Trends and factors associated with opioid prescribing from 2017 to 2023

Abstract Widespread misuse of prescription opioids has resulted in large numbers of opioid-related overdose deaths. It is critical to have a better understanding of the temporal patterns of opioid prescribing practices and associated clinical scenarios. We examined opioid prescription trends over 7 years in a large medical system using electronic health record data. Between 2017 and 2023, we identified 1,019,706 patients from 13 hospitals within a large health system in the northeastern United States, who had at least 1 opioid prescription. In total, there were 3,877,281 associated encounters with 18,225 prescribers. We examined the overall monthly opioid prescription rates and observed an average decrease during the 84 months of study period and discovered 4 distinct stages. A decrease was seen between January 2017 and January 2020 (monthly rate change: −0.70%, 95% CI: −0.89% to −0.41%), followed by a sharp decrease and a fast rebound between February 2020, April 2020, and July 2020 (monthly rate change: −10.60%, 95% CI: −14.73% to −2.52%; 13.06%, 95% CI: 3.42%-18.47%), then back to a gradual decrease from August 2020 to December 2023 (monthly rate change: −0.46%, 95% CI: −0.67% to −0.29%). When prescriptions were further classified by prescribing setting, patient demographics, and patient visit encounter types, we observed variations among these subgroups. We also identified significant associations between patient characteristics and provider specialty with high morphine milligram equivalent dose prescriptions. These results highlight the complexity of opioid prescription practice trends indicating that all these issues need to be considered in developing prescription guidance.