Liu 2009.

Methods	Single center, randomised, controlled trial 1. Allocation concealment: Adequate 2. Blinding of patients: Unclear/Not stated 3. Blinding of outcome assessor: No for the primary outcome, as the level of sedation was recorded by the designated nurse responsible for drug delivery (which was not blinded). For other outcomes Unclear/Not stated‐likely no blinding. 4. A priori calculation of sample size: Yes 5. Groups treated identical other than the named intervention: Yes
Participants	Number of eligible patients: 186 Number of patients enrolled:178 Drop outs/Withdrawals: 0 Incl. Criteria: 1) Patients scheduled for out patient colonoscopy only 2) Age 18‐65 years 3) ASA class I or II Excl. Criteria: 1) ASA class III or above 2) Pregnancy 3) Allergy to any sedative agents 4) Inpatient status 5) History of colonic or rectal resection 6) Cognitive deficit or intellectual disability 6) History of difficult endotracheal intubation 7) Inability or reluctance to give consent. Country: Hong Kong Setting: Outpatient Age (median(range))(yrs): Group A: 48.5 (20‐64) Group B: 49.5 (18‐65) N (%) males: Group A: 43 (49) Group B: 55(61) Groups comparable at baseline: Yes (age, gender, ASA class, previous history of colonoscopy, indications for colonoscopy and baseline hemodynamic parameters)
Interventions	Group A (n=88): Nurse administered propofol sedation. A loading dose of 40‐60 mg or 0.8 mg/kg propofol, whichever was the higher, was given to patients 1 minute before the commencement of the procedure.. Then a PCA pump was used by the designated nurse to deliver a bolus dose of 1,5 ml mixture containing 14.3 mg propofol admixed with 35 microgram of alfentanil, with zero lock‐out for a goal OSSA score of 3 (assessed every 30 seconds). Median (range) dose; propofol 165 mg (52‐292); alfentanil 0.175 mg (0.035‐0.595) Group B (n=90): Intravenous injection of Diazemuls 0.1 mg/ Kg and Pethidine 0.5mg/ Kg per bolus 1 minute before the commencement of the procedure. Additional drugs administered as half‐dosage bolus, upon request of the endoscopist. Median (range) dose; diazemuls 5 mg (5‐10); pethidine 25 mg (25‐50) Mode of administration: Group A: PCA pump bolus; Group B: Intravenous bolus injections Administered by: Five designated nurses Goal level of sedation: Group A: OSSA 4 Gropu B: as requested by the endoscopist Supplemental oxygen: All patients received supplemental oxygen (2 L/min) through nasal cannula
Outcomes	The primary end‐point of the study was the patient's level of sedation (measured by OSSA score) during colonoscopy intubation and when the cecum was reached. Secondary outcomes: 1) Time to cecal intubation and total procedure time 2) Patient, endoscopist and nurse satisfaction scores with regard to sedation 3) Overall pain score 4) Patients' willingness to repeat colonoscopy with the same sedation 5) Complications related to sedation 6) Cost comparison
Notes	Nurses and endoscopists involved in the study were first trained by an anaesthesiologist regarding propofol delivery, patient monitoring using OSSA score and airway management. Concealment of allocation is judged to be adequate as first written informed consent was obtained and then randomizations carried out by a designated nurse, using computer generated numbers inside concealed envelopes. Definitions: 1) Patients' level of sedation, measured by OSSA score 2) Complications related to sedation: Hypotension (systolic BP < 90 mm Hg); Oxygen desaturation<90%; Bradycardia (pulse< 50/min). 3) Full recovery= Fully alert (able to calculate serial subtraction in 7s from 100), hemodynamically stable (BP and HR within 20% of baseline and oxygen saturation >90% at room air) and ambulant. 4) Patient, nurse and endoscopist satisfaction: After full recovery of the patient, satisfaction scores regarding sedation from the patient, nurse and endoscopist were documented, using a 10 cm VAS, ranging from 0 (very unsatisfied) to 10 (very satisfied) 5) Pain: as reported by the patients on a similar scale. The continuous outcome variables (procedure time, recovery time, patient satisfaction, pain score, level of sedation) were reported as median and range. As per the Cochrane collaboration handbook, ranges should not be used to calculate SDs and hence these were not included in the meta‐analyses. For the review, we extracted the level of sedation as dichotomous outcome, with "failure to sedate" considered as those with OSSA score of 5 during colonoscopy intubation.