Grant: Travel Awards - Institute Community Support

Objective: To test a population-based method to analyze small area variations in prescription drugs Data sources/study setting: All prescription drugs covered by the public insurance plan for all senior residents of the province of Quebec, Canada Study Design: A population-based study design using small area variation analysis methods to estimate geographical variations and rank filled prescription drug classes by their level of variation. Data collection/extraction methods: Prescription drug data were extracted for the fiscal year 2016-2017 for all seniors enrolled in the public drug insurance plan and for each local health area. Drugs were grouped according to therapeutic indications based on the World Health Organization ATC system. Age-sex adjusted rates of filled prescription systematic components of variation and extremal quotient were calculated for each drug category in every local health area. Principal Findings: Drugs were classified into 47 therapeutic targets. There was low variation in the top 6 prescribed drug classes, and while the five drug classes with the highest variations were consumed by less than 5% of the study population. Three therapeutic targets were identified as highly used and as having high utilization variations: benzodiazepines, laxatives, and non-ORL corticoids. Large variations were also observed in therapeutic targets for which medical uncertainty is very low. Low variations were found in the most prescribed drugs classes. Conclusions: Our findings show variations in the consumption of many classes of prescription drugs across small areas in the senior population of the province of Quebec. Such an approach provides decision makers with objective measures of practice variations and uncertainties in order to identify and prioritize practice improvement interventions.