Chang 2010.

Study characteristics
Methods	Non‐blinded, parallel‐group randomised controlled trial with approximately 2 weeks duration of treatment and 20 days duration of follow‐up
Participants	Setting: tertiary referral centre, China, no dates provided Sample size: Number randomised: 48 Number completed: 48 Primary/secondary therapy: secondary therapy after failure of primary therapy Primary therapy: not described Participant (baseline) characteristics: age/sex (male, female)/hearing loss at start of therapy/start of treatment/start of initial treatment/start of treatment: Group I (intratympanic): 48.6 (14 to 67) y/14 m, 10 f/37.55 ± 8.32 dB HL/not reported/not reported Group II (no treatment): 51.6 (16 to 62) y/16 m, 8 f/39.31 ± 8.66 dB HL/not reported/not reported Inclusion criteria: diagnosis of ISSNHL and non‐responsive to conventional therapy within 20 days, resident in the region of Qinghai plateau (> 3000 m over sea level) Exclusion criteria: evidence of retrocochlear and central neuropathy
Interventions	General comparison: intratympanic corticosteroid therapy versus no treatment Intervention group (n = 24*): "intratympanic therapy": intratympanic injections of dexamethasone 5 mg/mL, 0.2 to 0.5 mL, 4 to 5 injections in total, 1x every 3 to 4 days Comparator group (n = 24*): "no therapy": none Use of additional interventions (common to both treatment arms): 30 mL Ginkgo biloba, 40 mg ATP, 100 U of coenzyme A intravenous and vitamin B‐complex oral daily *intention‐to‐treat analysis
Outcomes	Primary outcome measure: Change in hearing threshold with pure tone audiometry Secondary outcomes: None Primary endpoint for hearing threshold evaluation: 20 days Used PTA: 7PTA (0.125, 0.25, 0.5, 1, 2, 4, 8 kHz)
Funding sources	No information available
Declarations of interest	No information provided
Notes	—
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method of random sequence generation is not described.
Allocation concealment (selection bias)	Unclear risk	Method of concealment is not described.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Open‐label trial, no blinding.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Insufficient information to permit judgement (not mentioned).
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Insufficient information to permit judgement. For the meta‐analysis it was assumed that the number of patients analysed per treatment arm is identical to the respective randomised number of patients per treatment arm.
Selective reporting (reporting bias)	Low risk	No indication of selective reporting in the outcome parameters addressed by the review.
Other bias	High risk	No sample size calculation performed. Treatment delay from onset to secondary treatment in patients not reported.