BACKGROUNDOrthopedic surgery has been highlighted as a contributor to the opioid epidemic. There is a paucity of literature focused specifically on opioid prescribing trends following shoulder surgery. Our aims were to describe national and regional trends in opioid prescribing following total shoulder arthroplasty (TSA), proximal humerus fracture fixation, and rotator cuff repair (RCR), and to correlate opioid prescribing trends after shoulder surgery with regional heroin use and new opioid misuse diagnoses. METHODSThis is a retrospective study of 218,577 patients from a commercial insurance dataset who had undergone orthopedic shoulder procedures from 2014-2020. Opioid prescription data within 90-day postop were converted to morphine milligram equivalents (MME). Multivariate modelling was used to evaluate independent associations between MME’s prescribed and subsequent new postoperative diagnosis of opioid dependence. Additionally, longitudinal mixed effects modeling was employed to understand aggregated prescriber habits by state over time and the subsequent effects on estimates of state heroin use and drug overdose deaths. RESULTSFrom 2014 to 2020, the mean number of MMEs prescribed nationally decreased by over 50% for TSA, proximal humerus fracture fixation, and RCR. Over the study period, most states demonstrated an overall downtrend of 90-day opioid prescribing for these procedures.Individuals with a new opioid misuse diagnosis following shoulder surgery were found to be younger (SMD=0.27, p<0.001), have increased comorbidities (SMD=0.38, p<0.001), and increased MME’s prescribed (SMD=0.94, p<0.001). Increased comorbidity score (OR=3.28, p<0.001) and increased 90-day MME’s prescribed (OR=1.05, p<0.001) were all found to increase the odds ratio of a postoperative opioid misuse diagnosis in patient’s undergoing shoulder surgery, while increased age (OR=0.97, p<0.001) and patient’s undergoing RCR (OR=0.79, p=0.001) were found to have a decreased odds ratio of developing a new opioid misuse diagnosis.The predictors of regional heroin use included the within-state 90-day MMEs prescribed (p=0.002). The predictors of drug overdose death included year (p<0.001), but did not include type of surgery or 90-day MMEs. CONCLUSIONOrthopedic surgeons successfully reduced the number of opioids prescribed nationally after shoulder surgeries by over 50% during our study period with similar trends seen at the state level. Our data indicate a relationship between increased opioid prescribing after shoulder surgery and heroin use. This study highlights that while progress has been made, there is evidence that opioid prescribing following shoulder surgery is associated with the more widespread opioid epidemic, and that more research is needed to further lower postoperative opioid misuse in the future.