Desmet 2015.

Methods	Parallel group RCT.
Participants	In Belgim, 120 participants aged 18 years and older undergoing shoulder rotator cuff repair or subacromial decompression with interscalene brachial plexus block were included. Participants less than 18 years old, those with diabetes, brachial plexus neuropathies, severe bronchopulmonary disease, systemic glucocorticoid use, or pregnancy were excluded.
Interventions	Block All participants received nerve‐stimulator/ultrasound‐guided interscalene block with 0.5% ropivacaine. Dexamethasone/placebo Dexamethasone group: dexamethasone 10 mg intravenously. Placebo group: normal saline intravenously. Intraoperative anaesthesia/analgesia Oral lorazepam 2.5 mg 1 hour before surgery + intravenous midazolam 2 mg and sufentanil 2‐5 micrograms before block placement. General anaesthesia was induced with target‐controlled propofol infusion 3‐5 micrograms/ml, remifentanil (loading dose 1 microgram/kg, continuous infusion 0.05‐0.3 microgram/kg/min) and cisatracurium 0.5 mg/kg. Postoperative analgesia Paracetamol for moderate or severe pain or on participant request to a maximum 4 grams/24 hours supplemented with diclofenac to a maximum of 100 mg/24 hours and intramuscular piritramide 15‐20 mg.
Outcomes	Outcomes of interest for the review Duration of sensory block defined by the interval between the time block was done and the time to first analgesia request. Arm weakness at 24 hours. Incidence of sleep disturbance, postoperative nausea and vomiting. Satisfaction. Other outcomes Number of participants with no/mild pain at 24 and 48 hours.
Notes	This was a four‐arm study which included a placebo arm and three doses of dexamethasone: 1.25 mg. 2.5. mg and 10 mg. In order to avoid unit of analysis issues, we chose to include the dexamethasone 10 mg arm and placebo and exclude the other arms. Funding: supported through a grant from the Belgian Association for Regional Anesthesia. Conflicts of interest: none.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random sequence was computer‐generated.
Allocation concealment (selection bias)	Low risk	Allocation was concealed in sealed, opaque envelopes.
Blinding of participants (detection bias)	Unclear risk	No indication whether participants were blinded.
Blinding of personnel (detection bias)	Low risk	Assume operating room personnel were blinded since study drugs prepared by staff member not involved the study and delivered in unidentifiable syringes, however, no indication whether other personnel were blinded (surgeon, recovery room and ward nurses) was blinded.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Outcome assessors were blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Only one participant was lost to follow‐up.
Selective reporting (reporting bias)	Low risk	All outcomes reported as stated in the methods section.
Other bias	Low risk	Appears to be free of any other bias.