When the late Harvard Business School Professor Clayton Christensen would appear before new audiences curious about his deeply-researched book on disruptive innovation in healthcare, The Innovator’s Prescription, he would begin his talks by drawing concentric circles. Then he would explain through example after example how the history of healthcare has been a tale of these circles.
Medical knowledge and procedures, he would show, would start in the center – the most highly trained professionals and the most prestigious institutions. Then, as patterns would develop from the experience they would acquire, and eventually ironclad rules would take shape, care would migrate toward community hospitals, ambulatory practices, lower levels of medical licensure, and even ultimately to the home.
The Role of AI
AI looks set to turbocharge that long-standing story of disruptive innovation. Through finding patterns in diffuse, often unstructured data, AI can help to mainstream care typically provided by sub-specialists in academic medicine. The impacts on both academic medical centers and care providers at the periphery will eventually be enormous.
For elite referral hospitals, AI can free up time for scarce sub-specialists and help to ensure that beds are occupied by the most complex cases. Focus can go toward expanding expertise in delivering emerging types of medicine, such as new types of immunotherapies against cancer that can have hard-to-predict impacts on patients’ immune responses. Academic medical centers will need to ensure that they are reimbursed fairly for treating these challenging, costly patients, and also for their highly-trained staff providing remote support for patients seen at other facilities. They will need to think twice before expanding their numbers of beds, and they will also need to consider the ways that AI will impact how these complex patients are cared for day-to-day.
Community hospitals and ambulatory care sites, by contrast, will need to consider how they take on patients previously seen at the center of health systems. How will AI infuse the way they diagnose, treat, and follow-up with patients? When and how will their care teams involve more highly-trained sub-specialists? What are the implications for facility capacity, skills required, and profitability?
The impacts are likely largest where the supply of specialists is most constrained, such as in many developing countries. Provided that institutions and physicians have incentives to collaborate (and in some countries this is a major proviso), care can be lightly directed by elite medical teams and, with the assistance of AI, take place in secondary or tertiary cities and towns.
When Will Change Come?
Are these scenarios just science-fiction? For certain, the rate of change will be uneven. Physicians will be risk-averse about keeping patients at the periphery, and financial incentives will need to align with what technology enables. But medicine will not need to be entirely computer-driven for many of these impacts to occur. Through tracking patients better, identifying who to refer to the center and at what times, and identifying which patients are most at risk of complications, AI can be a major boost in making human physicians more productive and focused on the right patients for the capabilities that they have.
Change in medicine takes time, but the decade-plus cycles of innovation commonly seen in moving technology from lab bench to bedside may, in AI’s case, be substantially compressed. AI solutions are improving extremely fast, institutions are finding ways to reduce AI-driven error (such as through having separate AI systems do quality checks), and the scarcity of care providers raises the urgency of finding productivity gains.
Actions to Take Today
For existing care providers, there is urgency in determining the business model and resources required to thrive in the coming era. What investments – such as new construction – should be re-thought? What capabilities should the institution recruit and cultivate? Which job descriptions need editing? How does culture need to change?
For would-be insurgents, opportunity is immense. As with disruption in other industries, new entrants are unencumbered by old patterns of behavior and the business models they have already optimized around. They can take a fresh approach to the delivery of care. Already we have seen this occur absent AI – witness the fast rise of Oak Street Health, which zoomed in 11 short years from new company to $10 billion+ acquisition by CVS. AI makes these rapid successes all the more feasible, given that it can remove several constraints traditionally imposed by needing to find scarce providers and sites of care. The insurgents should plot out how AI can combine with new care models and reimbursement strategies to make disruption happen, fast.
Christensen’s study was of medical history, but he used the patterns of the past to predict the future, such as the rise of precision medicine. AI fits those patterns, and the de-centralization it unleashes will profoundly change the structure of care.
Stephen Wunker leads New Markets Advisors, where he consults for leading institutions such as Microsoft and the Mayo Clinic on innovation strategy and capabilities. He was a longtime collaborator with Clayton Christensen, leading his healthcare consulting practice.
