Summary of findings 1. Betahistine compared to placebo/no treatment for Ménière’s disease.

Betahistine compared to placebo/no treatment for Ménière’s disease
Patient or population: Ménière’s disease Setting: outpatients Intervention: betahistine (total daily dose ranging from 24 mg to 144 mg) Comparison: placebo/no treatment
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo/no treatment	Risk with betahistine
Improvement in vertigo frequency Assessed with: self‐rated improvement in either frequency or severity of vertigo Follow‐up: range 6 months to ≤ 12 months	Study population		RR 1.50 (0.98 to 2.29)	70 (1 RCT)	⊕⊝⊝⊝ very low1,2,3,4,5	The evidence is very uncertain about the effect of betahistine on improvement in vertigo frequency at 6 to ≤ 12 months.
	457 participants per 1000 would report that their vertigo had improved	686 participants per 1000 would report that their vertigo had improved (from 448 to 1000)
Improvement in vertigo frequency Assessed with: AAO‐HNS 1995 class A, B or C Follow‐up: range > 12 months	Study population		RR 1.11 (0.93 to 1.32)	62 (1 RCT)	⊕⊝⊝⊝ very low4,5,6,7	The evidence is very uncertain about the effect of betahistine on improvement in vertigo frequency at > 12 months.
	844 participants per 1000 would report that their vertigo had improved	937 participants per 1000 would report that their vertigo had improved (from 785 to 1000)
Vertigo global score Assessed with: geometric mean of monthly imbalance score (range 0 to 63, higher scores = worse symptoms) Follow‐up: range 3 months to < 6 months	The mean vertigo global score was 6.2 points	MD 0.7 points higher (6.67 lower to 8.07 higher)	—	34 (1 RCT)	⊕⊝⊝⊝ very low4,8,9	The evidence is very uncertain about the effect of betahistine on change in vertigo (using a global score) at 3 to < 6 months.
Change in vertigo frequency Assessed with: number of attacks per month Follow‐up: range 3 months to < 6 months	The mean vertigo frequency was 4.68 attacks per month	MD 1.90 attacks per month lower (3.05 lower to 0.74 lower)	—	117 (2 RCTs)	⊕⊝⊝⊝ very low4,5,10,11	The evidence is very uncertain about the effect of betahistine on change in vertigo (using the frequency of attacks) at 3 to < 6 months.
Change in vertigo frequency Assessed with: average number of attacks in 30 days Follow‐up: range 6 months to ≤ 12 months	The mean vertigo frequency was 3.084 attacks per 30 days	MD 0.63 attacks per 30 days higher (4.07 lower to 5.33 higher)	—	214 (1 RCT)	⊕⊝⊝⊝ very low4,9,12	The evidence is very uncertain about the effect of betahistine on change in vertigo (using the frequency of attacks) at 6 to ≤ 12 months.
Serious adverse events	Study population		RR 1.20 (0.63 to 2.29)	220 (1 RCT)	⊕⊝⊝⊝ very low4,9,12	The evidence is very uncertain about the effect of betahistine on serious adverse events.
	149 per 1000	178 per 1000 (94 to 340)
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; RCT: randomised controlled trial; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

¹High risk of bias for 5 domains in this study, and unclear risk of bias for remaining 2 domains.

²The criteria used for the diagnosis of Ménière's disease were poorly defined, therefore the population may not be appropriate.

³This outcome was reported as an improvement in either frequency or severity of attacks, not only frequency.

⁴Optimal information size was not reached (taken as < 300 events for dichotomous outcomes or < 400 participants for continuous outcomes, as a rule of thumb).

⁵Confidence interval ranges from a likely trivial effect to potential benefit.

⁶Unclear risk of bias for several domains, and high risk of bias due to differential use of intratympanic steroids in the intervention and control group.

⁷All participants also received intratympanic dexamethasone injections throughout the trial.

⁸Multiple domains at unclear risk of bias leading to an overall concern about the risk for this trial.

⁹Confidence interval ranges from potential harm to potential benefit.

¹⁰Multiple bias domains rated at unclear risk of bias. High risk of selective reporting bias due to incomplete outcome data for this result.

¹¹Numeric data used in this analysis were estimated due to incomplete reporting in the article.

¹²High risk of attrition bias, and potential for selective reporting.