Background: Booster COVID-19 vaccines were first authorized for certain populations on September 22, 2021. Administrative healthcare databases are valuable tools for assessment of the safety and effectiveness of vaccines targeting SARS CoV-2 in the US, however differentiation of primary series and booster doses can be challenging when non-specific codes are applied.
Objectives: To develop an algorithm for the identification of mRNA-based COVID-19 booster vaccines and to describe corresponding clinical and demographic characteristics in a US administrative claims database.
Methods: Using HealthVerity claims, frequency distributions of time between 1st and 2nd vaccine doses within the primary series, and time between completion of a primary series and the 1st observed booster dose of a mRNA COVID-19 vaccine were generated for both homologous and heterologous series and stratified by age and vaccine manufacturer. The observed median time +/- 5 days between doses was used to determine the minimum required interval to identify completion of a primary series. Aligned with CDC recommendations, a booster dose was required to occur a minimum of 42 days after an observed primary series, and to occur on or after September 22, 2021. This approach may mitigate potential misclassification of a 3rd dose in a primary series as a booster dose among immunocompromised patients. Clinical characteristics of recipients of a booster dose within the following windows since completion of a primary series were described: 42-89 days (early adopters), 90-349 days, and 350-450 days (late adopters). The absolute standardized difference (ASD) was calculated comparing the frequency of characteristics within each time window to the overall frequency; ASD >=0.1 was considered different.
Results: The median time observed between a 1st dose and a 2nd dose was 28 days for a homologous Spikevax series, 21 days for a homologous Comirnaty series, and 28 days for a heterologous series. Among ~5 million patients with 3 observed vaccine administrations, ~3.7 million (75%) met the booster dose specifications. The median time from a primary series to booster dose ranged from 230-250 days. Early adopters of booster vaccines were more likely to be immunocompromised and have comorbidities related to lung function; late adopters were more likely to have chronic conditions including cardiovascular disease, hypertension, diabetes, and cancer.
Conclusions: Our claims-based algorithm for booster COVID-19 vaccine administration identified patients and vaccine patterns aligned with CDC recommendation and clinical expectations and may mitigate misclassification of primary series vaccinations in claims data.
