The hype around AI in healthcare is undeniable. AI startups are dominating the digital health investment market, companies like Abridge and Ambience Healthcare are surpassing unicorn status, and the White House recently issued an action plan to advance AI’s use in critical sectors like healthcare.
However, one digital health executive thinks AI progress in the field could stall soon unless one key problem is fixed.
The issue is data infrastructure, according to Mitesh Rao, CEO of OMNY Health, a national data ecosystem that facilitates medical research. In Rao’s view, scalable AI in healthcare depends on access to data that is expansive, representative and interoperable.
The Hidden Administrative Tasks Draining Small Practices
Small practices play a critical role in healthcare delivery, but they cannot continue to absorb ever-increasing administrative demands without consequences.
Current systems are fragmented and difficult to access, which makes meaningful AI development difficult, he said in an interview last week.
These systems are siloed because information is stored across multiple platforms that lack standardized formats and shared data exchange channels. Additionally, incumbent vendors have little incentive to facilitate better data sharing.
CMS launched a new interoperability initiative last week — and while efforts like this are well-intentioned, they fail to have much impact if they don’t include tangible incentives for incumbents like Epic or Cerner, Rao noted.
This all creates a patchwork map of locked-down data infrastructures — and this makes it challenging for developers to access the data they need to create advanced AI solutions.
The Power of Real World Data to Study Women’s Health at Scale
Veradigm examines key clinical trends, comorbidity profiles, and treatment trends across adolescence, reproductive years, and peri-/post-menopause. Download it today!
“Most of the AI work that is successful in healthcare today is focused on data that is not that hard to access — things like documentation or revenue cycle management, areas that are not necessarily tied to proprietary and deep patient data,” Rao stated.
Data limitations are a difficult hurdle for innovators to overcome as they begin to explore more complex AI use cases in healthcare, especially applications that touch the clinical side of things, he remarked.
Overall, Rao believes healthcare tech leaders need to “build the roads before the Ferraris.” In other words, ambitious AI projects require basic infrastructure first.
Photo: Filograph, Getty Images