ABSTRACT
Background. COVID-19 has caused over 6 million deaths worldwide. While testing was initially centered in healthcare facilities with automatic reporting to health departments, it is increasingly being performed in the home with rapid antigen testing. Many at-home tests are self-interpreted and not typically reported. In December 2021, nirmatrelvir/ritonavir and molnupiravir were authorized for the treatment of COVID-19 in high-risk patients. In New York City (NYC), these antivirals are available for free via pickup or home delivery, allowing for early expanded access. We sought to identify if prescribing of antivirals might act as an early indicator of increasing COVID-19 incidence in the setting of decreased case reporting. Methods. Confirmed and probable COVID-19 tests are reported to the NYC Health Department. All pharmacies participating in the federal COVID-19 therapeutics program are required to report daily utilization data, which are made available to jurisdictions through the Tiberius platform. From December 27, 2021 to April 18, 2022, we analyzed citywide weekly case counts, using patients 65 years and over as a proxy for high-risk eligible patients, and weekly oral therapeutic prescription fills. We calculated weekly case rates and prescribing rates per 100,000 residents. Weekly case rates and prescription fill rates were trended to assess for concordance over time. Results. During this period, 17,522 prescriptions were filled. Prescription fills and case rates were concordant until the weeks of February 28 and March 7, when an increase in fill rate was noted from the week prior (1.76 to 2.34 to 4.10), while reported case incidence continued to decrease (6.16 to 5.08 to 4.61). This discordance coincided with the emergence of the BA.2 subvariant in NYC. After these weeks, an uptick in case incidence was noted for the remainder of the study period, in concordance with prescription data. Conclusion. Increasing rates of oral therapeutic prescribing could be an early indicator of increasing COVID-19 transmission. Limitations include the unclear impact of public and prescriber education efforts on prescribing trends. Next steps include exploring the addition of prescription data into modeling to predict trends in transmission as reported cases becomes a less reliable indicator.