Breaking the Kubernetes Kill Chain: Host Path Mount
Participatory OpenmHealth
1. Participatory mHealth:
an opportunity for innovation in healthcare, wellness, research
Deborah Estrin, destrin@cs.ucla.edu, http://cens.ucla.edu/Estrin
with collaborators from CENS, UCLA and UCSF (Ida Sim)
Patient self-care innovation happens outside the traditional enterprise and
clinical workflows
but it can still contribute to, and be, evidence-based
Monday, July 25, 2011 1
2. Why mobile/mHealth?
•3 lifestyle behaviors (poor diet, lack of exercise, smoking) cause
1/3rd of US deaths; 50% Americans have 1 or more chronic
diseases; age of onset getting younger.
•mHealth apps allow care support/data collection 24x7--chronic
disease prevention/management/research as part of daily life
•affordability/adoptability could support groundswell of medical
discovery, evidence-based practice about treatment/prevention
vision: support individuals, communities, clinicians to continuously
improve patient-centered, personalized, health and healthcare
mobile devices offer proximity, pervasiveness, programmability,
personalization
complementary to Internet interventions
Monday, July 25, 2011 2
3. Mobile devices can extend interventions and research beyond the
reach of traditional clinical care
168 hours a week...1440 minutes a day...(but not necessarily 365 days a year)
our actions our self report personal data repository
experience
sampling streams
context and
activity traces
Photo: Marshall Astor, WWW
aggregate measures,
trends, patterns
event detection
visualization processing
Monday, July 25, 2011 3
4. Whose mHealth?
•
A woman who is pre-diabetic tracks how eating/exercise habits
affect weight and fatigue; also explores effective, comfortable blood
pressure Rx dosage.
•
A young man who is struggling to find a treatment plan for
depression believes medication dose is ineffective; doctor blames
poor sleep habits/non-adherence. Patient self-monitoring includes
medication reminder/verifications, sleep survey, activity traces, to
guide adjustments in care plan, discussion of root causes.
•
A middle-aged woman who does not respond well to medication
for psoriasis monitors diet, stress, environmental factors; initiates
data campaign via social networking site for psoriasis sufferers. Each
volunteer runs mHealth app for 2-months to create large data set to
mine for patterns that precede flare-ups.
•
A group of high schoolers with asthma map their inhaler use and
make a case for shifting Track practice to an alternate location
farther from the freeway
Monday, July 25, 2011 4
5. Integrated personal data streams create Living Records
{ UI SKETCH } Jane Doe Messages (1) Create ODL
prev month May 2009 next month
Automatically prompted, geocoded, uploaded,
Interventions
May 1 May 31
analyzed:
Drug Prescribed
Patient prescribed new daily medication.
START DATE: May 26, 2009
Weight in pounds Measurements and trends for the month. AMOUNT: 200 mg
- physiological (weight, BP, glucose...)
FREQUENCY: One pill twice per day, once in the
160 morning and once in the evening.
- patient reporting (medication,
PREV. AMT.: 150 mg
symptoms, stress factors)
PREV. FREQ.: Same
NOTES:
Enter observations here.
- activity (location traces, exercise, sleep)
150
May 1 May 31
Blood Pressure Measurements and trends for the month.
- contextual, environmental,
{moves up and down
w/ scrolling, meta data
230 SYSTOLIC changes depending on
what is selected on left}
DIASTOLIC
social factors
0 Technical challenge to extract relevant
features, trends, patterns, anomalies
May 1 May 31
Map Traces and stationary times with points of interest highlighted
Gym Fast Food Park
switch to full screen view map
Processed/filtered personal data streams would become part
of emerging PHR/EHRs (complementary not duplicative)
5
Monday, July 25, 2011 5
6. Why focus on open architecture ?
broad applicability (diseases, demographics), heterogeneous/‘dual’ use (treatment,
engagement, evidence), evolving methodologies, need for innovation ecosystem
Stovepipe Architecture Open mHealth Architecture
Patient/Caregivers Patient/Caregivers Its not just a mobile app:
Analysis/
visualization/
• authoring prompts,
feedback Re-usable health triggers
A P P L I C A T I O N S
data and knowledge
services
Processing
• Individual feedback,
Storage
Standardized
personal data
tailoring
vaults and health
Data transport
specific data
exchange protocols • analysis and
Data capture
visualization
Mobile platforms Mobile platforms • Personal data vaults
iPhone/Android/ iPhone/Android/
Feature Phones Feature Phones
2
Monday, July 25, 2011 6
7. An open modular system is critical to foster rapid and meaningful
exploration and innovation
developers and data users: mobile app users:
clinicians, data analysts, etc. patients and healthy individuals
self
report
data dash- scripting reminders feedback
APPLICATION &
boards tools & &
LAYER automate
triggers messaging
d
measures
http://openmhealth.org
extract trends, feedback using
ANALYSIS
anomalies, data and social
LAYER
correlations media
import /
export
standardized data semantics
identity management EHR/PHR,
DATA
data security, privacy social media
SERVICES
configuration
Personal Data Vault
Swiernik, Estrin, Sim, et al 7
Monday, July 25, 2011 7
8. Open architectures enable privacy to be architected as well:
Personal Data Vault:
allow participants to retain control over their raw data
Mobile App Personal Third Party
Data Vault Services
- Data Capture / Upload - Analytics for Personal
(Prompted, Automated) - User Identity and Data Streams
Authentication
- Reminders - Interface to Clinical Care
- Long-term Data Plan, Personnel
- Feedback, Incentives Management
- Integration with EHR/PHRs
- Cross Patient Aggregation
Well defined interfaces allow Patient defined selective Well defined interfaces allow
mobile functions to be mixed, sharing with Open mHealth analytics functions to be mixed,
matched, and shared Server function matched, shared, compared
vault + filters = granular, assisted control over what/when you send
to whom, what data says about you, whether you reveal who you are or
share anonymously, ...
Monday, July 25, 2011 8
9. Closing remarks
“If you can’t go to the field with the sensor you want...go with the sensor you have!”
“The power of the Internet, the reach of the phone (Voxiva)”
Humans are in this loop--so HCI, privacy, visualization, bias, are part of research
agenda, and end to end systems that users can exercise are part of the process
It takes a healthy research ecosystem to bring information technology innovations to
meaningful societal use--Open architectures and platforms are a key building block.
24
Monday, July 25, 2011 9
10. Acknowledgments: Collaborators and Sponsors
Collaborators
Technology faculty, PIs:
Jeff Burke, Deborah Estrin, Mark Hansen, Ramesh Govindan, Martin Lukac, Nithya
Ramanathan, Mani Srivastava
Application/domain expert faculty/PIs (Health science, Education, Ecology):
Jacqueline Casillas, Patricia Ganz, Jeff Goldman, Eric Graham, Jerry Kang, Jenny Kim, Jane
Margolis, Maria Teresa Ochoa, Mary Jane Rotheram-Borus, Ida Sim (UCSF), , Dallas
Swendeman, Michael Swiernik
Students, staff:
Staff: Betta Dawson, Mo Monibi, Joshua Selsky, Eric Yuen, Ruth West,
Graduate students: Amelia Acker, Faisal Alquaddoomi, Peter Capone-Newton, Patrick Crutcher,
Hossein Falaki, Brent Flagstaff, John Hicks, Donnie Kim, Keith Mayoral, Min Mun, Sasank
Reddy, Jean Ryoo, Vids Samanta, Katie Shilton, Masanao Yajima, Nathan Yau,
Undergraduate students: Jameel Al-Azeez, Joey Degges, Gleb Denisov, Cameron Ketcham,
Ashley Jin, Chenyang Xia
Sponsors and Partners/Collaborators
UCLA centers: CENS, REMAP, Global center for families and children, Health Sciences
Federal funding: NSF: NETS-FIND Program, OIA, Ethics, BPC; NIH, NOAA
Corporate funding: Google, Intel, MSR, Nokia, T-Mobile, Cisco, Sun (RIP)
Foundations/NGOs: The California Endowment, Project Surya, Woodrow Wilson Center
Monday, July 25, 2011 10