(P-513) Trends in Hydroxychloroquine Drug Utilization and Safety in 2020

Background: On 19Mar2020, in a U.S. Coronavirus Task Force Press Briefing, hydroxychloroquine (HCQ) was introduced as an investigational treatment for COVID-19 following scientific reports of its potential. On the same date, media attention regarding the drug rose sharply, potentially influencing national HCQ utilization.

Objectives: Given the attention on the use of HCQ as a potential treatment for COVID-19, we assessed the impact on HCQ utilization within a large, commercially insured population in the U.S. and assessed safety outcomes.

Methods: Two mutually exclusive cohorts were created in the HealthCore Integrated Research Database: (1) new users of HCQ in the study period (1Jan2020-31Aug2020) without 2019 HCQ use, and (2) prevalent users with HCQ use in both 2019 and in the study period. Included patients had >364 days of continuous enrollment prior to first HCQ fill in 2020 (index). We characterized indications for use, and estimated proportion of days covered (PDC) for each cohort. Follow-up for safety events was from index through death, end of continuous enrollment, or 31Aug2020. We assessed the cohorts for three safety events: ventricular arrhythmia (VA), cardiac arrest (CA), and long QT syndrome (LQTS). Potential drug-drug interactions with QT-prolonging agents filled +/- 3 days of index were examined.

Results: We identified 19,484 new and 31,366 prevalent users. Of the new users, 3,146 (16%) had their index fill on 19Mar or 20Mar, corresponding with the 19Mar press briefing. New use returned to pre-March levels in April. In contrast, 1,289 (4%) of prevalent users had their index fill on either of these days, and prevalent use remained steady during the study period. Prevalent users appeared to have little to no disruption in access to therapy, with a median PDC of 91% (IQR: 71-99%) compared to 95% (IQR: 76-100%) in 2019. Overall, 27% of new and 74% of prevalent users had a diagnosis code for an autoimmune or arthritic condition suggesting HCQ indication. Safety events were observed during follow-up in 1.4% of HCQ users, including 1% and 1.6% of new and prevalent users, respectively: VA in 0.9% and 1.3%, and CA or LQTS in 0.2% and 0.3%. Concerning concomitant use of QT-prolonging agents, safety events were experienced by 1.5% of concomitant users and 1.4% of non-concomitant users.

Conclusions: While new use of HCQ rose sharply in association with COVID-19 media attention, we did not identify access issues for prevalent users as indicated by PDC. Many new users lacked a clear indication for therapy, suggesting use related to COVID-19. Ventricular arrhythmia, cardiac arrest, or long QT syndrome were infrequent, occurring in <1.5% of included HCQ users.