A groundbreaking AI-powered analysis has revealed that long COVID may affect more than 23% of adults in the United States, according to a new study published in the journal Med on November 8. This figure is significantly higher than the 7% prevalence estimated by earlier studies, suggesting a greater-than-anticipated burden of this complex and often debilitating condition.

Conducted by researchers from Mass General Brigham, the study utilized an advanced AI tool to parse through extensive electronic health records and detect symptoms of long COVID, which can manifest weeks or even months after initial recovery from a COVID-19 infection. Long COVID symptoms can vary widely, affecting multiple bodily systems, and may include chronic fatigue, persistent cough, cognitive issues known as "brain fog," and cardiovascular complications.

“Questions about the true burden of long COVID -- questions that have thus far remained elusive -- now seem more within reach,” said Dr. Hossein Estiri, head of AI research at Mass General Brigham and senior researcher on the study. By using this novel AI, the team aims to provide healthcare providers with a clearer diagnostic tool for an otherwise challenging condition to identify.

The AI tool is designed to spot patterns in long COVID symptoms that are often obscured within extensive and complex medical histories. By identifying symptoms that persist for over two months after COVID-19 infection and are not easily explained by pre-existing conditions, the tool can pinpoint cases likely to be long COVID. For example, the AI can assess if breathing difficulties might be due to an existing condition such as asthma or heart failure, rather than long COVID.

"Physicians are often faced with having to wade through a tangled web of symptoms and medical histories, unsure of which threads to pull, while balancing busy caseloads. Having a tool powered by AI that can methodically do it for them could be a game-changer,” explained Dr. Alaleh Azhir, lead researcher and internal medicine resident at Brigham and Women’s Hospital.

This AI-driven estimate of nearly 23% aligns more closely with rising reports of long COVID across the U.S. and reflects the urgency of addressing its impact on the healthcare system, researchers say.

To expand the reach of their findings, the team plans to make the AI tool publicly accessible, allowing physicians and healthcare systems nationwide to utilize and further test its efficacy. By offering it on an open-access platform, the researchers hope to support early and accurate diagnosis, which could be essential in managing and treating long COVID on a broader scale.