SINGAPORE: Groundbreaking Initiative

In a groundbreaking initiative, the Singapore General Hospital (SGH) is set to empower doctors with artificial intelligence (AI) to enhance antibiotic prescription for pneumonia patients. This innovative technology promises to reshape the approach medical professionals take in identifying and treating this serious respiratory condition, potentially saving countless lives.

AI System Functionality

The AI system will analyze symptoms reported by patients and determine whether their pneumonia is viral or bacterial in nature—an essential distinction as antibiotics are only effective against bacterial infections. By leveraging data and patterns, this intelligent system will not only recommend whether antibiotics are necessary but also suggest the most suitable type and dosage based on the individual’s clinical profile.

Development and Validation

Developed in partnership with DXC Technology and the national health tech agency Synapxe, this AI-driven solution, known as Augmented Intelligence in Infectious Diseases (AI2D), was trained on data from approximately 8,000 patients treated by SGH between 2019 and 2020. The training included various clinical parameters like X-rays, reported symptoms, and vital signs. A subsequent validation against 2,000 additional cases in 2023 indicated that the system could accurately assess the need for antibiotics in 90% of the cases studied.

Expert Insights

Dr. Piotr Chlebicki, a senior consultant in SGH’s Department of Infectious Diseases, highlighted that pneumonia can escalate rapidly, making timely intervention critical. “Tests confirming bacterial presence can take up to two days to return, and that delay can be dangerous. Our AI tool empowers doctors to make informed decisions more swiftly, especially in urgent cases where every moment counts,” he explained.

Impact on Antimicrobial Resistance

The implications of this new system extend far beyond efficiency. Reducing unnecessary antibiotic prescriptions is vital in combating the global crisis of antimicrobial resistance (AMR)—a phenomenon that threatens to render many treatments ineffective. During the pilot study, nearly 40% of antibiotics prescribed at initial diagnosis were later determined to be unnecessary. By minimizing this over-prescription, SGH aims to not only protect patients but also to prevent the emergence of drug-resistant pathogens that could complicate future healthcare efforts.

Further Research

As part of ongoing research, the SGH team is conducting a comparative study involving 200 inpatients to further assess the effectiveness of the AI system. This innovative approach may eventually be adapted for other common infections, such as urinary tract infections, creating a ripple effect throughout the healthcare system.

Addressing the Silent Pandemic

AMR is likened to a “silent pandemic” affecting global health, with the World Health Organization warning that the death toll could reach a staggering one person every three seconds by 2050 if the issue remains unaddressed. Associate Professor Andrea Kwa, who leads the project at SGH, underscored the urgency of the situation: “Our goal is to provide our doctors with rapid and reliable tools that reduce the misuse of antibiotics, thereby extending the efficacy of existing treatments and protecting patient health.”

Conclusion

This pioneering use of AI in antibiotic management not only aims to save valuable time—up to 20 minutes per case—but also signifies a monumental step towards ensuring safer, more accurate medical practices. As the healthcare industry continues to embrace technological advancements, SGH is at the forefront of a transformative movement that prioritizes both patient care and sustainable medical practices.