Division of Research investigators among key organizers and authors of JAMA AI Summit report
Artificial intelligence (AI) may offer massive benefits but also carries risks for health and health care, requiring “an ecosystem capable of rapid, efficient, robust, and generalizable knowledge about the consequences of these tools on health,” according to a national workgroup organized by JAMA Network.
A number of co-authors of the workgroup’s report, published Oct. 13, are from Kaiser Permanente and The Permanente Medical Group (TPMG), including several Division of Research (DOR) investigators who helped organize the summit.
“As AI rapidly gains traction in both our daily lives and in our health care workplaces, we thought this was the right time to take a broad look at how these technologies can be used most effectively, and what organizations need to be doing to ensure AI is effective and safe,” said Vincent Liu, MD, MS, DOR research scientist and TPMG chief data officer, and one of the conference organizers. “Among our goals is that AI ultimately improves patient care and augments our capabilities as doctors and care teams.”
The JAMA workgroup included more than 60 experts, representing multiple disciplines, who met over 2 days to discuss how health care organizations are using AI.
Their report noted the wide range of ways AI-related technologies are being applied in health and health care. These include clinical tools such as sepsis alerts, mobile health apps used by consumers, and tools to improve business operations as well as hybrid AI technologies that impact both operations and clinical care.
However, the workgroup concluded that evaluation of the safety and efficacy of these tools, along with their regulation, is inconsistent. Its report made 4 main recommendations:
- Involve end users in the development of AI tools and throughout their life cycle, and increase partnership of developers, regulators, and health care systems in evaluation as the tools are used.
- Develop and disseminate tools for rapid, efficient, and robust evaluation of AI methods’ effectiveness.
- Create a data infrastructure and learning environment for generalizable knowledge about health effects of AI tools in various settings.
- Establish an incentive structure using market forces and policy levels to drive these changes.
“The workgroup process was invaluable in giving us a moment to focus our considerations during a time of rapid change,” said Kristine Lee, MD, TPMG associate executive director. “Even as individual health care organizations may have their own process for evaluating safety and usefulness of this rapidly evolving technology, there are major operational, ethical, and logistic issues being raised that affect all of us. It is critical for us to take charge in working to ensure that AI benefits our patients, clinicians, and communities.”
DOR research scientist and JAMA deputy editor Tracy Lieu, MD, MPH, also helped organize the summit and co-author the report. “The JAMA Network was an ideal convener because the journals serve as a nexus for ideas and discussion among experts from the multiple disciplines involved in AI and health care from the U.S. and other countries,” Lieu said.
Additional co-authors include Andrew Bindman, MD, Kaiser Permanente executive vice president and chief medical officer; Elizabeth McGlynn, MD, former Kaiser Permanente vice president for research and quality measurement; and David Ouyang, MD, a DOR research scientist.
###
About the Kaiser Permanente Division of Research
The Kaiser Permanente Division of Research conducts, publishes, and disseminates epidemiologic and health services research to improve the health and medical care of Kaiser Permanente members and society at large. KPDOR seeks to understand the determinants of illness and well-being and to improve the quality and cost-effectiveness of health care. Currently, DOR’s 720-plus staff, including 73 research and staff scientists, are working on nearly 630 epidemiological and health services research projects. For more information, visit divisionofresearch.kp.org or follow us @KPDOR.
Comments (0)