Reexamining Social Determinants of Health Data Collection in the COVID-19 Era

SBDH-Reader: an LLM-powered method for extracting social and behavioral determinants of health from medical notes

Introduction

Social and behavioral determinants of health (SBDH) are increasingly recognized as essential for prognostication and informing targeted interventions. While medical notes contain rich SBDH details, these are unstructured and conventional extraction methods tend to be labor intensive, inaccurate, and/or unscalable. The emergence of large language models (LLMs) presents an opportunity to develop more effective approaches for extracting SBDH data.

Materials and Methods

We developed the SBDH-Reader, an LLM-powered method to extract structured SBDH data from full-length medical notes through prompt engineering. Six SBDH categories were queried including: employment, housing, marital relationship, and substance use including alcohol, tobacco, and drug use. The development dataset included 7,225 notes from 6,382 patients in the MIMIC-III database. The method was then independently tested on 971 notes from 437 patients at UT Southwestern Medical Center (UTSW). We evaluated SBDH-Reader's performance using precision, recall, F1, and confusion matrix.

Results

When tested on the UTSW validation set, the GPT-4o-based SBDH-Reader achieved a macro-average F1 ranging from 0.85 to 0.98 across six SBDH categories. For clinically relevant adverse attributes, F1 ranged from 0.94 (employment) to 0.99 (tobacco use). When extracting any adverse attributes across all SBDH categories, the SBDH-Reader achieved an F1 of 0.96, recall of 0.97, and precision of 0.96 in this independent validation set.

Conclusion

A general-purpose LLM can accurately extract structured SBDH data through effective prompt engineering. The SBDH-Reader has the potential to serve as a scalable and effective method for collecting real-time, patient-level SBDH data to support clinical research and care.

A Scoping Review and Assessment of the Area-Level Composite Measures That Estimate Social Determinants of Health Across the United States

OBJECTIVES

Evidence-informed population health initiatives often leverage data from various sources, such as epidemiologic surveillance data and administrative datasets. Recent interest has arisen in using area-level composite measures describing a community's social risks to inform the development and implementation of health policies, including payment reform initiatives. Our objective was to capture the breadth of available area-level composite measures that describe social determinants of health (SDH) and have potential for application in population health and policy work.

METHODS

We conducted a scoping review of the scientific literature from 2010 to 2022 to identify multifactorial indices and rankings reflected in peer-reviewed literature that estimate SDH and that have publicly accessible data sources. We discovered several additional composite measures incidental to the scoping review process. Literature searches for each composite measure aimed to contextualize common applications in public health investigations.

RESULTS

From 491 studies, we identified 31 composite measures and categorized them into 8 domains: environmental conditions and pollution, opportunity and infrastructure, deprivation and well-being, COVID-19, rurality, food insecurity, emergency response and community resilience, and health. Composite measures are applied most often as an independent variable associated with disparities, risk factors, and/or outcomes affecting individuals, populations, communities, and health systems.

CONCLUSIONS

Area-level composite measures describing SDH have been applied to wide-ranging population health work. Social risk indicators may enable policy makers, evaluators, and researchers to better assess community risks and needs, thereby facilitating the evidence-informed development, implementation, and study of initiatives that aim to improve population health.

Trends in publication and levels of social determinants of health reporting in Journal of Clinical and Translational Science from 2017 to 2023

Social determinants of health affect clinical and translational research processes and outcomes but remain underreported in empirical studies. This scoping review examined the rate and types of social determinants of health (SDoH) variables included in the JCTS translational research studies published between 2017 and 2023 and included 129 studies. Most papers (91.7%) reported at least one SDoH variable with age, race and ethnicity, and sex included most often. Future studies to inform the role of SDoH data in translational research and science are recommended, and a draft SDoH data checklist is provided.

Artificial intelligence, ChatGPT, and other large language models for social determinants of health: Current state and future directions

This perspective highlights the importance of addressing social determinants of health (SDOH) in patient health outcomes and health inequity, a global problem exacerbated by the COVID-19 pandemic. We provide a broad discussion on current developments in digital health and artificial intelligence (AI), including large language models (LLMs), as transformative tools in addressing SDOH factors, offering new capabilities for disease surveillance and patient care. Simultaneously, we bring attention to challenges, such as data standardization, infrastructure limitations, digital literacy, and algorithmic bias, that could hinder equitable access to AI benefits. For LLMs, we highlight potential unique challenges and risks including environmental impact, unfair labor practices, inadvertent disinformation or "hallucinations," proliferation of bias, and infringement of copyrights. We propose the need for a multitiered approach to digital inclusion as an SDOH and the development of ethical and responsible AI practice frameworks globally and provide suggestions on bridging the gap from development to implementation of equitable AI technologies.

The Impact of Health Geography on Public Health Research, Policy, and Practice in Canada

The link between geography and health means that the places we occupy-where we are born, where we live, where we work, and where we play-have a direct impact on our health, including our experiences of health. A subdiscipline of human geography, health geography studies the relationships between our environments and the impact of factors that operate within those environments on human health. Researchers have focused on the social and physical environments, including spatial location, patterns, causes of disease and related outcomes, and health service delivery. The work of health geographers has adopted various theories and philosophies (i.e., positivism, social interactionism, structuralism) and methods to collect and analyze data (i.e., quantitative, qualitative, spatial analysis) to examine our environments and their relationship to health. The field of public health is an organized effort to promote the health of its population and prevent disease, injury, and premature death. Public health agencies and practitioners develop programs, services, and policies to promote healthy environments to support and enable health. This commentary provides an overview of the recent landscape of health geography and makes a case for how health geography is critically important to the field of public health, including examples from the field to highlight these links in practice.

A Novel Framework to Address the Complexities of Housing Insecurity and Its Associated Health Outcomes and Inequities: "Give, Partner, Invest"

The association between housing insecurity and reduced access to healthcare, diminished mental and physical health, and increased mortality is well-known. This association, along with structural racism, social inequities, and lack of economic opportunities, continues to widen the gap in health outcomes and other disparities between those in higher and lower socio-economic strata in the United States and throughout the advanced economies of the world. System-wide infrastructure failures at municipal, state, and federal government levels have inadequately addressed the difficulty with housing affordability and stability and its associated impact on health outcomes and inequities. Healthcare systems are uniquely poised to help fill this gap and engage with proposed solutions. Strategies that incorporate multiple investment pathways and emphasize community-based partnerships and innovation have the potential for broad public health impacts. In this manuscript, we describe a novel framework, "Give, Partner, Invest," which was created and utilized by the University of Pittsburgh Medical Center (UPMC) Insurance Services Division (ISD) as part of the Integrated Delivery and Finance System to demonstrate the financial, policy, partnership, and workforce levers that could make substantive investments in affordable housing and community-based interventions to improve the health and well-being of our communities. Further, we address housing policy limitations and infrastructure challenges and offer potential solutions.