De Deyn 2005.

Study characteristics
Methods	Study design: randomised controlled trial Study grouping: parallel group Study duration: 10 weeks Rescue medication: allowed for NPS, not for EPS
Participants	Number randomised: 208 (106 aripiprazole 102 placebo) Mean age: 81.5 years Sex (female): 72% Type of dementia: Alzheimer's disease Severity of dementia: moderate Indication: Psychosis (Baseline Psychosis NPI Psychosis: 12.4) Setting: assisted living facilities, adult communities or living with a caregiver Country: USA
Interventions	Intervention characteristics Aripiprazole dosage: flexible, 2 mg to 15 mg per day Placebo Eligible patients were randomised to aripiprazole 2 mg/d or placebo, administered once daily for 10 weeks, in this multicentre, double‐blind study. Aripiprazole could be titrated to higher doses (5, 10 mg, and 15 mg/day) at 2‐week intervals (or more rapidly based on investigator’s judgement) if the patient showed insufficient clinical response. At end point, the mean daily dose of aripiprazole was 10.0 mg.
Outcomes	Psychosis: Neuropsychiatric Inventory (NPI) Psychosis subscale Extrapyramidal symptoms Somnolence Death Any serious adverse event Discontinuation (any reason) Discontinuation due to adverse events Cognitive function: Mini‐Mental State Examination (MMSE)
Identification
Notes	Sponsorship source: Bristol‐Myers Squibb and Otsuka Pharmaceutical Co., Ltd.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No comments
Allocation concealment (selection bias)	Unclear risk	No comments
Comparability of groups (selection bias)	High risk	Judgement Comment: No baseline information in main article. A clinical study report provides data on age, sex, race and weight only: small differences that might or might not be in favour of the drug. They are not adjusted for. Baseline scores for outcomes are given for analysed (not all randomised) patients only, but are adjusted for.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	No comments
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No comments
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Judgement Comment: Drop‐out was just below 20% and did not differ >5% between groups.It seems that not all patients were included in the analyses (see table with results on outcomes). Missing data were imputed with LOCF.
Selective reporting (reporting bias)	High risk	Insufficient information about outcomes in the protocol. Lots of secondary outcomes.
Other bias	High risk	"Following screening and a minimum 7‐day washout period for previous psychotropic medication (...)"