Method: The Maryland State Emergency Department Dataset created as part of the Healthcare Cost and Utilization Project spanning the years 2016 through 2020 was utilized. The outcome variable, opioid overdose admission, was constructed by consolidating data from all opioid poisoning codes among the ICD-10 classification of mental and behavioral disorders. The multilevel Poisson regression was employed due to the hierarchical structure of the data and the nature of the outcome variable.
Results: An exposure of the population size was used to model the rate of opioid overdoses in the respective counties. Counties with more RSS were associated with a 139% higher rate of opioid overdose admissions (IRR = 1.39, SE = .17, p < .05), relative to counties with fewer RSS. Drug overdose death rate (IRR = 3.60, SE = .99, p < .001), patient capacity rate (IRR = 1.01, SE = .01, p < .05), single-parent household rate (IRR = 1.04, SE = .01, p < .001), and non-Hispanic White rate (IRR = 1.02, SE = .00, p < .001) exhibited a positive association with yearly opioid overdose admissions. There was a 35% reduction in yearly opioid overdose admissions observed in 2020 (IRR = .65, SE = .13, p < .05) when compared to the reference year, 2016.
Conclusions and Implications: The findings suggest that RSS are strategically deployed in Maryland communities characterized by elevated rates of opioid overdose admission, thus implying a purposeful targeting of resources where they are most acutely required. They also suggest that communities that invest more in RSS may have the infrastructure to accommodate higher numbers of opioid overdose-related emergency department admissions. The findings suggest that the prevalence of fentanyl significantly aggravated both fatal and non-fatal overdose incidents in Maryland, especially starting around 2017. However, it didn’t seem to maintain the skyrocketing trend until 2020 possibly because of various harm reduction strategies that Maryland adopted. This study harnessed a unique source of data, the recently acquired dataset through a qualitative study detailing the provision of RSS. By incorporating this novel dataset, the study ventured into uncharted territory, shedding new light on the relationship between the provision of composite RSS and opioid-related health outcomes or health service utilization. In future research, expanding the scope of data will enhance our ability to establish more robust cause-and-effect relationships.