Meet care.coach, the Health Avatar Automating Care for Seniors

When we first met Victor Wang and heard about care.coach, we were duly impressed by his MIT + NASA pedigree. But in the end, it was his personal passion to solve one of healthcare’s most intractable problems that had us hooked. Victor has been working behind the scenes for years creating the tech framework for a product that could redefine remote care as we know it, and the industry is catching on in a big way. Want to know how? Read on!

Origin Story

When he was a baby, Victor Wang immigrated from Taiwan to the United States with his family. Among the relatives who stayed behind was his maternal grandmother. She had no nearby relatives and as the years went on, became increasingly lonely and isolated. Victor remembers how, as a child, he watched his mother struggle to care for her mother from halfway around the world. His grandmother had a hard time with technology and so even a Skype call was challenging to arrange.

Victor grew up and he carried this experience with him, albeit somewhere in his subconscious, as he headed off to grad school at MIT. He’d received a scholarship to study robotics and human-robot interactions, and he was realizing his dream of working in the lab on cutting-edge machines, bringing science fiction to life.

Then two things happened that shifted his trajectory, and his perspective. First, he took on a research assistantship at the MIT Human Systems Laboratory, where he helped coordinate a clinical trial, a sleep study at Brigham and Women’s Hospital exploring the impact of blue light therapy. This opened his eyes to the patient experience and the importance of validating and measuring any hypothesized benefits of technology.

Wang’s second awakening came while he was working on his own thesis research for NASA’s telerobotics program about astronaut performance aboard the International Space Station. As he got deeper into the subject, he began to read clinical research around loneliness and isolation in astronauts and learned how these psychosocial factors could hurt astronaut health and put the space program at risk. He was reminded of his grandmother in Taiwan — who his family still struggled to reach via Skype. His grandmother had suffered a fall in the intervening years, and her loneliness had severely reduced her motivation to self-care, which hindered her recovery outcomes. Suddenly it hit Victor that his work in robotics had applications way beyond NASA and that perhaps his understanding of artificial intelligence could improve senior care in a way that would help millions of people like his grandmother.

When Wang finished his Master’s degree, he put his PhD on indefinite hold and decided to leverage his tech skills to transform senior care.

The Challenge

Wang quickly discovered a number of well-documented but persistent problems. The first was a workforce shortage. There’s a number called the “Caregiver Support Ratio” that represents the number of family caregivers potentially available to care for seniors. AARP calculates that ratio to be around 5 currently, heading rapidly to 3 in the coming decades. The fundamental truth in the United States is that every day, more people age into conditions like Alzheimer's and related dementias, or are diagnosed with multiple chronic conditions such as diabetes, hypertension, heart failure, and COPD, or struggle with social isolation, loneliness, depression, and anxiety, each of these factors requiring attentive, day-to-day management. Meanwhile, the senior care workforce fails to keep pace. Wang describes it as a “fundamental limitation of resources.”

He also discovered that even when the senior care workforce is available, it’s often a frustrating, fragmented experience for patients and caregivers alike. It’s not that senior caregivers aren’t knowledgeable and supportive; they are often heroes. It’s just that 24/7 support requires a constantly rotating cast of nurses and health aids to care for a single individual. That’s a recipe for a fragmented, disconnected patient experience, which is the last thing you want when your patient needs to feel connected, supported, and motivated to self-care.

So Wang discovered a workforce shortage, poor patient experience, and then the trifecta: rising costs. We spend the vast majority of healthcare dollars on the oldest, most complex patients. When millions of seniors lack the care they need, their conditions worsen more quickly and compound the cost to the system. The care gap and the cost burden go hand in glove.

Thankfully, for Victor Wang, these challenges weren’t dealbreakers, but rather the prompts he needed for his next act: the launch of care.coach.

Under the Hood

When you think about the care.coach product, there’s what the patient sees, then there’s what’s happening behind the scenes. For the patient — let’s say an elderly woman admitted to the hospital after a fall — care.coach arrives as a specially designed tablet attached to an IV pole or set on the bedside table. Onto the screen pops a friendly dog, named “Buddy” by default (though patients often personalize their avatar with names like “Pony,” “Sparkle,” and “El Capitan”). As the nurse leaves and the patient settles into her hospital stay, Buddy starts a two-way conversation. He asks the patient how she’s doing and gives helpful coaching designed to prevent delirium and falls: “Is the hospital too noisy for you to sleep well? I can ask someone to bring you earplugs,’’ or, “Are you ready for your bed exercises to keep you strong? I can talk you through them,” for example. Then, when the patient is discharged, she’s invited to take Buddy home with her, as a care companion. A Stanford paper published in the Journal of Medical Internet Research shows that this care.coach avatar program mitigates delirium and falls among hospitalized elders while reducing loneliness 3x more than a daily nurse student visit.

At home, Buddy sits on a side table and checks in with the patient about everything from pain to medications, to how the grandkids are doing. The University of Washington has published their care.coach avatar research showing that seniors valued having a “friend who is always available and ready to talk when one is lonely,” as well as “the significance of the reminders that the digital companion would provide (for medication adherence or dietary restrictions).” UW’s research found that the seniors had reduced levels of depression thanks to Buddy. And care.coach’s health plan customers have published whitepapers showing how this daily, in-home, relationship-driven support has reduced members’ need for nurse visits to the home, reduced ED/hospital visits, and improved their providers’ ability to deliver care and support the member experience despite the COVID pandemic and self-isolation protocols.

“The user experience is absolutely critical,” says Wang. “We’re dealing with the most complex patient population out there.”

So far so good, but the real magic happens behind the scenes. First, there’s the avatar. Wang chose a dog (or a cat) because the benefits of pet therapy — even when the pets are virtual — have been studied and proven again and again. Then, Wang used his understanding of human-robot interactions to develop the “brain” behind the avatar. By talking naturally with older adults and even people with Alzheimer’s and related dementias, they’ve been able to build speech recognition systems that work better for this population than those created by Google, Amazon, and Microsoft. They use these unique machine learning models to help automate and scale the types of conversations and relationship-building interactions that would traditionally require human staff. For this work, they’ve received a major National Institute on Aging award.

One reason that care.coach chose an avatar over a human face is because they are working to scale care globally. Video telemedicine will always be one-on-one, and thus limited by workforce shortages. By using an avatar (and a synthesized voice), care.coach lets the patient form a bond with a single entity while a diverse team of care.coach health advocates work behind the scenes. It also enables Wang to end-run the workforce shortage in the United States through what he calls “geo-arbitrage.” care.coach employs people in Asia and Latin America — regions that don’t have the same caregiver shortages as the United States — to be the eyes, ears, and empathy behind the avatar.

It’s a human-in-the-loop model, where portions of dialogue with the avatar are automated. As Buddy asks basic questions, he’s gathering data points and getting smarter through machine learning. Then, as issues inevitably get more complex, they’re put into a triage queue for one of the many health advocates on call. That allows these people, who can be located anywhere in the world, to make the most efficient use of their time. Finally, when a situation warrants it, a remote health advocate can use the avatar to suggest a direct 1:1 telemedicine visit. At that point, and with no need to understand Skype or Zoom, the patient is brought face-to-face with their doctor through care.coach’s very own, ultra-easy Video Visits product, to deal with the issue at hand.

Why We’re Proud to Invest

There are lots of reasons to be excited about Victor Wang and the progress he’s making at care.coach. First, he’s built a novel way to tackle an intractable problem at scale. This is health moonshot thinking at its best. The caregiver shortage is one of the biggest problems facing healthcare today, and Alzheimer's and related dementia are some of the most costly and complex issues in the world. care.coach is leveraging the absolute best technology in artificial intelligence and machine learning to drive directly at these problems. It’s a bold, brave move, and Victor Wang has the passion and know-how to make serious progress.

Second, we love that Victor has employed a truly global perspective to solve this problem. The United States has a caregiver workforce problem, but globally, there is no shortage. There are people willing and able to work if the technology can connect them to the need. We’ve seen this geo-arbitrage work in many other industries, and it’s time to leverage it in health, for the good of the seniors we know and love.

Finally, we’re bullish on care.coach because they are seeing serious market traction and validation, and every patient encounter makes them smarter. care.coach is already working with population health companies, Medicare Advantage plans, and other large organizations and their sales quadrupled this year. The icing on the cake is that every new partnership means thousands of new conversations, and every patient encounter makes the machine learning algorithm smarter.

Join us in welcoming Victor Wang and his care.coach team to the StartUp Health family. We can’t wait to see what they do next.

Original Article