Can AI and digital health solve health care's biggest challenges? Insights from the Conference on Health IT and Analytics

What if AI could make the health care system work better? Would people have better access to care, greater insights into their health, and an easier time navigating the process? These possibilities, and more, are part of the discussion as AI creates more and more opportunities in health care.

The Center for Digital Health and Artificial Intelligence at the Johns Hopkins Carey Business School recently hosted the 14th Annual Conference on Health IT and Analytics, bringing together top researchers, policymakers, and industry experts to shape the future of digital health and AI in business.

"CHITA serves as a vital forum for advancing our understanding of strategy, policy, and systems in health IT and analytics," said Ritu Agarwal, CDHAI co-director and conference co-chair. "By fostering collaboration among academia, government, and industry, we aim to drive innovations that make tangible impacts in business and policy." 

"AI/ML applications in health care must be designed keeping equity considerations front and center," said Gordon Gao, co-director of CDHAI and CHITA conference co-chair. "Without intentional design informed by diverse perspectives, we run the risk of amplifying societal biases and exacerbating health disparities."

Driving innovations

Throughout the two-day conference, attendees engaged in 70 cutting-edge research presentations covering telemedicine, algorithmic bias, health disparities, online platforms, and AI implementation in clinical settings. The presentations showcased innovative solutions and sparked critical conversations around equity, privacy, and the responsible deployment of AI in health care. 

Joan Horenstein, managing director at Accenture Federal Services, emphasized in her keynote that effectively leveraging data will be key to realizing AI's potential in health care. Drawing from her extensive experience leading large-scale IT programs, including the backbone of healthcare.gov, Horenstein highlighted the power of data-driven design. 

"Data domain-driven design is incredibly powerful and adaptive," she said. "If there are capabilities that could help better understand data and be able to detect those [anomalies], those would be extremely powerful." Such data-savvy approaches, she suggested, could help make AI solutions more robust and impactful.

David Sontag, Professor of Electrical Engineering and Computer Science at MIT and CEO of Layer Health, shared insights into his MIT group's focus on Large Language Model (LLM)-based models for enhancing patient-clinician interaction, stating, “We have the incredible chance to improve patients' understanding of their own health data.” Sontag highlighted simplicity and real patient validation as key tenets of their approach, emphasizing the fusion of multiple LLMs to enhance the intelligibility and accuracy of outputs, with a focus on patient needs and outcomes. 

A panel discussion on human capital explored how AI will reshape health care labor markets and education. The panelists included Michael Rosenbaum, CEO of Arena Analytics; Laurie Buis from the University of Michigan; and Prasanna (Sonny) Tambe from the University of Pennsylvania. Buis emphasized the need to intentionally transform clinical processes and roles to realize AI's full potential. 

"Until we start thinking more thoughtfully about well-implemented new technologies and new processes and being very intentional about how we transform clinical care, it's really difficult to understand how to train the people who will provide that care," she said.

Transformation, revolution, and empowerment

Aneesh Chopra, president of CareJourney and former U.S. CTO, played a key role in the $35 billion project under the Obama administration to digitize medical records across hospitals and physician groups nationwide. At CHITA, Chopra highlighted the significant progress made in recent years to open up health care data and improve interoperability between systems.

"We've gone from having no [data sharing] standards to [having them] in production."

Looking ahead, Chopra envisions a future where generative AI can provide hyper-personalized health care guidance to individuals, much like TurboTax simplifies personal taxes. He predicted that "risk-bearing primary care [and] ACL networks" are best positioned to leverage AI in this way.

However, Chopra cautioned that dwindling public trust in the health care system could hinder people's willingness to share the personal data that fuels AI and digital health innovation. "It is on the systems shapers to win back that trust," he stressed, underscoring the urgent need to address this challenge.

As AI and digital health evolve rapidly, forums like CHITA play a vital role in grappling with the immense potential and thorny challenges on the horizon. By bringing together diverse stakeholders to share cutting-edge research and engage in candid discussions, CHITA is helping to light the path forward—toward a future where technology and data revolutionize healthcare access, experience, and outcomes for all.

CDHAI organized CHITA in partnership with the University of Michigan School of Public Health and Harvard Medical School. Learn more about CHITA.