'AI will dazzle but also disappoint': 6 health system exec predictions on artificial intelligence


Artificial intelligence will play a larger role in healthcare organizations in the future, predict six health system leaders; but not all promises about the technology will come to fruition.

Here are six bold predictions about how AI will evolve in healthcare over the next five years.

Eric Yablonka. CIO and Associate Dean of Technology and Digital Solutions at Stanford Health Care and School of Medicine (Palo Alto, Calif.): A combination of wearables and other biomedical devices, combined with machine learning and artificial intelligence will continue to transform clinical research, treatment protocols and increase the virtual care capabilities of health providers. This will challenge traditional healthcare organizations to compete with emerging retail and virtual providers in ways we have not experienced before. It will also enable healthcare delivery science and bring data scientists to the forefront of improving patient care outcomes.

Alexa Kimball, MD. President and CEO of Harvard Medical Faculty Physicians at Beth Israel Deaconess Medical Center (Boston) and President of Physician Performance: AI will dazzle but also disappoint. A lot of people are banking on AI to transform medicine and substantially solve challenges in healthcare, but I believe the impact will be far less dramatic than we think or hope, in large part because we aren't targeting the right use case. We will certainly learn a lot from this novel technology and it will power us to continue to improve our understanding of health, disease and healthcare delivery.

However, using AI as a diagnostic tool has significant limitations that aren't based in technology, but epidemiology. Common things will remain common and won't require AI, and rare things require a precise knowledge that we don't have in hand for AI to master – because they don't happen very often. As a result, the potential clinical impact in the diagnostics space is limited and it won't matter as much as other forms of testing and diagnostics that get at the underlying biology.

Edward Lee, MD. CIO of The Permanente Federation (Oakland, Calif.): In five years, physicians will no longer need to manually document their notes into the EHR. Instead, artificial intelligence will capture all the pertinent information from the patient-physician encounter. This will enable physicians to spend quality time with their patients instead of worrying about writing their notes or placing orders in a computer system. Joy and meaning for physicians will increase, physician burnout will decrease, and above all, patient care will improve.

Michael Pfeffer, MD. Assistant Vice Chancellor and CIO of UCLA Health: Health IT will enable each patient to have a unified, interactive view of their health information regardless of place of care or type of clinical data (i.e. phenotypic, genomic, imaging). AI-based apps will help make sense of their data, taking into account social determinants of health and potential health disparities to improve health equity and health literacy. This intelligence will be paired with personal health preferences and data on health provider quality, access, pricing and satisfaction to help patients make truly informed decisions about their care.

Patrick McCarthy, MD. Vice President of Northwestern Medical Group at Northwestern Medicine (Chicago): The next five years will demonstrate a sharp increase in the routine use of AI in clinical practice. The applications will be widespread from transforming our most basic tool the stethoscope, to wider distribution of advanced imaging acquisition and interpretation, and applying machine learning for individual patients to identify rare diseases, predict procedural risk, and determine best care pathways.

A real-world example of AI-enhanced tools for clinical use is Eko. Eko's suite of advanced stethoscopes are powered by advanced machine learning algorithms that are FDA-cleared to detect AFib and heart murmurs. This type of practical AI analysis will greatly improve patient care, specifically by the early detection of the leading cause of death in the U.S., heart disease.

Rich Temple. Vice President and CIO of Deborah Heart and Lung Center (Brown Mills, N.J.): I see artificial intelligence in many forms really gaining a significant toehold in healthcare information technology. Robotic process automation will be a much bigger player to streamline the intense manual efforts by healthcare staff to perform functions such as prior authorization management and referral management. I also see AI being used to perform detailed analyses of diagnostic images, and supporting physicians in identifying patterns in images through pattern-matching with image databases from all over the world. While I know AI presents a concern to many in terms of 'computers replacing humans doing jobs', I don't see AI as a threat to human employment, but rather an enhancer of humans doing their jobs by providing predictable and dependable tools that humans can do to perform their job duties even better.

Copyright © 2021 Becker's Healthcare. All Rights Reserved. Privacy Policy. Cookie Policy. Linking and Reprinting Policy.


Featured Whitepapers

Featured Webinars