Using the Texas All-Payer Claims Database (TX-APCD, ~60% of insured Texans), this study examined the temporal, demographic, and geographic patterns of Long COVID incidence in relation to SARS-CoV-2 variant periods (Omicron BA.1, BA.4/BA.5, XBB, EG.5) from October 2021 to October 2023, comparing ED vs. non-ED COVID-19 cohorts and adjusting for age, sex, payer, and HHS region.
The median time from a COVID-related ED encounter to first Long COVID diagnosis was 22 days (vs. 26 days for non-ED encounters); lag between weekly ED visit peaks and Long COVID claims was 2 weeks during BA.1 and 3 weeks during later variants. Crude incidence was highest in Medicare Fee-for-Service beneficiaries (155/10,000), adults over 70 (156/10,000), and females vs. males (81 vs. 59 per 10,000); geographic clustering was seen in High Plains and Northwest Texas.
- Administrative claims capture only *diagnosed* Long COVID — likely an undercount of true burden. - Coverage limited to ~60% of insured Texans, excluding uninsured populations. - Ecological lag analysis cannot establish individual-level causation between ED visits and Long COVID onset.
Clinicians and health systems should anticipate Long COVID diagnoses roughly 3–4 weeks after acute COVID ED surges, particularly among Medicare patients and adults over 70. Post-acute care capacity planning should prioritize older adults, female patients, and high-incidence regions like High Plains and Northwest Texas.
Explore related topics
