A SDoH Recap

In May 2018 this blog (Social Risk Adjustment for Provider Quality Measures), reviewed how Social Determinants of Health (SDoH) (data on a person or household’s income, education, housing, environment, transportation etc.) can be used to adjust performance scores for practices that serve disadvantaged populations.

In October 2018, our follow-up (A New Idea for Financing Key Investments in Healthy Communities) reviewed a paper proposing a funding and implementation structure to involve all the healthcare stakeholders (including insurers, government agencies, and care deliverers) in the data mission to fairly and accurately share the costs and benefits of practical programs to encourage the crucial investments that have been proven to proactively improve a population’s health. Recently, several popular articles have highlighted potential downsides of using Social Determinants of Health for population-health applications.

The SDoH Concerns

While the opportunities for positive and powerful use of this type of information can be tantalizing, there are also concerns. Several big-data aggregation firms offer healthcare providers — ACOs and payers — risk profiles upon individuals that report background information such as income, housing and education, but also criminal records, bankruptcies and burglaries. Recent articles in Politico.com’s subscription content, and in Pro Publica discuss these offerings, including the viewpoints of some of the data providers themselves, as well as opinions from the data buyers, and public policy professionals.

The concern is: would this data, these SDoH, be used not to manage the care delivery or level the playing field for the providers who take care of high risk populations, but rather to raise premiums on high-risk populations based on their social determinants? Would SDoH make it easier for health plans to cherry-pick and lemon-drop (picking healthy members for coverage, while dropping high-risk high-cost members)? How can we prevent inappropriate or undesirable use of Social Determinants?

HIPAA protects personal health data, but not personal lifestyle data. Europe has already legislated protection for both. And while here in the USA the ACA prohibits insurers from using pre-existing conditions as the basis to deny care, short-term health plans have recently been exempted from this regulation, and in any case, nothing in HIPAA or ACA prevents insurers from using patients’ personal information to evaluate individual risks and set prices.

Using SDoH Wisely

Reviewing the big picture on population health, it becomes clear that correlating SDoH data with health outcomes, to make practical policy changes that improve both health and care-delivery efficiency, will save a lot of money and result in happier and healthier citizens.

If we believe that 50% of our health is determined by social, behavioral, and environmental factors, compared to just 10% determined by healthcare, then it’s inevitable, necessary, and appropriate to incorporate this data into the delivery and analysis of care.

And the result should not be to exclude people from coverage or raise premiums on populations with high-risk social determinants — but to improve healthcare delivery, understand what works and what doesn’t work, recognize varying risk levels for providers, and deliver cost-effective care to various target populations.

SDoH Usage Ground Rules

To use SDoH Wisely in Population Health Management, I suggest the following ground rules:

  • Aggregate and de-identify: rather than using data on individuals or households, we should collect SDoH data at the neighborhood or census tract level. This gives a general indication of population risk factors — but does not point to individuals and families.
  • Use a composite risk score: by combining a number of social determinants (housing, transportation, food, income) into one risk score, we get a good, quantitative sense of the impact of social determinants on a population’s health — without pinpointing it to a particular individual or cause.
Pick your Social Determinants — income, education, environment, housing, food — great. Bankruptcies, felonies, police records — maybe that’s information a loan officer evaluates when deciding on a loan, but that’s not what a healthcare provider, care manager, or population-health manager should be taking into account.

And finally, as an expert in the Pro Publica article suggests: independently collected personal data should be treated the same way as credit scores. Anybody can contact the credit bureaus and request their credit report. The same rule should apply to any data that the health system uses to evaluate the delivery, quality, or cost of your medical care.

Bottom Line

Using Social Determinants in Health Care Analytics and Population Health Management is the right, smart, and obvious thing to do. Let’s agree to use this data to help improve care, reduce cost of care to populations, and better manage interventions (transportation options, food security initiatives). No one should lose coverage or see premium increases based on their personal social determinants, or their community’s vital signs.

Please contact me for further discussion on ways to actively — and appropriately — utilize Social Determinants in your region.


Links to Articles Cited Above:

ProPublica: Health Insurers Are Vacuuming Up Details About You — And It Could Raise Your Rates

Politico: The Potential Downside of Social Determinants

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.