Parlak 2006.

Methods	Stratified, randomized, blinded trial
Participants	2 adult groups: 18‐65 years and ≥ 65 years old, requiring sedation for cardioversion
Interventions	Group 1: < 65 years: fentanyl 1 μg/kg IV, followed by midazolam 2 mg, then midazolam 1 mg every 2 minutes Group 2: < 65 years: fentanyl 1 μg/kg IV, followed by propofol 20 mg IV, then 20 mg every 2 minutes Group 3: ≥ 65 years: fentanyl 0.5 μg/kg IV, followed by midazolam 2 mg, then 1 mg every 2 minutes Group 4: ≥ 65 years: fentanyl 0.5 μg/kg, followed by propofol 20 mg IV, then 20 mg every 2 minutes
Outcomes	Level of sedation (modified RSS) Participant reactions to cardioversion (recorded using subjective scale) Participant satisfaction (questionnaire including Likert‐type questions)
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "randomization was achieved by first doing a stratification on age and then using computer software to generate random numbers"
Allocation concealment (selection bias)	Low risk	Quote: "a study nurse obtained the randomization scheme from a computer and prepared the medications"
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Comment: data for 4 participants in propofol groups not collected (4/33)
Other bias	Low risk	Comment: no obvious "other bias" identified
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "the researcher who collected the data was blinded to patient treatment allocation. Blinding was achieved by obscuring the patient's arm from the person collecting information"
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: it is unclear whether the researcher who recorded the data was blinded to the outcome measures