Statistical Analysis of County-Level Contributing Factors to Opioid-Related Overdose Deaths in the United States
Code and data for for the paper doi: 10.1109/embc44109.2020.9176465
This paper examines county-level characteristic factors contributing to opioid-related overdose deaths in the United States. We categorized factors into three groups: demographic, socio-economic, and health care environmental group. These features were used as predictors to model the overdose deaths from all types of opioids including prescription (e.g., oxycodone and hydrocodone) and illicit opioids (e.g., heroin and fentanyl)toinvestigategeneraltrend,aswellasseparatemodels for heroin and fentanyl. Multilevel mixed-effect regression was adopted to adequately model grouping effect across counties. Keywords— Opioid Epidemic, Social Determinants, DemandSide Analysis, County-Level Features Abstract—This paper examines county-level characteristic factors contributing to opioid-related overdose deaths in the United States. We categorized factors into three groups: demographic, socio-economic, and health care environmental group. These features were used as predictors to model the overdose deaths from all types of opioids including prescription (e.g., oxycodone and hydrocodone) and illicit opioids (e.g., heroin and fentanyl)toinvestigategeneraltrend,aswellasseparatemodels for heroin and fentanyl. Multilevel mixed-effect regression was adopted to adequately model grouping effect across counties. Keywords— Opioid Epidemic, Social Determinants, DemandSide Analysis, County-Level Features