Healthcare Systems Must Shift From Crisis Care to Prevention, Says Nebius Executive
Asia Pacific healthcare systems are drowning in workforce shortages and aging populations. Artificial intelligence offers a way out-but only if hospitals stop treating it as a software purchase and start redesigning how care flows.
That's the core argument from Dr. Ilya Burkov, Global Head of Healthcare and Lifesciences at Nebius, who sees the fastest wins in chronic disease management and primary care triage. The opportunity is straightforward: prevent patients from reaching crisis point.
Where AI Makes the Biggest Difference Right Now
Reducing avoidable hospitalizations is the clearest target. Heart failure, COPD, asthma, and diabetes patients often end up in emergency departments because they deteriorated in the community. Earlier intervention and stronger outpatient monitoring can stop that cycle.
AI can identify high-risk patients before they worsen, prioritize follow-ups, and route referrals more accurately. The result: fewer emergency admissions and less pressure on hospital systems already stretched thin.
Burkov uses a traffic control analogy. AI spots congestion and risk before the system becomes overwhelmed, allowing clinicians to act sooner rather than react later.
Building Trust in AI-Guided Diagnostics
Non-clinical staff conducting routine scans with AI support sounds disruptive. It works only if clinicians remain accountable and in control.
Consider AI-guided echocardiography: technicians capture images and measurements with AI assistance, but clinicians review results and make diagnoses. The distinction matters. AI augments capacity; it doesn't replace judgment.
Four safeguards build clinician and regulator trust:
- Clinician oversight at every step
- Traceable outputs showing what evidence informed each recommendation
- Bias monitoring across patient populations
- Version control so providers know how current the model is
Healthcare also needs clearer regulatory pathways that allow innovation to move at technology speed without sacrificing patient safety or privacy. Aviation achieved this with strict protocols and audit trails. Healthcare can do the same.
The Reimbursement Problem
Most healthcare systems still reward treatment over prevention. Hospitals get paid for treating acute illness, not for stopping it from happening.
An AI system that reduces avoidable admissions improves outcomes and cuts costs downstream. But those savings may appear in a different part of the system entirely, leaving the hospital that invested in the technology with no financial benefit.
Policymakers need to shift toward outcome-based reimbursement models that reward fewer hospitalizations, faster diagnosis, and better chronic disease management. Shared savings models-where hospitals, primary care providers, and payers all benefit from system-wide cost reductions-can align incentives across the entire network.
Why Most AI Deployments Fail in Healthcare
The biggest mistake is treating AI as a technology implementation rather than a system redesign. AI changes workflows, responsibilities, and decision-making across the entire patient journey.
When hospitals approach deployment as a purely technical exercise, they underestimate clinician trust, nursing workflows, compliance, ethics, and operational integration. Multidisciplinary teams-clinicians, nurses, AI specialists, ethics teams, compliance leaders-need to be involved from day one.
Another common failure: adding complexity to already overstretched environments. If AI creates more clicks, more friction, or unclear accountability, clinicians won't use it. The strongest deployments are almost invisible. They fit naturally into existing workflows, surface relevant insights at the right moment, and reduce administrative burden.
Which Countries Are Best Positioned
Singapore combines strong digital infrastructure, policy alignment, and a national focus on preventive healthcare. That combination lets predictive AI and risk stratification scale quickly.
South Korea has similar foundations in digital maturity and healthcare technology adoption. Southeast Asian systems have a different advantage: they can leapfrog legacy infrastructure and move directly toward AI-enabled community diagnostics and distributed care models.
Countries that align regulation and infrastructure to patient needs will scale successfully. Without both moving together, AI stays stuck in pilot programs.
The Five-Year Outlook: Clinicians and AI as Co-Pilots
AI will fade from the spotlight and become background infrastructure in everyday clinical care. Clinicians will remain central decision-makers, but AI will function as a clinical co-pilot: summarizing records, identifying risk patterns, supporting guideline-directed care, and cutting administrative workload.
Allied healthcare workers will take on broader responsibilities in diagnostics, monitoring, and follow-up, working under clinician supervision with AI support.
The relationship will become more collaborative and team-based. AI handles repetitive pattern recognition and data synthesis. Clinicians focus on judgment, communication, and patient trust-the parts only humans can provide.
The goal is not less human care. It's giving healthcare professionals more time for the work that matters most.
Related: AI for Healthcare and AI Data Analysis resources for professionals implementing these systems.
Your membership also unlocks:
