Mintzer 2006 RIS USA 232.

Study characteristics
Methods	Study design: randomised controlled trial Study grouping: parallel group Study duration: 8 weeks Rescue medication: allowed
Participants	Number randomised: 473 (235 risperidone 238 placebo) Mean age: 83.3 years Sex (female): 77.0% Type of dementia: Alzheimer's disease, vascular dementia Severity of dementia: moderate Indication: psychosis (although 50 (included) patients with psychosis at screening were found not to have psychosis anymore at baseline) (baseline BEHAVE‐AD psychosis: 7.4) Setting: nursing homes, long‐term care facilities Country: USA
Interventions	Intervention characteristics Risperidone dose: 1.0 mg to 1.5mg per day Placebo Medication was initiated at 0.50 mg (0.25‐mg tablets twice daily) and increased after three days to 1.00 mg (0.5‐mg tablets twice daily). For patients whose clinical response remained inadequate by day 13, medication was increased to 1.50 mg (0.75 mg twice daily). Subsequent adjustments were allowed in patients who experienced adverse events. The minimum treatment dosage was 0.50 mg daily for patients who achieved and maintained efficacy at this dose. After randomisation, the mean daily dosage of risperidone was 1.03 ± 0.24 mg (median: 1.00 mg/day, range: 0.40 mg to 1.90 mg/day), with a maximum daily risperidone dosage of 2.5 mg, which exceeded trial recommendations.
Outcomes	Psychosis: Behavioural Pathology in Alzheimer's Disease (BEHAVE‐AD) psychosis subscale Number of responders for psychosis Extrapyramidal symptoms Somnolence Death Any adverse event Any serious adverse event Discontinuation (any reason) Discontinuation due to adverse events
Identification
Notes	Sponsorship source: Johnson & Johnson Pharmaceuticals
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No information
Allocation concealment (selection bias)	Low risk	"Investigators received sealed envelopes for each patient containing coded details of the treatment in this phase."
Comparability of groups (selection bias)	Low risk	(Small) baseline differences, f.i. in sex, type dementia and psychosis, but some are in favour of risperidone and others not. Only psychosis at baseline was adjusted for.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Placebo‐controlled with identical tablets, double‐blind, but no reference to blinding of participants and personel specifically.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No information
Incomplete outcome data (attrition bias) All outcomes	High risk	Drop‐out was high (>20%). Modified ITT (excluding patients without assessment after baseline). Handling of missing data not reported.
Selective reporting (reporting bias)	Unclear risk	Many subanalyses with primary outcome, but adverse events reported in more detail than usual.
Other bias	High risk	During the run‐in phase, all patients received placebo for 1 week to wash out previously used psychotropic medications. The run‐in length was reduced for patients not using psychotropic medications and for patients whose psychosis or agitation worsened.