Lemone 1996.

Methods	1. Randomized 2. Double Blind 3. 4 Parallel groups: ‐Buspirone 30mg/d steady dose after first week maintained for 2 weeks and withdrawn abruptly and replaced with placebo for 4 weeks ‐Clobazam 30mg with abrupt withdrawal followed by placebo for 4 weeks ‐Clobazam 30mg/d maintained for 2 weeks and withdrawn over 3 week period replaced with placebo for 1 week ‐Lorazepam 3mg after first week, maintained for 2 weeks and withdrawn over 3 week period and replaced with placebo for 1 week 4. Duration: 7 weeks active and withdrawal and one week observation for full study period of 8 weeks 5. Analysis: LOCF
Participants	1. Diagnosis: GAD (DSM‐III) 2. N = 128 3. Age (mean, SD): 39.9 (11.1) Sex: not stated Setting: outpatients History: excluded comorbid psychiatric or medical disorders
Interventions	1. Buspirone (N = 33) 2. Lorazepam (N = 33) 3. Clobazam (abrupt withdrawal) (N = 30) 4. Clobazam (slow withdrawal) (N = 27)
Outcomes	1. dropout rates 2. HAM‐A 3. Tyrer Scale
Notes	(a) HAM‐A data not included as end point means and SDs not included
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment (selection bias)	Unclear risk	B ‐ Unclear