Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.
Enabling Precision Behavior Change
@ehekler
Dr. Eric Hekler
Arizona State University
November 19, 2015
Talk given at the U...
Outline
• Precision behavior change
• Requirements for precision
behavior change
• Agile science
• UbiHealthy Cup
@ehekler
@ehekler http://www.nih.gov/precisionmedicine/
Behavior at the center
Hovell M, Wahlgren D, Adams M. Emerging theories in health promotion practice and research. 2009;2:...
Behaviors explain most variability in health
Flickr – Stuck in Customs@ehekler
40
15
5
10
30
Sub-Optimal Health behaviors
...
Why now? Personal, pervasive, &
powerful technologies
Flickr – Stuck in CustomsPatrick, Hekler, Estrin, Godino, Crane, Rip...
Why now? Behavioral meteorology
Flickr-Bart Everson
Patrick, Hekler Estrin, Godino, Crane, Riper, & Mohr, & Riley, Manuscr...
Why now? The world needs us…
Flickr – Stuck in Customs
http://youtu.be/QPKKQnijnsM
Flickr – just.Luc
Flickr-meanMrmustard
Just in Time Adaptive Interventions
Just in Time
• State of opportunity
or vulnerability
• Receptive
• Key target behavior...
Just in Time: State of vulnerability
Flickr - Rob Marquardt
@ehekler Nahum-Shani, Hekler, & Spruijt-Metz, (2016) Health Ps...
Just in Time: State of opportunity
Flickr - Miroslav Petrasko
@ehekler Nahum-Shani, Hekler, & Spruijt-Metz, (2016) Health ...
Just in Time: Receptive
Flickr-Jonathan Powell
Nahum-Shani, Hekler, & Spruijt-Metz, (2016) Health Psychology@ehekler
Adaptive: Series of “Just in Time” moments
@ehekler Flickr - Dave Gray
Precision behavior change spectrum
Individual/User
Controlled
System
Controlled
Individual/System
Balanced Control
@ehekler
System controlled
“Giving the fish”
NSF IIS-1449751: EAGER: Defining a Dynamical Behavioral
Model to Support a Just in Tim...
System model
Riley, Martin, Rivera, Hekler, et al. 2016; Martin, Riley, Rivera, Hekler, et al. 2014@ehekler
Idiographic trajectory models
Hekler, et al. 2013 Health Education and Behavior@ehekler
Martin, Rivera, & Hekler Manuscript Submitted for Publication
Model-predictive controller
@ehekler
Individual controlled
“Teaching to fish”
Eric Hekler, Jisoo Lee, Erin Walker, Winslow Burleson, Arizona State University; ...
Measure
success
towards
goal
Results
Self-experimentation
Plan
+ Implement for 1 week
@ehekler
@ehekler
Requirements for precision behavior change
• Interoperability/communication
– Robust system architectures
• Ecologically-v...
Interoperable systems
@ehekler
LeadSecondary
Secondary
Secondary
SecondarySecondary
Interoperable systems
https://www.apple.com/ios/whats-new/health/ http://researchkit.github.io/ http://sagebase.org/
Interoperable systems
www.openmhealth.org
Ecologically-valid data streams
@ehekler
Lead Secondary
Secondary
Co-Lead
Turning “noise” into information
https://ubicomplab.cs.washington.edu/
Data standardization
@ehekler
LeadCo-Lead
Secondary
Secondary
Secondary
Data standardization
www.openmhealth.org
Agile science targets
• Interoperability/communication
– Robust system architectures
• Ecologically-valid data streams
– S...
Pre-agile software “waterfall”
@ehekler
Agile (XP Scrum) development
@ehekler
Agile science philosophical assumptions
https://en.wikipedia.org/wiki/Philosopher#/media/File:The_School_of_Athens.jpg@ehe...
Target =“Idiographic generalization”
Analytic Perspective Focus Mixed Model Analogy
Between-person On-average effects acro...
Evidence & logic defines truth
https://en.wikipedia.org/wiki/Phases_of_Venus#/media/File:Phases-of-Venus.svg@ehekler
Rigor achieved via trial & error
https://en.wikipedia.org/wiki/Incandescent_light_bulb#/media/File:Edison_incandescent_lig...
Knowledge accumulation via effective curation
www..google.com@ehekler
Agile science products
• Modules
• System models
• Personalization algorithms
@ehekler
Modules
Smallest, meaningful, repurposable,& concrete
“Perfect” intervention package Components
Flickr - Paul Swansen Flic...
Modules
APIs
www.yelp.com@ehekler
IFTTT
http://www.ifttt.com
Modules
Templates
www.ifttt.com@ehekler
Modules
http://www.ifttt.comwww.ifttt.com@ehekler
System models: Meteorology analogy
Flickr-Bart Everson
Patrick, Hekler, Estrin, Godino, Crane, Riper, & Mohr, & Riley Manu...
System models
Riley, Martin, Rivera, Hekler, et al. 2016; Martin, Riley, Rivera, Hekler, et al. 2014@ehekler
Dynamic hypotheses- “Sweet Spot”
Hekler (PI), Rivera (Co-PI), NSF IIS-1449751
-15
-10
-5
0
5
10
15
20
0
2000
4000
6000
800...
Personalization algorithms
www.netflix.com@ehekler
Martin, Rivera, & Hekler Am. Control Conference (2015)
Personalization algorithms
@ehekler
Agile science process
@ehekler
Sprint
• GOAL: Discovery and resource-efficient
vetting of promising new approaches
@ehekler
Amy Luginbill; Samantha Quagliano; Sepideh Zohreh
S=Stop
M=Move
I= I statement; I can do it!
L=Love (positivity)
E=Exhale
...
Phoenix Proposition 104
John Harlow, Erik Johnston, Zoe Yeh@ehekler
Phoenix Proposition 104
John Harlow, Erik Johnston, Zoe Yeh@ehekler http://movephx.org/get-the-facts/maps/
Phoenix Proposition 104
John Harlow, Erik Johnston, Zoe Yeh@ehekler http://movephx.org/get-the-facts/maps/
Optimization
• GOAL: Translation of a promising resources
into useful & evidence-based tools for
real-world use.
@ehekler
Linda M. Collins
The Methodology Center
Penn State
methodology.psu.edu@ehekler
Micro-randomization design
• Sequential, full factorial designs
• Randomize intervention component
• Each time we might de...
System identification experiments
-100
100
300
500
700
900
1100
1300
1500
0
2000
4000
6000
8000
10000
12000
14000
1 8 15 2...
Release
• GOAL: Share useful resources via
effective curation.
@ehekler
Shared test-beds
@ehekler
Secondary
Secondary
Secondary
Lead
Co-Lead
Co-Lead
Paco (thank you, Bob Evans!)
www.pacoapp.com@ehekler
Paco-read this on the website
www.pacoapp.com@ehekler
Paco
www.pacoapp.com@ehekler
Paco
www.pacoapp.com@ehekler
Fundamental problem
@ehekler
We each build “optimized”
packages for one-off
problems
We need to build inter-operable
modul...
RoboCup
@ehekler
https://upload.wikimedia.org/wikipedia/commons/2/22/Robocup_Bremen_2006_-_four_legged.JPG
RoboCup
@ehekler
https://c2.staticflickr.com/8/7410/9238794627_4be245177e_b.jpg
RoboCup Structure
• Target: “developing by 2050 a Robot
Soccer team capable of winning against
the human team champion of ...
What is mHealth’s RoboCup?
@ehekler https://upload.wikimedia.org/wikipedia/commons/e/e3/13-06-28-robocup-eindhoven-099.jpg...
UbiHealthy Cup v.2
• Target:
– Actionable tool the community needs (e.g.,
passive measure of consumption, user
burden, goa...
UbiHealthy Cup Bracket
https://www.whitehouse.gov/assets/images/brackets2009c.jpg
Final four
RCT (36m, 4)
Opt. 2
(24m, 8)
...
RECAP
@ehekler
TARGET: Precision behavior change
Individual/User
Controlled
System
Controlled
Individual/System
Balanced Control
@ehekler
Why now? Behavioral meteorology
Flickr-Bart Everson
Patrick, Riley, Estrin, Hekler, Godino, Crane, Riper, & Mohr, Manuscri...
Why now? The world needs us…
Flickr – Stuck in Customs
http://youtu.be/QPKKQnijnsM
Flickr – just.Luc
Flickr-meanMrmustard
First step…
@ehekler
Stop building “perfect”
packages…
Start building interoperable
modules
Flickr - Paul Swansen Flickr -...
Next step, organize and share!
Dr. Eric Hekler, Arizona State University
ehekler@asu.edu, @ehekler
Enabling Precision Behavior Change
Enabling Precision Behavior Change
Enabling Precision Behavior Change
Upcoming SlideShare
Loading in …5
×

Enabling Precision Behavior Change

955 views

Published on

This talk was given at the University of North Carolina and describes a an open scientific research agenda for the development of more personalized and precise digital health interventions.

Published in: Health & Medicine
  • Be the first to comment

Enabling Precision Behavior Change

  1. 1. Enabling Precision Behavior Change @ehekler Dr. Eric Hekler Arizona State University November 19, 2015 Talk given at the University of North Carolina, Chapel Hill
  2. 2. Outline • Precision behavior change • Requirements for precision behavior change • Agile science • UbiHealthy Cup @ehekler
  3. 3. @ehekler http://www.nih.gov/precisionmedicine/
  4. 4. Behavior at the center Hovell M, Wahlgren D, Adams M. Emerging theories in health promotion practice and research. 2009;2:347-85.@ehekler
  5. 5. Behaviors explain most variability in health Flickr – Stuck in Customs@ehekler 40 15 5 10 30 Sub-Optimal Health behaviors Social Circumstances Environmental Exposures Healthcare Genetics McGinnis, et al. 2002 Health Affairs
  6. 6. Why now? Personal, pervasive, & powerful technologies Flickr – Stuck in CustomsPatrick, Hekler, Estrin, Godino, Crane, Riper, & Mohr, Riley, Manuscript in Prep@ehekler
  7. 7. Why now? Behavioral meteorology Flickr-Bart Everson Patrick, Hekler Estrin, Godino, Crane, Riper, & Mohr, & Riley, Manuscript in Prep@ehekler
  8. 8. Why now? The world needs us… Flickr – Stuck in Customs http://youtu.be/QPKKQnijnsM Flickr – just.Luc Flickr-meanMrmustard
  9. 9. Just in Time Adaptive Interventions Just in Time • State of opportunity or vulnerability • Receptive • Key target behavior does not have to happen now Adaptive • Responsive to: – micro-scale changes (e.g., weather, stress) – Meso-scale changes (e.g., season, motivational waves) – Macro-scale life transitions (e.g., retirement, becoming a parent) @ehekler Nahum-Shani, Hekler, & Spruijt-Metz, (2016) Health Psychology
  10. 10. Just in Time: State of vulnerability Flickr - Rob Marquardt @ehekler Nahum-Shani, Hekler, & Spruijt-Metz, (2016) Health Psychology
  11. 11. Just in Time: State of opportunity Flickr - Miroslav Petrasko @ehekler Nahum-Shani, Hekler, & Spruijt-Metz, (2016) Health Psychology
  12. 12. Just in Time: Receptive Flickr-Jonathan Powell Nahum-Shani, Hekler, & Spruijt-Metz, (2016) Health Psychology@ehekler
  13. 13. Adaptive: Series of “Just in Time” moments @ehekler Flickr - Dave Gray
  14. 14. Precision behavior change spectrum Individual/User Controlled System Controlled Individual/System Balanced Control @ehekler
  15. 15. System controlled “Giving the fish” NSF IIS-1449751: EAGER: Defining a Dynamical Behavioral Model to Support a Just in Time Adaptive Intervention, PIs, Hekler & Rivera @ehekler
  16. 16. System model Riley, Martin, Rivera, Hekler, et al. 2016; Martin, Riley, Rivera, Hekler, et al. 2014@ehekler
  17. 17. Idiographic trajectory models Hekler, et al. 2013 Health Education and Behavior@ehekler
  18. 18. Martin, Rivera, & Hekler Manuscript Submitted for Publication Model-predictive controller @ehekler
  19. 19. Individual controlled “Teaching to fish” Eric Hekler, Jisoo Lee, Erin Walker, Winslow Burleson, Arizona State University; Bob Evans, Google Flickr Juhan Sonin @ehekler
  20. 20. Measure success towards goal Results Self-experimentation Plan + Implement for 1 week @ehekler
  21. 21. @ehekler
  22. 22. Requirements for precision behavior change • Interoperability/communication – Robust system architectures • Ecologically-valid data streams – Smartphone, wearable, data and digital trace inference • Data standardization – Schemas, ontologies, and other knowledge structuring tools • Behavior change tools – Codified evidence-based and usable behavior change modules • Predictive computational models – Multi-level & multi-time scale mathematical models about health and behavior • Personalization algorithms – Recommender system, model-predictive controller, or other translations of data into useful adaptation decisions • Test-bed for iterative optimization – Data , “ground truth” definitions, and participants @ehekler
  23. 23. Interoperable systems @ehekler LeadSecondary Secondary Secondary SecondarySecondary
  24. 24. Interoperable systems https://www.apple.com/ios/whats-new/health/ http://researchkit.github.io/ http://sagebase.org/
  25. 25. Interoperable systems www.openmhealth.org
  26. 26. Ecologically-valid data streams @ehekler Lead Secondary Secondary Co-Lead
  27. 27. Turning “noise” into information https://ubicomplab.cs.washington.edu/
  28. 28. Data standardization @ehekler LeadCo-Lead Secondary Secondary Secondary
  29. 29. Data standardization www.openmhealth.org
  30. 30. Agile science targets • Interoperability/communication – Robust system architectures • Ecologically-valid data streams – Smartphone, wearable, data and digital trace inference • Data standardization – Schemas, ontologies, and other knowledge structuring tools • Behavior change tools – Codified evidence-based and usable behavior change modules • Predictive computational models – Multi-level & multi-time scale mathematical models about health and behavior • Personalization algorithms – Recommender system, model-predictive controller, or other translations of data into useful adaptation decisions • Test-bed for iterative optimization – Data , “ground truth” definitions, and participants @ehekler@ehekler
  31. 31. Pre-agile software “waterfall” @ehekler
  32. 32. Agile (XP Scrum) development @ehekler
  33. 33. Agile science philosophical assumptions https://en.wikipedia.org/wiki/Philosopher#/media/File:The_School_of_Athens.jpg@ehekler
  34. 34. Target =“Idiographic generalization” Analytic Perspective Focus Mixed Model Analogy Between-person On-average effects across participants Fixed effect (centered) Within-person On average effects over time Fixed effect (daily variation) Idiographic Individualized responses Random effect (error term) @ehekler
  35. 35. Evidence & logic defines truth https://en.wikipedia.org/wiki/Phases_of_Venus#/media/File:Phases-of-Venus.svg@ehekler
  36. 36. Rigor achieved via trial & error https://en.wikipedia.org/wiki/Incandescent_light_bulb#/media/File:Edison_incandescent_lights.jpg@ehekler
  37. 37. Knowledge accumulation via effective curation www..google.com@ehekler
  38. 38. Agile science products • Modules • System models • Personalization algorithms @ehekler
  39. 39. Modules Smallest, meaningful, repurposable,& concrete “Perfect” intervention package Components Flickr - Paul Swansen Flickr - Benjamin Esham @ehekler
  40. 40. Modules APIs www.yelp.com@ehekler
  41. 41. IFTTT http://www.ifttt.com Modules Templates www.ifttt.com@ehekler
  42. 42. Modules http://www.ifttt.comwww.ifttt.com@ehekler
  43. 43. System models: Meteorology analogy Flickr-Bart Everson Patrick, Hekler, Estrin, Godino, Crane, Riper, & Mohr, & Riley Manuscript in Prep @ehekler
  44. 44. System models Riley, Martin, Rivera, Hekler, et al. 2016; Martin, Riley, Rivera, Hekler, et al. 2014@ehekler
  45. 45. Dynamic hypotheses- “Sweet Spot” Hekler (PI), Rivera (Co-PI), NSF IIS-1449751 -15 -10 -5 0 5 10 15 20 0 2000 4000 6000 8000 10000 12000 14000 AveChangeSelfEffficacy ActualDailySteps Recommended Goal Actual Steps Δ Self-Efficacy @ehekler
  46. 46. Personalization algorithms www.netflix.com@ehekler
  47. 47. Martin, Rivera, & Hekler Am. Control Conference (2015) Personalization algorithms @ehekler
  48. 48. Agile science process @ehekler
  49. 49. Sprint • GOAL: Discovery and resource-efficient vetting of promising new approaches @ehekler
  50. 50. Amy Luginbill; Samantha Quagliano; Sepideh Zohreh S=Stop M=Move I= I statement; I can do it! L=Love (positivity) E=Exhale SMS: “If you are stressed today, try one of the following options, Deep breathing, Stretching, get up move around.” MOBILECAR MAIDSERVICES GREEN CLEAN Prototype 1: S.M.I.L.E. Prototype 2: Facial Wave Prototype 3: SMS Intervention Prototype 4: De-stress your carScrappy Trials @ehekler
  51. 51. Phoenix Proposition 104 John Harlow, Erik Johnston, Zoe Yeh@ehekler
  52. 52. Phoenix Proposition 104 John Harlow, Erik Johnston, Zoe Yeh@ehekler http://movephx.org/get-the-facts/maps/
  53. 53. Phoenix Proposition 104 John Harlow, Erik Johnston, Zoe Yeh@ehekler http://movephx.org/get-the-facts/maps/
  54. 54. Optimization • GOAL: Translation of a promising resources into useful & evidence-based tools for real-world use. @ehekler
  55. 55. Linda M. Collins The Methodology Center Penn State methodology.psu.edu@ehekler
  56. 56. Micro-randomization design • Sequential, full factorial designs • Randomize intervention component • Each time we might deliver component • Multiple components can be randomized • Randomized 100s or 1000s of times Klasnja, Hekler, Shiffman, Boruvka, Almirall, Tewari, Murphy, Health Psych, 2016@ehekler
  57. 57. System identification experiments -100 100 300 500 700 900 1100 1300 1500 0 2000 4000 6000 8000 10000 12000 14000 1 8 15 22 29 36 43 50 57 64 71 78 85 92 99 Points Stepsperday Days Points Provided (100, 300, 500) Fictionalized actual steps per day Daily step goal ((Baseline Median) to (Baseline Median+100% Baseline Median)) NSF IIS-1449751: Defining a Dynamical Behavioral Model to Support a Just in Time Adaptive Intervention, PIs, Hekler & Rivera@ehekler
  58. 58. Release • GOAL: Share useful resources via effective curation. @ehekler
  59. 59. Shared test-beds @ehekler Secondary Secondary Secondary Lead Co-Lead Co-Lead
  60. 60. Paco (thank you, Bob Evans!) www.pacoapp.com@ehekler
  61. 61. Paco-read this on the website www.pacoapp.com@ehekler
  62. 62. Paco www.pacoapp.com@ehekler
  63. 63. Paco www.pacoapp.com@ehekler
  64. 64. Fundamental problem @ehekler We each build “optimized” packages for one-off problems We need to build inter-operable modular resources Flickr - Paul Swansen Flickr - Benjamin Esham
  65. 65. RoboCup @ehekler https://upload.wikimedia.org/wikipedia/commons/2/22/Robocup_Bremen_2006_-_four_legged.JPG
  66. 66. RoboCup @ehekler https://c2.staticflickr.com/8/7410/9238794627_4be245177e_b.jpg
  67. 67. RoboCup Structure • Target: “developing by 2050 a Robot Soccer team capable of winning against the human team champion of the FIFA World Cup” • Rules: Change each year depending on state of the science http://www.robocup.org/about-robocup/regulations-rules/@ehekler
  68. 68. What is mHealth’s RoboCup? @ehekler https://upload.wikimedia.org/wikipedia/commons/e/e3/13-06-28-robocup-eindhoven-099.jpg Question generated by participants of the Schloss Dagstuhl Seminar on “Life-long Behavior Change Technologies:” June 21-26, 2015, http://www.dagstuhl.de/en/program/calendar/semhp/?semnr=15262
  69. 69. UbiHealthy Cup v.2 • Target: – Actionable tool the community needs (e.g., passive measure of consumption, user burden, goal-setting module, team module) • Bracket Science – Competing teams that are winnowed down at each stage (stop getting money) – Final four tested head-to-head in an RCT • Challenges change over time Thanks to Susan Murphy & Pedja Klasnja for co-developing this idea.@ehekler
  70. 70. UbiHealthy Cup Bracket https://www.whitehouse.gov/assets/images/brackets2009c.jpg Final four RCT (36m, 4) Opt. 2 (24m, 8) Opt. 2 (24m)Opt. 1 (18m,16) Opt. 1 (18m)Sprint 2 (12m, 32) Sprint 2 (12m)Sprint 1 (6m, 64) Sprint 1 (6m)
  71. 71. RECAP @ehekler
  72. 72. TARGET: Precision behavior change Individual/User Controlled System Controlled Individual/System Balanced Control @ehekler
  73. 73. Why now? Behavioral meteorology Flickr-Bart Everson Patrick, Riley, Estrin, Hekler, Godino, Crane, Riper, & Mohr, Manuscript in Prep@ehekler
  74. 74. Why now? The world needs us… Flickr – Stuck in Customs http://youtu.be/QPKKQnijnsM Flickr – just.Luc Flickr-meanMrmustard
  75. 75. First step… @ehekler Stop building “perfect” packages… Start building interoperable modules Flickr - Paul Swansen Flickr - Benjamin Esham www.agilescience.org
  76. 76. Next step, organize and share! Dr. Eric Hekler, Arizona State University ehekler@asu.edu, @ehekler

×