De Deyn 2004 F1D MC HGIV.

Study characteristics
Methods	Study design: randomised controlled trial Study grouping: parallel group Study duration: 10 weeks Rescue medication: allowed
Participants	Number randomised: 652 (520 olanzapine129 placebo) Mean age: 76.6 years Sex (female): 75% Type of dementia: Alzheimer's disease Severity of dementia: mild‐severe Indication: Psychosis (baseline Psychosis NPI‐NH: 9.7) Setting: long‐term nursing homes or continuing‐care hospitals Country: Europe, Australia, Israel, Lebanon, and South Africa
Interventions	Intervention characteristics Olanzapine dosage: fixed 1.0, 2.5, 5.0, or 7.5 mg per day Placebo Patients randomly assigned to receive olanzapine 1.0mg or olanzapine 2.5mg were respectively given a single 1.0 mg or 2.5 mg capsule of olanzapine daily (at bedtime) throughout the study period. Patients assigned to receive olanzapine 5.0 mg or olanzapine 7.5 mg began therapy on 2.5 mg/day for the first week and were titrated to their final dose by 2.5 mg/week increments. Patients unable to tolerate the assigned olanzapine or placebo dose were discontinued from the study.
Outcomes	Psychosis: Neuropsychiatric Inventory–Nursing Home version (NPI‐NH) Death Any serious adverse event Discontinuation (any reason) Discontinuation due to adverse events
Identification
Notes	Sponsorship source: Eli Lilly and Company
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	No information
Allocation concealment (selection bias)	Unclear risk	No information
Comparability of groups (selection bias)	High risk	Baseline characteristics were not reported per group and not for all randomized in the main results article (NPI). A limited set of characteristics (not the primary outcome) was reported in a clinical study report and showed clear differences in f.i. sex and smaller differences in other characteristics. Only baseline scores for outcomes were adjusted for.
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	No information
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	No information
Incomplete outcome data (attrition bias) All outcomes	High risk	Total number randomized 652 does not agree with numbers randomized per group (520 and 129).
Selective reporting (reporting bias)	High risk	Many subanalyses, for instance with the individual NPI items. Post‐hoc analyses. Each time for all four drug groups. No adjustment for multiple testing. Much missing data on adverse events (no information f.i. on somnolence, or just summary information, f.i. on cognitive function).
Other bias	High risk	"Following a placebo lead‐in phase of up to a maximum of 14 days (...)"