1. GRADE profile: intratympanic corticosteroids for Ménière's disease.
| Certainty assessment | Number of participants | Effect | Certainty | Comment | ||||||||
| № of studies | Study design | Risk of bias | Inconsistency | Indirectness | Imprecision | Other considerations | Intratympanic corticosteroids | Placebo/no treatment | Relative (95% CI) | Absolute (95% CI) | ||
| Improvement in vertigo frequency (follow‐up: range 6 months to ≤ 12 months; assessed with: AAO‐HNS 1995 Class A, B or C) | ||||||||||||
| 2 | Randomised trials | Seriousa | Not serious | Not serious | Seriousb | None | 30/31 (96.8%) | 28/29 (96.6%) | RR 1.00 (0.92 to 1.10) | 0 fewer per 1000 (from 77 fewer to 97 more) | ⨁⨁◯◯ Low | |
| Improvement in vertigo frequency (follow‐up: range ≥ 12 months; assessed with: AAO‐HNS 1995 Class A, B or C) | ||||||||||||
| 2 | Randomised trials | Seriousa | Not serious | Not serious | Seriousb | None | 31/31 (100.0%) | 26/27 (96.3%) | RR 1.03 (0.87 to 1.23) | 29 more per 1000 (from 125 fewer to 221 more) | ⨁⨁◯◯ Low | |
| Improvement in vertigo frequency: sensitivity analysis for complete/substantial improvement (follow‐up: range 3 months to < 6 months; assessed with: AAO‐HNS 1972 criteria: complete resolution of vertigo) | ||||||||||||
| 1 | Randomised trials | Very seriousc | Not serious | Seriousd | Seriousb | None | 15/18 (83.3%) | 13/18 (72.2%) | RR 1.15 (0.81 to 1.64) | 108 more per 1000 (from 137 fewer to 462 more) | ⨁◯◯◯ Very low | |
| Improvement in vertigo frequency: sensitivity analysis for complete/substantial improvement (follow‐up: range 6 months to ≤ 12 months; assessed with: AAO‐HNS Class A or B, or complete resolution) | ||||||||||||
| 3 | Randomised trials | Seriouse | Not serious | Not serious | Seriousb | None | 43/49 (87.8%) | 34/47 (72.3%) | RR 1.23 (1.01 to 1.48) | 166 more per 1000 (from 7 more to 347 more) | ⨁⨁◯◯ Low | |
| Improvement in vertigo frequency: sensitivity analysis for complete/substantial improvement (follow‐up: range > 12 months; assessed with: AAO‐HNS 1995 Class A or B) | ||||||||||||
| 2 | Randomised trials | Seriouse | Not serious | Not serious | Seriousb | None | 31/31 (100.0%) | 20/27 (74.1%) | RR 1.30 (1.02 to 1.65) | 222 more per 1000 (from 15 more to 481 more) | ⨁⨁◯◯ Low | |
| Change in vertigo (global score) (follow‐up: range 3 months to <6 months; assessed with: change from baseline in 'Gates Score'; scale from: 0 to 4) | ||||||||||||
| 1 | Randomised trials | Not serious | Not serious | Seriousf | Seriousb | Publication bias strongly suspectedg | 30 | 14 | — | MD 0.13 points lower (0.42 lower to 0.16 higher) | ⨁◯◯◯ Very low | |
| Change in vertigo (frequency) (follow‐up: range 3 months to < 6 months; assessed with: change from baseline in proportion of days with definitive vertigo episodes) | ||||||||||||
| 3 | Randomised trials | Not serious | Not serious | Not serious | Seriousb | Publication bias strongly suspectedg | 193 | 179 | — | MD 0.05 points (proportion of days affected by vertigo) lower (0.07 lower to 0.02 lower) | ⨁⨁◯◯ Low | |
| Change in vertigo (frequency) (follow‐up: range 6 months to ≤ 12 months; assessed with: change in number of episodes per month) | ||||||||||||
| 1 | Randomised trials | Serioush | Not serious | Not serious | Very seriousi | None | 11 | 9 | — | MD 0.1 episodes per month lower (0.79 lower to 0.59 higher) | ⨁◯◯◯ Very low | |
| Change in vertigo (frequency) (follow‐up: range > 12 months; assessed with: change in number of episodes per month) | ||||||||||||
| 1 | Randomised trials | Serioush | Not serious | Not serious | Very seriousi | None | 11 | 7 | — | MD 0.07 episodes per month lower (0.84 lower to 0.7 higher) | ⨁◯◯◯ Very low | |
| Serious adverse events | ||||||||||||
| 4 | Randomised trials | Not serious | Not serious | Not serious | Very seriousb,j | Publication bias strongly suspectedg | 9/296 (3.0%) | 9/204 (4.4%) | RR 0.64 (0.22 to 1.85) | 16 fewer per 1000 (from 34 fewer to 38 more) | ⨁◯◯◯ Very low | |
| Disease‐specific health‐related quality of life (follow‐up: range 3 months to < 6 months; assessed with: MDPOSI; scale from: 0 to 80, higher score = worse) | ||||||||||||
| 1 | Randomised trials | Not serious | Not serious | Not serious | Very seriousi,k | Publication bias strongly suspectedg | 14 | 30 | — | Quote: "No changes in quality of life as measured by the MDPOSI total score were observed." | ⨁◯◯◯ Very low | |
| Disease‐specific health‐related quality of life (follow‐up: range 6 months to ≤ 12 months; assessed with: AAO‐HNS Functional Level Scale; scale from: 1 to 6, higher score = worse) | ||||||||||||
| 1 | Randomised trials | Serioush | Not serious | Not serious | Very seriousi,j | None | 11 | 9 | — | MD 0.38 points lower (1.56 lower to 0.8 higher) | ⨁◯◯◯ Very low | |
| Disease‐specific health‐related quality of life (follow‐up: range > 12 months; assessed with: AAO‐HNS Function Level Scale; Scale from: 1 to 6, higher score = worse) | ||||||||||||
| 1 | Randomised trials | Serioush | Not serious | Not serious | Very seriousi,j | None | 11 | 7 | — | MD 0.45 points lower (2.03 lower to 1.13 higher) | ⨁◯◯◯ Very low | |
| Change in hearing (follow‐up: range 6 months to ≤ 12 months; assessed with: hearing threshold (dB) with pure tone audiogram) | ||||||||||||
| 1 | Randomised trials | Serioush | Not serious | Not serious | Very seriousi | None | 11 | 9 | — | MD 4.95 dB lower (16.5 lower to 6.6 higher) | ⨁◯◯◯ Very low | |
| Change in hearing (follow‐up: range > 12 months; assessed with: hearing threshold (dB) with pure tone audiogram) | ||||||||||||
| 1 | Randomised trials | Serioush | Not serious | Not serious | Very seriousi | None | 11 | 7 | — | MD 2.84 dB lower (9.61 lower to 3.93 higher) | ⨁◯◯◯ Very low | |
| Change in hearing: improvement in hearing (follow‐up: range 3 months to < 6 months; assessed with: change in PTA of > 10 dB) | ||||||||||||
| 2 | Randomised trials | Seriousl | Not serious | Not serious | Seriousb | Publication bias strongly suspectedg | 6/93 (6.5%) | 13/91 (14.3%) | RR 0.45 (0.18 to 1.15) | 79 fewer per 1000 (from 117 fewer to 21 more) | ⨁◯◯◯ Very low | |
| Change in hearing: improvement in hearing (follow‐up: range 6 months to ≤ 12 months; assessed with: change in PTA of > 10 dB) | ||||||||||||
| 2 | Randomised trials | Seriousm | Not serious | Not serious | Very seriousb,j | None | 14/38 (36.8%) | 19/38 (50.0%) | RR 0.73 (0.45 to 1.17) | 135 fewer per 1000 (from 275 fewer to 85 more) | ⨁◯◯◯ Very low | |
| Change in hearing: improvement in hearing (follow‐up: range > 12 months; assessed with: change in PTA of > 10 dB) | ||||||||||||
| 2 | Randomised trials | Seriousa | Not serious | Not serious | Very seriousb,j | None | 11/31 (35.5%) | 13/27 (48.1%) | RR 0.79 (0.46 to 1.35) | 101 fewer per 1000 (from 260 fewer to 169 more) | ⨁◯◯◯ Very low | |
| Change in tinnitus (follow‐up: range 3 months to < 6 months; assessed with: THI; scale from: 0 to 100, higher scores = worse) | ||||||||||||
| 1 | Randomised trials | Not serious | Not serious | Not serious | Seriousb | publication bias strongly suspectedg | 30 | 14 | — | MD 9.69 points lower (20.28 lower to 0.89 higher) | ⨁⨁◯◯ Low | |
| Tinnitus (follow‐up: range 6 months to ≤ 12 months; assessed with: THI; scale from: 0 to 100, higher scores = worse) | ||||||||||||
| 1 | Randomised trials | Serioush | Not serious | Not serious | Very seriousi,j | None | 11 | 9 | — | MD 1.12 points lower (24.75 lower to 22.51 higher) | ⨁◯◯◯ Very low | |
| Tinnitus (follow‐up: range > 12 months to 0; assessed with: THI; scale from: 0 to 100, higher scores = worse) | ||||||||||||
| 1 | Randomised trials | Serioush | Not serious | Not serious | Very seriousi,j | None | 11 | 7 | — | MD 6.6 points higher (7.79 lower to 20.99 higher) | ⨁◯◯◯ Very low | |
| Other adverse effects ‐ persistent tympanic membrane perforation | ||||||||||||
| 3 | Randomised trials | Not serious | Not serious | Not serious | seriousb | Publication bias strongly suspectedg | 12/203 (5.9%) | 0/117 (0.0%) | Peto OR 5.71 (1.56 to 20.96) | Not estimable | ⨁⨁◯◯ Low | |
| 1.0% | 45 more per 1000 (from 6 more to 165 more) with an assumed risk of 1% in the control group. | |||||||||||
| Other adverse effects ‐ ear pain | ||||||||||||
| 4 | Randomised trials | Not serious | Not serious | Not serious | Very seriousb,j | Publication bias strongly suspectedg | 15/295 (5.1%) | 7/205 (3.4%) | Peto OR 1.19 (0.47 to 3.04) | 6 more per 1000 (from 18 fewer to 63 more) | ⨁◯◯◯ Very low | |
| Other adverse effects ‐ post‐injection vertigo | ||||||||||||
| 3 | Randomised trials | Not serious | Not serious | Not serious | Very seriousb,j | Publication bias strongly suspectedg | 19/219 (8.7%) | 4/127 (3.1%) | Peto OR 1.78 (0.65 to 4.87) | 23 more per 1000 (from 11 fewer to 105 more) | ⨁◯◯◯ Very low | |
| Other adverse effects ‐ total hearing loss | ||||||||||||
| 2 | Randomised trials | Not serious | Not serious | Not serious | Very seriousb,j | Publication bias strongly suspectedg | 1/133 (0.8%) | 0/39 (0.0%) | Peto OR 3.47 (0.02 to 486.20) | 0 fewer per 1000 (from 0 fewer to 0 fewer) | ⨁◯◯◯ Very low | |
| 1.0% | 24 more per 1000 (from 10 fewer to 821 more) | |||||||||||
| Other adverse effects ‐ new onset of tinnitus in the affected ear | ||||||||||||
| 2 | Randomised trials | Not serious | Seriousn | Not serious | Very seriousb,j | Publication bias strongly suspectedg | 8/189 (4.2%) | 4/113 (3.5%) | Peto OR 1.09 (0.30 to 3.91) | 3 more per 1000 (from 25 fewer to 90 more) | ⨁◯◯◯ Very low | |
AAO‐HNS: American Academy of Otolaryngology ‐ Head and Neck Surgery; CI: confidence interval; MD: mean difference; MDPOSI: Ménière’s Disease Patient‐Oriented Symptom‐Severity Index; OR: odds ratio; PTA: pure tone audiometry; RR: risk ratio; THI: Tinnitus Handicap Inventory
aHigh risk of performance and detection bias in one study, and attrition bias the other study.
bSample size fails to meet optimal information size (taken as < 400 participants for continuous outcomes, < 300 events for dichotomous outcomes).
cRisk of performance bias and other bias.
dConcern over population included ‐ limited information on diagnosis of Ménière's disease. Outcome is complete resolution of vertigo, not improvement.
eHigh risk of performance and detection bias in the study with the largest weight in the meta‐analysis.
fAn unvalidated rating score was used to assess this outcome.
gWe are aware of two unpublished trials from the same pharmaceutical company that apparently showed negative efficacy results.
hRisk of attrition bias.
iExtremely small sample size.
jConfidence interval ranges from potential benefit to potential harm.
kNarrative description only, unable to provide an estimate of the effect.
lConcerns of performance bias and other bias in one trial, selective reporting bias in the other.
mRisk of performance bias in both studies, other bias in one study and detection bias in one study.
nTheI2 value is 32%; direction of effect varies between the trials.