The Growing Role of Social Determinants Data in Healthcare

Key Takeaways:

• Healthcare stakeholders recognize a social determinant of health when they see it, but organizations differ on which measures are important. Agreement about a core set of SDoH would help implementers better utilize this data type.
• SDoH and patient health are correlated, but determining causality remains challenging. Lack of knowledge about causality hampers providers’ ability to translate SDoH data into effective interventions.
• Leading public health stakeholders are using SDoH to fund organizations not traditionally involved in healthcare delivery to improve social drivers of cost.
• Data brokers are selling SDoH data. Widespread usage by payers and providers, coupled with interest from social media giants, could trigger increased market and regulatory scrutiny.

Marshalling aggregated EHR and claims data for use in applications is an ongoing challenge for most healthcare enterprises. Social determinants of health (SDoH) are a relatively new and amorphous data type that show great promise for contributing to a range of applications.

As healthcare shifts from volume to value, SDoH present opportunities at both population and individual levels. Patient cohort discovery in PHM programs could become more precise and accurate using relevant SDoH. SDoH offers the potential to “better predict potential healthcare outcomes across disparate populations.”

Implementers will need more experience with SDoH data before it becomes a routine inclusion in HIT applications. Otherwise, SDoH risk becoming just another unruly data source.

Providers and payers hope to achieve a better understanding of risk, better patient engagement, and more effective use of existing treatment resources. Questions remain about what qualifies as a SDoH, where to source such data, and how to use it. More experience will be needed before SDoH delivers broad-based benefits at both a patient and population level.

What is a Social Determinant of Health?

The CDC defined SDoH as “conditions in the environments in which 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.” While most organizations generally agree with this definition, different organizations report drastically different measures as important to health status.

A quick look at existing lists of SDoH confirms that there is no consistent, widespread acceptance for a single set of factors. For instance, the Kaiser Family Foundation reports 48 social determinants related to individual health, ranging from years of schooling to access to a full-service grocery store. Meanwhile, LexisNexis offers 442 measures that relate to patient health.

No governmental or commercial authority has established a definitive list with strong industry support. The CDC, Canadian government, and WHO have all produced reports outlining their take on which SDoH should be tracked, accompanied simply by policy and practice recommendations. For now, providers and payers are bombarded with different views of the sources, uses, and value of SDoH. Agreement on which SDoH are important would help implementers understand how and where to use this relatively new data type.

How Do SDoH Relate to Patient Health?

Despite uncertainty about which SDoH are relevant, many stakeholders believe that SDoH can help improve population and patient health. However, while most people accept that there is a correlation between SDoH and health outcomes, determining causality between specific kinds of SDoH and specific health outcomes remains challenging.

The correlation between an individual’s years of schooling completed and their probability of smoking provides a perfect example. While more years of education correlate to a lower probability of smoking, more education does not cause anyone to smoke less, and smoking does not cause anyone to drop out of school. Research shows that the differences in smoking behavior at age 24 are accounted for by differences in smoking behavior at age 17, implying that some third factor drives both the probability of smoking and the years of completed education. As a result, interventions intended to decrease the probability of smoking by increasing the years of education an individual completes would be ineffective.

While providers accept that SDoH are often correlated to health outcomes, lack of knowledge about causality hampers efforts to translate those correlations into effective interventions. In addition, SDoH data is not guaranteed to contribute in all contexts. A 2017 study showed that using SDoH does not enhance predictions about a patient’s need for social services beyond what EHR and claims data already provide. Implementers could use guidance on whether certain kinds of SDoH enhance an application.

SDoH Contributions to Public Health

SDoH is connecting healthcare stakeholders with organizations not previously thought of as directly involved in healthcare delivery. For example, after determining that better access to fresh produce, stable housing, and preventive screenings improves patients’ health, UnitedHealthcare awarded $1.95 million to organizations that could help. One recipient, Feeding Wisconsin, used the funds to expand support for local food banks.

Both the U.S. and Canadian governments have initiatives that mirror what UnitedHealthcare has done, but on a national level. Canada’s budget document, Growing The Middle Class, “detailed an unprecedented investment of $8.4 billion over five years in housing, education and child welfare for Indigenous peoples, and $2 billion to end longstanding boil-water advisories on reserves,” specifically citing SDoH as a reason for the investment. The U.S. initiative, Healthy People 2020, seeks to address SDoH by promoting economic stability, education, social and community context, health access and education, and built environment through funding provided by the Office of Disease Prevention and Health Promotion. These large initiatives fund smaller groups and programs that already have traction in addressing SDoH.

Accessing Relevant Social Determinants Data

Sourcing relevant SDoH data requires payers and providers to engage in a different process from the collection of traditional health data. In recent years, data brokers such as LexisNexis, Experian, and Axciom have been controversially selling SDoH data derived from a variety of consumer data sources. These companies collect vast amounts of data and create patient “health scores” similar to credit scores. Using these health scores, payers and providers can identify at-risk populations and prescribe personalized treatment options for individual patients.

Consumer understanding of the existence of this data remains low. SDoH data has not historically been considered directly pertinent to healthcare and is not subject to HIPAA. The recent usage could trigger more market or regulatory scrutiny of SDoH.

The number and variety of organizations offering social determinant data is increasing. The social media giants have an interest in further monetizing the data they have collected. Absent evidence that such data helps providers to make more informed decisions about patient health, market acceptance is not assured.

 

Ultimately, the interest level and desire to leverage SDoH in health IT is increasing rapidly. HIT vendors are responding slowly by including this data type in different products, but still in narrowly defined ways. Wider availability of a variety of SDoH is also fueling interest and experimentation. Incorporating SDoH into existing stores of EHR and claims data at the patient or cohort level introduces another layer of complexity for developers. Implementers will need more experience with this data before it becomes a routine inclusion in HIT applications. Otherwise, SDoH risk becoming just another unruly data source.

By Paul Nardone, 2018 Research Intern