AI in Singapore's healthcare requires regulatory framework
Singapore General Hospital highlights the potential of AI and large language models in healthcare, emphasizing the need for appropriate regulations before widespread adoption.
Singapore General Hospital (SGH) acknowledges the significant potential of artificial intelligence (AI) and generative large language models (LLMs) to enhance patient safety within Singapore's healthcare system. These advancements can improve safety through systemic capabilities or rapid data analysis.
AI can assist prescribing physicians by reducing unnecessary antibiotic prescriptions and personalizing optimal dosages for patients. SGH has developed the 'Augmented Intelligence in Infectious Diseases' model to promote more judicious antibiotic use. Furthermore, AI can enhance pharmacovigilance, the monitoring of drug effects post-licensure, by automating the analysis of vast datasets, including social media, to identify early adverse signals, such as potential risks associated with Covid-19 mRNA vaccines like myocarditis.
The technology also holds promise for detecting rare and undiagnosed diseases by analyzing electronic health records and clinical documents to identify patterns that might otherwise be missed. Generative AI is not intended to replace clinicians but to augment their decision-making and improve safety while retaining human oversight.
Singapore's Ministry of Health (MOH) has identified the use of AI in healthcare as an emerging regulatory issue. The MOH, along with the Health Sciences Authority (HSA) and the tech agency Synapxe, has developed the 'MOH Artificial Intelligence in Healthcare Guidelines' (AIHGle). These guidelines establish best practices for the safe and responsible development and implementation of AI in healthcare settings.
AIHGle, a living document, was initially published in October 2021, prior to the widespread introduction of generative AI tools like ChatGPT. The emergence of these technologies has introduced new considerations and risks that need to be captured in the guidelines. Despite AI's potential, significant challenges remain, including ongoing research and the need for rigorous validation in clinical environments.