Health systems aren’t short on digital health pilots. They’re short on digital health pilots that actually change outcomes. Across the industry, promising initiatives too often run into the same wall: technology that works in a demo but stalls against the realities of clinician workflows, patient behavior and change management.
During Becker’s 16th Annual Meeting, a panel of healthcare leaders discussed what separates digital health strategies that deliver from those that just look good on a slide.
The session featured:
● Tesa Anewishki, President and Chief Executive Officer, Loretto Hospital (Chicago)
● Steven Berkow, Head of Value-based Care Market Strategy, InterSystems
● Amaka Eneanya, MD, former Chief Transformation Officer, Emory Healthcare (Atlanta)
● Donna Roach, Chief Digital and Information Officer, University of Utah Health (Salt Lake City)
Thoughtful implementation trumps flashy technology
Too often, digital health vendors promise transformation, selling their technology solutions as seamless and intuitive. But patient feedback often suggests that many digital health initiatives actually end up fragmenting care.
Dr. Eneanya recounted how an institution she previously worked at entered a partnership with a global technology company to implement their solutions. “It was supposed to ‘revolutionize the patient experience,'” she said, recalling the vendor’s promise. But there was a gross underestimation of the budget, timeline, training and change management required. It didn’t work.
Flash isn’t everything. Thoughtful, nuanced approaches to technology implementation that focus on solving specific problems are essential. Ms. Roach explained that successful implementations of digital technology involve a combination of people, process and technology.
“Technology has become the easy part,” she said. “It’s the people and process that break down when we try to deliver solutions to our patient population.”
Early patient and clinician involvement is essential
To close care gaps, health systems recognize the importance of engaging with specific patient populations. Yet, connecting effectively is often difficult.
Loretto Hospital, in Chicago, struggles to engage with young patients about preventive care. For the University of Utah Health, a major challenge is enrolling more patients in clinical trials.
Every population has unique characteristics, and every hospital and health system has distinct challenges, so taking a one-size-fits-all approach to digital health strategies will likely fall short.
“When we considered how to address social determinants of health around transportation, we looked to partner with Uber for healthcare,” Ms. Anewishki said. “But, when we surveyed patients, we found our older population is used to our free transportation.”
The panelists agreed that a best practice is to involve patients early in the evaluation of any patient-facing digital health solution. “Patient advisory councils are often seen as a necessary checkmark at the end of an initiative,” Dr. Eneanya said. “But you need to involve patients in the design early on. Make sure they are at the center, not an afterthought.”
Similarly, engaging clinicians early is also important, because important drivers of patient experience are care team dynamics and communication. Yet, digital health solutions too often overwhelm front-line clinicians with information.
“We forget that less is more,” Mr. Berkow said. “Clinicians don’t have time to distill large amounts of data in the five minutes they’re interacting with the patient. We need to meet providers where they are technologically. Call out the one or two things they can influence at that moment in the modality that works best for their workflow.”
Healthcare AI solutions hold promise, but require clean data
AI has already become an efficient, effective way to address key patient and business issues. For example, after surveying patients, the University of Utah Health realized the organization wasn’t connecting with what people want from digital health.
Within 48 hours, the team created an internal AI tool called Swoop that uses patient survey data to identify sources of fragmentation and how those can be improved. “It’s shown us some disconnects we wouldn’t have seen without paying attention to what patients are telling us,” Ms. Roach said.
However, as healthcare organizations adopt generative and agentic AI solutions to advance the Quadruple Aim, the importance of clean and normalized data can’t be overlooked. Generative AI may experience drift and hallucinations. In addition, biases in data sets can show up in AI applications.
The need for structured, machine-readable data is also imperative as healthcare organizations move from generative AI to deploying AI agents to complete tasks. It’s a challenge Mr. Berkow sees firsthand at InterSystems, whose technology supports interoperability infrastructure for many large health systems, powers a majority of health information exchanges, and is increasingly used to enable AI applications across major U.S. providers.
“There are dramatic ramifications in terms of the quality of data assets needed to fuel AI,” Mr. Berkow said. “Let’s put the horse back in front of the cart, and the horse is the quality of your data assets.”
At the Becker's 11th Annual IT + Revenue Cycle Conference: The Future of AI & Digital Health, taking place September 14–17 in Chicago, healthcare executives and digital leaders from across the country will come together to explore how AI, interoperability, cybersecurity, and revenue cycle innovation are transforming care delivery, strengthening financial performance, and driving the next era of digital health. Apply for complimentary registration now.
