Artificial Intelligence

How Machine Learning Is Helping Consumers Take Action on Their Health

The Future is Now: How Machine Learning Is Helping Consumers Take Action on Their Health
Craig Wigginton, Chief Technology Officer at Icario

For years machine learning (ML) has been touted as a way for health plans to leverage the mounds of data they collect on their members, but practical use cases remain relatively uncommon. For some, the perception may be that ML is a futuristic and somewhat impersonal way to operate, but the exact opposite is true — ML is already being used and is here to stay, and when deployed correctly it actually creates a more personalized experience for members.

The first thought might be that ML and artificial intelligence (AI) can be used to help teams vastly improve mundane and tedious tasks. And while it is true that AI- and ML-powered technology such as robotic process automation or optical character recognition can do things such as quickly ingesting and processing claims and other paperwork, those aren’t necessarily groundbreaking game-changers. What we’re exploring here goes much deeper: deftly implementing a ML strategy that makes getting members to take action on their care a more effective and efficient process over time.

Here are three ways machine learning is helping plans connect members with the healthcare they need, improving both outcomes and plan performance in the process:

1. It reduces the number of communication touchpoints required. When health plans look to get members to take a specific action — whether an annual wellness visit, a cancer screening or an annual flu shot — one of the most basic tactics they use is to simply reach out to the member. This may come in the form of an email, a phone call, a text message, or any other communication method.

See also  Artificial intelligence services firm Sidetrade to open new tech hub

Even with the best intentions in place, the challenge of this approach is that every consumer is different and has their own individual preferences and expectations. Without knowing which communication method each individual member prefers, it might take plans 10 or more communication attempts per member to finally get them to take action, which quickly becomes inefficient and expensive considering the thousands of people health plans serve.

Machine learning eases this pain with its ability to reduce the required number of touchpoints over time. By using this technology to map to different personas — essentially segments of consumers who have similar preferences and tendencies — it can learn which communication methods are effective with each persona, as well as which don’t move the needle. This reduces the number of communication touchpoints required to get members to act over time, saving resources while also getting them the healthcare they need more rapidly.

2. It helps deliver communication that resonates. Similar to how we each have preferred vehicles of communication, we also have specific preferences when it comes to how we’re addressed. For example, some might respond to a lighter and somewhat breezy tone, while others prefer a more serious and straightforward approach. Or, some may require only basic high-level information to take action, while others need more detailed and in-depth information.

As the ML system learns the preferences of individuals over time, it helps determine which communication methods and messages work best to reach specific people. If something along the journey doesn’t work, ML quickly refines the process and finds a better way to engage. Communicating how and when someone prefers to be reached is vital — it adds a more human element to interactions with the health plan, which helps improve member satisfaction. 

See also  Newsroom | Howard Newsroom

3. It reduces member abrasion and boosts retention. Building on the previous point, it’s not hyperbole to say that member satisfaction is as important as it has ever been. For the contract year 2021, CMS increased measure weights for patient experience/complaints measures and access measures from two to four, meaning Consumer Assessment of Healthcare Providers and Systems (CAHPS) survey measures — which gauge satisfaction — now account for roughly a third of plans’ overall Star Rating.

This, again, is where deploying machine learning to create a more personal and consumer-like experience for members helps significantly. When plans use the technology to deliver relevant messaging on preferred communication platforms, members take notice. They remember when their health plan gives them a helpful tip or friendly reminder that helps their wellbeing, improving their level of satisfaction and reducing the risk that they’ll shop around for other plans.

Targeting members or segments of members is something that all health plans do in an attempt to get consumers the care that they need. Adding an ML strategy to the mix can improve these basic efforts exponentially — it could be the difference between a vague mass email sent to all members and a specifically tailored message to one individual based on their own health history, risk factors, gender, and location, all while communicating via their preferred channel at the optimal time.

When used smartly, ML humanizes healthcare and creates an experience that empowers people to take action based on a personalized engagement strategy. And that’s a win-win situation for any health plan.

See also  What a machine learning tool that turns Obama white can (and can’t) tell us about AI bias

About Craig Wigginton

Craig Wigginton is the Chief Technology Officer at Icario. With more than two decades of experience in technology leadership roles, Craig helps Icario’s clients achieve their digital transformation and consumer engagement strategies across the full spectrum of healthcare—from enrollment and quality to satisfaction and retention.


Leave a Reply

This website uses cookies. By continuing to use this site, you accept our use of cookies.