Keynote Address - University of Utah Health Sciences

Report
Informatics-Enabled Community
Engagement in the Washington
Heights Informatics Infrastructure
for Comparative Effectiveness
Research (WICER) Project
Suzanne Bakken, PhD, RN, FAAN, FACMI
School of Nursing and Department of Biomedical Informatics
Columbia University
1R01HS019853, R01HS022961
Community Engagement in Context of
the Learning Health Systems
• Building a strong fabric of trust among stakeholders
through communication and demonstration of value
• Principles as a foundation for fabric of trust:
– Build a shared learning environment
– Engage health and health care, population and patient
– Leverage existing programs and policies
• Difficulty of building a strong fabric of trust among
racial and ethnic minorities
– Low participation rates in research studies
– Low participation rates in biobanks
– Limited use of information technologies for health-related
purposes.
Goals of WICER
• Comprehensive understanding of the
Washington Heights/Inwood population
• Facilitate research with this population
• Demonstrate infrastructure capabilities for
comparative effectiveness research
Washington Heights/Inwood
• 5 zip codes: 10031, 10032,
10033, 10034, 10040
• Represents significant issues
in health care disparities
WICER Components
• Research data warehouse - Wilcox
• Research data explorer (RedX) and I2B2 for viewing
research data warehouse – Wilcox
• Informatics support for recruitment and integrating
clinical and research workflows – Weng
• Comparative effectiveness studies – Bigger,
Feldman (Visiting Nurse Service of New York)
• Community survey – Bakken, Boden-Albala, Fleck
WICER Data
In patient
Clinician
Clinical
Encounters
Out
patient
Clinician
Community Members
Integration
Platform
Research
Data Access
Research
Data
Warehouse
RedX
Home
care
Clinician
Long
term
Patient
Researcher
Clinician
Research
Encounters
Trial
Coordinator
Survey
Coordinator
Local
Databases
Data Access
Expert
Community Engagement
• Survey design and implementation
• Returning data to community
• Role of informatics
Community Survey
• Demographics
– Including socio-economic status
• Anthropometric measures and vital signs
– Blood pressure, height, weight, waist circumference
• Surveys (incorporate selected PROMIS measures) including
–
–
–
–
–
–
–
–
–
Nutrition
Physical activity
Social role performance
Health and illness perceptions
Self health assessment
Depression
Medication adherence
Quality of life
Health literacy
• Baseline and follow-up approximately 1 year apart
• Some overlap with BRFSS and NYC Community Health Survey – Yoon
eGEMS publication
Survey Populations
• Community
– 3,940
• Ambulatory Clinics
– 1,200
• Community
Outreach Center
– 800
Survey Design and Implementation
• CTSA-funded Columbia-Community Partnership for
Health (CCPH)
• Free community blood pressure screening and
education at CCPH
• Focus groups to inform survey content
• Data collection by bilingual community health workers
from Washington Heights/Inwood in homes,
community organizations, and local businesses as well
as CCPH and clinics
• Incorporation of snowball sampling methods
• Compensation for participant time with incentives of
value to residents (e.g., grocery coupons)
• Asked about top health concerns
Building a Strong Fabric of Trust
• Participation rates in research studies – increased
enrollment rates for WICER referrals to other
studies
• Participation rates in biobanks – exceeded
biospecimen collection goals
• Use of information technologies for healthrelated purposes – only 7% had used information
technology for health purposes, but most
participants agreed to linkage of survey data to
clinical data
Returning Data to the Community
• Survey respondents
• Community-based Organizations
Returning Survey Data to Participants
• Established WICER visualization working group that
includes investigators from a variety of disciplines
including nursing, medicine, public health, biomedical
informatics, engineering, human factors, communication,
art, fashion, theater
• Creation of infographics of varying levels of complexity
based on cognitive tasks
• Focus groups regarding preference and acceptability
• Basic research regarding comprehension
• Applied research related to actionability in consumerfacing and provider-facing applications
• Process reported in Arcia et al. AMIA 2013
Key Lessons from Focus Groups
 More is more
 Culture matters
Your Blood Pressure
Risks of High Blood Pressure
Vision loss
Hypertensive
retinopathy
Stroke
Cerebrovascular
accident
Heart
attack
Myocardial
infarction
Blood vessel
Damage
Kidney
failure
Atherosclerosis
Renal failure
Days with Adequate Physical Activity in
the Last Month
15
22
25
Jill Jones
54 years old
50-60 year-old females
in Washington Heights
50-60 year-old females
nationwide
Physical Activity
Jill Jones
54 years old
50-60 year-old females 50-60 year-old females
nationwide
in Washington Heights
Pictorial bar graph
Image from http://advgraphic2.blogspot.com/2011/01/icons-of-graphic-design.html
Physical Activity
Community-based Organizations
• Focus groups
• Elicitation of information needs and
desired formats
Informatics Approaches
• Simple topic modeling
• Community web site as mechanism of
return of survey data to respondents
• Electronic Tailored Infographics for
Community Engagement, Education, and
Empowerment (EnTICE3)
The EnTICE3 Framework
and Sample Infographic
EnTICE3 Requirements
 Construct individual- and community- level
visualizations
₋
adapt to change
“change the color & size”, “replace star icons with banana icons”,
“change the age group from 18-35 to 18-25”, ...
₋
reusable, generalizable
“same bar graph, but using different survey variables”, “generate the
bar graph for a different dataset”, “display this chart in another
website”, “a different institution wants a similar visualization”
EnTICE3 Requirements
• Construct individual- and community- level
visualizations
₋
adapt to change
■
■
₋
iterative approach
prototype → production
reusable, generalizable
modular, generic services
■ web standards (HTML, HTTP)
■ web components (encapsulated interactive
templates)
■
Governance Infrastructure
 Structures and processes that control data
Query and extract
₋ Store new data
₋ Access control
₋ Audit access
₋
 Underlies EnTICE3 and other components
Conclusions
• Significant barriers to research participation exist
for populations at high risk for health disparities
• Community engagement approaches essential to
reach such participants
• Ethical imperative to return data to those that
contribute it in a way that is comprehensible and
actionable for improving health
• Important component of a learning health system
• Requires continued engagement and iterative
refinement

similar documents