Integrating Social Determinants of Health into the Electronic Health Record
In this month’s Health Affairs, Michael Cantor and Lorna Thorpe wrote a piece titled: “Integrating Data On Social Determinants Of Health Into Electronic Health Records”. The authors suggest integrating Social Determinants of Health (SDoH) data into the patient’s Electronic Health Record (EHR) for improved population health. Policymakers in the Medicaid space have been considering this for some time, as the authors note, but the authors suggest expanding this practice to the wider population. The intent is to give the clinician a clearer picture of the patient’s overall health status and challenges, make more accurate referrals, and then use the data in the EHR to make more general policy changes that would benefit population health and reimbursement approaches.
The Cantor and Thorpe piece makes several recommendations related to data standardization and interoperability; measurement and reimbursement strategies; and improvements to infrastructure and workflows. Their approach is at a very high level, but could be operationalized by the Department of Health and Human Services (HHS) Office of the National Coordinator for Health IT (ONC).
HHS and ONC can learn from state efforts to integrate or analyze SDoH data. For example, importing data from other systems — into the EHR or another administrative system — involves clear patient and provider identification and enumeration. If you can’t clearly identify who a person is across multiple systems, you cannot link their data, and there is not always a unique key across systems. We wrote about this earlier in the year.
Furthermore, state administrative systems are rich with data — not just for individuals and families currently enrolled in a safety net program, but for most residents. A top-down federal approach to data standardization would likely be far-reaching, beyond just health IT, and would extend to agencies that govern vital records, motor vehicles, property tax assessments, education, employment, etc.
Another approach to consider, rather than individualized data feeds giving limited pieces of information, could be to aggregate the SDoH data and create a risk profile that alerts the clinician of a risk score for a patient, so the provider can ask more questions. For example, an individual recently appearing in the unemployment data might receive an alert in their EHR, as this can be a triggering event for numerous other health-related conditions and co-morbidities such as substance use, mental health, obesity, personal and family safety, etc.
