Recent research from Harvard Medical School highlights an innovative AI tool developed by Mass General Brigham to help identify long COVID, also known as post-acute sequelae of COVID-19 (PASC). This method uses electronic health record data to provide precise phenotyping, which improves the detection of long COVID without underestimating its prevalence. The study, involving over 295,000 patients, found that the AI tool achieved an accuracy rate of 79.9%, outperforming traditional diagnosis methods. With an estimated 409 million people affected globally, this research shows how AI can enhance precision medicine and lead to better healthcare outcomes for individuals suffering from long COVID symptoms.
Harvard Medical School has released exciting new research that highlights the potential of an artificial intelligence (AI) tool developed by Mass General Brigham. This AI tool can accurately identify cases of long COVID, also known as PASC (post-acute sequelae of COVID-19), by analyzing electronic health record (EHR) data.
Dr. Hossein Estiri, the senior author of the study and a professor at Harvard Medical School, emphasized that this new method offers superior precision in identifying long COVID cases without undercounting them. He noted that it also minimizes biases in diagnosing patients across different demographic groups.
The impact of long COVID is significant. A recent study estimates that about 409 million people globally experienced PASC in 2023. In the U.S. alone, approximately 6.9% of adults and 1.3% of children were affected by long COVID in 2022, as reported by the CDC.
Long COVID is a chronic condition where symptoms persist for at least three months after infection with the coronavirus. Common symptoms include fatigue, cough, issues with taste and smell, and various cognitive challenges. These symptoms can severely impact patients’ lives and may be indicative of other health conditions, such as anxiety or cardiovascular problems.
The research team used data from over 295,000 patients across community health centers and hospitals in the Mass General Brigham healthcare system. Their AI-based tool demonstrated an accuracy rate of 79.9%, which surpasses the traditional diagnostic methods that rely on specific codes. Notably, their method identified four times more long COVID cases than conventional diagnosis codes.
This study shows how AI and real-world data can work together to uncover important health patterns, paving the way for more personalized and effective medical care.
Tags: long COVID, PASC, artificial intelligence, Harvard Medical School, precision medicine, healthcare innovation, COVID-19 research
What is the AI Precision Phenotyping Tool?
The AI Precision Phenotyping Tool is a computer program that helps identify signs of Long COVID in patients by analyzing their symptoms and health data.
How does the tool work?
It uses advanced technology to look at patterns in a person’s health information, helping doctors understand if someone may have Long COVID.
Who can benefit from using this tool?
Patients who are experiencing lingering symptoms after a COVID-19 infection can benefit because it helps doctors make better treatment plans.
Is the tool safe to use?
Yes, the tool is designed to be safe and is used alongside other medical check-ups and evaluations to ensure patient care.
How can I get tested with this tool?
To be tested, you should talk to your doctor. They can use the tool as part of your medical assessment if they think you might have Long COVID.
