Opioid overdose hospitalization trajectories in states with and without opioid-dosing guidelines

Publication type
Journal article
Authors
Sears JM Fulton-Kehoe D Schulman BA Hogg-Johnson S Franklin GM
Date published
2019 Sep 01
Journal
Public Health Reports
Volume
134
Issue
5
Pages
567-576
Open Access?
No
Abstract

OBJECTIVES: High-risk opioid-prescribing practices contribute to a national epidemic of opioid-related morbidity and mortality. The objective of this study was to determine whether the adoption of state-level opioid-prescribing guidelines that specify a high-dose threshold is associated with trends in rates of opioid overdose hospitalizations, for prescription opioids, for heroin, and for all opioids. METHODS: We identified 3 guideline states (Colorado, Utah, Washington) and 5 comparator states (Arizona, California, Michigan, New Jersey, South Carolina). We used state-level opioid overdose hospitalization data from 2001-2014 for these 8 states. Data were based on the State Inpatient Databases and provided by the Healthcare Cost and Utilization Project (HCUP), Agency for Healthcare Research and Quality, via HCUPnet. We used negative binomial panel regression to model trends in annual rates of opioid overdose hospitalizations. We used a multiple-baseline difference-in-differences study design to compare postguideline trends with concurrent trends for comparator states. RESULTS: For each guideline state, postguideline trends in rates of prescription opioid and all opioid overdose hospitalizations decreased compared with trends in the comparator states. The mean annual relative percentage decrease ranged from 3.2%-7.5% for trends in rates of prescription opioid overdose hospitalizations and from 5.4%-8.5% for trends in rates of all opioid overdose hospitalizations. CONCLUSIONS: These findings provide preliminary evidence that opioid-dosing guidelines may be an effective strategy for combating this public health crisis. Further research is needed to identify the individual effects of opioid-related interventions that occurred during the study period