Using Analytics to Unlock the Value of Social Determinants of Health Data

3StepsToUnlockValueofSocialDeterminantsofHealthData

Over the past 10 years, the shift to value-based care (VBC) and accountable care organizations has been slow, but fruitful.  Today, more than half of healthcare organizations participate in an alternative payment contracts, and 40 healthcare payments in the U.S. are tied to alternative payment methods. These models aim to reduce the cost of care while fostering appropriate patient care and positive outcomes.   

Understanding and addressing social determinants of health (SDoH) are emerging as critical factors in achieving the goals of VBC. SDoH are the social and economic conditions, such as education, employment status, housing stability, physical safety, neighborhood conditions and racial and ethnic bias, that have a bearing on the health of individuals, communities and populations. Research shows that SDoH account for up to 80% of health outcomes, with clinical care accounting for only 20%.  

While there is growing appreciation for the importance of SDOH in well-being and in achieving greater value from health care and services, we face a contradiction, as noted in a recent white paper published by the Leonard Davis Institute of Health Economics at the University of Pennsylvania 

While patients in ACOs tend to have a higher clinical risk, providers in communities serving populations with social risk factors are less likely to participate in ACOs. This may be, in part, because care for populations with high social risk is typically concentrated within a subset of providers, such as Federally Qualified Health Centers (FQHCs). More troublingly, some ACO models may emphasize reducing wasteful utilization, but for many populations the primary problem is under-utilization and underspending. 

As the study points out, in some cases, high-value care may mean more services and more spending, and SDOH data can hold the key to understanding and prescriptively targeting care. 

SDoH can do nothing to improve health outcomes, however, if data can’t be gathered and analyzed for insights. Unfortunately, electronic health record (EHR) systems weren’t originally designed to capture SDoH data. Even if staff wanted to collect patient SDoH data at check-in, there was no place to easily record it. If it was captured, it went into notes fields that were free text and often buried four-screens deep, providing little clinical or analytical value. 

As the importance of SDoHs has become apparent, EHRs and their data schema have, in many cases, evolved to enable data to be captured regarding a patient’s living situation, employment, access to healthy foods, sense of physical safety, access to transportation and similar SDoH. Having that information captured in the EHR can be invaluable, helping knit together risks and complexities that would have otherwise gone undetected and not discussed.  

In the medical realm, then, there is marked progress in collecting and recording data around social determinants in a way that can enable whole-person care. However, we haven’t fully figured out how to marry this medical model – whether it involves primary care or behavioral health – to social services.  Until we do, we may be inadvertently sub-optimizing care in our quest for value.  As the Penn study found, some ACO models may emphasize reducing wasteful utilization, but for many populations the primary problem is under-utilization and underspending. 

Organizations such as Federally Qualified Health Centers (FQHCs), community health centers and social service agencies are struggling with capturing more SDoH data, such as referrals to Meals on Wheels, and tying that that data to medical and behavioral health outcomes will help reinforce the value of more services and more spending for targeted populations. 

Building a broad data set for social determinants is a bit of a catch-22, though, because we don’t know what SDoH data is the most important to collect. Care providers and agencies are collecting a lot of data, without understanding the data and their overall potential utility. However, we are beginning to establish a cumulative baseline of evidence to discern what data, interventions and social determinants matters most in terms of outcomes across disparate populations.  

Without SDoH information, a patient would simply be diagnosed as diabetic or having a behavioral health condition. Layering in social factors, such as employment status, nutritional issues and /or safety paints a fare more nuanced picture of the whole patient rather than one based strictly on a specific health condition or challenge and allows providers to more effectively and appropriately maximize the value of care and services  

SDoH in Action 

ED utilization by homeless patients is three times the US norm and has increased 80% over the last 10years.  While not directly causational, numerous studies show housing is key to reducing ED utilization.  A $1,000 monthly rent is the equivalent of 45 minutes in an ER and healthcare organizations are taking a more active position in housing vulnerable members of the community.  For example, hospital systems in Portland have collaborated to fund $21.5M in  housing for high-risk individuals because they know in the long run it saves costs and improves life quality for those individuals.  Recent data from the Portland “Housing for Health” show a savings of  $20,000-to-$30,000 per client per year in emergency room and criminal justice-related expenses. 

3 Keys for Leveraging SDOH Data to Drive Value  

For providers to take full advantage of SDoH in support of VBC, they must do three things.  

First, they must continue to build out systems to capture this data and develop analytics at the patient and population level to support socially derived care. It’s no good if a provider’s frontline staff asks questions about safety and travel or mobility and nutrition if nothing happens to it. You need the analytics, then you’ve got to have the informatics and the health management capabilities on the back end for early interventions. 

The second thing providers must do is tap into or help build a network for community referrals and care coordination. Our healthcare system is very good at creating links from Point A to Point B and Point C to Point Q, but life isn’t linear, especially for patients with complex conditions. We need to create a safety web of interconnectedness across the healthcare ecosystem. Innovation around SDoH can be the framework upon which we can build these webs.  

Finally, providers must focus on proactive care and outreach, which are the essence of VBC. In some cases, this may mean more care and services, not less. This again gets back to the importance of SDoH and analytics to extract meaningful information about socio-environmental drivers that impact health of. This knowledge, obtained through robust data and information across the community of care providers, is key to producing better patient and population health outcomes while reducing care utilization costs.  

Share:
Share on facebook
Share on twitter
Share on linkedin