Our Evolution: Becoming the OnStar for Care Management
“Sombit, before you connect that box, I want you to try some of my homemade scones…and coffee. I have the Olympics on in the living room. Those bobsledders are so fast! Come watch with me...”
~Joan*, QMedic user (*name anonymized)
3 hours later, I left Joan’s home—the service install itself only took 5 minutes but between Joan’s thoughts on bobsledding, her son, Cape Cod, kidney dialysis and Walter Isaacson’s biography of Steve Jobs, I kinda lost track of time…it was totally worth it.
When we started QMedic, my cofounder Dave and I would rent Zipcars and install our medical alert service all over Massachusetts for people like Joan. Similar to others, Joan had recently been discharged from skilled nursing and was in recovery from a stroke.
As a New York native and lifelong NY Giants fan, I reluctantly transplanted myself into the heart of Patriots nation 7 years ago. Since then, I’ve traversed the Bay State—from Gloucester to Marlborough to Uxbridge—to install our service for people struggling to live independently at home.
While we now ship our service nationally, these in-person installs have been the most rewarding and immersive exercise in product usability I’ve ever engaged in. And I still occasionally do them because I cherish these moments and want to make sure we stay in direct touch with customer/end user needs.
What we’ve learned:
- People hate Comcast and Verizon. I’ve been asked to fix cable wiring and boxes on numerous occasions because Comcast/Verizon installers have failed to show up in their designated 4-hour windows, or have otherwise provided terrible service.
- End-to-end, authentic customer service is enduring and sorely needed in a world of transactional, emotionless services. If you help someone take out the garbage or bring their groceries inside for them, the gesture goes a long way. They’ll value you more for going the extra mile, in whatever Zappos-like form that may be.
- Service trumps tech. Touting your gizmo will fall flat if your service stinks. A fancier mousetrap gets old and it probably won’t match Joan’s outfit tomorrow even if it does today.
- Affordability trumps tech. The people in greatest need are not living like kings and queens. Most live alone in modest 1-bedroom apartments, have either spent down their assets to children or don’t have much to begin with, and are making ends meet with the help of pensions, adult children, and government-subsidized Medicaid/Medicare coverage. Services need to be affordable and simple. Our end users aren’t buying Apple Watches.
- Distress calls/button presses are often triggered by social isolation and non-emergency health issues—these people should not be routed to the emergency room. Rather, it’s way more helpful to route them to some other care manager, service provider or family member who can better meet their needs.
- Simplicity is key. Users hate recharging or managing devices. If there’s effort involved in managing a service, or the service requires complex instructions, people will likely abandon quickly. We are proud of building a medical alert service that can be shipped and set up by a user or family member in less than 5 minutes. Importantly, the service and device require zero user maintenance or recharging.
- Shelf space in the home is usually limited; clutter is common. For those of lower income levels, many of whom receive Medicaid, there is usually very limited shelf space in the home to connect a base station or place a cable box. We decided to shrink the form factor of our base station to account for this. Many people are sensitive about moving their things around.
- Speaker/microphone quality and device range matter A LOT (duh). We decided we would blow the competition out of the water here, and we have. In Version 3 of QMedic (coming in early 2016), we’re introducing even better speaker quality and range to our service.
- Tracking behaviors proactively (e.g. sleep/activity) is primarily valuable insofar as it provides timely context clues to the remote care network. Outside the bubbles of Silicon Valley and Boston—for example, in Taunton, MA—people and their families don’t care about activity and sleep tracking and will look at you with puzzled glances if you try to sell them aggressively on it. This data is simply not actionable unless it’s set in the proper context and routed to a channel that improves service delivery.
- Fall detection is valuable but only if you can reduce the incidence of false alarms to a reasonable level (Full disclosure: Fall detection is not included in our commercially available service today. It’s something we’re working on).
- Mobile medical alert services can be useful for people with greater ambulatory function who are more likely to leave the home unsupervised. It’s important to differentiate these users’ needs from those whose risk is more confined to the home. The two use cases are quite different.
The Road to OnStar…
Historically, medical alert has been a transactional service. Press the button, route to the emergency room at the slightest hint of a problem, rinse and repeat…
This is stupid. As we’ve listened to the distress calls made through our service and talked to users and families directly, we’ve realized that over 50% of the calls have been triggered by non-emergency concerns. Some people just want someone to talk to, while others complain of mild pain, bruises or the common cold. In a nutshell, the majority of calls involve a person self-indicating something that would be poorly serviced by the emergency room.
And so we’ve started to partner with care plans and risk-bearing care providers, who bear the brunt of the cost for unnecessary emergency room visits, to route their members to the right services at the right time at lower cost. This generates enormous cost savings for the care plan and is way better for the member.
Here’s how it works…
If the member absolutely needs the emergency room, we’ll still route them there as soon as possible. We estimate this is around 30% of cases.
If they need help for an issue serviceable through remote or in-person triage—like a mild fever, infection or sprain—we’ll connect them to the appropriate triage team. Our two-way base station enables real-time routing and direct communication with any third party care manager or triage team.
If the member needs other services scheduled like a social worker visit, telepsychiatry, physical therapy, or transportation pickup for a primary care doctor visit, we can help a care plan route them to the appropriate team in the most seamless way possible to handle their needs. Importantly, lack of access to transportation is a common reason people either forego primary care services or call 911/ambulance services and go directly to the ER. This is entirely preventable and costs care plans a ton of money, both in the short- and long-term.
Another proactive aspect of QMedic’s service is that if we detect something is awry in the home, based on mobility or sleep changes captured from our medical alert device, we’ll prompt our call center to check on the person before escalating the case. This can be particularly important for members such as those with CHF or COPD, where monitoring mobility/sleep anomalies tends to be more relevant in their day-to-day care management.
Taken together, this is what makes QMedic the OnStar for care management. We can scalably and intelligently route and help provision third party services for at-risk people living at home.
For a care plan, which is typically bearing $1,500 and up per unnecessary ER visit, our turnkey routing service requires zero data integration and is priced affordably on a per-member-per-month basis. The service can be set up and shipped easily for populations at high-risk of ER utilization, where cost savings are immediately evident when members are appropriately re-routed.
The future of care management revolves around helping members like Joan live at home and get the right services at the right time at lower cost. At QMedic, it’s this mission that has driven our evolution—it’s also what gets me excited to come to work everyday.