Mira 2003.

Study characteristics
Methods	Allocation: 2 randomised lists (one for MD and one for BPPV) generated by the pharmaceutical company that supplied the drug and placebo tablets, using Fisher and Yates random number tables Design: multicentre, parallel‐group
Participants	Number: 144 randomised Age: range 18 to 65 Gender: (M:F) betahistine 33:42 placebo 27:42 Setting: 11 university hospitals Eligibility criteria: Ménière's disease (probable‐possible, AAO‐HNS (n = 81); benign paroxysmal positional vertigo (n = 63)Exclusion criteria: infections, cerebrovascular disease, drugs that act on cerebral circulation, antihistamines, calcium antagonists, anti‐aggregants, thiazide diuretics, corticosteroids and benzodiazepines, having any major medical or surgical condition likely to interfere with the absorption, distribution, metabolism or excretion of the drug used in the study or having a terminal disease Baseline characteristics: percentages of patients who had used anti‐vertigo drugs slightly higher in the betahistine group. Baseline data for dizziness scales not given
Interventions	Betahistine 16 mg twice a day for 3 months versus placebo Intervention group: n = 75 Comparator group: n = 69 Use of additional interventions: none
Outcomes	Primary outcome: number of vertigo attacks per monthSecondary outcomes: Dizziness Handicap Inventory, GISFaV self rating scale, dizziness assessment rating scale, patient and physician global assessment, adverse events
Funding sources	Pharmaceutical company funded, interest declared
Declarations of interest	Pharmaceutical company funded, interest declared
Notes	—
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"according to the random list", "Randomisation in groups of 4". 2 randomised lists (one for MD and one for BPPV) generated by the pharmaceutical company that supplied the drug and placebo tablets, using Fisher and Yates random number tables
Allocation concealment (selection bias)	Low risk	As above
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"drugs supplied in identical packages with a fantasy name"
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Double‐blind, but no further information
Incomplete outcome data (attrition bias) All outcomes	Low risk	Randomised patients all accounted for; low rate of attrition Participants lost to follow‐up: 8
Selective reporting (reporting bias)	High risk	Raw data frequently not given, only percentage change scores which are hard to interpret without baseline data. Dix‐Hallpike test results not given as an outcome for patients with benign paroxysmal positional vertigo
Other bias	Unclear risk	—