National Trends in Use of Regional Anesthesia and Postoperative Patterns of Opioid Prescription Filling in Shoulder Arthroscopy: A Procedure-Specific Analysis in Patients With or Without Recent Opioid Exposure

Background: There are few large database studies on national trends in regional anesthesia for various arthroscopic shoulder procedures and the effect of nerve blocks on the postoperative rate of opioid prescription filling. Hypothesis: The use of regional nerve block will decrease the rate of opioid prescription filling after various shoulder arthroscopic procedures. Also, the postoperative pattern of opioid prescription filling will be affected by the preoperative opioid prescription-filling history. Study Design: Cohort study; Level of evidence, 3. Methods: Patient data from Humana, a large national private insurer, were queried via PearlDiver software, and a retrospective review was conducted from 2007 through 2015. Patients undergoing arthroscopic shoulder procedures were identified through Current Procedural Terminology codes. Nerve blocks were identified by relevant codes for single-shot and indwelling catheter blocks. The blocked and unblocked cases were age and sex matched to compare the pain medication prescription-filling pattern. Postoperative opioid trends (up to 6 months) were compared by regression analysis. Results: We identified 82,561 cases, of which 54,578 (66.1%) included a peripheral nerve block. Of the patients who received a block, 508 underwent diagnostic shoulder arthroscopy; 2449 had labral repair; 4746 had subacromial decompression procedure; and 12,616 underwent rotator cuff repair. The percentage of patients undergoing a nerve block increased linearly over the 9-year study period ( R 2 = 0.77; P = .002). After matching across the 2 cohorts, there was an identical trend in opioid prescription filling between blocked and unblocked cases ( P = .95). When subdivided by procedure, there was no difference in the trends between blocked and unblocked cases ( P = .52 for diagnostic arthroscopies; P = .24 for labral procedures; P = .71 for subacromial decompressions; P = .34 for rotator cuff repairs). However, when preoperative opioid users were isolated, postoperative opioid prescription filling was found to be less common in the first 2 weeks after surgery when a nerve block was given versus not given ( P < .001). Conclusion: An increasing percentage of shoulder arthroscopies are being performed with regional nerve blocks. However, there was no difference in patterns of filled postoperative opioid prescriptions between blocked and unblocked cases, except for the subgroup of patients who had filled an opioid prescription within 1 to 3 months prior to shoulder arthroscopy. Future research should focus on recording the amount of prescribed opioids consumed in national databases to reinforce our strategy against the opioid epidemic.