Healthcare and Cognitive Computing Team Up for Big Changes

May 18, 2015


Article ImageData, data, data-it's everywhere. Everything, including your watch, your grocery store, and even your car, can collect it from you. But what do companies do with all this data they collect--and, more importantly, how can they make the best and most efficient use of it? That's where cognitive computing comes in. It uses a variety of techniques--such as image and speech recognition and machine-learning algorithms--to sense, predict, and even think.

It's not about simply making lives easier--it's about making them better. And as cognitive computing spreads to the healthcare industry, everyone from doctors to researchers to patients may soon reap its benefits.

Among the companies that are active in cognitive computing when it comes to healthcare is the Denver-based Welltok, Inc. Welltok describes itself as "a fanatical team of healthcare enthusiasts who are on a mission to transform the nation's healthcare system from one of sickcare to optimized health." The company's ?CaféWell Health Optimization Program, powered by IBM's Watson, offers consumers a way to identify personalized activities, health content, and condition management programs designed to optimize their health. It even tempts consumers with rewards for positive behavior change.

Once in the program, people can "interact and engage in an anonymous environment," says Jeff Cohen, co-founder and principal healthcare market solutions executive at Welltok. "We want to assess, we want to get to know the consumers at all different levels-behaviorally, psychologically, health interests, influencers, health conditions, hobbies, all those types of attributes," he says, "and drive them to programs, drive them to content, drive them through use of a personalized intelligent health itinerary, and then guide that consumer in terms of what types of things they can do, how they can get rewarded for their activities, and really make it easy for them to optimize their health."

With CaféWell, Cohen states, Welltok is able to apply cognitive technology attributes and natural language processing "in a fun and engaging way for consumers to ask questions in their own vernacular." And consumers are able to get responses "that are intelligent," he says, and get smarter all the time, based on repeated interactions with the consumer.

Over time, says Cohen, CaféWell can even recommend activities "that you're most likely to do based on everything we know about you." He says healthcare providers also benefit from CaféWell. "We're all about behavior change and engagement," Cohen says, "so anything we can do to help a provider or a health plan identify and change behavior [whether it's through social, gamification, behavioral science, or incentives] ... by providing that anonymous mutual platform, it benefits both providers and especially health plans in getting those patients and members to engage in an environment that feels safe, that they can trust, and that's filled with technology and engagement tools that are easy for them to use and they can integrate with their daily lives."

Cohen says he believes the long-term impact of cognitive computing on the healthcare industry will be "tremendous" with regard to "cost savings, to helping triage patients from a sick care perspective, as well as being able to align the best, most effective behaviors with the most appropriate resources that are out there."

Indeed, Forrester Research analyst Skip Snow believes cognitive computing has a "massive potential to improve healthcare." In February, he wrote a report stating just that. In it, Snow says Forrester believes that cognitive healthcare applications will help providers improve clinical automation and care management, payers improve population management and marketing and also reduce fraud and waste, and consumers read signals coming from their bodies and diets to improve health.

Snow singled out Welltok, and its partnership with IBM's Watson, when writing about companies' experiments with cognitive healthcare applications. "By combining domain expertise about health and nutrition and understanding a member's benefits and health status, Welltok learns from a consumer's behavior to advise her on exercise, food, and other activities in order to optimize her health," Snow writes. "In a crowded market, Forrester likes Welltok's views of healthcare as consumer-centric. Its strategy and approach to optimizing a population's health coupled with its savvy management make it a firm to watch."

Snow tells EContent some benefits of using cognitive computing with healthcare. A big one is doctors' and hospitals' development of a patient's clinical path--a process that involves a lot of work and, he says, is "overwhelmingly complex." But "with cognitive computing, you can greatly, greatly, greatly shorten the amount of time it takes to develop a clinical path, and you can rely on evidence from the charts of your patients," Snow states. "That is a huge asset to something which is very, very complex and very, very expensive. ... That's clearly a medical benefit.

"That's a huge potential benefit, because people with lesser skills are able to practice medicine at a higher level," Snow says, adding the "complexity" is that "you also have the ability to go down-license and have nurse practitioners, physician assistants, and other non-M.D. professionals practice medicine with these clinical recommendation engines, and that's probably where most healthcare is going in the long-term feature. ... Whether or not the consumers are going to be given the same level of care [with these down-license physician assistants and nurse practitioners] is TBD."

In a nutshell, Snow says, "More access, more efficiency? Yes. Better quality? TBD. Let's see the peer-reviewed research as this stuff matures." For doctors, he believes cognitive computing could lead to "better access to knowledge and more scientifically-based decision capabilities," as doctors won't have to read thousands of journal reports published every month; rather, computers will consume those reports for them, and doctors could ask questions "in a sort of two-way fashion, and get answers quickly."

Snow adds that he believes the process of ultimately implementing cognitive computing throughout all of healthcare will, in the end, be a slow one. "I will tell you that I'm a believer," he says. "I'm not one of those people who say it's the future. My thesis is cognitive computing is already embedded in the research part of healthcare and will not be operationalized for the next 10-15 years."

Why the wait? Snow explains, "The way in which healthcare works is very slowly. Something is invented or discovered, ... then you have to do peer-reviewed research in order to make that thing invented or discovered proven, then it has to be embedded into clinical practice, and that takes usually a decade or 20 years.