If you ask experts, the next big technology to revolutionize health care will be artificial intelligence (AI). “Health care professionals should get very interested in AI and machine learning. It is such a disruptive technology and already embedded in the many ways that health care is delivered,” says Saurabha Bhatnagar, MD, faculty director of Global Executive Education and faculty member in the Trends in Health Care Technology workshop of the Safety, Quality, Informatics, and Leadership program at Harvard Medical School. 

“Leaders need to understand this technology. And we need quality and patient safety experts who are able to provide the right guardrails around that technology in order to develop AI from an idea to a scaled solution,” he explains.

But even for someone who is willing to engage with AI, important questions remain: how to address skepticism and cost, ways to introduce the technology more effectively, and (crucially) what implementing it will do for the organization. Most important, what does AI mean for health care quality and patient safety?

The Difficulty of Introducing AI to Health Care

The primary challenge inherent to AI is not the technology’s potential; it's reticence based on the organization’s prior experience. “Health care is one of those industries where they are uniformly unhappy with whatever has occurred around technology. Clinicians are burned out with existing technology, which isn’t significantly helping their workflow. For the patients, AI hasn’t made health care much easier to understand or interact with,” says Bhatnagar.

“People are somewhat disenchanted with a ‘new’ technology coming out, and they worry that they’ve seen this before,” he adds. “They’ve seen how the story plays out, and it hasn’t really helped health care, their work, or what they’re passionate about—improving the lives of patients.”

This is even more of a barrier when the CEO or other organizational leaders have had an adverse experience with implementation in the past—incurring significant unseen costs, for example, or accidentally sustaining lags in patient care—which can sometimes mean they are less inclined to buy in on a new, expensive venture. 

Avoiding Pitfalls to AI Implementation in a Health Care Environment

According to Bhatnagar, health care leaders don’t necessarily know what a successful AI implementation could (or should) look like. “People think AI is something ambiguous that you buy off the shelf like a box of cereal, and then you just go and implement it,” he says.

“One of the mistakes that I see people make is they feel like they need to hire a team. They start building it, they don’t start seeing the return on investment (ROI), and then they reduce the team size and say, ‘Okay, it wasn’t worth it.’ But you don’t need to have a large team to implement this technology.”

Citing cases where hospital leaders deploy pilots that don’t scale well or are cost-intensive early on, Bhatnagar notes that leaders must consider investment and ROI in the right way if they’re using this model for AI—picking a pilot that enables greater cost saving than spending and allows for quicker, more visible scalable solutions.

“AI absolutely should be applied, but most often, it doesn’t reduce your need for workers or for humans in the workflow loop,” he notes. “Generally, it makes your existing workforce more productive in what health care leaders really care about quality improvement and patient safety. It helps your workforce better understand the gaps in quality, or better close the gaps in patient safety, rather than totally replacing your workforce with AI.” 

The Potential of AI in Health Care Organizations

When used correctly, Bhatnagar explains that AI can significantly augment the workflow, which makes it different from any other technology that has been introduced in the past several years. Regardless of one’s job in the organization, AI has the potential to improve the worker and their work. 

This is particularly relevant for providers and professionals working toward a value-based care model and determining the resources they need. As the health care system moves to an outcomes-based model, it introduces significantly more variables, with many more data points and location-based specifics,  all factor into value-based care. Bhatnagar highlights how difficult and costly it can be to pull and analyze data and then attempt to make investment decisions from it while trying to balance outcomes with ROI. 

“But if I use AI, I can reduce my time to aggregate data—it could be almost real-time or near real-time. And I don’t need to hire 10 data scientists to pull data and make spreadsheets. That task can be supplemented and enhanced, with one person potentially doing it better and in a more holistic format. Now I have more time and resources for my team to focus on doing the work of value-based care and improving outcomes,” Bhatnagar explains. 

“If you’re a quality and patient safety leader, for example, you might be focused on environmental management services: how to clean rooms, disinfect them, and rotate them for bed flow and optimization to best utilize your hospital system,” he says. “If I were trying to understand all the data from a large hospital system, it would take me a long time just to do data aggregation and analysis. AI allows health care leaders to shift their time toward the more important step of gaining knowledge and insights from the data.” 

In the larger sense, it’s important for health care professionals to educate themselves about these new developments. In Bhatnagar’s experience, many of those he works with falsely believe that technology is stagnant and thus irrelevant to their work. Instead, he encourages them to understand the cycles of innovation in health care technology to make smarter long-term decisions. 

“Technology is here to stay in health care. I guarantee you that it will continue to become more and more relevant in every nook and cranny: delivery, services, clinical delivery, patient care, and patient engagement,” he says. “For folks who spent all this time becoming health care experts, spending a bit more time understanding the technology landscape and the opportunities available is important, especially if they want to be health care leaders for the next several decades.”