Abstract:
|
Comprehensive longitudinal electronic health records are useful for studying post-COVID conditions (PCC). However, due to the heterogeneity of patients with and without COVID-19, it is challenging to find an appropriate comparison group. Further, dramatic decreases in healthcare utilization during the COVID-19 pandemic could cause strong secular confounding. Therefore, we proposed a difference-in-difference design to compare the change in utilization for select PCC outcomes in the COVID-19 test positive population with the change in PCC outcomes in a matched COVID-19 test negative population during a 6-month follow-up period, before and after the start of the COVID-19 pandemic. Comparing healthcare utilization within the same individuals during two time periods (2019 and 2020) controlled for variables such as demographics, health status, and healthcare seeking behavior. Comparing to a COVID-19 test negative group during the same period adjusted for secular confounding as well. We applied this approach to 255,718 test positive and test negative patients from 8 integrated healthcare organizations across the US to evaluate the excess healthcare utilization associated with PCC.
|