Community Use of Repurposed Drugs Before and During COVID-19 Pandemic in the Netherlands: An Interrupted Time-Series Analysis

Background: Repurposing registered drugs could reduce coronavirus disease (COVID-19) burden before novel drugs are authorized. Little is known about how the pandemic and imposed restrictions changed their dispensing. We aimed to investigate the impact of COVID-19 pandemic on repurposed drugs dispensing in the Netherlands. Methods: We performed interrupted time-series study using University of Groningen prescription database IADB.nl to evaluate dispensing trends of 24 repurposed drugs before (2017-February 2020) and after (March 2020-2021) the pandemic' start. Primary outcomes were monthly prevalence and incidence rates. An autoregressive integrated moving average model assessed the effect of pandemic and stringency index (measuring strictness of government's restriction policies). Results: Annual number of IADB.nl population ranged from 919,697 to 952,400. Generally, dispensing of common long-term-used drugs was not significantly affected by pandemic. The prevalence of antibacterials (-4.20 users per 1000 people), antivirals (-0.04), corticosteroids (-1.29), prednisolone (-1.32), calcium channel blocker (-0.41), and diuretics (-1.29) was lower than expected after the pandemic's start, while the prevalence of ivermectin (0.07), sulfonylureas (0.15), sodium-glucose co-transporter-2 (SGLT2) inhibitor (0.17), and anticoagulants (1.95) was higher than expected. The pandemic was associated with statistically significant decreases in the incidence of antibacterials (-1.21), corticosteroids (-0.60), prednisolone (-0.64) and anticoagulants (-0.02), and increases in ivermectin (0.02), aggregated antidiabetic drugs (0.13), and SGLT2 inhibitors (0.06). These trends were positively associated with pandemic and negatively associated with stringency index. Conclusion: Dispensing of most drugs was not significantly associated with pandemic and government's response. Despite some statistically significant disruptions, these were not necessarily clinically relevant due to small absolute differences observed.