Image: Spine AI – Developed by the Department of Diagnostic Imaging at the National University Hospital, it slashes the time needed for radiologists to interpret MRI scans by more than half, enabling faster diagnosis and treatment of lumbar spinal stenosis.

Modern healthcare systems face a trio of common challenges: ageing populations, workforce shortages, and rising expectations coupled with escalating costs. Many countries have begun to shift to preventive, community-based care models to optimize resources (i), however, it will take time for these changes to take effect. The integration of artificial intelligence (AI) into healthcare presents a transformative solution with immediate potential. By optimizing workflows to support healthcare workers and enhance patient care, AI offers a scalable, rapid solution to address critical gaps (ii), (iii).

The National University Health System (NUHS) in Singapore is an example of how AI-driven innovation can revolutionize healthcare delivery to meet rising demands, setting a global benchmark for efficiency and quality in patient care. We asked Ms Sandy Ho, Assistant Professor Ian Mathews, and Professor Aymeric Lim to share insights into Singapore’s approach.

The context: Singapore’s healthcare system

Singapore’s healthcare system is widely recognized as a global leader in efficiency and outcomes, delivering high-quality health indicators at a fraction of the expenditure seen in comparable countries. Ranked first worldwide in 2023 with a health index score of 86.9 (iv), Singapore’s success is driven by a long-term strategy built on three pillars: a culture of personal responsibility for health, a financing model that promotes the most appropriate use of services, and a high-quality workforce supported by strategic planning (v), (vi), (vii).

Singapore’s careful planning and practical approach delivers strong performances across key health metrics such as infant mortality, life expectancy, and disability-adjusted life years (viii). This forward-looking model delivers exceptional care that is also cost-effective. Achieving these outcomes at less than half the cost of comparable countries reflects strong government oversight, effective cost-control measures, and the public sector’s role in setting benchmarks for quality and efficiency.

Building the foundation: The creation of NUHS’ AI platform

As one of the three main healthcare clusters and its first academic medical centre, NUHS plays a critical role in upholding Singapore’s high standards of care. By embedding AI into its operations, NUHS is addressing rising healthcare demands while maintaining quality and cost-effectiveness. This demonstrates how technology, as part of strategic planning, can enhance care without compromising patient-centred values (ix).

The NUHS AI platform was designed with scalability, data integration, and interoperability in mind. The platform seamlessly fits into Singapore’s healthcare ecosystem and meets the nation’s rigorous data privacy and governance requirements.

AI platforms used in NUHS

Discovery AI and Endeavour AI are two distinct platforms with complementary roles in NUHS’s AI ecosystem.

• Discovery AI is a research-focused platform that supports predictive analytics and precision medicine. It enables clinicians to make use of large datasets for early diagnosis and disease prevention.
• Endeavour AI is a production-oriented platform, designed to optimize daily clinical and operational processes.

While Discovery AI provides insights for long-term patient outcomes, Endeavour AI enhances real-time decision-making, resource allocation, and workflow efficiency within NUHS facilities.

Synchronization with electronic health records

NUHS’s AI platforms are fully synchronized with Singapore’s National Electronic Health Record (NEHR) system, powered by Epic. This ensures seamless data sharing across healthcare providers – providing a comprehensive, unified view of patient records. By synchronizing with the NEHR, NUHS can provide continuity of care across multiple institutions, allowing clinicians to make informed decisions based on real-time, standardized health information. This integration also supports cross-institutional research, enabling data-driven insights that enhance Singapore’s healthcare system.

AI tools in action – Supporting NUHS’s healthcare vision

1   The NUHS App: Comprehensive patient care in one platform

The NUHS App is central to NUHS’s patient engagement strategy, providing a single platform that covers all aspects of patient care. With over 1.1 million downloads and over 25,000 patients using it daily, the app allows patients to consult an AI-enabled chatbot, manage appointments, access test results, request prescription refills, and consult with clinicians through telemedicine services. It empowers patients to take charge of their healthcare journey while also providing NUHS with valuable data for personalized care. The app’s functionality aligns with NUHS’s mission of patient-centric care, offering convenience and accessibility to Singapore’s multi-ethnic population.

2   Bots available for staff efficiency

To further improve operational efficiency, NUHS uses bots with different functions:

a) Clinical Bot: RUSSELL – designed to support clinicians by providing quick access to summaries of referrals, medical reports, and other essential documents, allowing them to make faster and more informed decisions.

b) Operations Bot: BOT-NUHS Productivity Suite – assists non-clinical staff with administrative tasks such as drafting emails, summarizing meeting minutes, consulting corporate policies, analysing user-uploaded documents and data, streamlining day-to-day operations, and allowing staff to focus on high-priority activities.

3   Manpower efficiency tools

AI solutions like automated scheduling and predictive analytics dashboards for patient flow help optimize staff allocation, ensuring that healthcare providers are deployed effectively based on demand patterns. This minimizes staff burnout and improves overall productivity, allowing NUHS to address workforce shortages more effectively.

4   AI for operational and clinical efficiency

The integration of MediVoice, a speech-to-text tool for clinical consultations, helps automate documentation, reducing the administrative burden on clinicians. AI-powered decision-support tools, such as the AI Tumor Board and medical imaging analysis algorithms, enhance diagnostic accuracy and facilitate early detection of diseases. By integrating these tools into clinical workflows, NUHS enables clinicians to provide evidence-based care, improving patient outcomes and reducing the risk of medical errors.

Proactive AI governance at NUHS

The use of AI in healthcare carries significant risks, where errors can have life-threatening consequences. Studies highlight that poorly implemented AI in clinical settings can lead to diagnostic inaccuracies, biased treatment recommendations, and unintended harm (x).

Recognizing these risks, NUHS has adopted a proactive approach to AI governance, ensuring that all AI applications align with principles of fairness, responsibility, transparency, trust, and patient-centeredness. Guided by national frameworks, NUHS’s governance policies include oversight by the NUHS Governance Committee. They follow standards set by the Personal Data Protection Commission (PDPC) Model AI Governance Framework (2020), the Ministry of Health’s AI in Healthcare Guidelines (2021), and Health Sciences Authority (HSA) regulations on software as medical device (xi), (xii).

Key principles in AI governance

At the heart of NUHS’s AI governance approach is a commitment to clinician accountability and patient safety. Clinicians retain full legal responsibility for patient outcomes, with AI designed to support – not replace – their judgment. This “human-in-the-loop” approach ensures that AI supports clinical decision-making while safeguarding professional standards and ethics (xiii). Policies within NUHS are adaptable, starting with lighter controls that can be scaled as needed to balance regulatory compliance with innovation. This maintains a safe and engaging environment for both clinicians and patients.

NUHS AI Governance Committee and policy implementation

The NUHS AI Governance Committee includes representatives from both clinical and non-clinical areas, overseeing the integration of AI across the organization. This committee ensures that all AI implementations align with NUHS’s core principles, regulatory standards, and ethical considerations, helping to create a robust governance structure that supports safe and effective use of AI.

Key governance measures

NUHS has established comprehensive governance measures that uphold transparency, trust, and patient-centred care within a regulated framework. These measures not only prioritize patient safety and ethical AI use but also encourage ongoing innovation, ensuring that AI is deployed responsibly to benefit patients and healthcare providers alike (xiv), (xv).

Collaborative forces behind NUHS AI initiatives

NUHS’s AI journey is supported by a strong network of stakeholders:

• KROI (Kent Ridge Office of Innovation) facilitates training in innovation and AI while working with all levels of staff to develop AI tools aligned with clinical and operational objectives.
• GCTO (Group Chief Technology Office) drives the majority of AI development, oversees the implementation and integration of AI systems and manages the technical infrastructure across NUHS.
• GCMIO (Group Chief Medical Informatics Office) ensures that AI applications like Discovery AI and Endeavour AI are clinically viable and adhere to strict governance standards for patient safety.
• GCDO (Group Chief Data Office) ensures robust data governance frameworks to support the ethical and safe use of AI.

A vision for the future: Personalized and preventative care

NUHS’ ongoing commitment to AI extends to envisioning a future of personalized, preventive care. AI-driven predictive analytics and precision medicine are part of this future, promising earlier diagnoses and customized treatment plans based on patient specific data. NUHS is also exploring how AI can support Industry 5.0’s vision of humancentric healthcare, where cognitive system work alongside healthcare providers to enhance – not replace – human judgement and empathy.

References

(i) Jones, K. (2020, January 15). The Problem of an Aging Global Population, Shown by Country. Visual Capitalist. https://www.visualcapitalist.com/aging-global-population-problem/

(ii) Spatharou, A., Hieronimus, S., & Jenkins, J. (2020, March 10). Transforming healthcare with AI: The impact on the workforce and organizations. McKinsey & Company. https://www.mckinsey.com/industries/healthcare/our-insights/transforming-healthcare-with-ai

(iii) Garrett, R. C. (2024, January 17). How AI can transform patient care and treatment. World Economic Forum. https://www.weforum.org/stories/2024/01/how-ai-can-transform-patient-care-and-treatment/

(iv) Statista (2024, September 24). Ranking of health and health systems of countries worldwide in 2023 https://www.statista.com/statistics/1376359/health-and-health-system-ranking-of-countries-worldwide/

(v) Tikkanen, R., Osborn, R., Mossialos, E., Djordjevic, A., & Wharton, G. A. (2020, June 5). International Health Care System Profiles Singapore. The Commonwealth Fund. https://www.commonwealthfund.org/international-health-policy-center/countries/singapore

(vi) Tan, C. C., Lam, C. S. P., Matchar, D. B., Zee, Y. K., & Wong, J. E. L. (2021, September 18). Singapore’s health-care system: key features, challenges, and shifts. The Lancet, 398(10305), 1091–1104. https://doi.org/10.1016/s0140-6736(21)00252-x

(vii) Healthier SG. (2022, September 21). The White Paper on Healthier SG. https://www.healthiersg.gov.sg/resources/white paper/

(viii) Ramesh, M., & Bali, A. S. (2019 S ept ember ). The remarkable healthcare performance in Singapore. In Great Policy Successes (pp. 42–62). https://doi.org/10.1093/oso/9780198843719.003.0003

(ix) Bowen, M. (2022, May 9). NUHS Singapore operationalizes AI to transform healthcare. Intelligent CIO. https://www.intelligentcio.com/apac/2022/05/09/nuhs-singapore-operationalizes-ai-to-transform-healthcare/

(x) Topol, E. J. (2019, January 7). High-performance medicine: the convergence of human and artificial intelligence. Nature Medicine, 25(1), 44–56. https://doi.org/10.1038/s41591-018-0300-7

(xi) PDPC, Personal Data Protection Commission Singapore. https://www.pdpc.gov.sg/help-and-resources/2020/01/model-ai-governance-framework

(xii) MOH, Ministry of Health Singapore. (Last updated 2024, November 21). https://www.moh.gov.sg/others/health-regulation/emerging-regulatory-policy-issues#d3e1e274f4936d6ecafc82c483cec12b

(xiii) United States Government Accountability Office, Technology Assessment, Artificial Intelligence in Healthcare. (2022, September). https://ddei5-0-ctp.trendmicro.com/wis/clicktime/v1/query?url=https%3a%2f%2fdigirepo.nlm.nih.gov%2fmaster%2fborndig%2f9918506288306676%2f9918506288306676.pdf&umid=81C4A0A6-2763-E006-8EFC-1881011C0F18&auth=d80153f3f03c6c2919581dd96b360379d3b69ad6-d215c56c836e1ef43509817eb2a733f473806da0

(xiv) Thomas Wiegand, Ramesh Krishnamurthy, Monique Kuglitsch, Naomi Lee, Sameer Pujari, Marc el Salathé, Markus Wenzel, Shan Xu, The Lancet (2019, July 6). https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(19)30762-7/abstract

(xv) WHO Guidance, (2021, June 28). Ethics and governance of artificial intelligence for health, World Health Organization. https://www.who.int/publications/i/item/9789240029200

With thanks to Ms Sandy Ho, Assistant Professor Ian Mathews, and Professor Aymeric Lim of National University Health System (NUHS) in Singapore.