Knowledge Center

Written by Samantha Holvey, MHL, Director, Workgroups & Communication

As the national spotlight on health equity intensifies, social risk (e.g., poverty, race and/or ethnicity, limited community resources) screenings are growing in popularity with health systems. Currently, there are no established standards for collecting social risk data or administering treatment plans based on this information. Instead of adding more work for primary care providers and exacerbating an already volatile physician burnout situation, some health systems are utilizing community-level data to identify patient-level social risks.

Although leveraging community-level data seems like a common-sense approach, research shows that 40% of patients that would benefit from social risk-informed care would be excluded and 57% would be incorrectly targeted by only using the community-level screening approach. Furthermore, patients themselves may not want their clinicians to intervene. This JAMA study showed that 64.5% of patients said they did not want help addressing identified risks. These results are mirrored in a 2019 study, which concluded that patients and caregivers believed social risk screening was important and acceptable, but they did not expect their health care teams to address their social challenges.

If patients report that social risk screening is important, but community-level data is not a sufficient tool for collecting social risk information, then where do we go from here? Clearly more research is needed to explore social risk screenings and best practices for addressing social risks without offending patients. However, new technology and standards can be utilized to collect and analyze patient-level social risks without adding to the physician workload. Tools are now available for efficiently gathering patient self-reported data that can lead directly to improved outcomes for patients with social risks.

Pre-visit screening digital tools for clinical and social risk factors can give clinicians critical patient self-reported data before they even walk into the exam room (virtual or in-person). This information saves time by automating the flow of data into the electronic health record (EHR), thus improving workflow and helping organizations meet value-based care requirements. Most importantly though, it improves care quality and outcomes for patients.

Natural Language Processing (NLP) is the key technology to extract patient social risk information from the clinical notes in a patient’s EHR. NLP technology can aid in the development of screening tools, risk prediction models, and clinical decision support systems.

Z-codes (diagnostic codes) were introduced in 2015 to better capture social determinants of health (SDoH) standardized data. Although Z-codes can be collected by a patient’s care team today, they are not widely utilized throughout the U.S. health system. However, we have seen that when they are utilized, Z-codes can help improve quality and care coordination. Leveraging SDoH data can lead to targeted and improved care plans, trigger referrals for social services to act on unmet needs and tracking of those referrals so providers and social organizations can better harmonize care delivery.

As technology and health care interoperability continue to advance, further research is necessary to find areas in which patient-level and community-level data overlap so that limited social resources can improve patient outcomes while also reducing costs.