Workflow headaches, patient fears and lack of regulation among barriers to artificial intelligence in healthcare

AI
Artificial intelligence tech can cause a number of challenges for providers looking to use it.

Artificial intelligence technology has promise for healthcare, but providers looking to deploy the technology could run into some significant challenges. 

AI tools are being applied to cybersecurity, precision medicine, wellness and a variety of other major initiatives in the industry. But for them to truly work, providers must analyze the costs effectively and tie the tech to a particular issue, write Jennifer S. Geeter and Dale C. Van Demark, healthcare attorneys at McDermott Will & Emery, in an article for Hospitals & Health Networks. 

RELATED: DOD researchers want to peek into the black box of AI, which could impact adoption in healthcare 

Additional challenges could include: 

  • Workflow concerns: If the AI takes over the role of a person, it will require significant "socialization" for staff members with the tech, they said. 
  • Patient pushback: Effective adoption of artificial intelligence will require addressing consumer concerns, including their fears. Patients may feel "forced" to interact with the tech. 
  • Regulatory delays: Use of AI is just now being explored from a policy and regulatory perspective, and that should be monitored. 

RELATED: Plenty of buzz for AI in healthcare, but are any systems actually using it? 

Experts have warned that the sky-high expectations for AI should be tempered. Stanford researchers said that the tech has reached the “peak of inflated expectations,” and looking at the future of it in that way could actually derail its effectiveness.

Instead, providers should focus on an approach that takes advantage of the best that both the tech and humans can do together. 

Other concerns include that AI could make already-existing healthcare disparities worse, as the poorest patients would have limited access to the technology. For it to truly succeed, the industry must solve its "data problem," which includes reaching these underserved populations to gather more information on them. 

Data collection and interoperability are significant shortcomings for one of the most high-profile AI technologies, IBM Watson.