Tramadol use before total shoulder arthroplasty: patients have lower risk of complications and resource utilization than those using traditional opioids.

Background: Evidence continues to mount for the deleterious effects of preoperative opioid use in the setting of total shoulder arthroplasty (TSA). Tramadol, a synthetic opioid with concomitant neurotransmitter effects, has become a popular alternative to traditional opioids, but it has not been well studied in the preoperative setting of TSA. The purpose of this study is to evaluate postsurgical outcomes in TSA for patients with preoperative tramadol use compared with patients using traditional opioids and those who were opioid naïve.

Methods: Using the IBM Watson Health MarketScan databases, a retrospective cohort study was performed for patients who underwent TSA from 2009 to 2018. Filled pain prescriptions were collected, and prescribing trends were analyzed. Outcomes were compared between 4 patient cohorts defined by preoperative analgesia use-opioid naïve, tramadol, traditional opioids, and combination (opioids and tramadol). Multivariate analysis was used to account for small variations in cohort demographics and comorbidities. Analysis focused on resource utilization and complications. Revision rates at 1 and 3 years postoperatively were also compared.

Results: A total of 29,454 TSA patients were studied, with 8959 available for 3-year postoperative follow-up. Of these, 10,462 (35.5%) were prescribed traditional opioids and 2214 (7.5%) tramadol only. From 2009 to 2018, prescribing trends in the United States demonstrated a significant decrease in the number of patients prescribed preoperative narcotics, whereas the number of patients prescribed preoperative tramadol and those who were opioid naïve significantly increased. Compared with opioid-naïve patients, the traditional opioid cohort had significantly increased odds of resource utilization and complications, whereas the tramadol cohort did not. Specifically, the traditional opioid cohort had an increased risk of prosthetic joint infection compared with both opioid-naïve and tramadol cohorts. The traditional opioid cohort had higher revision rates than opioid-naïve patients at 1 and 3 years, whereas the tramadol cohort did not.

Conclusion: Despite a decrease in opioid prescriptions over the study period, many patients in the United States remain on opioids. Although tramadol is not without its own risks, our results suggest that patients taking preoperative tramadol as an alternative to traditional opioids for glenohumeral arthritic pain had a lesser postoperative risk profile, comparable with opioid-naïve patients.