Parlak 2006.

Methods	RCT, parallel design
Participants	74 adult participants Elective and emergency procedures for participants with atrial fibrillation ASA status not given Emergency department and coronary care unit, Turkey
Interventions	< 65 years. Fentanyl (1 µ/kg iv). 3 minutes later 2 mg midazolam (1 mL = 1 mg) over 20‐30 seconds until reached RSS‐5. Then 1 mg of midazolam every 2 minutes < 65 years. Fentanyl (1 µ/kg iv). 3 minutes later 20 mg propofol (1 mL = 10 mg) over 20‐30 seconds until reached RSS‐5. Then 20 mg propofol every 2 minutes ≥ 65 years. Fentanyl (0.5 µ/kg iv). 3 minutes later 2 mg midazolam (1 mL = 1 mg) over 20‐30 seconds until reached RSS‐5. Then 1 mg midazolam every 2 minutes ≥ 65 years. Fentanyl (0.5 µ/kg iv). 3 minutes later 20 mg propofol (1 mL = 10 mg) over 20‐30 seconds until reached RSS‐5. Then 20 mg propofol every 2 minutes All groups given fentanyl citrate for preprocedural analgesia Aim: participant sedated to Ramsey Sedation Scale level 5
Outcomes	Unintended apneic episodes Patient awareness or recall Time to loss of consciousness Time to awakening Patient satisfaction
Results	< 65 years Midazolam vs < 65 years Propofol vs ≥65 years Midazolam vs ≥65 years Propofol Apnoea: 1/12 vs 1/11 vs 6/25 vs 2/22 Recall: 0/12 vs 1/11 vs 4/25 vs 1/22 Patient satisfaction: satisfied 12/12 vs 11/11 vs 23/25 vs 20/22 Unsure if satisfied: 0/12 vs 0/11 vs 2/25 vs 2/22
Notes	Four groups for 2 drug comparisons divided by age group (< 65 years, ≥ 65 years) Anaesthetic administered by 2 final year medical residents from Emergency department and Anesthetic department
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Stratification by age and then computer software to generate random numbers
Allocation concealment (selection bias)	Unclear risk	No details of how allocation was concealed
Blinding of participants and personnel (performance bias) Major adverse events	High risk	Medical resident who administered anaesthetic aware of group allocation
Blinding of participants and personnel (performance bias) Patient reported outcomes (recall/satisfaction)	High risk	Medical resident who administered anaesthetic aware of group allocation
Blinding of participants and personnel (performance bias) Success of cardioversion	High risk	Medical resident who administered anaesthetic aware of group allocation
Blinding of participants and personnel (performance bias) Minor adverse events	High risk	Medical resident who administered anaesthetic aware of group allocation
Blinding of participants and personnel (performance bias) Time to induction etc/need for resedation.	High risk	Medical resident who administered anaesthetic aware of group allocation
Blinding of outcome assessment (detection bias) Major adverse event	Low risk	Researcher who collected data was blinded to participant treatment allocation. Blinding achieved by obscuring participant's arm
Blinding of outcome assessment (detection bias) Patient reported outcomes (recall/satisfaction)	Low risk	Researcher who collected data was blinded to participant treatment allocation. Blinding achieved by obscuring participant's arm
Blinding of outcome assessment (detection bias) Success of cardioversion	Low risk	Researcher who collected data was blinded to participant treatment allocation. Blinding achieved by obscuring participant's arm
Blinding of outcome assessment (detection bias) Minor adverse events	Low risk	Researcher who collected data was blinded to participant treatment allocation. Blinding achieved by obscuring participant's arm
Blinding of outcome assessment (detection bias) Time to induction etc./need for resedation	Low risk	Researcher who collected data was blinded to participant treatment allocation. Blinding achieved by obscuring participant's arm
Incomplete outcome data (attrition bias) All outcomes	Low risk	Four participants excluded “because of difficulties in data collection”. Only 5%
Selective reporting (reporting bias)	Unclear risk	All outcomes from methods section reported Prepublished protocol not sought
Other bias	Low risk	Baseline characteristics largely comparable – given stratification by age