Seventy-two patients undergoing total shoulder arthroplasty were randomized to receive a continuous interscalene catheter with bupivacaine (n=24), a single-shot interscalene block with bupivacaine and dexamethasone plus saline infusion (n=24), or a single-shot interscalene block with liposomal bupivacaine plus saline infusion (n=24). The primary outcome was the worst numeric rating scale (NRS) pai...