Opioids are acknowledged as a standard of care for post-operative analgesia but with their use comes associated harms. Persistent opioid use rates after surgery are as high as 10%. Overall, opioid prescribing patterns in our region have been declining but there is a relative scarcity of Canadian literature on the use of opioid analgesia following orthopaedic shoulder procedures to guide best practice guidelines. Our study evaluates regional prescribing patterns for opioids after elective orthopaedic shoulder procedures. For this retrospective cohort study, hospital Discharge Abstract Database and National Ambulatory Care Reporting System data were used to identify orthopaedic shoulder surgeries, specifically implant, excision, and repair, from October 2016 to March 2023. The provincial Drug Information System was utilized to calculate rates of any opioids prescribed and morphine milligram equivalency (MME) for prescriptions filled at discharge. Prescribing patterns were compared between hospitals and surgeons and variability was assessed. Three-level hierarchical regression (HLM) models were run to test for associations between patient circumstances and the probability of any opioid prescription at discharge as well as MME levels. Covariance estimates were used to calculate the intraclass correlation (ICC) and test for variability between and within hospitals while t-tests were analyzed to investigate associations with covariates. A 95% confidence level was used for statistical significance. The study included 6885 observations, with repairs being the most common of the three sub-groups and day surgery accounting for 77.7% CI 76.7%–78.6%. Overall, 84.8% CI 83.9%–85.6% of cases filled an opioid prescription (Rx) at discharge increasing over the study period from 79.1% CI 75.0%–83.1% to 84.8% 80.8%–88.8%. The MME where positive decreased from 276.1 CI 262.0–290.1 to 165.3 CI 155.4–175.1. Variation in opioid Rx varied across hospitals however ICCs indicate 8.9% of the total variation was at the surgeon level (p-value=0.0019) compared to 3.5% across hospitals (p-value=0.1142). Similarly, more of the total variation in the MME was explained by differences across surgeons (ICC=41.8% p-value < 0 .0001) compared to hospitals (ICC=6.8% p-value=0.3092). HLM models including covariates indicated older patients were less likely to fill an opioid at discharge as were emergency cases, but day surgery cases were more likely. MME models showed females and day surgery patients received lower levels of morphine when filling an Rx. Those who live in a rural level had higher MME levels as did those who filled an opioid Rx in the six months before surgery. Variations between surgeons remain in Rxs (ICC 7.5%- p-value 0.0045) and MME (ICC 41.8% p-value < 0.0001) after controlling for patient characteristics. For patients undergoing orthopaedic shoulder surgeries in our province, there was an increasing trend in opioid Rxs provided at discharge while MME decreased. This difference may reflect the tension between appropriate pain management for patients and increasing awareness of the potential harms of opioids by prescribers. Causes for variations in prescribing patterns amongst prescribers require further evaluation. Evidence-based multi-modal pain management pathways offer a strategy to reduce variability in prescribing patterns and further decrease MME levels while still appropriately addressing postoperative analgesia.
Jordan et al. (2026) studied this question.