Background: Even with many direct oral anticoagulant options, brand-name or generic warfarin is still widely used to prevent atherothrombotic events in cardiology. Federal standards regulate bioequivalence of generic vs. brand-name drugs through comparative bioavailability studies but does not regulate clinical equivalence nor tolerability in a “real-life” settings. Through public health surveillance, we have evaluated the impact of the generic warfarin commercialization on health care utilization: emergency room (ER) consultations or hospitalizations.
Methods: We used an interrupted time series analysis using the Quebec Integrated Chronic Disease Surveillance System, a surveillance system from the second populous province in Canada (~8.3 million in 2017). Rates of health care utilization for warfarin users (n=280,158) aged ≥ 66 years were calculated for 6-month periods, 5 years before up to 15 years after warfarin commercialization (from January 1996 to January 2016). Periods before and after generic warfarin commercialization were compared by negative binomial segmented regression models for all users with a specific variable for generic or brand-name users. Sensitivity analyses were also conducted.
Results: Generic warfarin analogs (n=5) were commercialized from January 2001. There was an approximated mean rate of 1134 ER or hospitalizations for 1000 brand-name and generic users per 6-month period, similar before and after generics commercialization. After generics commercialization, there was an immediate increase in rates of health care utilization for generic (9.9%) vs. brand-name users (0%), a statistically significant difference (9.9% [95% confidence interval: 4.4% to 15.5%], p = 0.0001). Rates of health care utilization remained stable and higher for generic vs. brand-name users throughout the period after generics commercialization.
Conclusion: Among generic warfarin users, we observed an increased rates of health care utilization soon after generics commercialization. Risk and survival analysis studies controlling for potential confounders are required to deepen this pharmacovigilance signal as stricter licensing process may be required.