In this factorial design study, 280 participants having outpatient, arthroscopic shoulder surgery will be randomized into 4 equal sized groups. All participants will receive a standardized interscalene brachial plexus block and 4mg or 8mg of dexamethasone given by the intravenous or perineural (by the nerve with the nerve block) route just prior to their surgery. The purpose of this study is to determine which dose and route of dexamethasone provides the longest duration of pain control and the fewest side effects after surgery. The investigators hypothesize that giving dexamethasone by the perineural route, and by higher doses, will result in the longer durations of pain control, without increased side effects.

BACKGROUND & RATIONALE Interscalene brachial plexus blockade (ISB) has emerged as the choice method of analgesia for arthroscopic shoulder surgery performed on an outpatient basis. The ISB is typically administered as a single injection of local anesthetic prior to surgery. Benefits include reduced pain, opioid consumption and postoperative nausea and vomiting on the first postoperative day. However, the limited duration of analgesia with a single injection of local anesthetic, typically 12 to15 hours, is inadequate for many patients. Recent interest in prolonging the analgesic duration of single injection ISB has been focused on the addition of various adjuncts to the local anesthetic, with varying degrees of success.

Dexamethasone is the most promising of these adjuncts, with both perineural and intravenous administration consistently resulting in increased analgesic duration, by as much as two-fold. Nevertheless, it would seem prudent to administer perineural adjuncts at the lowest effective dose and only if they offer significant benefit over other routes of administration.

The multiple small clinical studies published to date have used dexamethasone doses between 4 and 10mg for both intravenous and perineural routes. Attributing differences in results between any of these studies to the dose and route of dexamethasone administered is not possible due to differences in drugs, doses, study populations and methodology. Most importantly, no study has compared two different doses of dexamethasone and the only two studies that compared the intravenous and perineural routes arrived at different conclusions. Thus, the effect of dexamethasone dose and route on the analgesic duration of interscalene block remains unclear, with multiple recent editorials and meta-analyses calling for further study. In light of these gaps in the evidence, a trial elucidating the effect of dose and route of administration of dexamethasone on analgesic duration of ISB is required. This is the focus of our proposed study.

OBJECTIVES AND HYPOTHESES In this factorial design study of outpatients undergoing elective arthroscopic shoulder surgery with interscalene brachial plexus block (ISB), the investigators will compare perineural and intravenous administration of dexamethasone at high (8mg) and low (4mg) doses. The investigators hypothesize that high doses and the perineural route will provide significantly longer duration of analgesia, without significant statistical interaction between dose and route of administration.

METHODS This single-centre, factorial design, double-blinded, randomized controlled superiority trial has four parallel groups and 1:1:1:1 randomization. Consenting and eligible adult ambulatory patients undergoing arthroscopic shoulder surgery will receive preoperative, ultrasound guided ISB with 30 millilitres (mL) of 0.5% bupivacaine and 4mg or 8mg of preservative free dexamethasone either intravenous or perineural. The remainder of the intraoperative and postoperative care is at the discretion of the attending anesthesiologist and surgeon. Based on previous work, with 70 patients per group, the study is at least 90% powered to detect a 3 hour difference in analgesia for dose or route, and a 4 hour synergistic interaction, even with a 5% attrition rate. Outcomes will be assessed by chart review and telephone follow up on postoperative day one, postoperative day 2 (if necessary), and postoperative day 14.

An interim analysis of group standard deviations only will occur after 100 patients have completed the study to determine if the original power calculations were accurate, and if testing for the interaction effect is feasible within the planned sample size of 280 participants. If testing for the interaction of dose and route is not feasible, a new study sample size will be determined, based on the number of participants needed to have 90% power to test the two main effects.

The primary outcome analysis will analyze by intention to treat only those patients who were randomized and did receive an attempt at an interscalene block. A secondary analysis will exclude patients who had a "failed" inter scalene block in the post anesthesia care unit. A tertiary analysis will be a multivariable analysis adjusted for demographics, preoperative naproxen use, use of general anesthesia, cumulative analgesic use and failed inter scalene block.

Inclusion Criteria:

• Elective ambulatory surgery patients undergoing arthroscopic shoulder surgery
• Including rotator cuff repair
• Stabilization procedures
• Acromioplasty
• Debridement and distal clavicle excision

Exclusion Criteria:

• Patient refusal, diabetes
• Pregnancy
• Coagulopathy significant enough to be a contraindication to regional anesthesia as determined by the attending anesthesiologist
• Sensitivity to local anesthetics or dexamethasone
• Severe chronic obstructive pulmonary disease
• Contralateral vocal cord paralysis
• Contralateral diaphragmatic paralysis
• Surgical limb brachial plexus neuropathy
• Interscalene block site infection
• Systemic glucocorticoids in the last 2 weeks
• Epidural or intraarticular steroid injection in the past 3 months
• Chronic opioid use defined as daily use for the last two weeks
• International normalized ratio (INR) > 1.5
• Active peptic ulcer disease
• End-stage renal disease
• Cirrhotic liver disease
• Previous participation in the study.