How Mercy is using AI to improve patient care

Mercy, the Missouri-based hospital system, isn’t tentative about using artificial intelligence to try to improve patient care.

The health system has teamed with Aidoc, a healthcare technology company, to use its AI-powered platform to improve its diagnostic imaging. Mercy isn’t just using the Aidoc system for a small pilot program. The health system is in the process of adding the AI across the entire system, which includes 50 acute care and specialty hospitals.

Steve Mackin, president and CEO of Mercy, tells Chief Healthcare Executive® that he’s encouraged with how smoothly the rollout is proceeding. He says the AI platform will be in place across all of Mercy’s hospitals and clinics within a month.

“It's a significant implementation,” Mackin says, but he adds, “The early reads are positive. So first of all, the physicians have really embraced it.”

“It’s been a very positive transition and or an implementation of this scale, to have so few technical and operational challenges with it, is quite remarkable,” Mackin says.

• Read more: Two out of three doctors say they're using AI

Elad Walach, CEO of Aidoc, tells Chief Healthcare Executive® that Mercy’s commitment has enabled the ambitious implementation. He says installing the technology across all of an organization in a matter of months is “rapid deployment speed for health systems.”

“They wanted to go big,” Walach says.

‘An AI guardian angel’

Mercy is using the AI-powered tools to help support diagnostic imaging across the system, enabling the detection of conditions such as brain hemorrhage, pulmonary embolism, cervical spine fracture and lung nodules.

The Aidoc platform isn’t replacing the work of radiologists, but it’s helping Mercy’s radiologists in handling the large volume of images they examine. Aidoc’s platform is also helping Mercy’s radiologists identify patients that may be facing urgent problems, so they can receive attention, Mackin says.

The physicians using Aidoc’s platform appreciate the ability to help see patients who are at risk for greater complications more quickly, he adds.

For providers, “they have an AI guardian angel who can do a double check,” Mackin says.

“It helps them prioritize,” he explains. “They want to see and they want to be able to act on that patient who is in an emergency. And sometimes the emergency isn't necessarily flagged. You know, the emergencies are based on what you're finding. And if they don't have the opportunity to do that read until five other reads, that's a problem. So this is helping them do their highest and best work faster.”

Even with Mercy taking on an admittedly ambitious rollout of Aidoc’s platform, Mackin says he was comfortable with the approach.

“The benefit of Mercy is that we have, for the most part, a highly aligned health system,” Mackin says.

“We have an intelligent data platform, a data layer that we've created, and technology layers that allows us to implement at scale,” he says. “And so this is something we're accustomed to when it comes to implementing validated AI. And so we felt very comfortable with this type of rollout schedule, with this technology, with Aidoc, because these are validated tools that they have.”

He also points out that Mercy has been utilizing various AI technologies for more than a decade, giving the system experience in tackling a project of this scope. Mercy incorporated its first AI model to identify candidates for home health care 11 years ago, Mackin says.

“Today, every day, we have over 200 active AI models running in the background every day across Mercy. And so we have experience with these types of implementations,” he says.

Elad Walach, CEO of Aidoc, said Mercy's experience with AI and strong leadership helped with the implementation of the technology.

Aidoc has partnered with 150 health systems. Walach says Mercy’s experience in AI technology, and its focus on deploying AI in clinical uses, helped speed the implementation.

“It's just an incredible system to begin with, I will say. But also, I think they had the infrastructure and the strategy to implement …. We didn't come in for a chaotic chaotic environment. It was well organized, which would allow us to move fast and make this impact pretty quickly.”

Beyond the technology capabilities, Mercy’s leadership has driven the success of the project, Walach says.

“What Mercy has in abundance is clear strategic directive and vision, clear leadership and ability to make decisions, which I think is the biggest difference in technology innovation, more than the tech stack, more than anything else, is sheer force of will and leadership,” he says.

• Read more: Mercy, Alice Walton Foundation team up on $700M mission to change health care

‘Side-by-side support’

Mackin is optimistic that the wider adoption of Aidoc’s platform is going to benefit patients, providers and the entire Mercy organization.

“It helps the whole hospital and health system prioritize their work. That's a win for the providers, and makes them more efficient for the patients,” Mackin says.

The use of the AI platform is designed to support physicians, but won’t serve as a substitute for clinicians, he stresses.

“We're not doing any medical AI that replaces a physician,” Mackin says. “It's always augmented. So it's providing side-by-side support, insights, prompting, so they can put their eyes on the right issues. But it's not replacing them in the decision-making process. It's just bringing a whole new set of insights to help them better manage their workflows. And in the near term, that's a really big problem to solve.”

Many physicians are struggling with heavy workloads due to staffing shortages, which leads to burnout. Mackin hopes the Aidoc platform, and other AI tools, can help reduce the stress on clinicians.

“We're seeing AI really helping us do things we wouldn't otherwise do,” Mackin says. “And there's not a solution just to add more doctors to this, because there's just not enough.”

Mercy will be tracking different metrics to gauge the success of the Aidoc platform.

“We should be able to measure significantly the quality outcomes that will come from this, the number of incidents that have been escalated, the time savings to intervention, from diagnosis to intervention,” Mackin says. “That's a measurable outcome for our patients. There are quality outcomes that we'll be able to measure. We'll be able to measure the efficiency gains that are coming to the providers and the efficiency gains for the radiology teams.”