We have all seen the startling maps that highlight the gap in life expectancy (at-birth) between places that are just a few miles apart. For example, just seven L stops (elevated train) in Chicago can mean a difference in life expectancy of 16 years! And yes, this pattern is repeated in almost every major city in the US. Of course, the zip code by itself is not significant, but the social determinants of health (SDH) – factors such as education, jobs, affordable housing, availability of nutritional food, parks, clean air, etc. – associated with these neighborhoods are. Some observe that behavioral, social and environmental factors have much higher effect on health outcomes than medical care.
While there is growing awareness of the impact of the “upstream” socio-economic determinants (such as living and working conditions and social and economic opportunities) on the health outcomes of population, these determinants have been considered beyond the control of the individual or even the medical system to influence, and thus considered best addressed by policy makers on a larger scale.
There are also a range of stakeholders (e.g. schools, employer, business, government, healthcare professionals) that have varying levels of accountability for many of the social determinants, which may make it difficult if not impossible for healthcare professionals and health systems to consider any meaningful intervention based on SDH to improve outcomes.
However, that view may be changing.
The recent push to implement value-based care models and population health management is compelling health systems to look at innovative ways to design interventions that takes into account the medical needs with the unmet social and economic needs of their population, often with the help of community partners.
But before we consider what the health systems can do about SDH, let us look at what make up the social determinants. Social history has been captured during a patient’s clinic visits rather inconsistently over the years and not necessarily readily available for clinical use. In 2014, the Institute of Medicine (now known as “Health and Medicine Division” of the National Academies) recommended that 12 measures representing 11 social and behavioral domains be captured in EHR for incorporation into meaningful use guidelines.
Let us look at these 11 domains a little closely:
- financial strain
- physical activity
- tobacco use and exposure
- alcohol use
- social connections and social isolation
- intimate partner violence
- residential address/census tract-median income
All of these could be collected during patient visits, but the last one – residential address/census-tract level median income – stands out. It stands out because of the ability of the geocoded address to harness not just the census-tract level median income, but also, a wealth of publicly available data about the community in which the patient lives, i.e. Community Vital Signs.
OCHIN, a health information network with focus on safety net clinics, and its partner Robert Graham Center (an affiliate of American Academy of Family Physicians), have implemented a pilot to integrate these community vital signs. Their definition of these vital signs include
- environmental exposure
- neighborhood race/ethnic composition
- neighborhood socio-economic composition, and more
from a variety of publicly available sources. I should point out that although these types of data have been used extensively to help in community planning and policy decisions using tools like HealthLandscape, UDS Mapper (all from the Graham Center), and in economic and social science research for assessing neighborhood effects, OCHIN’s implementation is an innovative integration of publicly available community data at a clinical setting.
Now, I wondered what a clinic or physician might do with these data. They could be used in screening/counseling recommendations, identifying cohorts to be followed up for care coordination, or even used to refer patients to community partners to fulfill some of their unmet social needs that might affect the patient’s/caregiver’s ability to take care their health (I found two articles that describe the opportunities nicely here and here). I can see these being applied in many steps of the Cohort Management process of the ACO Navigator in the HIMSS Value Suite. Recently, the HIMSS Physician Community hosted a webinar on population health management by Dr. Brian Jacobs where he described how Children’s National Health System uses the community data along with geospatial analysis techniques for managing their specific sub-populations.
Being a data geek, it is exciting for me to imagine the opportunities that open up with these additional data. But there are some challenges and concerns to be overcome. As the HIMSS C&BI Committee Vice-Chair David Butler asks in his HIMSS blog post, “Data, Data Everywhere, and Not a Drop to Drink?”, how do we provide “drinkable data to quench the thirst”? Also, how best to integrate these data in clinical workflow, what is the best way to store/manage the integrated data, and how do we address the potential privacy concerns? These are topics for another blog (or three!).
How have you implemented these (or envision incorporating them) in your system? What challenges have you faced? Share your thoughts with us.