Wang 2008a.

Methods	Allocation: randomised. Blindness: no details. Duration: 8 weeks. Setting: not stated.
Participants	Diagnosis: schizophrenia (CCMD‐3). N = 64. Age: risperidone: 65±2 years; perphenazine: 64 ± 3 years. Sex: F29, M35 (risperidone: M18, F16; perphenazine: M17, F13) History: length of illness: risperidone: 0.9 ± 1.4; perphenazine: 1.2 ± 1.4 years. Included: schizophrenia; age > 60; no history of taking antipsychotic drugs; PANNS > 60. Excluded: severe physical disease; drug abuse. Consent: not stated.
Interventions	1. Perphenazine: start dose 2‐4mg/d, maximum dose 30mg/d, 23.5 ± 1.3mg/d, n = 30. 2. Risperidone: 5mg tablet, start dose 0.25‐0.5mg/d, maximum dose 3.5mg/d, 2.2 ± 1.2mg/d, n = 34. No other combinations permitted.
Outcomes	PANSS (marked improvement => 60 reduction; improved => 40%; no effect =< 40%). TESS ‐ adverse effects. Unable to use ‐ PANSS score – no means/ SD. TESS scores – no means/ SD. Not all adverse events reported by group.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	‘Randomised’ – no further details.
Allocation concealment (selection bias)	Unclear risk	No details.
Blinding (performance bias and detection bias) All outcomes	Unclear risk	No description.
Incomplete outcome data (attrition bias) All outcomes	High risk	Not all outcome measures reported; adverse events not reported fully by group.
Selective reporting (reporting bias)	Unclear risk	As above.
Other bias	Unclear risk	No details of funding/ who administered rating scales.