FIG 2.

Racial and ethnic disparities in EOL opioid access by patient sex: the adjusted absolute differences by race, ethnicity, and sex (A) in the probability of filling any opioid near EOL; (B) in the probability of filling a long-acting opioid near EOL; (C) in opioid dose in MMED among patients filling an opioid prescription near EOL; and (D) in the total dose of opioids filled per decedent near EOL in MMEs. White women are the reference group for all analyses. Circles reflect the adjusted correlation coefficients and the error bars reflect the 95% CIs from regression models. For all analyses, EOL was defined as the last 30 days before death or hospice. Bold text and asterisk indicates the presence of a statistically significant, negative interaction between Black race and male sex. For all outcomes, the negative effect of being Black on EOL opioid outcomes is disproportionately large for men. EOL, end of life; MME, morphine milligram equivalent; MMED, morphine milligram equivalent per day.