DyMand: An Open-Source Mobile and Wearable System for Assessing Couples' Dyadic Management of Chronic Diseases

Abstract

Married adults share illness management with spouses and it involves social support and common dyadic coping (CDC). Social support and CDC have an impact on health behavior and well-being or emotions in couples dyadic management of diabetes in daily life. Hence, understanding dyadic interactions in-situ in chronic disease management could inform behavioral interventions to help the dyadic management of chronic diseases. It is however not clear how well social support and CDC can be assesed in daily life among couples who are managing chronic diseases. In this work, we describe the development of DyMand, a novel open-source mobile and wearable system for ambulatory assessment of couples dyadic management of chronic diseases. Our first prototype is used in the context of diabetes mellitus Type II. Additionally, we describe our experience deploying the prototype in two pre-pilot tests with five subjects and our plans for future deployments.

Full text

DyMand: An Open-Source Mobile and Wearable
System for Assessing Couples’ Dyadic
Management of Chronic Diseases
George Boateng1, Prabhakaran Santhanam1, Janina L¨uscher2, Urte Scholz2,
and Tobias Kowatsch3
1ETH Z¨urich, Z¨urich, Switzerland
2University of Z¨urich, Z¨urich, Switzerland
3University of St. Gallen, St. Gallen, Switzerland
Abstract. Married adults share illness management with spouses and
it involves social support and common dyadic coping (CDC). Social sup-
port and CDC have an impact on health behavior and well-being or
emotions in couples dyadic management of diabetes in daily life. Hence,
understanding dyadic interactions in-situ in chronic disease management
could inform behavioral interventions to help the dyadic management of
chronic diseases. It is however not clear how well social support and CDC
can be assesed in daily life among couples who are managing chronic dis-
eases. In this work, we describe the development of DyMand, a novel
open-source mobile and wearable system for ambulatory assessment of
couples dyadic management of chronic diseases. Our first prototype is
used in the context of diabetes mellitus Type II. Additionally, we de-
scribe our experience deploying the prototype in two pre-pilot tests with
five subjects and our plans for future deployments.
Keywords: Mobile health ·Wearable computing ·Design science ·Dyadic
coping ·Social support ·Chronic diseases.
1 Introduction
Evidence suggests that for married adults, illness management is mainly shared
with their spouses and it involves social support and common dyadic coping
(CDC) [11, 9]. Social support and CDC have been shown to have some impact
on health behavior and well-being or emotions in couples’ dyadic management
of chronic diseases such as diabetes in daily life [6, 8, 4]. Hence, understanding
dyadic interactions in-situ in chronic disease management could inform behav-
ioral interventions via a digital coach to help the dyadic management of chronic
diseases. It is however not clear how well social support and CDC can be as-
sessed in everyday life among couples who are managing chronic diseases such as
diabetes mellitus Type II (T2DM). In this work, we describe the development of
a novel open-source mobile and wearable system for ambulatory assessment of
couples’ Dyadic Management of chronic diseases (DyMand). A first prototype
is used in the context of T2DM, a common chronic disease affecting 9.4% of the

2 G. Boateng et al.
U.S. population [2] and in Switzerland, 4.9% of the male Swiss population and
4.2% of the female Swiss population [10].
2 Design of the Artifact
The DyMand system as shown in Fig. 1 consists of a smartwatch app 4, smart-
phone app 5and a cloud-server system built on top of MobileCoach, an open-
source software platform for the design of behavioral interventions and ecological
momentary assessments [3,5]. In developing DyMand, experts from the field of
computer science, information system and health psychology used justificatory
knowledge from prior work [6,4,13,1] about social support, CDC, health be-
havior and well-being to derive a list of design requirements (DR) that are im-
portant for collecting corresponding data in-situ. We describe the requirements
below along with our implementation approach.
Fig. 1. Overview of the DyMand system
DR1: Track physical closeness of the couple during waking hours. Imple-
mentation: We used the Bluetooth Low Energy (BLE) signal strength of the
smartwatch. Each partner is given a smartwatch with one acting as the central
and the other acting as the peripheral. The central device continuously scans
and tries to find the peripheral device and then logs the BLE signal strength.
This data gives an estimate of the physical closeness of the couple throughout
the day.
DR2: Collect relevant multimodal sensor data during waking hours every
hour when the couple is close and speaking (Fig. 2). Implementation: The
smartwatch (Polar M600) collects five minutes of the following sensor data once
per hour within the morning and evening hours set by the couples: audio, heart
rate, accelerometer, gyroscope, ambient light and BLE signal strength between
each partner’s smartwatch. The smartphone (Nokia 6.1) collects video, audio
and ambient light for three seconds when the subjects are completing the self-
report on the phone. We ensure that there are at least 20 minutes between
subsequent data collection to reduce the burden of completing the self-reports.
To optimize the quality of data collected within that hour, we collect data when
4https://bitbucket.org/mobilecoach/dymandwatchclient/src/master/
5https://bitbucket.org/mobilecoach/dymand-mobilecoach-client/src/master/

DyMand: An Open-Source Mobile and Wearable System 3
the couple is physically close and when the partners are speaking. We use a
two-step process. First, we determine closeness using the BLE signal strength
between the smartwatches. We check if the signal strength is within a certain
threshold, which corresponds to a distance estimate. Then, we determine if the
couple is speaking by using a voice activity detection (VAD) algorithm, which
is implemented on the smartwatch. The VAD algorithm is a machine-learning
algorithm, which was trained to classify speech versus non-speech. In the case
in which this condition of closeness and speaking is not met in the hour, we do
a backup recording in the last 15 minutes of the hour.
Fig. 2. Closeness and Speaking Detection
DR3: Collect self-report data immediately after sensor data collection and at
the end of the day (Fig. 3). Implementation: After the five minutes of sensor
data collection, the smartwatch app triggers a self-report for each partner to
complete on their phones. Additionally, at the end of the day, the self-report is
also triggered. The self-report asks questions about social support, CDC, health
behavior and emotions.
3 Significance to Research
This work seeks to answer the following research question: How effectively can a
mobile and wearable system collect self-report and sensor data about a couple’s
dyadic interactions in everyday life? The development of DyMand entails the
use of novel approaches such as combining BLE signal strength and VAD to
optimize data collection among couples. The method used in developing DyMand
can be used by other researchers to develop similar artifacts to collect data to
understand various constructs among couples who are managing chronic diseases,
and also among other dyadic interactions such as friendships, and sibling and
parent-child relationships. Also, another potential research use case is to better
understand communication patterns in-situ and performance measures of teams
in organizations.

4 G. Boateng et al.
Fig. 3. Screenshots of self-report on DyMand app
4 Significance to Practice
The DyMand system results in the collection of subjective data from couples in-
situ about social support, CDC and their dyadic management of chronic diseases.
Additionally, the system collects objective data such as audio, heart rate, and
movements. Audio for example can be used to code constructs such as social
support and CDC. Also, audio together with heart rate and movements data
can be used to assess emotions, which are an important outcome of the social
support and CDC constructs. All the data collected by DyMand can then be used
to inform just-in-time-adaptive interventions by for example, a digital coach to
improve the management of chronic diseases [7].
5 Evaluation of the Artifact
Currently, we have run two pre-pilot tests of the DyMand system. We had five
subjects in each of the two tests with each test lasting one-week. At the end of the
testing, we had collected 900+ five-minute sensor data along with corresponding
triggered self-reports. We had two main technical challenges based on the results
of the testing which have been resolved: 1) Unavailability of Internet sometimes
resulted in self-reports not being triggered on the phone and 2) A bug in the
smartwatch app caused the battery of one of the watches to last for only six
hours, which resulted in the watch needing to be recharged during the day and
increased the user burden.
We plan to run a study starting in April 2019, funded by the Swiss National
Science Foundation [12], through which we will collect various self-report and
sensor data in everyday life using the DyMand system. The goal of this study is
to understand the relationship between the social support and CDC constructs,
and the health behavior and well-being or emotions of couples in which one
partner has T2DM. In this study, we aim at including N =180 couples (n =360

DyMand: An Open-Source Mobile and Wearable System 5
individuals), where one partner has T2DM. We will collect sensor and self-report
data from them for seven days when they spend time together in the mornings
and evenings during the weekdays, and the whole day during the weekends.
References
1. Boateng, G., Kotz, D.: Stressaware: An app for real-time stress monitoring on the
amulet wearable platform. In: 2016 IEEE MIT Undergraduate Research Technol-
ogy Conference (URTC). pp. 1–4. IEEE (2016)
2. Centers for Disease Control and Prevention: National Diabetes Statistics Re-
port, 2017 (2017), https://www.cdc.gov/diabetes/pdfs/data/statistics/national-
diabetes-statistics-report.pdf
3. Filler, A., Kowatsch, T., Haug, S., Wahle, F., Staake, T., Fleisch, E.: Mobilecoach:
A novel open source platform for the design of evidence-based, scalable and low-
cost behavioral health interventions: overview and preliminary evaluation in the
public health context. In: 2015 Wireless Telecommunications Symposium (WTS).
pp. 1–6. IEEE (2015)
4. Iida, M., Parris Stephens, M.A., Rook, K.S., Franks, M.M., Salem, J.K.: When
the going gets tough, does support get going? determinants of spousal support
provision to type 2 diabetic patients. Personality and Social Psychology Bulletin
36(6), 780–791 (2010)
5. Kowatsch, T., Volland, D., Shih, I., R¨uegger, D., K¨unzler, F., Barata, F., Filler,
A., B¨uchter, D., Brogle, B., Heldt, K., et al.: Design and evaluation of a mobile
chat app for the open source behavioral health intervention platform mobilecoach.
In: International Conference on Design Science Research in Information System
and Technology. pp. 485–489. Springer (2017)
6. Miller, D., Brown, J.L.: Marital interactions in the process of dietary change for
type 2 diabetes. Journal of Nutrition Education and Behavior 37(5), 226–234
(2005)
7. Nahum-Shani, I., Smith, S.N., Spring, B.J., Collins, L.M., Witkiewitz, K., Tewari,
A., Murphy, S.A.: Just-in-time adaptive interventions (JITAIs) in mobile health:
key components and design principles for ongoing health behavior support. Annals
of Behavioral Medicine 52(6), 446–462 (2017)
8. Prati, G., Pietrantoni, L.: The relation of perceived and received social support to
mental health among first responders: a meta-analytic review. Journal of Commu-
nity Psychology 38(3), 403–417 (2010)
9. Rintala, T.M., Jaatinen, P., Paavilainen, E., ˚
Astedt-Kurki, P.: Interrelation be-
tween adult persons with diabetes and their family: a systematic review of the
literature. Journal of family nursing 19(1), 3–28 (2013)
10. Schweizerische Diabetes-Gesellschaft: Diabetes mellitus. Das Wichtigste in Kurz
(Feb 2015), http://www.sdgshop.ch/media/pdf/197-de.pdf
11. Seidel, A.J., Franks, M.M., Stephens, M.A.P., Rook, K.S.: Spouse control and
type 2 diabetes management: moderating effects of dyadic expectations for spouse
involvement. Family relations 61(4), 698–709 (2012)
12. Swiss National Science Foundation: Measuring the Impact of Social Support and
Common Dyadic Coping on Couples Dyadic Management of Type II Diabetes
by a Novel Ambulatory Assessment Application for the Open Source Behavioral
Intervention Platform MobileCoach, http://p3.snf.ch/project-166348

6 G. Boateng et al.
13. Timmons, A.C., Chaspari, T., Han, S.C., Perrone, L., Narayanan, S.S., Margolin,
G.: Using multimodal wearable technology to detect conflict among couples. Com-
puter 50(3), 50–59 (2017)