Zachriat 2004.

Methods	Randomised controlled trial. Randomisation by throwing dice.
Participants	77 (of 83 initially recruited) patients (66.6% male, mean age 53.8 years) allocated to 3 groups Inclusion criteria were: (1) Duration of tinnitus more than 3 months (2) Absence of a treatable organic cause of tinnitus (3) Absence of Ménière's disease (4) Hearing capacity for communication within groups (5) Tinnitus disability score > 25 on TQ (6) No ongoing psychotherapy or masker treatment
Interventions	Three experimental groups: (1) Tinnitus Coping Training (TCT = 29 patients) (2) Habituation‐based treatment (HT = 31 patients) (3) Educational intervention (EDU = 23 patients) TCT comprised 11 sessions of 90 to 120 minutes duration, each group consisted of 6 to 8 patients. There was a 4‐week recess between the first and second session of TCT and HT to assess the effect of education alone, and then TCT and HT continued. HT was conducted in 5 sessions of 90 to 120 minutes (spaced over 6 months) to a group of 6 to 8 patients, where education, noise generator and counselling was conducted. Education consisted in a single session informing about the physiology and psychology of tinnitus. This session was identical to the first session for TCT and very similar to the HT one. Patients in EDU group were also offered a further treatment after 15 weeks should they wish. All groups were conducted by 5 qualified psychologist therapists. Assessments were carried out at 7 measurement periods: at pre‐treatment, post‐treatment, 6, 12 and 18 (21 for TCT) months follow up
Outcomes	The outcome measures included: (1) Self‐monitoring tinnitus diary (3 times per day during 1‐week period): loudness, hours of tinnitus awareness and subjective control of tinnitus (2) Psychometric questionnaires: TQ, Tinnitus Coping Questionnaire, QCC, QDC, JQ, a German questionnaire in changes in well‐being and adaptive behaviour (VEV), SSR, SCL‐90R and Minimal Diagnostic Interview of Psychological Disorders (DSM‐III‐R) Most variables were assessed at pre‐ and post‐treatment periods; the TQ was the only one applied at every time period Objective tinnitus parameters (pitch masking and masking measurements) were excluded from the study
Notes	The number of patients that were lost before the post‐treatment period was 6 (2 in the TCT group, 1 in HT group and 3 in EDU group). A further 2 drop‐outs (one in TCT and one in HT group) occurred at 18 months follow up (21 months for the TCT group). Total drop‐out = 8/77 = 10.38%. There were no adverse effects reported.
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Low risk	Randomisation by throwing dice

ATQ = Automatic Thoughts Questionnaire; BDI = Beck Depression Inventory; DB = double‐blind; IDCL = International Diagnostic Check‐List; ISI = Insomnia Severity Index; JQ = Jastreboff Questionnaire; LCB = Locus of Control of Behavior Scale; QCC = Questionnaire of Catastrophizing Cognitions; QDC = Questionnaire of Dysfunctional Cognitions; QS = quality score; R = randomisation; SCL‐90R = Symptom Checklist; SSR = Questionnaire of Subjective Success; STI = Structured Tinnitus Review; TCQ = Tinnitus Cognitions Questionnaire; TCSQ = Tinnitus Coping Strategies Questionnaire; TDQ = Tinnitus Disability Questionnaire; TEQ = Tinnitus Effect Questionnaire; THI = Tinnitus Handicap Inventory; THQ = Tinnitus Handicap Questionnaire; TKQ = Tinnitus Knowledge Questionnaire; TQ = Tinnitus Questionnaire; TRCS = Tinnitus Related Control Scale; TRQ = Tinnitus Reaction Questionnaire; TRSS = Tinnitus‐related Self‐Statement Scale; W = withdrawals