The promises and pitfalls of AI innovations in enhancing patient care

AI-enabled technologies have demonstrated enormous potential for health care, fueling advances in areas across nearly every medical specialty area and as varied as drug development, software-as-a-service, and autonomous and assistive algorithmic analysis of medical images. Importantly, AI systems can address pressing issues such as health care workforce shortages and access challenges for underserved communities. However, these technologies also raise vexing questions, including:

Data: Where does it come from and is it of sufficient quality to be of use for AI algorithms? Who owns and / or controls the data, who has permission to use or exchange it, and for what purpose? And most importantly, how will sensitive patient health information be protected in view of escalating privacy and cybersecurity risks?

Regulatory: Should the AI-enabled product or service be regulated as a medical device (whether or not the implicated algorithm is iterative / self-updating), under which approval scheme(s) and in which jurisdiction(s)? What complexities will arise in view of additional industry-agnostic frameworks?

Coverage and reimbursement: Which government and / or private payers will partner with innovators and under what valuation models? Are there non-traditional payment or other incentive models which may ensure appropriate coverage and reimbursement by, for example, by permitting temporary transitional coverage or prioritizing preventive care?

Stakeholders with AI-enabled systems and tools currently under development must stay on the alert across jurisdictions in this rapidly developing area.

Additional resources