Finding Long-COVID: Temporal Topic Modeling of Electronic Health Records from the N3C and RECOVER Programs

    September 2023 in “ medRxiv (Cold Spring Harbor Laboratory) ”
    Shawn T. O’Neil, Charisse Madlock‐Brown, Kenneth J. Wilkins, Ryan McGrath, Hannah Davis, Gina Assaf, Hannah Wei, Parya Zareie, Evan French, Johanna Loomba, Julie A. McMurry, A Zhou, Christopher G Chute, Richard A. Moffitt, Emily Pfaff, Yun Jae Yoo, Peter J. Leese, Robert F Chew, Michael D. Lieberman, Melissa Haendel
    Image
    Studysummary This study found that many conditions in patients with COVID-19 were significantly increased compared to controls, with specific phenotypes identified across different demographic and diagnostic attributes. Our plain-language summary of this paper — not a Tressless recommendation.
    The study utilized electronic health records from the N3C and RECOVER programs to identify and analyze conditions associated with Long-COVID (PASC) using temporal topic modeling. By examining data from over 14 million patients, the researchers identified 213 significant conditions, including non-scarring alopecia and telogen effluvium, particularly in pediatric and adolescent PASC patients. The study highlighted the diverse and long-term health impacts of COVID-19, emphasizing the importance of comprehensive data analysis and improved diagnostics for understanding and treating Long-COVID.
    Discuss this study in the Community →

    Research cited in this study

    1 / 1 results