Authors: Michael N Cantor and Lorna Thorpe
Reference: HEALTH AFFAIRS 37, NO.4 (2018): 585-590.
Summarised on: 13 July 2018
This commentary piece discusses how incorporating data on social determinants of health into electronic health records (EHR) can provide a broader perspective on potential drivers of a patient’s health status and identify approaches to improving the effectiveness of care.
Social determinants of health account for between 25-60% of deaths in the United States (depending on the scope of the definition used). They can be divided into two main categories: individual-level determinants specific to a patient (eg, education level or employment status); and community-level determinants (eg, pollution levels or housing quality).
Some EHR vendors have begun developing tools for capturing and addressing the determinants of health, and using them for population health management. However, there are technical and implementation issues that need to be addressed.
Technical issues include a lack of a uniform, accepted data model for representing these determinants in EHRs. This is important for valid data aggregation across practices, EHR systems and communities.
It should be relatively easy to incorporate information on community-level determinants into an EHR where that information is already available and where patient addresses are accurate and geo-coded. However, a key implementation challenge will be in capturing individual-level determinants, because this generally relies on clinics collecting that information when patients come in for treatment and also because these can change rapidly.
The authors suggest that before integrating social determinants data into their EHR, health services must ask what actions needs to be taken once data are collected. Addressing the determinants requires both additional infrastructure (to support tasks such as referring patients to community services and tracking the outcomes) and a stronger evidence base.
Studies have already shown that data related to social determinants of health can improve predictive models and give more complete understanding of a patient’s life circumstances. Yet stronger evidence is needed to demonstrate that referral to community services for social needs will lead to better clinical outcomes.
The authors conclude that integration of determinant data into EHRs has tremendous potential, but achieving the goal of improving outcomes for patients and populations will only be achieved if these issues can be overcome.