LO65: Assessing opioid-prescribing patterns for low back pain patients before and after the implementation of clinician performance indicators in the emergency department

Abstract

Introduction: Canada is in the midst of an opioid crisis. The number of apparent opioid-related deaths between January and March 2018 increased by 44% compared to the same period in 2016. The increasing use of prescription opioids and higher doses of opioids can lead to opioid addiction, toxicity and even death. Opioids are commonly prescribed for low back pain management in the ED, but the variability in opioid-prescribing patterns suggested an opportunity for improvement. Our centre implemented Clinician Performance Indicators (CPI) in 2015. CPIs were reported to each ED physician every 3 months and included the percentage of patients who were prescribed opioids. The intent was to raise awareness of opioid-prescribing patterns at our institution. Therefore, we evaluated opioid-prescribing patterns for patients with low back pain (LBP) before and after the CPI implementation. Methods: Data were obtained retrospectively for patients discharged from the ED from July 2015 to December 2018 with LBP-associated ICD 10 codes. We excluded admitted patients, those with specialist consultations, and patients who left without being seen. The primary outcome was opioid prescribing patterns for patients with LBP before and after CPI implementation. We performed a descriptive analysis of the data and compared the prescribing rates pre-implementation (July-Dec 2015) to post-implementation (July-Dec 2016) following a 6-month wash-out period. Moreover, we analyzed opioid-prescribing patterns over an extended period until December 2018. Results: After the exclusion criteria were applied, 8993 patients were included in the analysis. 53.5% were female and the mean (SD) age was 48.3 (19.78). During the three years of the study period, the percentage of LBP patients who received opioids showed a decreasing trend. Comparison of the pre and post CPI implementation periods showed a decrease in opioid prescriptions (42.0% vs 35.5%, 95%CI 2.9% to 10.2%). There was variation in opioids prescribed at our institution, the most common being hydromorphone (29.9%), followed by acetaminophen-oxycodone (24.2%) and acetaminophen-tramadol (20.0%). Conclusion: The implementation of CPIs positively impacted physicians' opioid-prescribing patterns for patients presenting with LBPs at our institution. Future studies are required to further improve the effectiveness of CPIs in influencing opioid-prescribing patterns.