Social Determinants of Health Data Collection Needs Improvement


Collecting data on social determinants of health carries a number of issues and needs to be improved to create learning health systems.

During the COVID-19 pandemic, the United States’ insufficient data infrastructure exposed the most socially marginalized and vulnerable populations, showing the true extent of existing structural health care disparities. These populations are at a greater risk for physical, social, and financial harm; however, our understanding of these harms often comes from weak data sources, anecdotes, experiences of single institutions, and indirect data taken from the United States Census.

In a recent study,1 researchers used data from the American Society of Clinical Oncology’s COVID-19 in Oncology Registry to compare cancer treatment delays with various factors like demographic factors, disease-related factors, and social determinants of health (SDOH).

Black race, Hispanic ethnicity, multimorbidity, and timing of COVID-19 diagnosis were all associated with delays of treatment. Other SDOH,such as United States Census tract-level racial and ethnic make-up, median household income, health insurance level, and education level were studied, but no significant relation was found between SDOH factors and treatment delays.

SDOH data are often collected by public health agencies, the UnitedStates Census Bureau, and other groups. However, these data are often incomplete due to limited regulations that govern data collection and inconsistent use of data standards. Health care professionals find themselves lacking this information and trying to pursue it raises a number of problems.

First, SDOH data, at its best, offers indirect inferences about what patients are experiencing. Certain disparities can be concluded, but little is said about exact proximal or distal cause of the disparity. Attenuation bias, inaccuracies, and misclassification bias are often a result of the unit of measurement becoming more distant from individual patients.

Second, there are delays between collection and reporting. Additionally, data use agreements, anonymization, and institutional review board approvals that need to be put in place cause further delays. With this slow process, many situations caused by SDOH, like employment, change before data can even be registered, making it even harder to study this data.

“Further, efforts at value-based care require data to be translated into actionable insights far sooner than in the encounter-based fee-for-service business model. Indeed, by replacing fee-for-service with episode-based payments, historic value-based payment models inherently

incentivize health systems to address the negative externalities brought on by SDOH,” researchers wrote.

New value-based models are prioritizing a focus on health equity, a necessity for securing accurate data on population groups and social needs. Innovations in health information technology are making the collection of real-time data more attainable. Regulations that limit researchers from analyzing data andcirculating findings need to be reduce by architects of healthy systems data.

Securing patient-level data are essential as these data are vital to understanding the interaction between SDOH factors of different levels. Understanding these interactions can result in more precisely targeted interventions. Routine, prospectively collected, systematic, patient-reported assessments is the future of understanding patient health, with both patients and physicians exchanging critical safety information and valuable data.

“Collecting more comprehensive and longitudinal SDOH data and making it concurrently available for use in both designing clinical care programs and informing research implies the creation of a learning health system… An architectural design that can support clinical care, quality improvement, and research as integrated activities will require harmonizing terminology, a shared vision across stakeholders, strong partnerships, and funding, but the payoff will be well worth it,” researchers concluded. 


1. Mullangi S, Aviki E, Hershman D. Reexamining social determinants of health data collection in the COVID-19 era. Jama Oncol. Published online October 27, 2022. doi:10.1001/jamaoncol.2022.4543

Related Videos
fake news misinformation | Image Credit: Bits and Splits -
Dr. Charles Lee
Related Content
© 2024 MJH Life Sciences

All rights reserved.