UTHealth Houston database fills gap in long COVID surveillance

During the pandemic, surveillance systems tracked COVID-19 cases, emergency department visits, hospitalizations and deaths. No similarly comprehensive surveillance system exists for long COVID. (Photo by UTHealth Houston)

UTHealth Houston researchers have found that long COVID diagnoses varied significantly across populations and virus variants from 2021 to 2023, revealing patterns that could help public health agencies and health systems better target follow-up and care, monitor long-term impacts, and allocate healthcare resources where they’re needed most.

The analysis was published in the International Journal of Infectious Diseases.

Researchers used the Texas All-Payor Claims Database—which covers nearly 60% of insured Texans and is maintained by UTHealth Houston and the Center for Health Care Data at UTHealth Houston School of Public Health—to examine patterns in long COVID diagnoses across different populations and major COVID-19 variant periods from October 2021 to October 2023.

According to the Centers for Disease Control and Prevention, long COVID is a chronic condition that occurs after SARS-CoV-2 infection and lasts at least three months. Although long COVID is more common among patients who were severely ill, it can affect anyone who has been infected. Because each  COVID infection carries a risk of long COVID, repeated infections may continue to contribute to its overall burden. While fewer new cases are being reported than early in the pandemic, long COVID remains a significant public health concern, affecting millions of U.S. adults and children.

“We still do not have a clear picture of how often it is being diagnosed, which groups are most affected, or where the burden is highest,” said Boya Peng, a doctoral candidate in biostatistics at UTHealth Houston School of Public Health.

During the pandemic, surveillance systems tracked COVID-19 cases, emergency department visits, hospitalizations, and deaths. No similarly comprehensive surveillance system exists for long COVID. Population-level estimates often rely on self-reported surveys, which are affected by who chooses to participate, how well people recall their symptoms, and differences in how long COVID is defined.

The Texas All-Payor Claims Database was established in 2021 by the 87th Texas Legislature. The database includes medical, pharmacy, and dental claims, along with eligibility and provider files, from private and public payers. It contains administrative claims information for approximately 60% of insured Texans and represents nearly all medical claims from health plans regulated by the state. Using this large, population-level resource, researchers were able to examine how documented long COVID differed by age, sex, insurance type, and geographic region across multiple COVID variant periods.

They found that long COVID diagnoses generally followed waves of acute COVID-19 infection. Peaks in COVID-19-related emergency department visits typically preceded increases in long COVID claims by two to three weeks. Older adults exhibited substantially higher rates of long COVID, especially those 70 years and older. Medicare fee-for-service beneficiaries also had particularly high rates, and women had higher rates than men.

Researchers also identified notable geographic differences in documented long COVID rates and variation in the time between an initial COVID-19 encounter and a subsequent long COVID diagnosis. Higher rates were observed in parts of the High Plains and northwest Texas regions, and rural counties generally had higher documented long COVID incidence than urban counties. Among patients who visited an emergency department for COVID-19, the median time from that encounter to a long COVID diagnosis was 22 days, compared with 26 days among patients whose COVID-19 care did not involve an emergency department visit.

“Our main conclusion is that large healthcare claims databases such as the TexasAll-Payor Claims Database can help researchers understand the documented burden of long COVID at the population level,” Peng said. “Claims data can complement surveys, electronic health records, and traditional public health surveillance systems to support future planning for long COVID care.”

UTHealth Houston School of Public Health authors include Kaiming Bi, PhD; Dei'sharrah Allen-Benson; Yashar Talebi, MD, PhD; Samiran Ghosh, PhD; Ashraf Yaseen, PhD; Melissa Valerio-Shewmaker, PhD; Eric Boerwinkle, PhD; Stacia M. DeSantis, PhD; Michael Swartz, PhD; and Cecilia Ganduglia Cazaban, MD, DrPH.