healthcare delivery institute · set glucose, weight, and activity goals, and sugar’s feedback is...

Healthcare Delivery Institute

mHealth Apps: Designing for Patient Engagement and Behavior Change

Diane M. Strong

Professor of Information TechnologyRobert A. Foisie Business SchoolWorcester Polytechnic Institute

MassMEDIC Patient Engagement SummitJanuary 31, 2017

Worcester Polytechnic Institute


2

Problem:• Growth in chronic diseases and conditions, e.g., obesity, type 2

diabetes, and aging population

• Resulting lower quality of life and higher healthcare costs

“Individuals’ daily decisions and health behaviors are the most critical factor in how they fare”

Solutions: mHealth technologies are now available to help

• Mobile Devices (Smartphones)─ “The Future of Medicine is in your Smartphone”, Topol, WSJ, Jan 9, 2015.

─ “Pocket Doctor”, Strong, Agu, Pedersen, Tulu, Practical Patient Care, Oct 2012.

• Wearable, wireless health sensors and devices

• Online health information (e.g., web sites, patient portals)

• Online health communities (e.g., disease-specific support groups)

Why mHealth Technologies?



Key Design Challenges formHealth Solutions

3

Discontinued use before benefits occur• Within months, many mHealth apps are no longer used

• Many mHealth apps provide little or no user feedback, leading to discontinued use

• Effort burden (e.g., manual data entry) key in discontinued use

• Smart systems and automation can reduce burden but also reduce user control, which promotes engagement and use

mHealth devices are often viewed as data collection devices, rather than patient engagement devices

Lack of behavior change• Double adoption problem: Must adopt the app or device and adopt

new behaviors, both over the long-term

• Supporting healthy behaviors differs from supporting task completion, and requires voluntary usage assumption

WPI’s mHealth Research Activities

• Our goals and methods

• Overview of our Apps

4



5

Use mHealth technologies to design, develop and test apps that

• Engage patients in managing their health

• Motivate patients to adopt healthy behaviors

• Support self-management of chronic conditions, and wellness

Conduct research that

• Seeks to understand and generate predictions from collected health and usage data (e.g., using data mining and machine learning)

• Tests ideas, techniques, and predictive algorithms

Employ best clinical and technological practices in

• Disease management, Behavioral medicine, Wellness

• User engagement, Designing the user experience

Focus has been on

• Smartphone apps – design, develop, and test

• Patient portals – how and why patients use them

Our Goals and Methods


Healthcare Delivery Institute Our mHealth Ecosystem

6

Patients

•Technology & behavioral expertise•HDI & its Living Lab

•UXDM Lab

http://www.fhcw.org/

http://www.fhcw.org/



7

• TJR-Decision─ Physician-patient decision

making about Total Joint

Replacement, based on pain

and activity data

─ AHRQ funding

─ http://tjrapp.wpi.edu

• Sugar─ Advanced type 2 diabetes

and diabetic foot ulcers (image processing)

─ NSF funding─ http://sugar.wpi.edu

WPI’s Health Apps (sample)

For chronic disease, primarily elderly

http://tjrapp.wpi.edu/

http://sugar.wpi.edu/



• Relax─ Stress-induced eating

─ NIH funding

• SlipBuddy─ Overeating episodes

─ NSF REU Site funding

8

• Habit─ Weight loss─ NIH funding─ http://habit.wpi.edu/

• Mom-o-meter─ Control gestational

weight gain


For weight control

http://habit.wpi.edu/



9

• Socialoscope─ Senses loneliness by

analyzing smartphone activities of college students

─ At risk freshmen, seniors, internationals


For college student health

• SleepHealth─ College student

sleep health• Alcogait

─ Senses intoxication levels from gait

─ Put smartphone in your pocket and it tracks as you walk

Designing to Meet mHealthChallenges

• Minimize manual data entry

• Predict discontinued use and intervene

• Support goal setting and other behavioral medicine principles

• Follow user experience and behavior change support system design principles

10



Design for Minimal Manual Data Entry

11

SleepHealth will learn from data and reduce data entry over time

SlipBuddy only collects data on

problem episodes

Sugar uses Bluetooth glucometer and scale; footbox

for wound image capture



Predict and Intervene to Avoid Discontinued Use

12

Sugar provides feedback on measure-ment frequency and activity levels (e.g., goal of measuring weight once a week)

Sugar provides a summary of usage activity to care team, for purposes of human interventions

Socialoscope captures measures of smartphone usage to predict loneliness

SleepHealth will continually monitor usage and predict when usage is declining



13

For Sugar, users can set glucose, weight, and activity goals, and Sugar’s feedback is relative to those goals

Support Goal Setting, Other Behavioral Medicine Principles

HABIT’s design is based on behavioral medicine principles that are highly effective for weight management



Follow User Experience and BCSS Design Principles

14

Design Principles

• Delight the user with an excellent user experience

• Generate smart feedback and context-appropriate interventions─ Based on prior user input, time, location

─ SleepHealth will suggest different interventions depending on point in the educational cycle (e.g., near end of term) and location of the user

• Use Behavioral Change Support Systems (BCSS) design recommendations─ Use system features (e.g., personalization)

that aid behavioral change

─ Use system features that are persuasive and motivating

Allow some app personalization

For Mom-o-meter, provide pink, blue backgrounds; Display baby names


Healthcare Delivery Institute Conclusions and Next Steps

15

Lessons Learned

• Design for patient engagement – to attain continued use and behavior changes

• Development team needs technical, clinical, and behavioral expertise; also patient involvement

Next Steps

• Smarter feedback

• Design and test interventions

• Add more group and community level features

Questions, Comments, or Suggestions?

Funding for our research comes from the National Science Foundation, NSF Grants IIS-1065298 (Sugar app) and CNS-1560229 REU SITE (SlipBuddy app), National Institutes of Health, NIH Grants R21 DK098556-01 (Habit app) and R01 HL122302-01A1 (Relax app), and Agency for Healthcare Research and Quality, AHRQ Grant R21 HS024003 (TJR app)

Any opinions, findings, conclusions or recommendations expressed in this presentation are those of the authors and do not necessarily reflect the views of these funding organizations.


Healthcare Delivery Institute References

17

Apps

• TRJ-Decision: http://tjrapp.wpi.edu Sugar: http://sugar.wpi.edu Habit: http://habit.wpi.edu/• Socialoscope: https://web.wpi.edu/Pubs/ETD/Available/etd-042716-222034/unrestricted/pulekar.pdf (MS thesis)• Alcogait: https://www.bostonglobe.com/business/2016/12/19/app-make-drunk-drivers-toe-

line/jTpGccVnyn7upXgKpSJD5N/story.html

Our Publications (Sample re Diabetes App):

• Wang, L., P. C. Pedersen, D.M. Strong, B. Tulu, E. Agu, R. Ignotz, and Q. He (2016) “An Automatic Assessment System of Diabetic Foot Ulcers based on Wound Area Determination, Color Segmentation and Healing Score Evaluation”, Journal of Diabetes Science and Technology 10(2), pp. 421-428.

• Wang, L., P.C. Pedersen, D.M. Strong, B. Tulu, E. Agu, and R. Ignotz (2015) “Smartphone-Based Wound Assessment System for Patients With Diabetes”, IEEE Trans. on Biomedical Engineering 62(2), pp. 477-488.

• Strong, D., B. Tulu, E. Agu, Q. He, P. Pedersen, L. Wang, R. Ignotz, R. Dunn, S. Pagoto, and D. Harlan (2014) “Design of the Feedback Engine for a Diabetes Self-care Smartphone App.”, Proceedings of the Americas Conference on Information Systems.

• He, Q., E. O. Agu, D. M. Strong, B. Tulu, P. C. Pedersen, and L. Wang (2013) “The Design, Architecture and Implementation of Sugar: An Android Smartphone App for Advanced Diabetes”, Proceedings of the Diabetes Technical Meeting.

• Agu, E. O., P. C. Pedersen, D. M. Strong, B. Tulu, Q. He, L. Wang, and Y. Li (2013) “The smartphone as a Medical Device: Assessing Enablers, Benefits and Challenges”, Proceedings of the Workshop on Design Challenges in Mobile Medical Device Systems.

Other Literature:

• Kelley, H., M. Chiasson, A. Downey, and D. Pacaud (2011) “The Clinical Impact of eHealth on the Self-Management of Diabetes: A Double Adoption Perspective”, Journal of the Association of Information Systems, 12, pp. 208-234.

http://tjrapp.wpi.edu/

http://sugar.wpi.edu/

http://habit.wpi.edu/

https://web.wpi.edu/Pubs/ETD/Available/etd-042716-222034/unrestricted/pulekar.pdf

https://www.bostonglobe.com/business/2016/12/19/app-make-drunk-drivers-toe-line/jTpGccVnyn7upXgKpSJD5N/story.html

healthcare delivery institute · set glucose, weight, and activity goals, and sugar’s feedback is...

Documents