B103.

Methods	Double‐blinded, 3 arm, parallel‐group randomised controlled trial 13 week treatment followed by 2 weeks of washout with placebo with an optional double‐masked extension
Participants	Setting: Europe and UK; 54 centres, between March 1991 and March 1992 Sample size: 402 participants (226 female, 176 male), 133 in the 6 mg/day group, 136 in the 4 mg/day group and 133 in the placebo group Age: range 50 to 90 years, mean age 69.4 years Inclusion criteria: Age 50 to 90 years Had a diagnosis of DSM‐III for mild to moderate dementia, NINCDS‐ADRDA criteria for probable AD, MMSE score of at least 16 points and able to perform 3 out of 4 other tests of the psychometric battery Medication for non‐cognitive aspects of AD or concomitant conditions was allowed Exclusion criteria: cognitive enhancing medications were discontinued for 3 weeks before entry
Interventions	1. Rivastigmine: 4 mg/day divided into 2 doses (titrated to target dose in 1 week) 2. Rivastigmine: 6 mg/day divided into 2 doses (titrated to target dose in 3 weeks) 3. Placebo (identical) taken twice daily Doses maintained for 10 weeks after titration period to week 13. All patients then had a 2 week washout period with placebo (single‐blinded)
Outcomes	Outcomes measured at baseline and at 13 weeks 1. Cognitive function Mini Mental State Examination (MMSE) Fuld Object‐Memory Test (OME) Benton Visual Retention Test (VRT) Trail Making Test (TMT) Digitial symbol substitution test (DSST) Nurses' Observation Scale for Geriatric Patients (NOSGER) 2. Activities of daily living Nurses' Observation Scale for Geriatric Patients (NOSGER) Performace of three individual activities of daily living 3. Behavioural symptoms Nurses' Observation Scale for Geriatric Patients (NOSGER) 4. Physician rated global impression tests Clinical Global Impression of Change (CGIC) 5. Incidence of adverse events reported as incidence of most frequent events: nausea, vomiting, diarrhoea, abdominal pain, headache and dizziness 6. Discontinuation total withdrawals withdrawal due to adverse events
Source of funding	Novartis Pharma Ltd
Declaration of interest	Study sponsored by Novartis Pharma
Notes	Primary hypothesis:to assess short term (3 month) symptomatic efficacy of rivastigmine 4 and 6 mg/d compared with placebo in patients with AD
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Patients were assigned a randomisation number by the investigator in chronological order according to a list generated by study sponsor (Novartis)
Allocation concealment (selection bias)	Low risk	Method not described
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Active medication and placebo capsules had the same physical appearance, and the number of capsules for each dose intake was the same in all three groups
Blinding of outcome assessment (detection bias) All outcomes	Low risk
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	A total of 346/402 (86%) patients completed study. Analyses done with ITT population with imputations for missing values and an observed case population
Selective reporting (reporting bias)	Low risk	Outcomes listed in protocol were reported