Harnessing Novel Data Sources and Technologies for the Study of Social Determinants of Health in Heart, Lung, Blood, and Sleep Disorders

The National Institutes of Health’s (NIH) National Heart, Lung, and Blood Institute (NHLBI) hosted a virtual workshop on September 29–30, 2020, to explore digital technology to create novel data sources and leverage that technology for research on social determinants of health (SDOH). Healthy People 2030 defines SDOH as “…the conditions in the environments where people are born, live, learn, work, play, worship, and age that affect a wide range of health, functioning, and quality-of-life outcomes and risks.” Workshop participants also examined unintended consequences from technology use and discussed implementation strategies to put the new technologies into action. The workshop represented a continuation of NHLBI’s long-standing work to understand and address the drivers of racial and geographic disparities related to heart, lung, blood, and sleep (HLBS) disorders. Per the 21st Century Cures Act of 2016, both the Precision Medicine Initiative and the NIH-Wide Strategic Plan incorporate consideration of the influence of SDOH as well as biological factors on health that contribute to health disparities. Investigating the factors that account for differences in health among populations and leveraging emerging opportunities in data science are part of the NHLBI’s Strategic Vision.

Agenda

Background

Health encompasses many facets of our lives and is more than physical well-being. The disproportionate burden of coronavirus 2019 (COVID-19) on minority communities has brought U.S. health disparities—and the imperative to promote health equity—into sharp focus. Data science and novel data sources and technologies such as machine learning and mobile apps that can provide insight into SDOH and potential evidence-based approaches to eliminating health inequities.

Evidence indicates that communities and neighborhoods (e.g., ZIP Codes) drive health outcomes such as life expectancy and quality of life. A range of new digital tools and data sources present an opportunity to improve health outcomes by facilitating measurement (e.g., sensor sampling in real time), diagnosis (e.g., portable imaging), treatment (e.g., remote care), and global responses (e.g., disaster responses and disease surveillance). Challenges to realizing the enormous potential of novel technologies and data sources for improving individual and population health include those related to the digital divide—that is, inequities (particularly socioeconomic) in access to these technologies. Workshop presenters discussed solutions and presented examples of projects that overcome such challenges; overarching themes included the importance of community engagement and partnerships and the crucial need to work seamlessly across multiple sectors. A key element of community engagement is that it provides insight into how SDOH impacts health, which in turn drives health outcomes where people live, work, and play.

Workshop Objectives

1. Review the use of digital technology to enhance and create novel data sources (e.g., sensors, geospatial mapping, social media);
2. Determine how to leverage digital technology and novel data sources in the study of SDOH;
3. Determine the influence of the digital divide and unintended consequences of novel technologies widening health disparities by geography, race/ethnicity, etc.; and
4. Discuss implementation strategies to enhance the use of novel data sources by the research community to advance the science of SDOH research.

Overarching Themes

The following overarching themes emerged from the discussion:

• Community-based (participatory) approaches are crucial to technology development for acceptance and usability of the technology for the intended population. They involve an iterative process influenced by stakeholders and the information they provide to researchers, which ensures that SDOH identification and tracking are aligned with individual and community needs.
• Community members should also be involved in decision-making and priority setting related to SDOH data and technology-based solutions, and the process should include the specific experiences of people of color.
• In-home and mobile technologies offer a wealth of data that can be leveraged for addressing SDOH and improving health outcomes, but the information must be summarized and presented in a way that is actionable.
• Digital health applications (with coaching support via chat or telephone) must be customized to meet specific needs and priorities of underrepresented minorities, with the involvement of community members (particularly young people) in their development and attention to training people in their use.
• In the current technology world, many health applications used by consumers have not been properly tested or developed with patient or clinical input. That is, the marketing of technology outpaces the science.
• The lessons learned over the past decade and rapid technological advancements can be applied to create a digital learning health system for the future. In the context of the shift from fee-for-service or fee-for-volume toward fee-for-value, the technology and new data sources present opportunities to address SDOH.
• There is a need to enhance the ability to integrate patient-generated health data and information on SDOH into the digital health ecosystem.
• Interoperability is essential for leveraging data on SDOH for intervention in real-world settings.
• Technology can be used to bring expertise to bear to deliver solutions for communities, with an increasing focus on responding to disasters and addressing mental health needs.
• Industry stakeholders and policy makers need education on the value of leveraging SDOH data to improve health and healthcare, and the importance of addressing privacy and security concerns.
• By linking community resources and data on SDOH, predictive analytics can help advance community health and eliminate disparities, particularly in limited-resource settings.
• Novel data resources are needed to understand the problem or the human and physical characteristics of places and environments (also called the Uncertain Geographic Context problem).
• Geospatial approaches and layering spatial data over health data can support the delivery of personalized medicine to scale to address population health, resulting in precision community health. Data from studies using geospatial approaches needs to be uploaded in ways that are accessible and actionable.

COVID-19 Themes

Participants identified the following key considerations to address SDOH and COVID-19:

• Regarding data:
  • Real-time data are needed (e.g., testing, how information was gathered, who was gathering it, and how it was being used).
  • Currently, there are no measures that directly address COVID-19 risks or impacts.
  • Generally, national datasets have long time lags and are therefore less relevant for action on COVID-19 or other health emergencies.
  • City Health Dashboards that support the development of COVID-19 local-risk indexes, and other related metrics (e.g., economic indicators of well-being and social mobility measures) can also be generated.
• Regarding clinical practice:
  • Medical group visits moved online during the pandemic and recently billable, but it took time to get clinicians up to speed.
  • Patients’ desire to use tele-visits could be increased further and leveraged for better care and connections to social and community services.
  • For the Medicare population, patients need information on coverage for COVID-19 services and how to obtain care.
  • There is a need to counter social isolation for individuals without family members or those living alone.
  • A population vulnerable to the COVID-19 impact is service workers, who have experienced an aggravation of clinical symptoms (e.g., diabetes, anxiety, blood pressure) and increased need for health and social services (e.g., meals on wheels, rent assistance, and stable employment).
• Regarding enhancing the care system:
  • A military-sector model is being applied to provide critical care tele-health from anywhere via smartphones and over cellular networks, which may work in low-resource areas including rural areas with few hospital or ICU beds.
  • As capabilities for tele-health grow, the architecture may support the management of patient groups, shifting of resources from one region to another based on need, and scaling to a national level.
  • Capabilities for the remote collection and monitoring of device data (e.g., ventilator and infusion pumps) are available and may reduce the need for local clinicians.

Challenges

Workshop participants identified the following challenges:

• The digital divide remains—that is, not all individuals and communities have access to the relevant technologies.
• Most data on health and drivers of health are not organized at the city or neighborhood level, rather they are organized at the county, state, or national level.
• Many studies on SDOH and health outcomes use national or regional data and analyze data independent of rurality.
• Major data lags limit opportunities for neighborhood-based research.
• Researchers and practitioners face a number of challenges related to recording and capturing SDOH, including a lack of standardization, knowledge and buy-in from providers and staff members, and time to discuss SDOH.
• Currently, most providers lack awareness of Z Codes and are not using them possibly due to a lack of reimbursement as a barrier.
• The need for Wi-Fi, data, and physical devices can limit the use of these technologies.
• Significant privacy concerns remain (e.g., whether voice-based assistants are recording all conversations).
• There is a need to develop a framework for digital research that centers around people and providers, supports healthcare delivery, and encourages the desire to participate and share information.
• Processes for coordinating social service organizations and healthcare stakeholders are needed to evaluate and track results and adjust interventions.
• Effective research requires better coordination between basic and applied science.

Workshop Outcome Strategies

Customize intervention:

1. Use geospatial tools to map SDOH needs (e.g., education, food, health care, etc.) and target resources in those areas related to SDOH for HLBS disorders
2. Use mobile data to measure health and environmental conditions for HLBS disorders (e.g., sensors, smartphone, smartwatch, etc.) in various locations where people live, work, and play
3. Use in-home technology (e.g., telemedicine, voice-activated home devices, etc.) to manage chronic HLBS disorders in high-risk communities impacted by SDOH
4. Involve community members in decision making most impacted by SDOH (e.g., needs assessment, community boards, etc.)
   1. Determine the risks and ethics of customized research solutions
   2. Address accessibility of research solutions in the communities most impacted by SDOH (e.g., Internet access, cost of technology, literacy, etc.)

Make SDOH data in electronic health records (EHRs) actionable:

1. Provide apps to map and link community resources via an application programming interface (API) to EHR records that provides information and location for resources available in a targeted area
2. Study adoption of standardized SDOH data based on accepted standards for local EHRs and collaborate between beneficiaries, community groups, health care providers and other collaborators to improve health.

Workshop Chair

Garth Graham, M.D., M.P.H., Vice President, Chief Community Health Officer, CVS Health

Workshop Leaders

Erin Iturriaga, DNP, M.S.N. NHLBI, Division of Cardiovascular Sciences (DCVS)
Rebecca Campo, Ph.D. (DCVS)
Alison Brown, M.S., Ph.D. (DCVS)
Marishka Brown, Ph.D. NHLBI, National Center on Sleep Disorders Research (NCSDR)
Patrice Desvigne-Nickens, M.D. (DCVS)
Michelle Freemer, M.D., M.P.H. NHLBI, Division of Lung Diseases
Melissa Green Parker, Ph.D., NHLBI, Center for Translation Research and Implementation Science (CTRIS)
Michelle Olive, Ph.D. (DCVS)
Nicole Redmond, M.D., Ph.D., M.P.H. (DCVS)
Sharon Smith, Ph.D. NHLBI, Division of Blood Diseases and Resources
Catherine (Kate) Stoney, Ph.D. (CTRIS)

Workshop Speakers/Moderators in Order Listed in Workshop Agenda

Gary H. Gibbons, M.D., Director, NHLBI
Mark Savage, J.D., Center for Digital Health Innovation at UCSF
Abhinav Sharma, M.D., Ph.D., McGill University Health Centre
Marie Lynn Miranda, Ph.D., Charles and Jill Fischer Provost of the University of Notre Dame
Peter James, Sc.D., Harvard Medical School Environment Exposures
Lorna E. Thorpe, Ph.D., M.P.H., NYU School of Medicine
Sarah DeSilvey, DNP, APRN-C, Northwestern Medical Center/Pediatric Faculty, Larner College of Medicine at the University of Vermont
Paula Gardiner, M.D., University of Massachusetts
Meagan Khau, M.H.A., Office of Minority Health, Center for Medicare & Medicaid Services, Director, Data and Policy Analytics Group
Jennifer Covich Bordenick, M.H.R.M., Executive Director, eHealth Initiative and Foundation (eHI)
Santosh Kumar, Ph.D., University of Memphis
Wendy Nilsen, Ph.D., National Science Foundation
David R. Wilson, Ph.D., Director, Tribal Health Research Office/NIH
Matthew Quinn, M.B.A., Science Director for the U.S. Army’s Telehealth & Advanced Technology Research Center
Stacy Lindau, M.D., M.A.P.P., NowPow’s founder and Chief Innovation Officer, Professor, University of Chicago
Taylor Justice, M.B.A., Co-Founder and President of Unite Us
Jennifer DeVoe, M.D., Oregon Health & Science University
Charles R. Jonassaint, Ph.D., M.H.S., University of Pittsburgh
Andrew Hamilton, M.S., Chief Informatics Officer and Deputy Director, AllianceChicago

Speaker Bios