Gwynne 1962.

Methods	Allocation: randomly assigned, no further details. Blindness: double (identical capsules). Design: 3 parallel groups, single centre. Duration: 16 weeks (preceded by at least 1 month washout). Analysis: intention‐to‐treat not performed in the trial. Country: USA.
Participants	Diagnosis: schizophrenia, no further details. N=78. Age: mean 49 years. Sex: M 39, F 39. Setting: hospital. History: chronic illness (average time in hospital 20 years); withdrawal for at least 1 year; poor response to previous therapy.
Interventions	1. Trifluoperazine: range 10‐40 mg/day, mean at week‐16, 11.2 mg. N=26. 2. Chlorpromazine: range 100‐400 mg/day, mean at week‐16, 146mg. N=26. 3. Placebo: N=26.
Outcomes	Leaving the study early. Side effects. Unable to use ‐ Mental state: (Lorr Multidimensional Scale for Rating Psychiatric Patients ‐ only "p" value).
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment?	Unclear risk	B ‐ Unclear