While wearable gadgets from Fitbit Inc. and Apple Inc. can track how much people are moving, a three-year-old New York company says it can figure out what exactly those movements are — from smoking to taking medication.
The technology from privately held Somatix Inc. takes data collected by smartwatches and other wearables and discerns which actions the wearer is performing, the company’s CEO and co-founder Eran Ofir, said in an interview. It was able to correctly determine the wearer's actions 80% of the time in a study published in November 2017 in the journal Nicotine & Tobacco Research.
With those results in hand, Somatix has launched its first application, SmokeBeat. The app will enable wearers and smoking cessation programs to tell when and how much people are smoking and pinpoint times when they tend to smoke more — such as after they meet with their boss.
While wearables are expected to remain popular for encouraging fitness, Somatix's technology illustrates how interest is growing in using them to encourage patients to adopt a wide array of healthy behaviors. If successful, the technology has the potential to reduce costs for insurers and hospital systems, according to a report also released in November 2017 by market intelligence firm IQVIA Holdings Inc.
The number of health apps worldwide has increased five-fold since 2013, to more than 318,000, according to the research by IQVIA's Institute for Human Data Science. While most remain associated with general wellness, those focused on helping manage health, such as detecting medical conditions and monitoring whether patients are really taking their medication, grew by 48% between 2015 and 2017, the report found.
Trials for such devices have also increased. While 82% of trials are sponsored by hospitals and patient-care institutions, those backed by the U.S. government have tripled since 2015, the report said. In addition, industry-sponsored studies have risen by more than half over that time, including those backed by medical device-maker Medtronic Inc., pharmaceutical company Roche Holding AG, software developer Posit Science Corp. and Proteus Digital Health Inc.
Among the new wearable products is iRhythm Technologies Inc.'s wearable patch, which can detect silent heart abnormalities before they trigger strokes and other events.
In addition, a University of California, San Francisco, proof-of-concept study published March 21 in the journal JAMA Cardiology found that smartwatch data coupled with a type of machine learning algorithm, deep neural networks, can detect atrial fibrillation, a leading cause of stroke.
Somatix's SmokeBeat can detect when and how much a person is smoking. The company says the information can be used to send messages of encouragement to people who are trying to quit at times they are likely to smoke.
Fitness trackers like those that count the number of steps a person takes, only take advantage of a small portion of the data wearables like smartwatches can collect, Somatix's Ofir said.
The SmokeBeat technology developed in 2014 by Somatix co-founder and Chief Technology Officer Uri Schatzberg goes beyond detecting how many motions people are making to determine "if you are smoking, eating, brushing your teeth, biting your nails or taking medication," Ofir said. "And all of this has many uses in healthcare."
Ofir sees the program helping with elder care and with managing patients who have returned home after surgeries. "Remotely the caregiver can know what's going on, how much the person is walking, eating and taking medication, or if they fall," Ofir said.
Encouraging hospitals to use the technology, Ofir said, are federal payment policies that penalize facilities if patients have to be readmitted.
"Hospitals have started approaching us because now hospitals have a big problem," Ofir said. By better monitoring patients after they are released from the hospital, they can act if there are signs the patient is not doing well, he said.
Alison Buttenheim, assistant professor of health policy at the University of Pennsylvania's Perelman School of Medicine, sees possibilities in SmokeBeat. Buttenheim is running a clinical trial that is looking less at whether it can eliminate smoking than at whether pregnant, low-income rural women would be open to using smartwatches to encourage them to quit.
One immediate way to use the data is in a micro-incentive program, Buttenheim said. SmokeBeat could let smoking cessation and public health programs know if a woman had refrained from smoking so they could design a program giving the woman cash or credit on a gift card on the days she did not smoke or otherwise met her goals.
Ofir said smoking cessation programs can go further. They can synchronize SmokeBeat to a person's calendar and discover, for example, that the person tends to chain-smoke after meeting with a manager, he said.
"Next time the system sees you have this meeting with your manager coming up at noon, it might send a message: 'Hey, why don't you do meditation or do whatever will help you refrain from smoking after having that meeting?'" he said.