Use & Optimization News

Leveraging SDOH EHR Data, Health IT to Advance Population Health

Linking SDOH EHR data with community-level data through artificial intelligence tools like machine learning could help boost population health and health equity.

Leveraging SDOH EHR Data, Health IT to Advance Population Health

Source: Getty Images

By Hannah Nelson

- Health systems should leverage health IT solutions that capture and analyze social determinants of health (SDOH) EHR data to improve population health, according to an article published in JAMA Network Open.

Health IT could provide opportunities to improve health equity by addressing the underlying factors that lead to care disparities, Elham Hatef, MD, MPH, from the Johns Hopkins School of Medicine and Bloomberg School of Public Health, wrote in the JAMA op-ed.

Studies have found that type 2 diabetes is more common among Black adults compared with non-Hispanic White adults, Hatef wrote, offering type 2 diabetes disparities as an example. Lower socioeconomic status has also been linked to higher frequencies of type 2 diabetes.

Limited research has found a link between food insecurity and hypoglycemic events in patients with diabetes. However, findings are limited to the direct screening of food insecurity among a select number of patients.

Hatef said that the use of EHR data in population health research will help provide more accurate findings compared to survey and claims data.

“The use of real-time EHR data on a large population of patients, compared with the use of survey data with limited scope and claims data with the time lag, provides a source of high-volume data, the potential of which has not been fully exercised in health care systems,” Hatef emphasized.

Additionally, few studies have examined the association of “place-based determinants of health” with the risk of type 2 diabetes. Such factors include living in a food desert or having access to fast food establishments.  

A recent study leveraged EHR data and health IT tools to examine the association between neighborhood food environment and the risk of incident type 2 diabetes across different community types (high-density urban, low-density urban, suburban, and rural), Hatef referenced.

The authors used Veterans Administration (VA) national EHR data for the VA Diabetes Risk cohort, a cohort of veterans without type 2 diabetes. They also used VA EHR data related to the patient’s military service history, disability, income, and eligibility for Medicaid or other VA benefits. From this data, researchers created a low-income or disability marker that they used as a proxy for socioeconomic status.

Neighborhood-level covariates included social and economic environment, which researchers measured based on US Census–derived measures from the American Community Survey, Hatef said, still citing the study as an example.

Then, the researchers linked the individual-level data and neighborhood-level attributes with the food environment, including the proportion of total food-serving establishments that were fast-food establishments and the ratio of the total retail food outlets that were grocery stores.

The study found that the neighborhood food environment was associated with increased type 2 diabetes risk among veterans in multiple community types.

Non-Hispanic Black adults had the highest incidence compared with other racial and ethnic groups. Additionally, adults with disabilities and those with low income but no disability had a higher incidence of type 2 diabetes than those with neither disability nor low income.

“The study is a great example of the capabilities of HIT to provide a comprehensive assessment of a person’s health, which goes beyond just documenting clinical diseases and medical interventions,” Hatef pointed out.

Other IT advancements, such as machine learning, could help health systems promote population health.

“The integration of these HIT tools in a health system’s EHR can help to identify at-risk patients, link their EHR data to publicly available data on place-based SDOH, and evaluate the associations between social needs and place-based SDOH and type 2 diabetes incidence and management,” Hatef noted.

“This approach could foster collaborations between the health systems and at-risk communities they serve and help to reallocate health system resources to those in most need in the community to reduce the burden of type 2 diabetes and other chronic conditions among racial minority groups and socioeconomically disadvantaged patients and to advance population health,” Hatef concluded