Beyond the Clinic: How SDOH Data Informs Actionable Insights and Improves Health Equity

In our Ecosystem Explorer Series, we interview leaders from organizations who are advancing access to health data. Today’s interview is with Christine Lee, Head of Health Partnerships at AnalyticsIQ. 

Christine Lee is Head of Health Partnerships for AnalyticsIQ. Christine has over a decade of experience in the data and analytics space. She is a member of HIMSS, RISE, and ISPOR and has worked with industry leaders across verticals like healthcare, pharma, non-profits, and more.

AnalyticsIQ is a leading people-based data creator and predictive analytics innovator. Their mission is to fuel better outcomes for all by creating reliable people-based data that provides a robust view of individuals and the SDOH factors impacting population health. AnalyticsIQ’s PeopleCore consumer data, BusinessCore B2B data, and Connection+ B2B2C linkages provide insight into individuals that empower organizations to achieve better outcomes.

Introduction to SDOH data

To set the stage, can you give an overview of what social determinants of health (SDOH) data are, where they come from, and why they’re important for healthcare?

Social Determinants of Health (SDOH) are the social and economic factors that have a major impact on a population’s health. SDOH data are variables that help fill in the gaps of what impacts health outcomes, beyond the four walls of the clinical setting. These data correspond to key domains such as social and community contexts (e.g., age, race/ethnicity, gender, veteran status), economic context (e.g., income, employment status, education), physical infrastructure or neighborhood and built environment (e.g., housing, transportation, food insecurity, access to green spaces) and healthcare or wellness context (e.g., access to health insurance, access to care). Though SDOH is expanding to more adjacent areas such as Health Literacy, Social Isolation, Digital Divide, Maternal Care, and more.

SDOH data can come from a wide variety of places, reflecting the broad range of factors that influence health outcomes. Traditional medical sources include things like electronic health records (EHRs), patient questionnaires, public health data, healthtech and wearable devices, claims and billing data, and more. Today, however, thanks to things like predictive analytics, data science, and organizations on a mission to fuel better outcomes, SDOH data and insights can also come from external, third-party sources like people-based marketing data and custom research such as that created by AnalyticsIQ. These predictive data sources can provide deep insight into aspects of life that are crucial yet tough to measure while also filling in coverage gaps — essentially giving providers a comprehensive view into a patient’s life outside the clinic.

With approximately 80% of a patient’s health influenced and determined outside of the doctor’s office, SDOH data is vital. Healthcare providers only have so much control when it comes to ensuring the wellness of the populations they serve, but with the insight provided by SDOH data, they are better equipped to convey the right messages and provide the right services — and ultimately, improve health outcomes.

It sounds like SDOH is quite a broad category. Does AnalyticsIQ focus on a specific sub-category of SDOH? Tell us a bit about AnalyticsIQ and how your company fits in.

At AnalyticsIQ, we understand the importance of truly understanding social determinants of health when it comes to improving patient outcomes and powering healthcare research. That's why we are proud to offer a comprehensive, scalable, and predictive Social Determinant of Health Data solution covering 12 distinct, actionable categories. By blending cognitive research and data science to power data creation, our proprietary consumer database, PeopleCore, covers over 263+ million individuals and 127+ million households to create a robust view of individuals and populations that goes beyond the chart. These are: Access to Care/Health Behaviors, Access to Technology, Core Demographics (including REL), Economic Insecurity, Education, Food Insecurity, Geography (Rural/Urban), Housing Insecurity, Language (including Health Literacy), Social Isolation, Substance Use and Transportation.

Our SDOH data is leveraged by a variety of healthcare organizations to power a variety of use cases like outcomes research, health equity initiatives, patient engagement efforts, and more.

What are some of the common use cases healthcare organizations use SDOH data for? Are you seeing any recent trends in the application of SDOH data to solve a specific business or research need?

By leveraging partners like Datavant to link AnalyticsIQ’s proprietary SDOH data with claims, EHR, labs, mortality, and other sources of health data, teams across the healthcare ecosystem gain value and greater utility out of the more traditional sources of real-world data. These are research, clinical trial design, real-world evidence, public sector,Life Sciences, and many other teams who aim to understand population trends, create more diverse and decentralized clinical trials, increase medical adherence, improve outcomes and patient experience, and reduce healthcare costs or unnecessary utilization. More specifically, use cases include:

• Health equity initiatives still top the list of use cases needing greater SDOH intelligence. For example:some text
  • Improving disparities in maternal care and mortality rates among African American/Black mothers.
  • Reducing barriers to care by linking SDOH data with claims datasets and understanding patient experiences.
  • Enhancing provider understanding of national health disparities with Race, Ethnicity, and Language (REL) data.
• Connecting employer data with health plans to support modeling and segmentation that enhances member engagement and improves healthy lifestyles/behaviors.
• Researchers enhancing predictive analytics with SDOH data, EHR data, and mortality data to identify at-risk communities, predict future health outcomes, and implement preventative measures.

Beyond those use cases, how does SDOH data complement other types of clinical or real-world data?

Fortunately, there is an abundance of healthcare data available today. With so much of what impacts the population’s health and wellness happening outside of the four walls of a clinical setting, the type of data we create around SDOH, which is people-based data, complements the real-world data that is available. Our data helps fill in the missing pieces about where a patient lives, their income, education level, job status, transportation barriers, access to technology, health literacy, primary language spoken, level of assimilation, and more.

Any memorable success stories or case studies you can share?

Our team worked with Dr. Joshua Stein, Associate Professor, Ophthalmology and Visual Sciences at the University of Michigan’s Michigan Institute for Computational Discovery & Engineering and principal investigator of the Sight Outcomes Research Collaborative (SOURCE), on research related to ophthalmology patient outcomes. Dr. Stein and the researchers at SOURCE used AnalyticsIQ’s SDOH data in conjunction with clinical data to get a holistic view of the patients in their consortium and how non-healthcare factors may impact a sight-threatening eye disease.

They leverage several key SDOH variables from AnalyticsIQ including annual income, estimated net worth, number of persons in the household, education level, and employment status. They found a direct connection to a patient’s risk of experiencing blindness to several of AnalyticsIQ’s SDOH variables. This type of insight allows HCPs to identify and understand risks outside of the clinic so they can better address them and drive improved outcomes.

The Constraints and Challenges with SDOH data

Let’s pivot to talk about some of the challenges with SDOH data. The White House recently released The U.S. Playbook to Address Social Determinants of Health, wherein they note that SDOH data is “frequently not collected in a standardized and interoperable format,” if at all. Given the breadth and diversity of SDOH data, what are the specific challenges with collecting and curating SDOH datasets?

I think you’re spot on in that the main challenge when it comes to healthcare providers being able to successfully leverage SDOH data in a consistent and scalable manner comes from the fact that such data is not collected nor shared in a standardized way. I believe the standardization of SDOH data — from the definition to the categories to the data format — is something that stakeholders of all types across the ecosystem should be advocating for and collaborating on to achieve. This will facilitate far greater data and insight sharing resulting ultimately in better outcomes.

This challenge is something we’ve tried to help address at AnalyticsIQ. The format of AnalyticsIQ’s predictive data attributes can be incorporated into electronic health records (EHR), allowing users to generate prompt insights leading to prompt action. In addition to the powerful insights and ease of use, AnalyticsIQ’s data attributes are not only compatible with the International Classification of Disease, 10th edition (ICD-10) SDOH Z codes Z55-Z59, Z63, and Z65, but they paint a fuller picture of patients’ and beneficiaries’ SDOH relative to claims data alone. For example, AnalyticsIQ’s HousingIQ and Community-Level Environmental Risk data attributes can provide greater context for the intervention points in one’s environment that providers or payers cannot identify using only the code Z58 “Problems related to physical environment.”

Other considerations for collecting and utilizing SDOH data include privacy compliance and reliability. Working with trusted partners is absolutely crucial as we work towards more standardization.

We’ll circle back to privacy in a bit. Before we do, are there specific kinds or sources of SDOH data that are more challenging to source and manage than others? How do you think about prioritization and determining which SDOH sources are most valuable to end users?

I think there are a couple of ways to look at this — collecting SDOH data at scale is a challenge, so attempting to establish a comprehensive set of first-party data is a nearly insurmountable task. Also, people may be hesitant to fully share information around sensitive topics which can create gaps. While understanding individuals at scale is key and increasing in importance, it can be difficult to achieve without collaboration - which reinforces the point from the previous question around the need for standardization.

On the user side, what difficulties do researchers or analysts typically encounter that are unique to SDOH data? Do you have any lessons learned to overcome these challenges?

The team at AnalyticsIQ is truly a group of curious researchers at heart. In a more literal sense, we are primarily made up of cognitive psychologists and data scientists who drive our proprietary research and data creation. We’re actually the first data company to have PhD cognitive psychologists on staff leading our research department — which is the foundation of our data. And this is not by accident. Our team learned through experience that in order to achieve reliable research results, it is best left to professional researchers who know the ins and outs of best practices and also the human cognitive process.

Another consideration is understanding that when conducting research, you are often missing portions of the population, and it is often portions of the population with the greatest needs who are often most underrepresented such as those near or under the poverty line. Ensuring you have rigorous processes in place to capture reliable, robust, and fully representative data must be a priority.

As you know, data privacy and security are essential, especially with sensitive SDOH information. What protocols and practices does AnalyticsIQ implement to ensure high standards of data privacy and security?

AnalyticsIQ is committed to protecting privacy and the ethical creation and use of data, and our dedicated team continues to monitor and adapt to the ever-evolving privacy landscape to ensure compliance across the many industries we serve — including healthcare. We are continuously assessing the impact of evolving regulations on AnalyticsIQ’s data and individuals, and we are closely watching regulatory developments that affect our obligations. Much more information can be found on our Your Privacy page and in our complete privacy policy.

The Future of SDOH Data

Over the past several years, we’ve seen growing private and public sector interest in the value of SDOH data, particularly in public health research and health equity. What new trends do you expect to see in the coming years with SDOH data?

One major trend we have seen is simply the increased recognition of the importance of SDOH data and its continued adoption. Five or ten years ago, I don’t think as many folks had an understanding of the power of SDOH data — let alone leveraging it effectively. Another trend I expect to see is an increased need for reliable SOGI data — or sexual orientation and gender identity data. The White House even acknowledges there may be an increased need to collect this type of data in order to better understand populations in a way that can inform policies looking to address inequality.

Beyond increased adoption of SDOH data, how might future advancements in data analytics and AI/ML impact the uses or value of SDOH data?

I believe some of the key themes we’ve discussed today — such as data standardization challenges, scale and reliability of data, privacy and more — can be positively impacted with advancements in data analytics and AI/ML. Leveraging predictive analytics can help when it comes to filling in gaps and respecting individual privacy, and technologies like AI/ML, I believe, will only accelerate progress.

Are there any other trends or innovations that you’re particularly excited about?

I’m mostly excited about the progress we’re making as an industry and the fact that every step along the way only increases positive outcomes for population health and wellbeing. There’s so much data for good going on behind the scenes, and I wish more people knew about it. It seems almost daily I learn about a new type of data or innovative use case, where teams are thinking about how and when to integrate SDOH data into their project or mission. The potential is vast, and with the passionate and dedicated stakeholders — like Datavant and AnalyticsIQ — I know the future is bright.

Thanks for the interview! For readers who would like to learn more, what resources would you recommend?

Absolutely, I welcome all to check out our SDOH data white paper, Achieving a Full View of Patients: How Data is the Missing Piece in Solving Health Equity Struggles.

I also want folks to be on the lookout for our upcoming SDOH Research Report directly from our Cognitive Sciences Department. The Cognitive Sciences team at AnalyticsIQ conducted a large, nationwide survey of US adults who consented to participate. We assessed their demographics, healthcare utilization, subjective physical and mental wellbeing, as well as assessments measuring aspects of Healthy People 2030’s five dimensions of SDOH:

• Economic stability
• Education access and quality
• Social and community contexts
• Neighborhood and built environment
• Healthcare access and quality

This comprehensive report is coming very soon!