BackgroundOver-prescription of opioids in the United States increases risks of opioid dependence, overdose, and death. Increased perioperative and postoperative opioid use during orthopedic shoulder surgery is a significant risk factor for long term opioid dependence. The authors hypothesized that a multidisciplinary perioperative pain management program (Transitional Pain Service or TPS) for major shoulder surgery would lead to a reduced amount of opioids required postoperatively. MethodsA TPS was implemented at a Veterans Affairs Medical Center focused on non-opioid pain management and cessation support. Opioid consumption during the implementation of the TPS was compared to a historical cohort. All patients undergoing shoulder arthroplasty or rotator cuff repair were included. The primary outcome was the proportion of patients continuing opioid use at 90 days postoperatively. Secondary outcomes included postoperative pain scores, time to opioid cessation, and median opioid tablets consumed at 90-days. A multivariable model was developed to predict total opioid use at 90-days postoperatively. Kaplan Meier curves were calculated for time to opioid cessation. ResultsThe TPS group demonstrated decreased persistent opioid use at 90 days post-discharge (12.6% vs. 28.6%; p=0.018). Independent predictors associated with increased total opioid tablet prescriptions at 90 days included length of stay (β=19.17), anxiety diagnosis (β=37.627), and number of tablets prescribed at discharge (β=1.353). Shoulder arthroplasty surgery (TSA) was associated with decreased 90-day opioid utilization (β= -32.535) when compared to cuff repair (RCR). Median time to cessation was shorter in TSA (6 days) when compared with RCR (8 days). Pain scores were reduced compared to population mean by post-discharge day 2 for TSA and by post-discharge day 7 for RCR. Median number of post-discharge opioid tablets (oxycodone 5 mg) consumed under TPS management was 25 in both RCR and TSA surgery groups (180 MME). Discussion and ConclusionsThis study demonstrates that a TPS reduces the amount of opioid use of patients undergoing shoulder arthroplasty or cuff repair at 90 days when compared with a historical control. Multivariable regression indicated that fewer opioid tablets at discharge was a modifiable factor that may aid in reducing opioid consumption and that anxiety diagnosis, increased length of stay, and cuff repair surgery were other factors independently associated with increased opioid consumption. This data will assist surgeons in counseling patients, setting narcotic use expectations, and minimizing overprescribing. Use of a similar multidisciplinary perioperative pain management program may greatly reduce opioid over prescriptions nationally.