(351) Preliminary Findings from a Feasibility and Capacity Assessment of Truveta® Studio using Flu Vaccine as Example: A Near Real-Time Electronic Health Records Database with Potential to Monitor Post-Approval Vaccine Safety

Background: Timely availability of data for monitoring post-approval safety of drugs and vaccines is vital to protect patient safety. Truveta, an electronic health records (EHR) database, collects near real-time data in the US. However, Truveta’s capability to conduct near real-time active vaccine safety surveillance has not been formally assessed.

Objectives: To assess the availability and completeness of sociodemographic, clinical, and influenza (flu) vaccination data in Truveta EHR from 01/2017 to 05/2022 and to compare it to data from 2020 US Census, 2021 CMS Chronic Conditions Data Warehouse (CCW) and CDC FluVaxView estimates.

Methods: In a retrospective cohort study, 2 cohorts were identified in Truveta EHR: Cohort 1, patients ≥ 65 years, and Cohort 2, patients aged 18-64 years at high risk for severe flu (i.e., those with asthma, diabetes, heart disease, COPD, non-skin cancer), each with ≥1 healthcare (HC) encounter in a flu season (July 1-May 31). Patients were followed from start of season until the earliest of receipt of flu vaccine (outcome of interest), last HC encounter, or end of season. Patient characteristics of Cohort 1 ascertained in the 6 months prior to start of season were compared to the Census and CCW data. Vaccination coverage (i.e., crude proportion of vaccinated patients in a season) in both cohorts was compared to CDC FluVaxView estimates.

Results: Cohort 1 was comprised of 12,408,874 patients, with more females compared to Census data (55.7% vs. 50.9%), a comparable race distribution to Census data (White, 77.9% vs. 76.6%, Black, 8.5% vs. 9.2%) In Cohort 1, less patients had comorbidities compared to CCW estimates (e.g., Atrial fibrillation: 10.1% vs. 13.7%; Anemia: 12.0% vs. 21.1%). Cohort 2 was comprised of 18,291,209 patients (59.3% females, 66.2% White, 12.3% Black). Missing data for race ranged from 9.1-14.2% in both cohorts.
Cohort 1 vaccination coverage decreased from 28.9 to 20.6% from fluseason 2017/2018 to 2021/2022, respectively, while CDC estimates increased from 59.6 to 73.9% and were higher compared to Truveta. Cohort 2 vaccination coverage ranged from 55.5 to 57.2% compared to lower CDC estimates ranging from 38.8% to 51.4% in seasons 2017/2018 to 2019/2020. In the last 2 seasons, Truveta vaccination coverage was lower compared to CDC estimates (2020/2021: 33.3% vs. 51.0%, 2021/2022: 36.1% vs. 50.4%).

Conclusions: The findings suggest that sociodemographic/clinical data from Truveta are comparable to the Census and CCW data and Truveta has the potential to support near real-time active safety surveillance studies. Further exploration of differences in vaccination coverage between Truveta and CDC data is warranted, which may be due to patients obtaining vaccinations outside of their primary health systems via pharmacies or free clinics.