Patterns, Changes, and Trends in Prescription Opioid Dispensing in Canada, 2005-2016

Levels of prescription opioid (PO) dispensing have been rising in Canada - also in global comparison - since the mid-2000s, and are co-occurring with extensive PO-related morbidity and mortality. Previous analyses have demonstrated correlations between PO dispensing and related harm levels, yet also distinct heterogeneous interprovincial PO-dispensing patterns, in regards to quantities and individual PO formulations. Several system-level interventions have been implemented recently (since 2012) to address high PO-use levels and related harms in Canada; the effects of these interventions on PO-dispensing levels remain largely unexamined.

Our aim was to examine over-time patterns and trends of levels of PO dispensing quantitatively (in defined daily doses [DDDs]) for 'strong' and 'weak' opioids and qualitatively (by individual PO formulations) by province and Canada total, for the period of 2005-2016.

We examined annual PO-dispensing levels, by 'weak' and 'strong' POs (individual PO formulations, but excluding methadone), by province and for Canada total, from 2005-2016. Raw dispensing information for POs were obtained from IMSQuintiles CompuScript [new name: IQVIA], based on monthly retail dispensing data from a representative sample of community pharmacies covering about 80% of all dispensing episodes in Canada. These data were converted into annual dispensing values in DDDs (DDD/1,000 population/day), based on standard methodology, for the PO formulation groups of interest. Patterns and trends of 'strong' and 'weak' POs and individual PO formulations were examined descriptively, aided by segmented regression analyses to identify significant break-points in over-time trends. In addition, changes in 'strong'/'weak' PO dispensing ratios between 2005 and 2016 were examined.

'Weak' PO use remained largely stable across Canada over the study period. For 'strong' PO dispensing, half of the provinces featured consistent increases, while remaining provinces presented initial increases with subsequently reverting downward trends at divergent levels. Dispensing of individual 'strong' PO formulations varied interprovincially; specifically, substantial decreases for oxycodone co-occurred with increases in other 'strong' PO formulations. The dispensing ratios for 'strong'/'weak' POs increased significantly across jurisdictions between 2005 and 2016 (P < .05).

Retail pharmacy-based data do not cover the total - but the large majority - of PO dispensing in Canada. There are limitations to DDD/1,000 population/day as a comparative measurement unit for PO dispensing. The causal contribution of interventions associated with changes in PO dispensing observed cannot be verified with the data available.

Heterogeneous trends for PO dispensing, driven mostly by variations in 'strong' PO use, continue to be observed provincially across Canada. Recent changes in PO dispensing are likely influenced by recent intervention efforts (e.g., PO de-scheduling, monitoring, guidelines) aiming to reduce PO-related harms, which, however, have shown limited impact on PO-dispensing levels to date.

Opioids, prescribing, dispensing, interventions, policy, population, monitoring, Canada.