Click here to view
The mean score for methodology quality of all included studies was 3 ± 0.9 (mean ± standard deviation). Only 2 studies (4%) had a quality score of 6 points. The result is presented in [Table 3]. Most studies failed to report blinding method and concealment of allocation. There were only 7 RCTs that mentioned the blinding method and 2 RCTs mentioned the concealment of patient allocation. All studies did state that their patients were randomized; however, most of them did not reveal the specific method of randomization. Similarity in patients' baseline conditions was also mentioned in most, but not all, studies. A sensitivity test was performed by ruling out those studies, of which methodological scores are lower than 3, and the result still remained significant (OR = 3.11, P< 0.01). Since most studies were conducted in China while a few others were in European countries, an additional analysis was conducted to see the impact of location on study quality. The result showed a significant difference of methodology quality scores between locations where the studies were conducted that studies in European countries had a higher score distribution than those conducted in China (Independent-samples Mann–Whitney U-test, P< 0.01).
| Discussion|| |
The primary aim of this meta-analysis was to determine the efficacy of interventions in the management of blood viscosity for ISSNHL. The results suggested the significant efficacy of these interventions in improving hearing threshold, relieving tinnitus, and vertigo in patients with ISSNHL. Of them, fibrinolytic therapies, hemodilution, plasmapheresis, and anticoagulation were all effective to improve hearing ability in patients with ISSNHL. However, it should be pointed out that more studies are needed to strengthen the conclusions in certain subgroups as the number of RCTs included in these subgroups was small except for fibrinolytic therapies.
This is the first study that assesses all types of interventions in the management of blood viscosity for ISSNHL. It has long been assumed that the improvement of microcirculation could be obtained by applying vasodilators and/or interventions that reduce blood viscosity. Vasodilators were known to increase the caliber of blood vessels and therefore thought to be effective increasing local blood flow. However, Fisch et al. found there is no evidence that cochlear blood flow is increased by oral vasodilators, such as papaverine or nicotinic acid, which produce facial flushing and headache. It is now believed that nonselective widening of cranial blood vessels may lead to a condition called “vascular steal,” that dilation of one blood vessel network “steals” blood flow from another region within the organ that is already maximally dilated because of the presence of proximal lesions. In the case of ISSNHL, vasodilators may actually decrease the blood flow of the Corti organ and make the microcirculation worse. As a matter of fact, Wang et al. and Gong et al. have independently conducted systematic reviews and meta-analyses on the effect of vasodilators for ISSNHL; neither of them found enough evidence to support vasodilator was more effective than placebo for ISSNHL., Neither, Agarwal and Pothier did not find enough evidence to prove the efficacy of vasodilators and vasoactive substances for ISSNHL. However, in Agarwal's study, they included different types of vasoactive substances that provide multiple therapeutic effects (e.g., antioxidation, anti-inflammation) besides reducing blood viscosity. This may have confounded the effect of any individual type of intervention. To minimize such confounding factors from other therapeutic mechanisms, the individual effect of interventions that manage blood viscosity was uniquely in this study. Therefore, the results were more reliable with regard to its individual therapeutic effect for ISSNHL.
Several types of interventions that manage blood viscosity were identified and analyzed in this study. Defibrinogen or fibrinolytic therapy effectively reduces the plasma fibrinogen level. Fibrinogen is a macromolecular component in the blood; reduction of its level leads to decreased blood viscosity, and substantially improved blood flow. What is more, since fibrinogen is involved in blood coagulation, decreased plasma fibrinogen level also reduces the risk of microthrombosis, which may also cause disturbance of microcirculation. Anticoagulation prevents aggregation of red blood cells and platelets while lipoprotein reduction therapy decreased the plasma lipoprotein level, especially LDL level. Therefore, they together produce a similar effect to defibrinogen or fibrinolytic therapy, reducing the blood viscosity and minimizing the risk of clotting. Fibrinogen/LDL apheresis removes fibrinogen and LDL in the plasma and therefore also has a similar effect as the above two treatments. Hemodilution reduces the concentration of red blood cells and plasma constituents by increasing the plasma volume. It helps to reduce the plasma viscosity and contributes to improve microcirculation. Our results indicate all these interventions are effective for ISSNHL. This also indicates a possible role of microthrombosis in the pathogenesis of ISSNHL.
There are several strengths in this meta-analysis. First, all RCTs included in this meta-analysis showed great homogeneity. It minimized any unknown variance that may diminish the analytical power. Even though they were further divided into four subgroups based on their therapeutic mechanisms, subgroup analyses did not show any significant difference in their effect sizes. Second, there was no evidence for publication bias or small study effect in this meta-analysis; therefore, a risk of impact on mean effect sizes from studies with small sample sizes was minimal. Third, we conducted a sensitivity test by ruling out studies that had low methodological scores, but the result remained consistent and significant, indicating the effect of interventions that manage blood viscosity for ISSNHL is robust.
However, our study also has several limitations. First of all, although all included studies were RCTs, most of them were unclear in randomization sequence generations, blinding method as well as patient dropouts, and hence selection bias or confounding factors might have been present. Second, most of included RCTs were fibrinolytic therapies, and there were less than five studies in each of other categories. More RCTs of hemodilution, plasmapheresis, and anticoagulation should be needed in the future. Third, as PTA is the most commonly used outcome measure in terms of hearing gain assessment, most authors tended to report dichotomous outcome based on whether there was an improvement in hearing ability after treatment. Only a few RCTs reported the specific amount of hearing gain for both experimental and control groups. Notably, the latter should be preferred as continuous data has stronger analytic power than dichotomized data. In addition, the criteria of hearing improvement varied from study to study and from country to country. For example, most RCTs conducted in China had applied a national standard of at least 15 dB hearing gain in average PTA across all frequencies as a criterion for hearing improvement. To the contrary, a study in Germany used a different criterion for hearing improvement. Both drawbacks of reporting dichotomized outcome might compromise the reliability of a study. In this sense, reporting hearing gain in both experimental and control groups may be strongly encouraged for any future study. In addition, as the outcome measurement was limited to hearing gain in most RCTs, report of other outcomes such as relief of tinnitus and vertigo, as well as adverse events should also be necessary as they may provide additional information about the effect and side effect of these interventions for ISSNHL.
| Conclusion|| |
This meta-analysis provided evidence to support the effect of interventions that manage blood viscosity in the treatment of ISSNHL. There is a need for more high-quality randomized controlled trials in the future, especially for hemodilution, anticoagulation as well as plasmapheresis.
The author would like to thank Dr. Emily Tanner-Smith for her contribution to the study design and suggestion on statistical analysis. The expertise of Dr. Robert Labadie was invaluable. I acknowledge Dr. Daniel Ashmead and Dr. James Bodfish for their advice in manuscript write-up.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Agarwal L, Pothier DD. Vasodilators and vasoactive substances for idiopathic sudden sensorineural hearing loss. Cochrane Database Syst Rev 2009;;(4):CD003422. doi: 10.1002/14651858.
Hughes GB, Freedman MA, Haberkamp TJ, Guay ME. Sudden sensorineural hearing loss. Otolaryngol Clin North Am 1996;29:393-405.
Stokroos RJ, Albers FW. The etiology of idiopathic sudden sensorineural hearing loss. A review of the literature. Acta Otorhinolaryngol Belg 1996;50:69-76.
Li Y, Rentmeester LA, Wang H. Antioxidants for idiopathic sudden sensorineural hearing loss. Cochrane Database of Systematic Reviews 2012:CD010058. DOI: 10.1002/14651858.CD010058.
Cole RR, Jahrsdoerfer RA. Sudden hearing loss: An update. Am J Otol 1988;9:211-5.
Shikowitz MJ. Sudden sensorineural hearing loss. Med Clin North Am 1991;75:1239-50.
Thurmond M, Amedee RG. Sudden sensorineural hearing loss: Etiologies and treatments. J La State Med Soc 1998;150:200-3.
Qiang Q, Wu X, Yang T, Yang C, Sun H. A comparison between systemic and intratympanic steroid therapies as initial therapy for idiopathic sudden sensorineural hearing loss: A meta-analysis. Acta Otolaryngol 2016:1-8. [Epub ahead of print].
Kim HA, Lee H. Recent advances in understanding audiovestibular loss of a vascular cause. J Stroke 2016. [doi: 10.5853/jos.2016.00857. [Epub ahead of print].
Miller JM, Dengerink H. Control of inner ear blood flow. Am J Otolaryngol 1988;9:302-16.
Suckfüll M; Hearing Loss Study Group. Fibrinogen and LDL apheresis in treatment of sudden hearing loss: A randomised multicentre trial. Lancet 2002;360:1811-7.
Marcucci R, Alessandrello Liotta A, Cellai AP, Rogolino A, Berloco P, Leprini E, et al.
Cardiovascular and thrombophilic risk factors for idiopathic sudden sensorineural hearing loss. J Thromb Haemost 2005;3:929-34.
Oreskovic Z, Shejbal D, Bicanic G, Kekic B. Influence of lipoproteins and fibrinogen on pathogenesis of sudden sensorineural hearing loss. J Laryngol Otol 2011;125:258-61.
Baskurt OK, Meiselman HJ. Blood rheology and hemodynamics. Semin Thromb Hemost 2003;29:435-50.
Labus J, Breil J, Stützer H, Michel O. Meta-analysis for the effect of medical therapy vs. placebo on recovery of idiopathic sudden hearing loss. Laryngoscope 2010;120:1863-71.
Conlin AE, Parnes LS. Treatment of sudden sensorineural hearing loss: I. A systematic review. Arch Otolaryngol Head Neck Surg 2007;133:573-81.
Guyatt GH, Sackett DL, Cook DJ. Users' guides to the medical literature. II. How to use an article about therapy or prevention. A. Are the results of the study valid? Evidence-Based Medicine Working Group. JAMA 1993;270:2598-601.
Fisch U, Murata K, Hossli G. Measurement of oxygen tension in human perilymph. Acta Otolaryngol 1976;81:278-82.
Haberkamp TJ, Tanyeri HM. Management of idiopathic sudden sensorineural hearing loss. Am J Otol 1999;20:587-92.
Gong Y, Liang C, Li J, Tian A, Chen N. Vasodilators for sudden sensorineural hearing loss: A systematic review of randomized controlled trials. Zhonghua Er Bi Yan Hou Ke Za Zhi 2002;37:64-8.
Wang M, Liu Y, Du Z, Zhu X, Luo G. A systematic review of vasodilators for sudden sensorineural hearing loss. Lin Chung Er Bi Yan Hou Tou Jing Wai Ke Za Zhi 2010;24:869-71.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6]
[Table 1], [Table 2], [Table 3]