As the coronavirus spreads rapidly across China and to other continents, employers are implementing health tools and practices to protect their employees and offset concerns over the outbreak.
Buoy Health — an AI-based platform for health — started collecting information about the coronavirus on Jan. 25, when there were only two confirmed cases in the United States.
Anticipating that the virus could potentially spread in the U.S., and hoping to ease growing fears over an outbreak, Buoy updated its algorithm to screen for symptoms and risk factors related to coronavirus. They also provide advice for next steps and methods for self-triage, as recommended by the Centers for Disease Control and Prevention (CDC). Additionally, a partnership with HealthMap, an outbreak tracking system operated out of Boston Children’s Hospital, identifies hotspots of disease in Bouy’s own data.
Buoy’s CEO Andrew Le says the upgrades were a way to mitigate any fears employers and workers have about contracting the virus.
“There's this almost rampant mob-like fear that's spreading well ahead of where the coronavirus is actually going,” CEO Andrew Le says. “Having a tool that gives you very specific personalized outcomes for your specific risk in your situation can do a lot to alleviate that fear.”
There are ways to protect employees’ health and address concerns, as well as safeguard employers’ business interests from virus-related impact and loss. Le shared how AI is playing a major role.
How are employers using Buoy’s updated AI to screen for symptoms and risk factors related to coronavirus?
Even within our small company, there's been a lot of fear around the coronavirus. Many companies have employees that are traveling, and just the idea of getting on the plane for work can be stress inducing, when you see everyone around you wearing a mask. We're telling our employer partners “Have people be able to turn to Buoy the moment they feel scared.” What they're really trying to do is give the [employees] peace of mind when appropriate. If they are truly at risk, then they need to seek care. But on the whole, the value that we're bringing for this particular response is to help people better understand that the risk isn't that high for them in this particular moment, based on what's going on with them.
How does the technology behind your algorithm work?
We have about 9 million active users and every single time someone uses it, the program learns and gets better. When we work with an employer, we combine that with what we call our benefits navigation. On the benefits side, we basically take your entire suite of benefits and then map them one to one, back to the clinical endpoints. Based on the time of day, your location, your benefit design, what health plan you're on, and your preferences, we show where the nearest doctor is and just get you to care when you need it at the right place at the right time.
As for the coronavirus, we basically took the CDC guidelines on what risk factors and symptoms are consistent enough to seek care specific for coronavirus versus not. It helps you understand whether coronavirus is a potential yes or no, and if it’s a no, we would help you understand what is actually going on.
Why is AI an efficient way to help reduce any fear or anxiety around coronavirus?
If people are worried about the coronavirus, right now they have to go to their doctor or call their doctor or a nurse line. At the end of the day, there's only so many doctors and nurses, and during these periods of an epidemic, there is way more demand for care than there are suppliers. So there's got to be something else that is readily available, is intelligent and can help people make good decisions about when to go and when not to go. And that's where we see AI being a really great front door to healthcare, because it’s not confined to the number of people that can sit on a phone, or a number of people that can see patients, especially in these times where there's just so many people that are worried. You just don't have enough skilled clinicians to do that.
What will the development of AI detection tools like Buoy’s look like going forward?
We can essentially start to find pockets of disease before they're ever identified. For example, we have 45% of the Orlando population on Buoy today. If we find that all of a sudden there are people with very specific symptoms only in Orlando, and the symptoms don't seem to be consistent with any particular disease in our database, that's a really strong signal that something else — something new — might be going on. Part of what we're excited about, working with public health officials, is not just being part of the response, but actually being an asset for them to potentially identify outbreaks before they ever even go into the doctor's office. If you do that you could potentially stop the spread of something even earlier than you would have otherwise.
