Skip to main content
Interview

Using Social, Behavioral Determinants of Health to Determine Hospitalizations Among High-Risk VA Patients


November 12, 2020

By Julie Gould

Recent research published online in JAMA Network Open, found that when patient-reported social information was added to electronic health records (EHR), the estimation of 90- and 180-day hospitalization risk among veterans was improved.

“Despite recognition of the association between individual social and behavioral determinants of health (SDH) and patient outcomes, little is known regarding the value of SDH in explaining variation in outcomes for high-risk patients,” the researchers of the study wrote. 

To better understand the study and its findings, we spoke with Donna Zulman, MD, and Matthew Maciejewski, PhD. They explained why screening veterans on SDH measures in routine clinical practice could help identify SDH challenges of high-risk veterans and enable the VA to address them, which could potentially improve certain health outcomes.  

What existing data led you and your co-investigators to conduct this research? 

The VA is the largest integrated health system in the US and has one of the longest standing EHRs. The current EHR includes rich information about a patient’s sociodemographics and clinical characteristics, including information such as residential address, race, and alcohol use. These data are routinely used by health services researchers to understand the factors influencing Veterans’ access, health care utilization, and health outcomes. We recognized that there were many SDH that were not available in the VA EHR, although VA has been a leader in integration of patient-reported outcomes into the EHR because self-reported pain has been available for over 20 years, self-reported alcohol use has been available for over a decade and homelessness is starting to be asked routinely. We were interested in determining whether augmenting this existing dataset with additional patient-reported SDH factors such as social support and transportation barriers, would improve predictions of hospitalization, particularly in Veterans with high risk for this outcome.   

Please briefly describe your study and its findings. Were any of the outcomes particularly surprising? 

The purpose of this study was to describe SDH factors among high-risk Veterans, and to determine whether adding patient-reported SDH measures to EHR measures improves estimation of 90-day and 180-day all-cause hospital admission.  We linked survey data on respondents to their VA medical record data.  As expected, we found that adding patient-reported SDH measures to a regression model with EHR data improved estimation of 90-day and 180-day hospitalization risk for high-risk veterans.  This improved ability to predict hospital admission was due to the addition of patient-reported health-related locus of control, resilience, smoking status, health literacy, and medication insecurity. 

What are the possible real-world applications of these findings in clinical practice? 

In addition to improving predictive algorithms, screening veterans on these SDH measures in routine clinical practice could identify SDH challenges of high-risk Veterans and enable VA to address them, which would could potentially improve certain health outcomes. Incorporating these types of factors into the EHR could also assist with population management and health system decisions, for example by informing partnerships with certain community agencies.  

Do you and your co-investigators intend to expand upon this research? 

We are expanding upon this research by conducting similar analyses for emergency department visits and functional independence in the community.  We are also investigating the possibility of using this unique survey data to examine other outcomes important to Veterans and the VA health system and to conduct a new survey on a more general population of Veterans. 

Is there anything else pertaining to your research and findings that you would like to add?

Social factors exert a substantially more potent impact on health than does health care, especially among disadvantaged patient populations. Incorporating information about key SDHs factors into the EHR can facilitate discussions with clinicians, referral for critical resources, and population-based interventions. This study provides insight into the value of incorporating patient-reported SDHs into predictive algorithms, and highlights several SDHs that might be particularly valuable in improving understanding of a patient’s risk of hospitalization. 

About the Authors:

Donna Zulman, MD, MS, is a general internist and Investigator at VA Palo Alto’s Center for Innovation to Implementation, and an Assistant Professor in the Division of Primary Care and Population Health at Stanford University.   

Matt Maciejewski, PhD, is a Senior Research Career Scientist at the Durham VA HSR&D Center of Innovation to Accelerate Discovery and Practice Transformation and Professor in the Department of Population Health Sciences at Duke University. 

Reference:

Zulman DM, Maciejewski ML, Grubber JM, et al. Patient-Reported Social and Behavioral Determinants of Health and Estimated Risk of Hospitalization in High-Risk Veterans Affairs Patients. JAMA Netw Open. 2020;3(10):e2021457. Published 2020 Oct 1. doi:10.1001/jamanetworkopen.2020.21457

Agree or disagree with an article? Share your professional thoughts on an article you read.

Your Name
6 + 0 =
Solve this simple math problem and enter the result. E.g. for 1+3, enter 4.
Back to Top