The opioid overdose crisis continues within the U.S., and the role of prescribed opioids and prescribing patterns in overdose deaths remains an important area of research. This study investigated patterns of prescription opioids dispensed in the 12 months prior to opioid-detected overdose death in Connecticut between May 8th, 2016 and January 2nd, 2018, considering differences by demographic characteristics.
The sample included decedents who had an opioid dispensed within 30 days preceding death. Using multilevel modeling, we estimated the slope of change in mean morphine equivalent (MME) daily dose over 12 months prior to death, considering linear and quadratic effects of time. We estimated the main effects of age, sex, race, and ethnicity and their interactions with time on MME. A sensitivity analysis examined how excluding decedents who did not receive long-term (≥90 days) opioid therapy affected mean MME slopes. Secondary analysis explored differences by toxicology results.
Among 1,580 opioid-detected deaths, 179 decedents had prescribed opioids dispensed within 30 days preceding death. Decedents’ mean age was 47.3 years (±11.5), 65.5% were male, 81% White non-Hispanic, 9.5% Black non-Hispanic, and 9.5% Hispanic. In the time-only model, linear (β=6.25, p<0.01) and quadratic (β=0.49, p=0.02) effects of time were positive, indicating exponentially increasing dose prior to death. Linear change in MME was significantly attenuated in men compared to women (β=-4.87, p=0.03); however, men were more likely to have non-prescription opioids in their toxicology results (p=0.02). Sensitivity analysis results supported primary findings.