“That is simply due to the fact that we have yet to find a product or service we felt represented sufficient benefit to justify the cost,” said Nicholas Nussbaum, MD, director of medical affairs and community services at Adams Medical Group. “To be frank, most of what is being marketed as ‘AI’ is either just an enhanced version of a search engine or machine learning, neither of which remotely justify the term ‘intelligence.'”
There has been a rush of technology platforms aiming to streamline operations and automate routine tasks, including revenue cycle and documentation functions. Many health systems have started pilots with ambient listening companies to support clinicians, but it takes time to refine these products and realize the benefits.
“I’m certain we will get to that point, but we aren’t there yet,” he said. “When the technology reaches that point, we want to be ready to move. Until then, we are keeping our powder dry and letting everyone else spend their money underwriting the R&D for the software companies.”
There are risks involved in early adoption. It takes time and resources to incorporate new technology platforms into existing workflows, training staff members (even for small pilots) and tracking the results. There are also AI bias, patient data ownership and cybersecurity concerns associated with moving too quickly on AI models. Health systems have valuable resources for technology companies looking to test and improve their products.
“Leverage always works in both directions, meaning that the same lever that you are using to lift something can also swing back and crush you,” said Dr. Nussbaum. “Overspending on a product that does not or cannot deliver at the level promised may prove to be exactly that kind of crushing leverage, especially if the funding for that product or project was in the form of borrowed money.”
