ARTIFICIAL INTELLIGENCE

The Role of AI and Machine Learning in Health Care Marketing

From farming to manufacturing to sports analytics, artificial intelligence is revolutionizing almost every industry and facet of life today. That said, it’s also transforming health care and health care marketing: specifically, how advertisers identify and target customers, and how they acquire and maintain digital media platforms.

But what exactly is AI? Do images of “The Terminator” or “I, Robot” pop into your head? Well, we’re not quite there yet, according to Daniel Fell, senior strategist with Optum Insight.

“My best definition of AI is probably the facsimile of human intelligence in computers and machines, which are designed to behave more like people,” Fell says. “And in time, especially as technology advances and our programming becomes increasingly sophisticated, there will be systems that will be able to make decisions and even ‘think’ on their own.”

However, AI is a very broad term that can encompass many things. Much of what the general public thinks of as AI today is actually something more appropriately called machine learning, according to Fell.

“The easiest way to define machine learning is people using math and data science to make algorithms that can process an extreme amount of data, extremely quickly,” he says. “Computers can literally process millions of bits of information really fast and rather effectively. If you’re looking at a huge data set, let’s say financial, and if you want to develop certain insights or a predictive model quickly, then that’s where you get machine learning. This can also be particularly useful in health care and marketing efforts.”

AI in Health Care Marketing

 Advertising has come a long way in the last 20 years, where marketers can now use data to inform, share and optimize their efforts. The first example of this has actually been around for many years, and it’s known as programmatic media buying. As the internet grew and became more complex, marketers found that they could shift from the straightforward — and manual — approach of subjectively purchasing ad space to organizing media placements in a more systematic way.  

“Today, most digital advertising is run by programs and not humans,” Fell says. “Machine learning now allows us to scan hundreds of thousands of websites and automatically decide what an ideal placement for millions of diverse users is. This approach is cheaper and more efficient, and includes all of the things that we’re looking for in marketing today, such as the ability to track and measure data in a meaningful way.”

Predictive analytics is another example of the use of machine learning in marketing today. This is essentially leveraging one’s computer’s ability to process millions of bits of information, analyze it and then identify patterns in the data.  

“In health care marketing specifically, this translates to taking countless data sets and health records, and then creating highly predictive models on what a consumer’s health care needs are,” Fell explains. “So, instead of targeting everyone in a designated area, or even concentrating on things like age or gender, marketers can now focus on specific people, such as someone with a high propensity for having heart disease, exponentially increasing an advertisement’s efficiency.”

According to Fell, another growing area of interest in health care marketing is the use of things like chatbots and intelligent platforms where people can type in a question and receive an immediate response. “It’s a really interesting application for the use of AI in real time, providing relevant and effective communication between providers and patients.”

Implementing Machine Learning

So, what’s the best advice for a health care marketer today? Should they be afraid of this rapidly growing technology, or should they fully embrace it without question? Well, according to Fell, they should probably slowly lean into it.

A good place for marketing departments to begin is by assessing all of the existing utilizations of AI in their hospital, including in clinical areas and operations, he said.

“I think it’s really important that marketers understand the technology because it will undoubtably not only help them understand how to apply it but also recognize what works and what doesn’t in health care today,” Fell says. “I wouldn’t say that this technology would make my job easier; in fact, it’s quite possibly harder in the short term, as it will require a learning curve and a new set of skills. However, in time it absolutely has the ability to make everything, including my productivity, more efficient and cost-effective.”

 Finally, regarding privacy and security, both are always critical issues whenever dealing with consumer information, especially regarding patient records and data, he said.

“Much like with everything else, marketers and executives should be balanced in their approach. It’s like with the internet 30 years ago or cryptocurrency now — there is a lot of hype, so we have to manage expectations and approach it with realism and practicality. Establishing proper protocols requires time, resources and money, and maintaining security requires more of those things. But in time, these concerns will be worked out and AI will become more routine, making all of our lives easier and more efficient.”

Improving Hospital Performance Using AI and Patient Data

Hospitals and health systems have struggled with quickly identifying life-threatening conditions, such as sepsis, as well as standardizing the care of patients at risk for adverse events, such as health care-associated infections and acute myocardial infarction.

Conventional methods, such as electronic health records, do have a role in surveillance, but they have limitations. Such tools lack the capability to continuously, and automatically, collect and analyze relevant data in order to alert health care providers of early warning signs.

The reliance on segmented data that’s frequently not being accessed when most needed is a major problem for health care providers, according to Marie Cleary-Fishman, vice president of clinical quality, American Hospital Association.

“One of the most difficult things for the clinician is that they’re getting so much information about a particular patient, and while we have electronic health records, they’re not always integrated in the system. So, some of the data is evaluated in a retrospective mode,” says Cleary-Fishman. “So, if we’re able to pull that data together across those silos, it’ll be much easier for physicians to make real-time decisions that are best for the patient.”

Conversely, AI and machine learning can help to do just that: develop surveillance algorithms programmed to understand disease risk factors, thus lowering patient risk.

New technologies like data science and AI save lives, and it’s that simple, according to Jonathan B. Perlin, MD, PhD, MSHA, MACP, FACMI, president, clinical operations and chief medical officer, HCA Healthcare.

“Take sepsis, which claims about a third of the deaths in America’s hospitals, where time is life,” Perlin says. “In sepsis, mortality increases by about 8% for every hour of delayed diagnosis. Being able to monitor the data 24 by 7 by 365, and providing indications to care providers at the bedside that their patient may have sepsis, means that we can invoke the appropriate interventions as early as possible.”

However, not only do these technologies have the ability to improve care, but they also offer tremendous opportunity to improve value, according to Perlin.

“Today, if you meander towards a diagnosis it doesn’t help the patient, rather it only wastes resources,” he says. “So, to the extent that these technologies can help us get the right diagnosis and the right therapy as efficiently as possible, we also achieve the best outcomes. That’s why we like to think of RIO not only as return on investment but also return on information. And what we’re really excited about is the dual bottom line: better care and better value.”

 

Learning More

• For more on the topic of AI and machine learning and to hear from Danny Fell, listen to this SHSMD Rapid Insights podcast episode and read a follow-up SHSMD blog post.
• Click here to watch AHA’s Transformation Talk episode  featuring Dr. Perlin and Marie Cleary-Fishman on “Improving Hospital Performance Using AI and Patient Data”
• Stay tuned for a future-forward article on AI in Futurescan 2023, to be released in November 2022! Download Futurescan 2022 today. (SHSMD members receive one digital copy for free!)

 

This article features interviews with:

Daniel Fell
Senior Strategist
Optum Insight

Marie Cleary-Fishman
Vice President, Clinical Quality
American Hospital Association

Jonathan B. Perlin, MD, PhD, MSHA, MACP, FACMI
President, Clinical Operations & Chief Medical Officer
HCA Healthcare

image credits: istockphoto.com/metamorworks, istockphoto.com/Natali_Mis, Blue Planet Studio/shutterstock.com