3 sensor startups collecting population health data

From the mHealthNews archive
By Ephraim Schwartz
08:03 am

With the Baby Boomer generation - the largest generation in U.S. history - reaching senior age, there is a real concern that their needs will overwhelm the nation's struggling healthcare network. But advances in mHealth are giving hope that today's (and tomorrow's) seniors will be able to age gracefully and in their own homes.

While 70 may be the new 60 for seniors, they can take comfort in the fact that wireless technology is maturing almost as fast as they are. This is especially true in the field of sensor technology. High-tech manufacturers are now putting cellular, Wi-Fi, GPS, accelerometers, magnetometers, laser lights and even speech recognition inside pill bottles and wristwatches, while sensors that stick to the wall can identify to caregivers where a senior is in his or her house and what they're doing.

This week we look at three technologies in incubation at Startup Health: AdhereTech offers a wireless pill bottle that lights up, buzzes and sends text messages. LifeAssist offers a watch that includes its own gyroscope and GPS. And CarePredict includes a unique room beacon that allows caregivers to know exactly what the patient is doing in real time. 

Information these companies collect will provide vital data for population health professionals and  payers, for the longer they can keep seniors living independently, the less it costs them.

Adhere Tech’s Smart Pill Bottle

Josh Stein, CEO and co-founder of AdhereTech, says he doesn’t expect consumers to pay for the wireless technology embedded in pill bottles that light up, buzz and send text or automated voice messages. That’s why AdhereTech is targeting pharmaceutical companies, who will see increased revenues from patient engagement and adherence. AdhereTech is in pilot trials with Boehringer Ingelheim to deliver specialty medications for such disease states as HIV, cancer and stroke.

A secondary target market is pharmacy benefit managers (PBMs), who often deliver medications via mail order pharmacies. Pharmacies will be incentivized to return the pricier bottles and patients will be given pre-paid envelopes for return.

Where adherence (or lack thereof) may have serious ramifications on the outcome of drug trials, the cellular technology inside the bottle will also aid contract research organizations.

The feature that pharmaceutical companies like best, says Stein, is the feedback request AdhereTech sends out to patients who haven’t taken their meds despite all the reminders. An automated telephone call asks patients to indicate why they didn’t take their meds. Possible answers are given to the patient, who only has to press the number that corresponds to the correct reason. The responses can be routed to the case manager at the PBM.

“Reasons for non-adherence vary wildly,” Stein says.

The smart pill bottle, an opt-in program, is FDA-registered as a Class 1 medical device.

CarePredict Tempo

CarePredict’s Tempo is aptly named, as it monitors and learns from seniors their unique tempo of life and sends alerts whenever that changes.

Like AdhereTech, CarePredict isn’t putting all of its eggs in one marketing basket. The target markets include business-to-consumer - adult children of seniors - and business-to-business: accountable care organizations, managed care companies, post discharge facilities and hospitals.

“It allows them to take care of a large population by looking at the outliers,” says Satish Movva, founder and CEO. “Those that follow the norms are not a target for concern.”

Tempo consists of three main components: a wearable wireless sensor worn on the wrist of the user's dominant arm, a communications hub for sending and receiving alerts and messages and a room beacon about the size of two AA batteries in length and width that works with the patient device to indicate what room the patient is in. The communications hub also includes environmental sensors to measure temperature, humidity, CO2 levels and particulate levels.

CarePredict’s unit works by analyzing arm movements to tell if a person is eating, drinking or brushing his or her hair or teeth. When a patient enters a room, the beacon uses a laser light to determine location. The room beacons stick on the wall and include a turn dial to select the name of the room. And the magnetometer and accelerometer sensors in the device measure steps and body orientation.

During the first seven days the sensor is a passive observer to determine what is normal activity for the user. This then becomes the baseline for algorithms to determine if or when something out of the ordinary is happening. At that point it would send a text message or e-mail notification to a smart device. When used by primary care providers, more detailed notifications are sent. For example, it might say Mr. Jones went to the bathroom seven times today when normally he goes twice.

One of the real strengths of the device, Movva says, is the treasure trove of backend data it can give to a pharmaceutical company. It can help them understand, for instance, how a typical 76-year-old woman behaves. This can be cross-correlated with the diagnosis and medication data.

LifeAssist’s AudibleAssist

Similar to CarePredict Tempo, AudibleAssist targets both the 20 million people 75 years and older and their adult children, as well as ACOs and Fortune 500 companies like AT&T (with the second largest insurance pool in the country, consisting of some 300,000 retirees).

Not unlike OnStar, whicht tracks a subscriber’s vehicle and sends alerts in case of an accident, AudibleAssist is also a subscriber service, with the sensors inside a watch rather than behind the steering wheel.

Unlike OnStar, this system learns from past behavior. During the learning period AudibleAssist records the user’s activity through the accelerometer and location data through GPS and cellular, building a database of where the user tends to go, when they go, and other vital data such as what time they wake up or lay down.

“This gives us a good picture of habits so that the system can deploy anomaly detection. We do this against each user’s personal database,” says Jean Anne Booth, CEO of LifeAssist.

The device includes an astounding number of sensors - cellular, Wi-Fi, GPS, a 9-axis accelerometer plus speech-to-text with the basic vocabulary included in the watch. It goes out to the cloud if it does not recognize a word, and if that doesn’t work, it escalates to a live operator.

AudibleAssist is currently in development and will be in trials by the end of the year. 

Population health payoffs
The longer payers can keep seniors living independently, the less it costs them.

All the aggregate data that these devices collect will also offer invaluable information for population health in general, especially by grouping seniors into various categories.

AdhereTech's data on adherence will also give healthcare professionals details about adherence patterns for patients based on drug, disease state and demographic. Caregivers can review data points to gain a better understanding of which interventions were successful or unsuccessful, based on those same metrics.

Data gathered by CarePredict and LifeAssist, meanwhile, correlated with demographics, diagnoses, number and type of activities of daily living plus rates of motion and behaviors can be used in a de-identified manner to seek answers at a population level and even address the question of whether  seniors with these devices are living independently longer than the general population.

Ephraim Schwartz is a freelance writer based in Burlington, Vt. Schwartz is a recognized mobile expert and columnist, having spent 15 years as Editor-at-Large for InfoWorld, half of them covering the mobile space. Prior to that he was Editor-in-Chief of Laptop Magazine.


The latest news in digital health delivered daily to your inbox.

Thank you for subscribing!
Error! Something went wrong!