Canada is experiencing an ongoing opioid-related public health crisis, including persistently rising opioid (e.g., poisoning) mortality. Previous research has documented marked correlations between population-levels of opioid dispensing and deaths. We examined possible correlations between annual population-level dispensing of specific opioid formulations and related poisoning deaths in Ontario (Canada), for the period 2005-2016. Annual coroner statistics-based numbers of poisoning deaths associated with six main opioid formulations (codeine, fentanyl, hydromorphone, methadone, morphine, and oxycodone) for Ontario were converted into annual death rates (per 100,000 population). Annual dispensing data for the opioid formulations under study were based on commercial retail-sales data from a representative, stratified sample of community pharmacies (IMSQuintiles/IQVIA CompuScript), converted into Defined Daily Doses (DDD/1,000 population/day). Possible relationships between the annual death and dispensing rates were assessed by Pearson's correlation coefficient analyses. Death rates increased for almost all, while dispensing rates increased for half of the opioid categories. A significant positive correlation between death and dispensing rates was found for hydromorphone (r = 0.97, 95% CI: 0.88-0.99) and oxycodone (r = 0.90, 95% CI: 0.68-0.97) formulations; a significant negative correlation was found for codeine (r = -0.78, 95% CI: -0.93 to -0.37). No significant correlations were detected for fentanyl, methadone, and morphine related deaths. Strong correlations between levels of dispensing and deaths for select opioid formulations were found. For select others, extrinsic factors - e.g., increasing involvement of non-medical opioid products (e.g., fentanyl) in overdose deaths - likely confounded underlying correlation effects. Opioid dispensing levels continue to influence population-level mortality levels, and need to be addressed by prevention strategies.
