van Eijk 2010.

Methods	RCT comparing rivastigmine vs placebo as an adjunct to haloperidol for treatment of delirium Study took place in 6 ICUs in the Netherlands
Participants	Participants included 104 (N = 54 rivastigmine, mean age 68.0 ± 11.4 years, 38/54 (70%) male; N = 50 placebo, mean age 70.0 ± 12.2 years, 29/50 (58%) male) critically ill patients diagnosed with delirium according to the CAM‐ICU and expected to remain in the ICU for at least 48 hours Study enrolment between November 2008 and January 2010
Interventions	Participants received rivastigmine or placebo twice daily. Rivastigmine was delivered in a 2‐mg/mL solution The dosing regimen for rivastigmine was as follows Study days 1 to 3: 0.75 mL twice daily Study days 4 to 6: 1.5 mL twice daily Study days 7 to 9: 2.25 mL twice daily Study days 10 onward: 3.0 mL twice daily. Once delirium was resolved or participants were discharged from hospital, the dose regimen was reversed and study drug was tapered off over 3 days. If a possible side effect occurred during treatment, study drug was reduced until the side effect was resolved, or was stopped if the side effect persisted for longer than 3 days. Participants with persistent side effects were followed until an endpoint was reached (end of delirium, discharge from hospital, or death) Delirium was assessed daily using the CAM‐ICU until 3 days after study drug cessation. The CAM was used if the patient was discharged to a regular ward. All participants received usual care including frequent orientation, physical therapy, and exercise. Participants ≥ 70 years of age received 1 mg intravenous haloperidol 3x/d, and those aged ≤ 69 years received 2.5 mg intravenous haloperidol 3x/d. Participants were allowed to receive 1 mg intravenous lorazepam at night (22:00). The treating physician could adjust treatment with haloperidol or a benzodiazepine, similar to usual care. Rescue haloperidol (2.5 mg if ≥ 70 years, and 5 mg ≤ 69 years) was recommended in the event of persistent agitation and was repeated every 30 minutes if needed. If haloperidol proved ineffective, 1 mg/kg per hour intravenous propofol was administered. In the event that propofol was contraindicated, 5 mg per hour intravenous midazolam was used instead. The dose of propofol or midazolam was increased until the participant was calm but was tapered every 12 hours thereafter. Study drug was continued during sedation
Outcomes	Primary Duration of delirium during hospital admission (i.e. in the ICU and hospital wards combined) Secondary Percentage of fixation days (i.e. proportion of study days on which patient was restrained by arms, legs, or both) Number of self‐removed catheters Severity of delirium Use of psychoactive drugs ICU and hospital length of stay Other outcomes Total dose of study drug Number of study days Mortality during treatment and at 90‐day follow‐up
Notes	Funding ZonMw, the Netherlands Brain Foundation provided funding Novartis supplied the study drug, information about the study drug, and empty bottles for placebo None of the funding sources had any role in the design or conduct of the study Registration Study was registered at clinicaltrials.gov under NCT00704301 Study authors were not contacted
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Participants were randomized in a 1:1 ratio. The randomization sequence was computer generated by the trial pharmacist and was stratified by study centre. The leading pharmacist held a list of study codes, which could be broken at any time if deemed necessary by the treating physician
Allocation concealment (selection bias)	Low risk	The trial pharmacist consecutively numbered the bottles according to the randomization sequence to conceal allocation of the next participant in the sequence
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Once eligibility of participants was confirmed, the investigator used bottles of the study drug consecutively to mask every patient and families, medical staff, and investigators from treatment allocation. All centres received batches of 10 identical bottles, 5 of which contained a solution of the study drug and 5 of which contained a placebo solution. The solutions had identical colour, smell, taste, and viscosity
Blinding of outcome assessment (detection bias) All outcomes	Low risk	This was a double‐blind study
Incomplete outcome data (attrition bias) All outcomes	Low risk	5 participants were withdrawn from the study by their families (1 on rivastigmine and 4 on placebo), leading to a modified intention‐to‐treat analysis of 54 participants on rivastigmine and 50 on placebo. Data were censored for 16 participants who died and for 19 participants who were discharged from hospital while still delirious
Selective reporting (reporting bias)	Low risk	All outcomes reported in methods were included in results. Data presented for all participants were included in the analysis
Other bias	Low risk	All included participants received usual care, which included frequent orientation, physical therapy, and exercise. The treating physician could adjust treatment with haloperidol or a benzodiazepine, similar to usual care. In case of persistent severe agitation, rescue haloperidol was recommended ZonMw and the Netherlands Brain Foundation did not contribute to study design, data collection, data analysis, data interpretation, writing of the report, or the decision to submit for publication. Novartis had no role in the decision to conduct the study or to stop it early; nor in study design, data collection, data analysis, data interpretation, writing of the report, or the decision to submit for publication

CAM‐ICU: Confusion Assessment Method for the ICU.

DI: Delirium Index.

D5W: 5% dextrose in water.

ICDSC: Intensive Care Delirium Screening Checklist.

ICU: intensive care unit.

IQR: interquartile range.

MAAS: Motor Activity Assessment Scale.

QTc: measure of time between start of the Q wave and end of the T wave in the heart's electrical cycle corrected for heart rate.

RASS: Richmond Agitation‐Sedation Scale.

RCT: randomized controlled trial.

REB: research ethics board.

SAILS: Statins for Acutely Injured Lungs from Sepsis trial.

SAS: Riker Sedation‐Agitation Scale.

SOFA: Sequential Organ Failure Assessment.