OBJECTIVES:

This study aims to examine the association between opioid overdose mortality and access to medical care while considering the impact of social determinants of health (SDOH).

METHODS:

This ecological study focused on the Maryland population. We collected county-level data on opioid overdose mortality rate in 2019, demographics, social vulnerability index (SVI), and access to medical care (pharmacy and physician density) via PolicyMap tool. Association between each predictor variable and opioid overdose mortality was measured via bivariate analysis using a negative binomial regression. Multivariate analysis was performed to examine association between access to medical care and opioid overdose mortality while controlling for the stated SDOH.

RESULTS:

The observed median county-level opioid overdose death rate was 36 deaths per 100,000 people. Findings of bivariate analysis demonstrated a statistically significant association between opioid overdose mortality and pharmacy density (P 0.02, % change 8.61, 95% CI: 1.31, 16.18). Borderline P-values were observed for SVI (P 0.07) and physician density (P 0.06) in the bivariate analysis. Pharmacy density lost its association with opioid overdose mortality (P 0.26) when we controlled for demographics, SVI, and physician density via multivariate-adjusted model. In contrast, physician density which was not strongly associated with opioid overdose mortality in the bivariate analysis, became statistically significant in the multivariate-adjusted analysis (P 0.002, % change -3.2, 95% CI: -5.15, -1.2). Demographics were not strongly associated with opioid overdose mortality in both bivariate and multivariate analysis.

CONCLUSIONS:

Controlling for demographics, social factors, and pharmacy density, lower physician density was associated with higher rates of opioid overdose mortality at the county level. The results suggest that in addition to considering the impact of social factors, efforts for prevention and treatment strategies should focus on areas with fewer physicians. Further studies are needed to collect more data in other states and to confirm the findings seen in this analysis.