Table 3.
Associations between previous year’s dispensary counts (natural logarithm) and log transformed yearly opioid related mortality rates at county level
| Log transformed opioid related mortality rate | Coefficient (95% confidence interval), Romano-Wolf P value | |||||||
|---|---|---|---|---|---|---|---|---|
| All opioid types | Prescription opioids | Heroin | Synthetic non-methadone | |||||
| Population control only | ||||||||
| All storefront dispensaries | −0.18 (−0.24 to −0.13), 0.002 | −0.10 (−0.15 to −0.06), 0.004 | −0.06 (−0.12 to −0.01), 0.07 | −0.26 (−0.32 to −0.20), 0.002 | ||||
| Medical dispensaries only | −0.18 (−0.25 to −0.11), 0.002 | −0.10 (−0.16 −0.04), 0.002 | −0.11 (−0.17 to −0.05), 0.002 | −0.20 (−0.28 to −0.13), 0.002 | ||||
| Recreational dispensaries only | −0.08 (−0.15 to −0.02), 0.04 | −0.08 (−0.15 to −0.02), 0.27 | 0.04 (−0.01 to 0.10), 0.27 | −0.21 (−0.28 to −0.13), 0.002 | ||||
| All covariates | ||||||||
| All storefront dispensaries | −0.17 (−0.23 to −0.11), 0.002 | −0.08 (−0.13 to −0.03), 0.03 | −0.08 (−0.13 to −0.02), 0.03 | −0.21 (−0.27 to −0.14), 0.002 | ||||
| Medical dispensaries only | −0.15 (−0.21 to −0.09), 0.002 | −0.06 (−0.11 to −0.00), 0.07 | −0.12 (−0.18 to −0.06), 0.002 | −0.14 (−0.21 to −0.07), 0.002 | ||||
| Recreational dispensaries only | −0.11 (−0.17 to −0.04), 0.01 | −0.05 (−0.11 to 0.02), 0.3 | 0.03 (−0.03 to 0.10), 0.30 | −0.21 (−0.29 to −0.13), 0.002 | ||||
Regression models specify random effects at county level, with yearly fixed effects. All storefront dispensaries reflects estimated effect of log transformed county level count of all storefront dispensaries (medical and recreational) on log transformed yearly age adjusted mortality rates among people aged 21 years and older for each type of opioid. Medical and recreational reflect effects of log transformed county level counts of medical storefront and recreational storefront dispensaries (estimated simultaneously) on log transformed yearly age adjusted mortality rates. Models under population control only are estimated with total population aged 21 years and older (natural logarithm (ln)) as the only control variable. Models under all covariates including county level time varying covariates for unemployment rate, average per capita income (ln), gross domestic product (ln), proportion white, total population aged 21 years and older (ln), proportion college educated, proportion men, population with disability (ln). State level time varying covariates: naloxone access law, pharmacist patient identification law, pain management clinic oversight, recreational legal. Models are estimated with robust standard errors on 812 counties in the 23 states with legal, operating storefront dispensaries by end of 2017 and District of Columbia.