One’s neighborhood socioeconomic status may not contribute much more to risk prediction than information already within electronic health record (EHR) data, according to new research in the Journal of the American Medical Association (JAMA).
Researchers from Duke University set out to study more than 90,000 patients 18 years or older who lived in North Carolina’s Durham County and had at least one healthcare encounter at the Duke University Health System and Lincoln Community Health Center, both of which are located in Durham.
Machine learning methods were used to develop risk models and determine how adding neighborhood socioeconomic status (nSES) to EHR data affects risk prediction. Neighborhood socioeconomic status was defined using the Agency for Healthcare Research and Quality (AHRQ) SES index, a weighted measure of multiple indicators—such as income, property value, education rate, employment rate, and number of people in each household—of neighborhood deprivation.
The data showed that among the 90,000 patients in the training set of the study, and the 122,000 patients in the testing set of the study, those living in neighborhoods with lower nSES had a shorter time to use of emergency department services and inpatient encounters, as well as a shorter time to hospitalizations due to accidents, asthma, influenza, myocardial infarction, and stroke. But when added to EHR variables, nSES did not improve predictive performance for any health outcome, the researchers concluded.
What’s more, the study found that, while the risk of clinical outcomes differs based on nSES, and although nSES is moderately predictive of clinical outcomes, nSES does not meaningfully improve risk prediction of clinical events above and beyond what is easily extractable from the EHR. One primary explanation for this finding, according to the researchers, could be that, for this specific population that was studied, demographic characteristics are highly associated with nSES. Indeed, in this study, knowledge of a patient’s age, sex, race/ethnicity, and insurance status explained more than 28 percent of the variability in nSES, the researchers noted.
The study’s authors stated that it’s well known that neighborhoods are significantly associated with the health of their residents through physical and social attributes. And, the mechanisms by which neighborhoods are associated with health include increased stress level, decreased physical activity, and poor nutrition, which in turn affect both proximal risk factors, such as blood pressure, diabetes control, and inflammation, and distal health outcomes, such as cardiovascular disease.
As such, while these results support prior research in this area by showing that patients who live in areas with lower nSES have poorer health outcomes than patients who live in areas with higher nSES, “information about the environment in which a person lives may not contribute much more to population risk assessment than is already provided by EHR data.” And, the researchers noted, “Although this result does not mean that integrating social determinants of health into the EHR has no benefit, researchers may be able to use EHR data alone for population risk assessment.”