Kay 1981.

Methods	A 4‐armed (betahistine, mexiletine, diazepam and placebo), double‐blinded, cross‐over randomised controlled trial with 28 days duration of treatment and 28 days duration of follow‐up
Participants	Location: England Setting: not reported Sample size: Number randomised: 42 allocated to 4 arms; it was not reported how many participants were allocated to the betahistine group and how many participants were allocated to the placebo group Number completed: 5 in betahistine group and 6 in placebo group Participant baseline characteristics: Age: not reported Gender: not reported Inclusion criteria: non‐treatable tinnitus Exclusion criteria: cardiovascular risk based on clinical history and electrocardiographic examination
Interventions	Intervention group: betahistine capsules, 16 mg daily initially (8 mg 2 times a day) followed by 24 mg daily (8 mg 2 times a day), 28 days Comparator group: placebo capsules, 2 capsules daily initially, followed by 3 capsules daily, 28 days Use of additional interventions: none reported
Outcomes	Change in tinnitus loudness measured using a visual analogue scale (range of 0 to more than 10; an upper limit was not reported) at 28 days Adverse effects at 28 days
Funding sources	No information provided
Declarations of interest	No information provided
Notes	The trial was stopped after the first cycle of medication because of an adverse event of mexiletine in an unrelated trial. In this first cycle, 5 test participants received betahistine and 6 control participants received a placebo.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Methods of randomisation not reported
Allocation concealment (selection bias)	Unclear risk	Methods of allocation concealment not reported
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	In the materials and methods section it was mentioned that the trial was double‐blind. However, the authors did not report who was blinded for what.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	In the materials and methods section it was mentioned that the trial was double‐blind. However, the authors did not report who was blinded for what.
Incomplete outcome data (attrition bias) All outcomes	High risk	21 out of 42 participants were excluded from the trial on medical grounds or because they quickly became non‐compliant.
Selective reporting (reporting bias)	Low risk	There is no protocol available. The outcomes listed in the materials and methods section of the article are all reported in the results section of the article.
Other bias	Unclear risk	Conflicts of interest and funding were not reported