Celecoxib significantly reduces opioid use after shoulder arthroplasty

The opioid crisis has illuminated the risks of opioid use for pain management, with renewed interest in reducing opioid consumption after common orthopedic procedures. Anti-inflammatory medication is an important component of multimodal pain management for patients undergoing orthopedic surgery. The purpose of this study was to evaluate the effect of celecoxib on pain control and opioid use after shoulder surgery.

Patients scheduled for either total shoulder replacement (group 1) or rotator cuff repair (group 2) were candidates for the study. The exclusion criteria included allergy to celecoxib, coagulopathy, use of anticoagulants, baseline use of long-acting opioids, and a history of medical conditions such as myocardial infarction or stroke. Consenting patients were randomized by type of procedure using block randomization to receive either placebo or celecoxib 1 hour prior to the procedure and for 3 weeks postoperatively. The primary outcome measure assessed was opioid utilization as measured by morphine-equivalent dose (MED). Secondary outcome measures included pain scores at 3 and 6 weeks postoperatively. Data were analyzed using multiple linear regression.

Of 1081 patients scheduled for either total shoulder replacement or rotator cuff repair from February 2014 to February 2018, 78 were enrolled for arthroplasty (group 1, with 39 receiving celecoxib and 39 receiving placebo) and 79 were enrolled for rotator cuff repair (group 2, with 40 receiving celecoxib and 39 receiving placebo). Compared with the placebo arm, patients prescribed celecoxib took fewer MEDs by -168 (95% confidence interval [CI], -272 to -64; P < .01) at 3 weeks in the total population and by -197.7 (95% CI, -358 to -38; P = .02) in the arthroplasty group. Similarly, at 6 weeks, total MEDs used was -199 (95% CI, -356 to -42; P < .01) in the total population and -270 (95% CI, -524 to -16; P = .04) in the arthroplasty group. No statistically significant differences in opioid consumption were found between study arms in the cuff repair group, at either 3 or 6 weeks. Of note, preoperative opioid use was statistically associated with higher levels of opioid use in the total population and group 1 at 3 and 6 weeks (P < .01 for all) but not in group 2 (P > .05 for both).

Use of morphine equivalents was statistically significantly less at 3 and 6 weeks in patients who took celecoxib in the total population and in the arthroplasty group. Patients prescribed celecoxib for 3 weeks after shoulder surgery took less opioid medication for pain at 3 and 6 weeks. Multimodal pain control using celecoxib is an effective way to reduce postoperative opioid use in shoulder arthroplasty patients. Preoperative opioid use is associated with higher levels of opioid use after shoulder arthroplasty.