Singapore General Hospital is at the forefront of developing innovative solutions to combat antibiotic resistance and improve patient outcomes. In collaboration with DXC Technology, the hospital has created the Augmented Intelligence in Infectious Diseases (AI2D) model, which leverages artificial intelligence to determine the necessity of prescribing antibiotics for pneumonia cases.
The AI2D model was developed using deidentified clinical data from over 8,000 SGH patients, including X-rays, clinical symptoms, vital signs, and infection response trends. It covers seven broad-spectrum intravenous antibiotics commonly prescribed for pneumonia. A pilot validation study conducted last year compared the AI model against 2,000 pneumonia cases and yielded promising results.
The study found that the AI2D model significantly reduced the number of cases that needed manual review and accurately identified cases requiring intervention. It demonstrated a 90% accuracy rate in determining the necessity of antibiotics, highlighting that nearly 40% of antibiotics prescribed for pneumonia cases may have been unnecessary.
Pneumonia accounts for a significant portion of infections treated at SGH, with antibiotic prescriptions being the most common. Overprescription of antibiotics can lead to antimicrobial resistance, a global concern that hospitals are actively addressing through antimicrobial stewardship programs. These programs aim to prevent the overuse of antibiotics and identify the most appropriate treatments for each patient, ultimately reducing hospital stays, deaths, and costs.
Using automation and AI in antimicrobial stewardship programs can provide real-time insights at the point of prescription, helping healthcare teams make informed decisions. SGH’s research team is now conducting a comparative study to evaluate the effectiveness of the AI model in reducing antibiotic use and plans to expand its application to other infections like urinary tract infections.
In a larger trend, hospitals worldwide are adopting AI-powered solutions to address antimicrobial resistance. For instance, China Medical University Hospital in Taiwan implemented an Intelligent Anti-Microbial System that utilises AI to identify drug-resistant strains, predict sepsis and death, recommend drug doses, and monitor drug interactions.
Dr. Piotr Chlebicki, a senior consultant at SGH Department of Infectious Diseases, emphasises the importance of striking a balance between the benefits and risks of antibiotic use. AI technologies like the AI2D model offer a valuable tool for healthcare providers to make more informed decisions, ultimately improving patient outcomes and combating antimicrobial resistance.
As hospitals continue to leverage AI and automation in antimicrobial stewardship programs, the future of infection treatment looks promising, with the potential to save lives, reduce costs, and mitigate the growing threat of antibiotic resistance.
