Liu 2009a.

Methods	RCT, parallel design
Participants	230 patients, scheduled to undergo outpatient shoulder arthroscopy under interscalene block Excluded: < 18 yrs, > 75 yrs, patient refusal, pregnancy, dementia, severe pulmonary disease, and known pre‐existing neurological disorders involving the operative limb
Interventions	Ultrasound (n = 115) vs nerve stimulation group (n = 115) Interscalene block of axillary, musculocutaneous, ulnar, radial and median nerves with mepivacaine 1.5% with 1:300,000 epinephrine and NaCO3 (1 meq/10mL); for participants < 50 kg, total dose of 45 ‐ 55 ml was used for patients ≥ 50 kg, total dose of 55 ‐ 65 mL was used Ultrasound: A linear 10 ‐ 13 MHz US probe was used to visualize the brachial plexus. Initial US visualization was at interscalene area. 5 cm 22 G needle placed through middle scalene muscle, into interscalene groove, in‐plane US guidance to visualize the entire needle. LA injected in divided doses with frequent aspiration under ultrasound visualization. Nerve stimulation (Stimuplex): 5 cm 22 G insulated needle placed into interscalene groove. Current decreased to range between 0.2 mA and 0.5 mA. If not still present then needle adjusted. LA injected in divided doses with frequent aspiration
Outcomes	Supplementation rate ("required conversion to general anaesthesia") Time to perform block Number of needle passes Complications
Notes	No details of experience of anaesthetists giving nerve blocks
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated random number table
Allocation concealment (selection bias)	Unclear risk	Used a “sealed envelope sequence”. No further details
Blinding of participants and personnel (performance bias):adequacy of block, need for supplementation	High risk	Not possible to blind anaesthetist
Blinding of participants and personnel (performance bias):Complications	High risk	Not possible to blind anaesthetist
Blinding of participants and personnel (performance bias): time outcomes	High risk	Not possible to blind anaesthetist
Blinding of outcome assessment (detection bias): adequacy of block, supplementation of block	Low risk	Adequacy of block assessed by an investigator who was unaware of group allocation
Blinding of outcome assessment (detection bias): Complications	Low risk	Neurological complications assessed and analyzed by blinded investigator
Blinding of outcome assessment (detection bias): time outcomes	Unclear risk	Time to perform block assessed by an investigator not performing block but no details as to whether they were blinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	Some losses to follow‐up but relatively small number and details given
Selective reporting (reporting bias)	Unclear risk	Pre‐published protocol not sought
Baseline characteristics	Low risk	All comparable
Funding sources	Low risk	None apparent
Operator expertise	Unclear risk	No details of who, how many and experience of anaesthetists giving blocks