To meet new blood pressure guidelines, we will need artificial intelligence
The following is a guest post from Julia Hu, CEO and co-founder of Lark.
Earlier this month, the American Heart Association, the American College of Cardiology, and a nine-member task force redefined the standards for what qualifies as high blood pressure for the first time in 14 years. In one stroke of the pen, 30 million more American adults will be classified as hypertensive; the percentage of American adults considered hypertensive will increase from 32% of that population to nearly half. The number of men under age 45 with a diagnosis of high blood pressure will triple, and the prevalence among women under age 45 will double.
The experts behind the new clinical guidelines cautioned that these newly-diagnosed patients will not have to immediately go on medication. Rather, they most likely will need to eat better and exercise more, which with the Thanksgiving holiday may not be seen as opportune.
As anyone who has tried to change their diet or exercise regimen knows, this advice sounds simple, but is incredibly hard – and especially during the holiday season. It may be why, according to research published earlier this year, fewer people are trying to lose weight and obesity rates continue to grow.
Yet, it’s essential that we prevent and manage these conditions as one in two U.S. adults has a chronic disease, one in four has at least two or more chronic conditions, and the costs of these diseases represent more than 86% of all healthcare expenditures.
But instead of relying solely on human willpower, we can now tap robot power to help these tens of millions of patients with high blood pressure.
Advances in artificial intelligence (AI) have made it possible to provide compassionate and personalized care to help people manage and prevent chronic conditions like hypertension and type 2 diabetes. Because of nursing shortages, increasing demand, and the sheer cost and time it takes to hire and train a health coach workforce, AI is the best bet to possibly meet the already huge – and now growing – need for health coaching and chronic disease management at scale.
This does not mean that an army of robots will be deployed to households across America. Think Siri and Alexa, but super specialized. AI health coaches or nurses are trained with sophisticated medical algorithms and equipped with machine learning that allows them to predict and emulate what the best health care experts and coaches would say to you at any moment to answer a question or help you through a difficult time. The response is both lifelike and available 24 hours a day, seven days a week with instantaneous response.
And according to the latest research, AI works.
For instance, a new peer-reviewed study found that the personalized AI health coach at Lark, the company I founded, had comparable results to nationally-recognized programs that are led in-person when it comes to weight loss, the key measure to preventing those with pre-diabetes from acquiring type 2 diabetes.
Specifically, over the longitudinal study of 103 individual coaching sessions, users lost an average of 2.4 kg (or 2.4% body weight), compared with an average of 2.32 kg reported in a meta-analysis of 22 lifestyle intervention studies with in-person components. Additionally, the percentage of healthy meals eaten increased by 31%.
Moreover, AI is an effective way to bring care to men, who have been less likely to enroll in in-person, one-on-one, and group counseling as well telephonic programs. In addition to data in the study, Lark has seen 250% more men enroll in its AI coaching, compared to programs led in-person. These men were just as engaged as women and had the same positive outcomes. There was a privacy about it that men seemed to appreciate.
And privacy concerns are just one of the potential barriers that can impact the adoption of any digital health tool, including AI. Privacy is of the utmost importance to health consumers, as is controlling how much data you share. You have to take that seriously as an AI platform, especially when using machine learning to figure out the patterns a patient may have, and to coach them through behavior change.
People are also often surprised to learn that you can systematically use AI to mimic very soft things, like the idea of compassion, of love, of no judgment. This kind of translation is what extends a personal, 24/7 experience from the comfort of home, in between all the other items on your to-do list. Another misconception is that utilizing AI requires a wearable, which can come with its own adoption issues. Instead, successful AI platforms aggregate user data through smart phones and app integration. In other words, you don’t have to use a new tool, you can just pick-up your phone.
This does not mean that AI will replace real-life nurses and doctors.
There’s just not enough doctors, nurses, coaches, and therapists to provide such high-touch personalized care. AI – whether our platform or the many others in development – can scale infinitely and complement our hardworking health care providers. The best AI platforms will be able to triage those who need medical care to physicians and nurses while, at the same time, lightening the load on overburdened physicians and nurses by helping millions manage their chronic conditions.
The new blood pressure standards will create a massive need for medical care, and we need a scalable way to address this growing epidemic that is cost-efficient, but still empathetic and supportive, and gives us a fighting chance to lower the risk for heart attacks and strokes.
For those who will get a hypertension diagnosis in the weeks to come, AI represents the most realistic and clinically-validated way to help millions of new patients without, well, raising their blood pressure.