Suddendeafness.files.wordpress.com

Treatment of Sudden Sensorineural Hearing Loss
II. A Meta-analysis
Anne Elizabeth Conlin, BA&Sc, MD; Lorne S. Parnes, MD, FRCSC Objective: To pool and meta-analyze the results of all
5 met inclusion criteria for meta-analysis. Pooling of data randomized controlled trials (RCTs) on treatment of sud- from 2 RCTs that compared steroids with placebo showed den sensorineural hearing loss (SSHL).
no difference between treatment groups (OR, 2.47; 95%CI, 0.89-6.84; P=.08). No difference existed between pa- Data Sources: A MEDLINE search and hand search were
tients treated with antiviral plus steroid therapy vs pla- conducted to identify RCTs published between January cebo plus steroid therapy (OR, 0.92; 95% CI, 0.29-2.92: 1966 and February 2006 in the English language on the P=.88). Finally, there was no difference between sub- treatment of SSHL. Search terms included hearing loss, jects treated with steroids vs subjects treated with any sensorineural (MeSH term), sensorineural hearing loss (text other active treatment (OR, 1.27; 95% CI, 0.64-2.55; words), and sudden deafness (text words).
Study Selection: Prospective RCTs on the treatment
Conclusions: Despite the traditional practice in North
of patients diagnosed as having sudden sensorineural hear- America of treating of SSHL with systemic steroids, a meta- analysis revealed no evidence of benefit of steroids over Data Extraction: A meta-analysis using the random ef-
placebo. There was also no difference in the addition of fects model was conducted when data existed for 2 or antiviral therapy to systemic steroids, nor was there dif- more studies. Odds ratios (ORs), 95% confidence inter- ference between systemic steroids and other active treat- vals (CIs), and tests for heterogeneity were reported.
Data Synthesis: Twenty RCTs were identified, of which
Arch Otolaryngol Head Neck Surg. 2007;133:582-586 SUDDENSENSORINEURALHEAR- temporalbonesexaminedatpostmortem
tent with viral deafness in patients with SSHL9,10; and finally, animal experiments have demonstrated viral penetration of the 100 000 persons per year,1 for which a va- riety of treatments exist. Many causes have tory disturbance and (2) viral infection. It agents, carbogen, vitamins, and more com- monly, oral or intratympanic steroids.
play a role. Evidence supporting the cir- culatory disturbance theory is circumstan- tial, based on case reports of sudden deaf- dence of SSHL, researchers are often faced vascular disease1 and animal models show- ing histopathological cochlear changes due to vascular occlusion.2 Both direct and in- with a sufficiently large sample size to per- Author Affiliations:
mit a highly statistically powered study.
tory reaction theory: SSHL has been asso- Meta-analysis presents a powerful tool to University of Ottawa, Ottawa,Ontario (Dr Conlin), and ciated temporally with active viral upper pool data across several RCTs and thus in- respiratory illnesses3-5; patients with SSHL crease the chance of detecting a real, sta- have antibody titers to several viruses6-8; tistically significant effect. We therefore (REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133, JUNE 2007 2007 American Medical Association. All rights reserved.
Total Events: 28 (Steroids), 20 (Placebo)Test for Heterogeneity: χ2 = 1.11 (P = .29), I2 = 9.5% Test for Overall Effect: Z = 1.74 (P = .08) Figure 1. Meta-analysis of steroids vs placebo. CI indicates confidence interval; OR, odds ratio.
undertook this review to (1) identify, evaluate, and re- neous distribution of effect sizes. Significant heterogeneity be- view all RCTs on the treatment of SSHL (see Treatment yond random fluctuation is known to exist if PϽ.05. A statis- of Sudden Sensorineural Hearing Loss: I. A Systematic Re- tic for quantifying inconsistency, I2, also was reported. The view13 [part I]) and (2) to pool and meta-analyze the re- inconsistency statistic describes the percentage of the variabil- sults of individual RCTs, when possible, to increase the ity in effect estimate that is due to heterogeneity rather than tosampling error or chance. A value greater than 50% may be con- statistical power and produce reliable estimates of the ef- sidered substantial heterogeneity, while a value lower than 30% ficacy of a given treatment. Implications for clinical prac- may represent inconsequential heterogeneity.16 tice and future research are discussed herein.
Graphic displays of the results, termed forest plots, were pro- vided for all meta-analyses to aid in interpretation. The forest plot is a widely used form of presentation that depicts pointestimates (black squares) and error bars (horizontal lines) foreach study.16 Each black square is proportional to the sample A MEDLINE literature search was conducted to identify RCTs size of the study it represents. The combined result of the pooled on the treatment of SSHL, published between January 1996 and data are depicted by a black diamond spanning the 95% CI.
February 2006. Details of this literature search are described When most or all of the 95% CIs in the individual studies con- elsewhere (see part I13). Data were extracted by one reviewer tain the combined rate difference, the studies are considered (A.E.C.) as per the Berlin method,14 whereby there was no blind- ing to authors, journals of publication, or results of the stud-ies. Characteristics and results of all included studies were re-viewed systematically. Studies were then categorized by treatment protocol and evaluated for suitability for meta-analysis.
Meta-analyses were performed using RevMan 4.2 (The Coch- CHARACTERISTICS OF INCLUDED STUDIES
rane Collaboration, Oxford, England). For continuous data, themeans and standard deviations were recorded for each study Twenty RCTs,17-36 reporting 21 treatment comparisons, arm. If continuous data were reported as a mean change or if met the inclusion criteria of this review. The character- standard deviations were not reported, the data were ex- istics, interventions, outcome measures, and results of cluded from pooled analysis. For dichotomous data (eg, sub- these studies are described elsewhere (see part I13).
jective report of tinnitus), absolute numbers were expressedas fractions. If the dichotomous data were expressed as a pro-portion, the data were converted to the original fraction. In stud- STEROID THERAPY VS PLACEBO THERAPY
ies in which continuous outcomes were translated into dichoto-mous variables (eg, pure-tone average scores reported Two studies investigated oral steroids vs oral placebo for categorically as “improvement” or “no improvement”), data were the treatment of SSHL.17,18 Figure 1 displays the data on
analyzed as dichotomous data. Studies in which only graphi- hearing recovery rates reported by Wilson et al17 and Cina- cal representations of data were used, and thereby raw data were mon et al.18 Data from a total of 88 patients were pooled not reported, were excluded from pooled analysis.
for meta-analysis. The test for heterogeneity was not sig- Statistical tools for the meta-analysis were chosen in recog- nificant (P=.29) and inconsistency was marginal (I2=9.5%), nition of the broad inclusion criteria within the studies and theexpectation of between-study variability. Data from indi- which indicated that pooling the data was valid. Pooled vidual studies were combined by means of a random effects analysis of the data revealed no statistically significant dif- model of meta-analysis,15 which assumes a population, or dis- ference between systemic steroids and placebo (odds ra- tribution, of true effect size with each source study represent- tio [OR], 2.47; 95% CI, 0.89-6.84; P=.08).
ing one member of the population. Using this model, weweighted studies by the inverse of variance, and a random ef- ANTIVIRAL PLUS STEROID THERAPY VS
fects estimate of the combined effect and 95% confidence in- PLACEBO PLUS STEROID THERAPY
terval (CI) were calculated. The random effects model of meta-analysis generates a wider 95% CI of the pooled result, whichtherefore generates a more conservative estimate of the treat- Four studies evaluated treatment of SSHL with antiviral therapy and steroid therapy vs placebo and steroid therapy.
A test of heterogeneity was performed using the ␹² statistic Among the 4 RCTs comparing antiviral agents plus ste- to evaluate whether the pooled studies represented a homoge- roids vs steroids alone, 2 were excluded from the meta- (REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133, JUNE 2007 2007 American Medical Association. All rights reserved.
Total Events: 53 (Antivirals + Steroid), 46 (Steroid)Test for Heterogeneity: χ2 = 1.94 (P = .16), I2 = 48.4% Test for Overall Effect: Z = 0.15 (P = .88) Figure 2. Meta-analysis of steroids vs antivirals plus steroids. CI indicates confidence interval; OR, odds ratio.
analysis on the basis of incomplete data reporting.19,20 Con- colleagues28 reported a greater rate of recovery among pa- sequently, pooled analysis of the data was possible for 2 tients treated with carbogen than without and provided studies that evaluated the utility of antiviral therapy com- percentages of patients demonstrating categorical im- bined with steroid therapy. Figure 2 displays the pooled
provement defined as fair, good, recovery, or no im- analysis of rates of improvement reported in these 2 stud- provement. Nageris et al29 also reported a greater rate of ies21,22 across 138 patients. The most similar outcome data recovery among patients treated with magnesium vs with- reported in each study, defined as at least 50% improve- out but provided only a graphical representation of re- ment over baseline,21 was pooled, as was the actual per- covery and not the actual number of patients demonstrat- centage of improvement.22 There was no significant ing recovery. Therefore, owing to the incomplete reporting heterogeneity (P = .16) and there was an acceptable de- of data in the study by Nageris et al,29 pooled analysis of gree of inconsistency (I2= 48.4%) between the 2 studies, studies evaluating magnesium was not possible.
suggesting validity of the pooled analysis. The forest plotindicated no significant difference between steroid plus OTHER THERAPIES
antiviral therapy vs steroid therapy alone (OR, 0.92; 95%CI, 0.29-2.92; P = .88).
Five RCTs assessed the utility of other treatment proto-cols.30-34 Because of the inherent heterogeneity of the treat- STEROIDS VS OTHER THERAPY
ment regimens in these studies (see part I13), pooled analy-sis of these studies was not applicable.
To determine whether steroid treatment might consti-tute the gold standard in treatment of SSHL, pooled analy-sis was completed across all studies that compared ste- roid therapy with any other active treatment. Two studieswere included in this category. Active treatments of car- There is a remarkable array of therapeutic approaches to bogen inhalation18 and fibrinolysis23 were compared with the treatment of SSHL. While the idiopathic nature of this condition inherently presents a therapeutic dilemma, it Figure 3 demonstrates a pooled analysis of the ob-
nonetheless is important to identify the treatment that jective outcome measure data across 183 patients in 2 studies.18,23 There was no significant heterogeneity (P=.30) Steroids have been used widely in the treatment of or inconsistency (I2= 5.9%). The data did not reveal any SSHL because of their proposed benefit in infectious, in- overall effect in favor of steroid therapy over other treat- flammatory, and other immune-mediated conditions and ment (OR, 1.27; 95% CI, 0.64-2.55; P = .50).
despite the fact that their specific mechanism of actionis unknown.37 Independently, the study by Wilson et al17 VASOACTIVE AND HEMODILUTION THERAPY
indicated that treatment with systemic steroids resultedin a statistically significant greater rate of recovery than Five RCTs assessed the utility of vasoactive and hemo- placebo. When the data in this study were pooled with dilution treatments, including pentoxifylline, dextran, the data of the RCT by Cinamon et al18 in a comparison Ginkgo biloba, nifedipine, and combinations thereof.24-27 of systemic steroid therapy vs placebo, there was no longer The treatment methods described in the studies varied a statistically significant treatment effect. To this extent, widely, and no 2 studies used treatment regimens that systemic steroids do not appear to represent an effective were sufficiently homogeneous to permit valid between- study comparisons. Therefore, pooled analysis of out- Viral infection has been proposed as a possible cause come measures of studies regarding vasoactive and he- of SSHL, and a unifying theory implicates both viral in- modilution therapies was not applicable.
fection resulting in an inflammatory response and cir-culatory disturbance as components of a causative cas- MAGNESIUM VS OTHER THERAPY
cade. Viral injury can cause direct vascular and erythrocyteinjury, resulting in secondary vascular insufficiency. Vi- Two RCTs compared the addition of magnesium with ruses also can cause direct inflammation, which simi- other active treatment for patients with SSHL. Gordin and larly causes secondary vascular insufficiency.38 In this (REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133, JUNE 2007 2007 American Medical Association. All rights reserved.
Total Events: 66 (Other Treatments), 59 (Steroids)Test for Heterogeneity: χ2 = 1.06 (P = .30), I2 = 5.9% Test for Overall Effect: Z = 0.68 (P = .50) Figure 3. Meta-analysis of steroids vs other treatments. CI indicates confidence interval; OR, odds ratio.
meta-analysis, among the 4 RCTs that compared antivi- This meta-analysis and systematic review13 does, how- ral plus steroid therapy with placebo plus antiviral therapy ever, have several important strengths. First, there was for the treatment of SSHL, none identified any statisti- no significant heterogeneity or inconsistency across any cally significant difference between the 2 treatment groups.
of the meta-analyses. Despite the diversity of the in- Moreover, in a pooled analysis of 2 of these trials, there cluded studies, use of the random effects model of meta- was no significant evidence that combining antiviral and analysis permitted valid pooling of the data. This is very steroid therapy is better than steroids alone.
beneficial for the clinician because the diversity of the While systemic steroids have been labeled the gold studies in this meta-analysis helps to support the valid- standard for treatment of SSHL,38 this conclusion was ity of the findings and increases the likelihood that the based on a comparison of systemic steroids with pla- results could be applied in a given clinical encounter.
cebo. To our knowledge, a review of systemic steroidsvs another active treatment protocol that targets 1 or more of the possible causes of SSHL never has been done. There-fore, pooled analysis was completed across all studies that Despite the traditional view in North America that sys- compared steroid treatment with another treatment. Only temic steroids are the standard of treatment for 2 RCTs18,23 evaluated steroids alone vs a single active treat- SSHL,1,12,37-39 this meta-analysis showed no benefit of sys- ment modality. Meta-analysis of the data reported in these temic steroids vs placebo when data were pooled across studies failed to support a statistically significant treat- 2 studies. Also, when compared with other forms of ac- ment effect favoring systemic steroids. This meta- tive treatment, steroids offered no greater treatment effect.
analysis was unable to definitively support systematic ste- These quantitative findings regarding the data, in con- roids as the gold standard for treatment of SSHL.
junction with the qualitative findings regarding the meth- Limitations of the literature search and the studies on ods (see part I13), considerably challenge the conven- which this meta-analysis are based are discussed else- tional view that systemic steroids constitute the gold where (part I13). Another limitation is that pooled analy- sis was only possible for a small number of RCTs. Of the At present, SSHL remains a medical emergency with- studies identified, only 5 were sufficiently similar in treat- out a scientific understanding of its cause or a rational ment protocol and reported sufficiently complete re- approach to its treatment. The low incidence of SSHL pre- sults to permit pooled analysis. Also, this meta-analysis sents a considerable challenge in designing any single RCT is inherently limited by the outcome measures reported with sufficient power to detect a real, statistically signifi- in the studies. There appears to be no universally ac- cant treatment effect. To review any condition with a con- cepted best outcome measure for defining success of treat- troversial or unclear treatment protocol, such as SSHL, ment in SSHL. Many of the outcome measures reported systematic review and meta-analysis are powerful tools in the RCTs were subjective outcome measures, such as to integrate prior research, identify research gaps, de- perceived improvement in hearing and presence of tin- fine priorities for future research, and guide clinical man- nitus, as well as indirect objective data, such as categori- agement. Therefore, it is imperative that future reports cal improvement by pure-tone average. Moreover, among describing research on the treatment of SSHL include all categorical outcome measures of improvement, the defi- pertinent data. Continuity in reporting outcome mea- nition of improvement that was used varied across stud- sures across studies will permit powerful calculations of ies, from 50% reduction in a symptom in some studies treatment effect, by means of meta-analyses, and will aid to 75% in others.17,21 Not only does conversion of con- the clinician in identifying the best treatment protocol tinuous outcome measures to dichotomous measures introduce the potential for bias, it necessitates using aweaker statistical measure (ORs) instead of weightedmean differences. For future research, hearing levels Submitted for Publication: August 8, 2006; final revi-
should be reported as decibels with means and stan- sion received December 1, 2006; accepted January 3, 2007.
dard deviations to permit more meaningful statistical Correspondence: Lorne S. Parnes, MD, FRCSC, Depart-
ment of Otolaryngology, University of Western On- (REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133, JUNE 2007 2007 American Medical Association. All rights reserved.
tario, 339 Windermere Rd, London, Ontario, Canada N6A 18. Cinamon U, Bendet E, Kronenberg J. Steroids, carbogen or placebo for sudden hearing loss: a prospective double-blind study. Eur Arch Otorhinolaryngol. 2001;258:477-480.
Author Contributions: Drs Conlin and Parnes had full
19. Stokroos RJ, Albers FWJ, Tenvergert EM. Antiviral treatment of idiopathic sud- access to all the data in the study and take responsibility den sensorineural hearing loss: a prospective, randomized, double-blind clini- for the integrity of the data and the accuracy of the data cal trial. Acta Otolaryngol. 1998;118:488-495.
analysis. Study concept and design: Conlin and Parnes. Ac- 20. Uri N, Doweck I, Cohen-Kerem R, et al. Acyclovir in the treatment of idiopathic quisition of data: Conlin. Analysis and interpretation of data: sudden sensorineural hearing loss. Otolaryngol Head Neck Surg. 2003;128:544-549.
Conlin. Drafting of the manuscript: Conlin. Critical revi- 21. Tucci DL, Farmer JC, Kitch RD, Witsell DL. Treatment of sudden sensorineural sion of the manuscript for important intellectual content: hearing loss with systemic steroids and valacyclovir. Otol Neurotol. 2002;23: Conlin and Parnes. Statistical analysis: Conlin. Study su- 22. Westerlaken BO, Stokroos RJ, Dhooge IJ, et al. Treatment of idiopathic sudden Financial Disclosure: None reported.
sensorineural hearing loss with antiviral therapy: a prospective, randomized, double-blind clinical trial. Ann Otol Rhinol Laryngol. 2003;112:993-1000.
Previous Presentation: This study was presented at the
23. Kubo T, Matsunaga T, Asai H, et al. Efficacy of defibrinogenation and steroid thera- Canadian Meeting of Otolaryngology Head and Neck Sur- pies on sudden deafness. Arch Otolaryngol Head Neck Surg. 1988;114:649- gery 59th Annual Meeting; June 22, 2005; St John’s, New- 24. Kronenberg J, Almagor M, Bendet E, et al. Vasoactive therapy versus placebo in the treatment of sudden hearing loss: a double-blind clinical study. Laryngoscope.
1992;102:65-68.
25. Probst R, Tschopp K, Ludin E, et al. A randomized, double-blind, placebo- controlled study of dextran/pentoxifylline medication in acute acoustic traumaand sudden hearing loss. Acta Otolaryngol. 1992;112:435-443.
1. Byl FM. Sudden hearing loss: eight years’ experience and suggested prognostic 26. Reisser CH, Weidauer H. Gingko biloba extract EGb 761 or pentoxifylline for the table. Laryngoscope. 1984;94:647-661.
treatment of sudden deafness: a randomized, reference-controlled, double blind 2. Perlman HB, Fernandez C. Experiments on temporary obstruction on the inter- nal auditory artery. Laryngoscope. 1959;69:591-613.
study. Acta Otolaryngol. 2001;121:579-584.
3. Van Dishoeck HAE, Biermann TA. Sudden perceptive deafness and viral infection.
27. Burschka MA, Hassan HAH, Reineke T, et al. Effect of treatment with Ginkgo bi- Ann Otol Rhinol Laryngol. 1957;66:963-980.
loba extract EGb 761 (oral) on unilateral idiopathic sudden hearing loss in a pro- 4. Jaffe BF. Viral causes of sudden inner ear deafness. Otolaryngol Clin North Am.
spective randomized double-blind study of 106 outpatients. Eur Arch Otorhinolaryngol. 2001;258:213-219.
5. Rowson KE, Hinchcliffe R. A virological and epidemiological study of patients 28. Gordin A, Goldenberg D, Golz A, et al. Magnesium: a new therapy for idiopathic with acute hearing loss. Lancet. 1975;1:471-473.
sudden sensorineural hearing loss. Otol Neurotol. 2002;23:447-451.
6. Veltri RW, Wilson WR, Sprinkle PM, et al. The implication of viruses in idio- 29. Nageris BI, Ulanovski D, Attias J, Tikva P. Magnesium treatment for sudden hear- pathic sudden hearing loss: primary infection or reactivation of latent viruses? ing loss. Ann Otol Rhinol Laryngol. 2004;113:672-675.
Otolaryngol Head Neck Surg. 1981;89:137-141.
30. Mann W, Beck C, Beck C. Calcium antagonists in the treatment of sudden deafness.
7. Wilson WR, Veltri RW, Laird N, et al. Viral and epidemiological studies of idio- Arch Otorhinolaryngol. 1986;243:170-173.
pathic sudden hearing loss. Otolaryngol Head Neck Surg. 1983;91:653-658.
31. Ogawa K, Takei S, Inoue Y, et al. Effect of prostaglandin E1 on idiopathic sudden 8. Wilson WR. The relationship of herpesvirus family to sudden hearing loss: a pro- sensorineural hearing loss: a double-blind clinical study. Otol Neurotol. 2002; spective clinical study and literature review. Laryngoscope. 1986;96:870-877.
9. Schuknecht HF, Kimura RS, Naufal PM. The pathology of sudden deafness. Acta 32. Mora R, Barbieri M, Mora F, et al. Intravenous infusion of recombinant tissue plasminogen activator for the treatment of patients with sudden and/or chronic 10. Schuknecht HF, Donovan ED. The pathology of idiopathic sudden sensorineural hearing loss. Ann Otol Rhinol Laryngol. 2003;112:665-670.
hearing loss. Arch Otorhinolaryngol. 1986;243:1-15.
33. Joachims HZ, Segal J, Golz A, et al. Antioxidants in treatment of idiopathic sud- 11. Woolf NK, Harris JP, Ryan AF, et al. Hearing loss in experimental cytomegalo- den hearing loss. Otol Neurotol. 2003;24:572-575.
virus infection of the guinea pig inner ear: prevention by systemic immunity. Ann 34. Suckfu¨ll M, Seidel D, Thiery J, et al. Finbrinogen and LDL apheresis in treatment Otol Rhinol Laryngol. 1985;94:350-356.
of sudden hearing loss: a randomized multicentre trial. Lancet. 2002;360:1811- 12. Rauch SD. Invited comment to: Haberkamp TJ, Tanyeri HM. Management of id- iopathic sudden sensorineural hearing loss. Am J Otol. 1999;20:587-595.
35. Topuz E, Yigit O, Cinar U, Seven H. Should hyperbaric oxygen be added to treat- 13. Conlin AE, Parnes LS. Treatment of sudden sensorineural hearing loss: I. a sys- ment in idiopathic sudden sensorineural hearing loss? Eur Arch Otorhinolaryngol.
tematic review. Arch Otolaryngol Head Neck Surg. 2007;133:573-581.
14. Berlin JA; University of Pennsylvania Meta-analysis Blinding Study Group. Does 36. Ho HG, Hung-Ching L, Min-Tsuan S, Cheng-Chien Y, Hsun-Tien T. Effectiveness blinding of readers affect the results of meta-analyses? Lancet. 1997;350:185- of intratympanic dexamethasone injection in sudden deafness patients as sal- vage treatment. Laryngoscope. 2004;114:1184-1189.
15. DerSimonian R, Larid N. Meta-analyses in clinical trials. Control Clin Trials. 1986; 37. Haberkamp TJ, Tanyeri HM. Management of idiopathic sudden sensorineural hear- ing loss. Am J Otol. 1999;20:587-595.
16. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat 38. Chandrasekhar SS. Updates on methods to treat sudden hearing loss. Oper Tech Otolaryngol Head Neck Surg. 2003;14:288-292.
17. Wilson WR, Byl FM, Laird N. The efficacy of steroids in the treatment of idio- 39. Koc¸ A, Sanisoglu O. Sudden sensorineural hearing loss: literature review on re- pathic sudden hearing loss. Arch Otolaryngol. 1980;106:772-776.
cent studies. J Otolaryngol. 2003;32:308-313.
(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133, JUNE 2007 2007 American Medical Association. All rights reserved.

Source: http://suddendeafness.files.wordpress.com/2010/01/conlin-treatment-of-sshl-meta-analysis.pdf