Hospitals aren’t ready for AI — but it’s ready for them

Only one in six US hospitals have oversight ready and policies in place to make use of artificial intelligence, according to research by the Center for Connected Medicine (CCM).

Just 16% of executives running healthcare systems said they had a system-wide governance policy in place, with many telling CCM they planned to wait until federal regulations showed the way before creating their own guidelines.

Why wait? Respondents highlighted concerns with both ethics and legal issues. They also cited a lack of internal expertise required to tell if AI actually works in clinical environments, according to the CCM report.

Benchmarking AI for hospitals

Answering such questions is key to implementing AI safely and effectively, says Jeffrey Jones, Senior Vice President of Product Development at UPMC Enterprises, a member of CCM.

“Before adopting generative AI technologies in healthcare, it’s crucial for executives to clearly define their objectives and establish measurable benchmarks,” Jones said in a statement.

“Regular evaluations are essential to adjust strategies as necessary,” he added. “Generative AI is not a one-time fix, but a dynamic tool that requires attention and calibration.”

One area where AI is proving popular is electronic health records (EHR), with more than 70% of respondents saying they would use AI if built into records software. That’s starting to happen, with EHR provider Epic teaming up with Microsoft last year to use generative AI in its healthcare software lineup. The AI is already helping to summarise notes and streamline clinical documentation.

AI healthcare startups

Hospitals may not feel prepared for the looming AI revolution, but that hasn’t stopped the technology industry from targeting healthcare. Established companies and startups alike are looking to disrupt the sector.

For example, Google has a medical version of one of its large language models, designed to sift through doctors’ notes. Meanwhile, startup Vuno uses AI for medical imaging and AI is even being used to diagnose cancers.

OpenAI’s ChatGPT can also answer medical questions, though you should still go see a GP for whatever it is that’s concerning you.

It’s clear that AI has utility in healthcare, especially amid budget constraints and with a looming demographic time bomb. Plus, AI is a technology well suited for sifting through data, meaning it could be particularly helpful in medicine, from drug development to diagnosis.

“There are many ways healthcare can and will benefit from AI, including freeing up our clinicians to focus more on caring for patients and helping systems more efficiently process a range of tasks,” said Rob Bart, MD, Chief Medical Information Officer for UPMC, a founding partner of the CCM, in a statement. “But it is essential that healthcare executives also take seriously the responsibility to protect our patients’ privacy and health data.”

Indeed, hospitals and the wider healthcare market may be wise to take on such technology more slowly. That way they will avoid falling for the AI hype, only rolling out systems that are proven to be safe and accurate.