Frontline Voices: What Artificial Intelligence and Value-based Models mean for Not-for-Profit Hospitals and Health Systems

Context: 

Although Artificial intelligence has been around for 50 years and has experienced several starts and stops, the last 5 to 10 years have seen a considerable uptick in adoption, especially in healthcare. It’s embedded now in machine learning that enables faster and more precise imaging studies, clinical decision support tools in electronic medical records systems and many more. In recent months, its potential to play a bigger role, possibly replacing physician judgement among others, has received added attention.

The November 2022, the announcement of OpenAI’s ChatGPT platform drew widespread attention with speculation it might displace clinicians in diagnosing and treatment planning for patients. On March 22, 2023, tech moguls Elon Musk, Steve Wozniak and Andrew Yang called for a 6-month moratorium on generative AI stating: “Should we develop nonhuman minds that might eventually outnumber, outsmart, obsolete and replace us? Should we risk loss of control of our civilization? We call on all AI labs to immediately pause for at least 6 months the training of AI systems more powerful than GPT-4.” (1)  To date, more than 13,000 have signed on to their appeal. Per Lumeris CTO Jean-Claude Saghbini “Putting aside our own opinions as to whether or not a moratorium should be implemented, our recent experience of the last three years in the inability to have effective cross-governmental alignment on policy to fight the COVID pandemic suggests that global alignment on AI policy will be impossible”.

There’s widespread belief generative AI and GPT-4 are game changers in healthcare.

How, what, when and how much ($$$) are the big questions. The near-term issues associated with implementation–data-security, workforce usefulness, regulation, investment costs—are expected to be resolved eventually. Thus, it is highly likely that health systems, medical groups, health insurers and retail and digital health solution providers will operate in a widely-expanded AI-enabled world in the next 3-5 years.  

Questions: 

What role will AI and ChatGPT play in hospitals/health systems and other provider settings? Will development of AI systems more powerful than GPT-4 be suspended in response to the appeal? How is your organization preparing for the next wave of AI?

Key Takeaways from Discussion:

• ‘Generative AI will not take the place of clinician judgement anytime soon. The processes of diagnosing and treating patients, especially complex conditions, will not be displaced. However, in primary and preventive health where standardization is more attainable, it will have profound impact perhaps sooner than in other areas.’
• ‘GPT-4 et al will have profound impact on the delivery of healthcare and hospital operations, but there are many unknowns and risks associated with its use beyond routine tasks that can be standardized based on pattern recognition. ‘
• ‘Continued development of platform solutions using GPT-4 and others in healthcare and other industries will accelerate. The moratorium will not happen. There’s too much at stake for investors and users.’
• ‘Non-profit hospitals and health systems are struggling financially as a result of the supply and labor cost increases, declining reimbursement from payers and negative returns on investing activities (non-operating income). Caution is key, so AI-related investing will be conservative in the near-term. An exception would be AI solutions that mitigate workforce shortages or reduce administrative costs for documentation.’