Uncovering Social Determinants of Health in your EHR data
There is increased recognition that health outcomes are influenced by social, economic and environmental factors. As defined by the World Health Organization, Social Determinants of Health (SDoH) are “conditions in which people are born, grow, live, work and age.”
Studies show that these factors contribute significantly to health and well-being. For example, average life expectancy is reduced by 15 to 20 years for people living in low-income communities, due to increased risk for stroke, chronic disease and other health concerns.
6 SDoH domains and how they affect patients
There are a wide range of conditions that are SDoH, and experts classify them into six major categories:
1. Economic stability. Factors such as employment, income and debt affect a patient’s ability to access and maintain healthcare services.
2. Neighborhood and physical environment. Where patients live, their access to transportation – as well as the safety and walkability of communities – will influence decisions that contribute to wellness.
3. Education. Access to good schools can improve literacy rates, provide early childhood education, vocational training options and more opportunities for higher education.
4. Food. Some communities have limited access to healthy food options, leaving patients to deal with hunger and food insecurity, which can create or complicate health issues.
5. Community and social context. Discrimination, dysfunctional support systems, poor social integration and a lack of community engagement can contribute to stress and other damaging health effects.
6. Healthcare system. Patients may face barriers when trying to access quality care, such as inadequate insurance coverage, or absence of providers with appropriate linguistic and cultural competencies.
These factors contribute to life expectancy, morbidity and mortality. They also can significantly affect healthcare expenditures, which is why more value-based payment models are focusing on SDoH. In the public sector, for example, the Centers for Medicare and Medicaid Services (CMS) launched an Accountable Health Communities model in 2016 to test whether identifying and addressing SDoH needs affect costs and utilization. Private payers are also investing in addressing SDoH to improve population health programs.
The value of data science in addressing SDoH
To succeed with value-based payments, healthcare organizations are paying more attention to SDoH in their own data. But it requires consistent data collection, comprehensive data sources and data analytics. This is the only way to capture rich data and effect meaningful change.
Allscripts clients are collecting data about SDoH risk factors by using the associated ICD-10 codes already implemented within their electronic health records (EHRs). We’re developing a comprehensive SDoH coding framework to align definitions between disparate SDoH codes within ICD-9, SNOMED and LOINC. We have aligned more than 300 risk factor codes and that number continues to grow.
Applying the SDoH framework enables us to capture data held in unstructured text and diagnosis codes. We have discovered more than 2 million SDoH-related diagnoses in our data lake of more than 40 million de-identified patient records. Top diagnoses include disruption of family by separation or divorce, disappearance or death of a family member and unemployment.
When providers gain insight into their patients’ background, they can better tailor healthcare services and help patients overcome barriers. Building on these insights with capabilities such as artificial intelligence and machine learning can ultimately deliver better outcomes and transform healthcare as we know it.