Panel, Conference Presentation
Bot or Not?: AI in Health Care | Future of Health Summit 2024
- The outlook anticipates a shift from reactive, one-size-fits-all care to proactive, personalized interventions by 2030, driven by AI that identifies patient subgroups (such as five distinct obesity clusters) and optimizes treatments based on specific variables like hormone levels.
- By 2025, the primary barrier to progress is identified as psychological behavioral change rather than technological constraints like storage or compute, necessitating global cooperation on data privacy and reciprocity to enable country-scale AI validation.
- Industry transformation relies on standardizing data collection and measurement; current clinical trial data is noted as insufficiently heterogeneous, predominantly representing urban professionals, while real-world evidence and multimodal data from wearables must be embedded directly into care pathways.
- AI applications are projected to significantly improve diagnostic accuracy and speed, with potential capabilities including fivefold improvement over humans in endoscopy, detecting stage-zero pancreatic cancer, and identifying physiological patterns missed by physicians, though three-quarters of current FDA-approved devices remain in radiology.
- A severe workforce crisis is expected, with a predicted shortage of 18 million healthcare professionals by 2030, alongside existing burnout and recruitment challenges, driving the need for AI solutions that provide efficiency, scale, and connectivity for care teams.
- Strategic goals include curating the world's data to touch four billion people by 2030 and reimagining care for the top 20 conditions by cost and prevalence to lower costs, reduce invasiveness, and empower individuals to manage their health behavior daily.
- Successful implementation requires healthcare professionals to cross-skill and upskill in digital data to participate meaningfully, while initial AI integration may be limited to efficiency gains over a one-to-two-year period before demonstrating broader stakeholder impact.
- Risks to realizing these outcomes include the lack of data standardization which creates hurdles for building reliable systems, and the risk that AI alone cannot transform healthcare without these foundational data practices in place.