Artificial Intelligence Identifies High-Risk Patients for Avian Flu

Strategic Insights -

Researchers from the University of Maryland School of Medicine developed a new application of artificial intelligence (AI) to quickly scan notes in electronic medical records and identify high-risk patients who may have been infected with H5N1 avian influenza or 'bird flu'. The AI tool, a generative AI large language model (LLM), analyzed 13,494 visits across University of Maryland Medical System (UMMS) hospital emergency departments from adult patients in urban, suburban, and rural areas in 2024. The model flagged 76 patients who mentioned a high-risk exposure for bird flu, such as working as a butcher or at a farm with livestock, and after a brief review by research staff, 14 patients were confirmed to have had recent, relevant exposure to animals known to carry H5N1.

Key Takeaways:

  • The AI tool, a generative AI large language model (LLM), analyzed 13,494 visits across University of Maryland Medical System (UMMS) hospital emergency departments from adult patients in urban, suburban, and rural areas in 2024.
  • The model flagged 76 patients who mentioned a high-risk exposure for bird flu, such as working as a butcher or at a farm with livestock, and after a brief review by research staff, 14 patients were confirmed to have had recent, relevant exposure to animals known to carry H5N1.
  • The LLM (GPT-4 Turbo) demonstrated strong performance in identifying mentions of animal exposure, with a 90% positive predictive value and a 98% negative predictive value when it was evaluated on a sample of 10,000 historical emergency department visits from 2022-2023.
  • The AI review required only 26 minutes of human time and cost just 3 cents per patient note, demonstrating high scalability and efficiency.
  • The study shows how generative AI can fill a critical gap in our public health infrastructure by detecting high-risk patients that would otherwise go unnoticed.
  • The researchers hope to next test the large language model for prospective surveillance and deployment within the electronic health record, for faster real-time identification of high-risk patients.
  • H5N1 has infected more than 1,075 dairy herds across 17 states, and over 175 million poultry and wild birds have tested positive during this outbreak period.
  • Identified human cases remain rare, with 70 confirmed infections and just one fatality in the U.S. by mid-2025, according to the Centers for Disease Control and Prevention (CDC).

Statistics:

  • 13,494 visits analyzed across University of Maryland Medical System (UMMS) hospital emergency departments from adult patients in urban, suburban, and rural areas in 2024.
  • 76 patients flagged by the AI model who mentioned a high-risk exposure for bird flu.
  • 14 patients confirmed to have had recent, relevant exposure to animals known to carry H5N1 after a brief review by research staff.
  • 90% positive predictive value and 98% negative predictive value for the LLM (GPT-4 Turbo) when evaluated on a sample of 10,000 historical emergency department visits from 2022-2023.
  • 26 minutes of human time required for AI review.
  • 3 cents per patient note for AI review.
  • 1,075 dairy herds infected with H5N1 across 17 states.
  • 175 million poultry and wild birds tested positive for H5N1 during the outbreak period.
  • 70 confirmed human infections and one fatality in the U.S. by mid-2025, according to the Centers for Disease Control and Prevention (CDC).

Sources:

  • Original Press Release: August 25 -- University of Maryland Medical System
  • University of Maryland School of Medicine
  • University of Maryland Medical System
  • Centers for Disease Control and Prevention (CDC)
  • Agency for Healthcare Research and Quality
  • UM Institute for Health Computing