Navigate the healthcare industry
Healthcare Brew covers pharmaceutical developments, health startups, the latest tech, and how it impacts hospitals and providers to keep administrators and providers informed.
It seems every day we hear a new prediction on how artificial intelligence (AI) is going to change the future, from eliminating millions of jobs to making contact with aliens. Many predictions are based on types of AI technology that don’t actually exist yet. But a recent study found that AI technology that does exist has the potential to save the US healthcare industry up to $360 billion per year, or 10% of annual healthcare spending.
The study, from Harvard University and consulting firm McKinsey, estimated that those cost savings are attainable within the next five years “and would not sacrifice quality or access.”
“Everything we are talking about here…are technologies that already exist today and are used in case studies that we have seen in our experience,” Nikhil Sahni, a partner at McKinsey and lead author of the study, told Healthcare Brew. “We’re not talking about fancy new use cases that no one has ever seen implemented.”
While the study laid out a number of administrative use cases—such as using AI to optimize staffing levels for patient demand, or to optimize an operating room’s schedule so more surgeries can be performed—the main cost-saving uses of AI are for clinical tasks, according to the study. These include using AI to predict readmissions and to interpret medical images and scans.
“The administrative spending in total for a payer, for example, is 15% to 20% of the operating expenses. The medical side is more like 80% to 85%,” he said. “So in the end, the bigger savings will come from the medical side use cases for both payers and providers.”
But scaling AI to the point that healthcare would see such huge cost savings remains a challenge, Sahni said. For one, it’s hard to get enough people to trust AI that they’d be willing to invest in scaling it, especially when it comes to using it for clinical rather than administrative tasks.
“It’s very easy to talk about, ‘Build the algorithm, put the data in, and it’s going to spit out the answer,’” Sahni said. “But how do you make your people trust that answer?”
Also, a lot of people in healthcare are busy. They don’t have the time it takes to invest in scaling AI, he added.
“While all these are attainable in the next five years, meaning you have the technology and you won’t affect quality and access, the actual ability to capture it is really difficult, given how much has to happen for it to take hold,” Sahni said.
And the industry has a long way to go—“fewer than 10% of healthcare organizations today fully integrate AI into their business processes,” the study said.
“As other industries have shown, AI as a technology could have an outsized financial and non-financial impact in healthcare, enabling patients to receive better care at a lower cost,” the study authors wrote. “The next few years will determine whether this promise becomes a reality.”