Collins 1967.

Methods	Allocation: randomised. Blindness: unclear. Duration: 10 weeks, preceded by 2 weeks wash‐out . Setting: hospital, single‐centre.
Participants	Diagnosis: chronic schizophrenia. N = 87. Age: mean ˜ 41 years. Sex: all male. History: chronic 'long‐stay' inpatient. Included: no details. Excluded: no details. Consent: not stated.
Interventions	1. Perphenazine: 4 mg/day for five weeks, then 8 mg/day for another five weeks, n = 29. 2. Placebo: one tablet daily for the first five weeks, then two tablets daily for another five weeks, n = 29. [3. Perphenazine and amitriptyline: 4 mg/day and 25 mg/day for five weeks, then 8 mg/day and 50 mg/day for another five weeks, n = 29 ‐ not included in data and analysis, combined with anti‐depressant].
Outcomes	Leaving the study early. Behaviour: Wing‐Scale B ‐ social withdrawal, socially embarrassing behaviour (SD imputed from standard error). Adverse events. Unable to use ‐ Global state: overall improvement; relative improvement (unusable data). Physical: body weight (no SD). Other side effects: nausea; dizziness; skin rash (no data reported).
Notes	*Night sedatives as needed, but no other medication administered.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Random – “allotted to treatment groups at random”.
Allocation concealment (selection bias)	Unclear risk	No details.
Blinding (performance bias and detection bias) All outcomes	Low risk	Double – quote, “neither nursing staff nor doctors concerned knew which tablets were which” (p1428).
Incomplete outcome data (attrition bias) All outcomes	Low risk	No losses ‐ N = 29 participants in each treatment group completed the trial.
Selective reporting (reporting bias)	Low risk	Not all adverse effects reported in full; not all means and SD reported.
Other bias	Unclear risk	Funding: Messrs. Allen & Hanburys Ltd supplied matching tablets of Triptafen Forte, Fentazin and placebo. Raters: not stated to be independent of treatment.