Students, researchers and health care leaders convened for the second annual Healthcare Analytics Symposium at the UCLA Anderson School of Management on May 16.
Hosted by the Morrison Family Center for Marketing and Data Analytics and the UCLA Anderson’s Master of Science in Business Analytics program, participants at the symposium shared perspectives on potential solutions to health care’s most complex data management problems.
The symposium sought to promote an “exchange of ideas” surrounding the potential solutions AI tools could provide to support clinical and business outcomes, said Andres Terech, faculty director of the Morrison Center for Marketing and Data Analytics and adjunct professor of marketing at the School of Management.
Cinthya Gomez, a business analyst in the cultural and linguistic equity department at AltaMed – a non-profit community health center – said she was drawn to attend the symposium to learn more about the technology aspect of health care.
Terech said he drew the inspiration for the theme of the symposium from a conversation he had with Dr. Nasim Afsar, chief health officer of Oracle, who told him 80% of the data that determines a person’s health originates from outside clinical settings.
He said he was struck by the challenge of making complex decisions with only 20% of the necessary data. This spurred him to think about the potential of AI to solve issues stemming from fragmented and inaccessible data.
“There are elements in the health care industry that you don’t see in other industries,” Terech said. “These underground or fundamental characteristics of the industries complicate the potential of AI to deliver a good solution in terms of the diversity of the data.”
Moderated by Howard Park, data and AI advisor at Moxy Analytics, the first panel explored the ways in which two startups, Netomi and Harrison.ai are employing AI-driven tools to provide a better patient experience and improve diagnostic capabilities in medical imaging, respectively,
Dimitry Tran, the co-founder of Harrison.ai, said a challenge the company faces is training and validating AI tools on patient populations that are diverse in socioeconomic and racial backgrounds, as there is limited access to data from underprivileged communities.
To obtain Federal Drug Administration regulation as a medical device, AI tools need to meet a rigorous “anti-bias” requirement to ensure they work across populations, he said in the panel. However, disadvantaged communities are more likely to not be represented in the data, Tran added.
“Having diversity of data is crucial in what we do because close enough is not good enough when it comes to cancer,” he said in the panel.
Robert Alger, a speaker in the third panel and the senior vice president of Health Plan Business Technology Solutions and Services at Kaiser Permanente, said he worked to implement an AI-driven ambient listening tool for clinical documentation across Kaiser Permanente’s 40 hospitals nationwide.
He said his team provided the tool to 1,000 doctors for 10 weeks, but allowed them to continue using it if they wanted to. The tool now has 15,000 users and has created clinical documentations for over 4 million patient encounters, he added.
“The only rule was you have to ask your patient if it is okay,” he said.
Gomez said AltaMed recently implemented a similar ambient listening tool. She added that she found the panels insightful and hopes to see AI-driven tools used to better the language interpretation services her department provides for patients.
As a former medical social worker, Perla Aparicio, a student at Anderson’s Executive MBA program, said that while the operational and business benefits of employing AI-driven tools made sense, she hoped to learn more about the steps organizations were taking to foster patient trust from the “top-down.”
“At the end of the day, we want faster medications, better diagnoses,” she said. “But if my personal information is out there and it’s being consumed by more than the regular doctor, physician and their staff, that’s something I would want to know.
Following the panel discussions, attendees were invited to participate in roundtable discussions with industry executives and data analytics researchers about their experiences implementing AI in their organizations.
Looking forward, Terech said he hopes this symposium series will continue and be a lifelong learning experience for UC and Anderson alumni.
“(The goal is) to bring people together to have the knowledge, to have the experiences to connect with each other and do better together,” Terech said. “That’s really the hope.”
Comments are closed.