Objective Measurement of Sociability and Activity: Mobile Sensing in the Community

We read the Berke et al. article on mobile sensing and the accompanying Stanley/Osgood editorial with great interest. The finding that a wireless mobile device’s objective measurement of physical activity correlates highly with aggregate Yale Physical Activity Survey (YPAS) score – the latter being subject to recall bias and the other limitations described in the article – is intriguing, as are findings regarding sociability correlations with traditional survey scores, though the study needs replication.

We are currently investigating a wireless mobile device, with a complementary focus on trending data but also with a similar goal of determining a picture of one’s “well being” via analysis of physical activity and location patterns. The device is a socially acceptable, non-cumbersome FDA-cleared wristwatch that monitors skin temperature, location, and activity [1]. Blood pressure, weight, glucose, oxygen saturation, heart rate, and survey responses can also be collected via a Bluetooth interface to the system. Initial studies have found that elderly adults (aged 75+) are willing to wear the watch and find it easy to use [2]. Unlike the device used by Berke et al., the watch sends data in real time to a remote secure server that processes the data using “intelligent” algorithms. The system then provides real-time feedback – both to users and designated caregivers – in the form of exception-based alerts aimed at decreasing caregiver burden. Such alerts can be sent to the watch itself and/or to a pager, a smartphone, etc. We fully agree with Stanley/Osgood that too much data, without time and capability to analyze them, are not helpful. The ultimate goal of monitoring is to support improved health care decision-making. Towards this end, the watch system provides graphical summary data through on- demand reports via a secure web-based interface. This rechargeable watch device was designed with 24/7 continuous monitoring in mind, with a battery life of approximately three days when collecting full activity, location- tracking, and impact-sensing data. Furthermore, the system is designed so that inputs from additional Bluetooth-enabled third-party devices, such as a microphone, can be easily added. This would enable the seamless convergence of technologies described by Berke et al. with the type we have been investigating; the result would be a more comprehensive, “intelligent” sensing system to promote health-enhancing behaviors by individuals as well as insight for health care providers as described by Stanley/Osgood. We are excited about the possibilities fueled by the ideas these authors have presented in the Annals of Family Medicine.

References

[1] Papadopoulos A, Crump C, Wilson B. Comprehensive home monitoring system for the elderly. Proceedings of Wireless Health 2010, October 2010.

[2] Charness N, Fox M, and Papadopoulos A. Effectiveness and usability of a portable aging survey system for data collection in 75 and older elderly population: a feasibility study. Final report from Florida State University to AFrame Digital, Inc., 2010.

Competing interests:   Trying to do similar and complementary work; have received funding from DARPA, NIH, and NIA for this work.