Fruensgaard 1978.

Methods	Allocation: randomised. Blindness: double. Duration: 3 weeks (for acute) and 12 weeks (for chronic). Setting: multi‐centre, Denmark.
Participants	Diagnosis: (group I) acute schizophrenia and “psychogenic (reactive) psychosis”, n = 22; (group II) chronic schizophrenia, n = 25. N = 47. Age: mean ˜ 37 years. Sex: 15 F, 32 M. History: acute and chronic. Included: no details. Excluded: known hypersensitivity to dibenzoxazepine compounds or butyrophenones, manic‐depressive illness, electro‐convulsive therapy or insulin coma or sub‐coma therapy within the preceding 8 weeks, alcoholism or drug dependence as a significant feature of clinical history, serious impairment of renal or hepatic function, a history of increased intra‐ocular pressure or history of narrow‐angle glaucoma or urinary retention, cardiovascular or metabolic disease (severe cases), pregnancy. Consent: not stated.
Interventions	1. Perphenazine: 8 mg per capsule, initially 2 capsules daily, could be increased to a maximum of 15 capsules daily according to the therapeutic efficacy and side effects, n = 24*. 2. Loxapine: 10 mg per capsule, initially 2 capsules daily, could be increased to a maximum of 15 capsules daily according to the therapeutic efficacy and side effects, n = 23**. Additional medication: chloralodol for severe insomnia, antiparkinsonian drugs (orphenadrine and/or biperiden) as necessary.
Outcomes	Leaving the study early. Global state: CGI – dichotomised data, no better or deterioration. Adverse events. Laboratory tests (no data). Unable to use ‐ Mental state: BPRS end score (no mean or SD). Behaviour: NOSIE end score (no mean or SD). Laboratory tests (not all data reported).
Notes	*Perphenazine mean dose for participants with acute schizophrenia = 34 mg after three weeks of treatment; mean dose for participants with chronic schizophrenia = 90.1 mg after two months of treatment. **Loxapine mean dose for participants with acute schizophrenia = 54.4 mg after three weeks of treatment; mean dose for participants with chronic schizophrenia = 81.1 mg after two months of treatment.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	“Randomly assigned” – no further details.
Allocation concealment (selection bias)	Unclear risk	Not stated.
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Double‐blind stated – no further details.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	21% (n = 10) dropped‐out of the study and excluded from analysis – n = 5 from perphenazine group and n = 5 for loxapine group (reasons only provided for n = 2 ‘inadequate therapeutic response, and n = 2 aversion to continued treatment). Not fully addressed (ITT used).
Selective reporting (reporting bias)	High risk	Not all adverse event data reported; no means or SDs reported for BPRS, CGI, NOSIE.
Other bias	Unclear risk	Funding: no details. Raters: not stated to be independent of treatment.