Implementing Social Determinants of Health (SDoH) in Translational Science
SDoH are the non-clinical covariates of how people live, grow, learn and age and relates to how people manage stressors or prevent worsening health outcomes. Progressively, data about patient-level SDoH are collected using clinical screening tools for social needs and social risk factors. CD2H has led multiple SDoH synergistic initiatives relating to overall improvement of SDoH encoding and informing care and research in the CTSA community. The first was a pre-pandemic effort to create a maturity model for CTSA hubs to understand their current capacity to capture and use SDoH data. The pandemic further extended the CD2H’s efforts to understand SDoH data and its use. This work was presented to the CTSA Steering Committee for discussion about how to best align such initiatives across the program.
Dr. Jimmy Phuong, co-lead of the National COVID Cohort Collaborative (N3C) Social determinants of health domain (SDoH) team, has led efforts to extract SDoH variables into the OMOP Common Data Model (CDM), transformations in the data life-cycles, and understanding the needs in data sharing limitations towards the use of SDoH in clinical research. In collaboration with Boston Medical Center (BMC), TriNetX, and N3C, Dr. Phuong explored the encoding needs for transmitting THRIVE SDoH screening tool responses and the technical hurdles experienced in data ingestion and harmonization (DI&H) into the N3C OMOP dataset. This work was presented at the 2021 AMIA Summit and will appear in the conference proceedings as a full paper.
Dr. Charisse Madlock-Brown, co-lead of N3C SDoH, has led efforts to study SDoH variables as associated with U.S. county-level COVID-19 cumulative cases and death incidence at various periods. The study identifies county-level SDoH factors and policy factors important to COVID-19 case and death burden, highlighting the importance of evaluating the choice of time interval and interpretation of SDoH variables as the pandemic progressed. This work helps guide efforts to feature selection and understanding the impact of SDoH variables. This work was published in BMC Public Health.
While these screening tools help care providers to identify individual-level SDoH for social health interventions, data engineering hurdles continue to prevent SDoH data from flowing into or retrieval from research CDMs. The challenges of SDoH data collection and harmonization impact research across multiple domains. Additionally, studies examining the relationship between SDoH, demographics, and health outcomes poorly represent these relationships – leading to misinterpretations, limited study reproducibility, and limited secondary research use capacity. To promote integrative research with SDoH data, a collaborative effort between the N3C SDoH and Environmental Health domain teams developed a developed a perspective article published in Advanced Genetics exploring a research frameworks that is inclusive of clinical findings, patient-level and community-level SDoH data, environmental data sets, genomic information, and anticipated role of AI/ML tools.
The N3C SDoH domain team continues to explore the role of SDoH data in collaborative translational science. Critical momentum has focused on data mapping and standardizations to fill data engineering gaps with clinical screening tools, quality control of downstream data flow, and exploring the feature engineering and modeling of SDoH and COVID-19 in cohorts of interest, such as the Long COVID RECOVER effort. Beyond COVID-19, the domain team continues to explore data considerations needed for research on climate change and concurrent crises, applications in health equity and fairness research, and approaches to understand SDoH that is inclusive of variations in human abilities.
Recent publications include:
1. Madlock-Brown C, Wilkens K, Weiskopf N, Cesare N, Bhattacharyya S, Riches NO, Espinoza J, Dorr D, Goetz K, Phuong J, Sule A. Clinical, social, and policy factors in COVID-19 cases and deaths: methodological considerations for feature selection and modeling in county-level analyses. BMC Public Health. 2022 Dec;22(1):1-3. Published 14 Apr 2022. Available at: https://doi.org/10.1186/s12889-022-13168-y
2. Phuong J, Riches NO, Madlock-Brown C, Duran D, Calzoni L, Espinoza JC, Datta G, Kavuluru R, Weiskopf NG, Ward-Caviness CK, Lin AY. Social Determinants of Health Factors for Gene–Environment COVID-19 Research: Challenges and Opportunities. Advanced Genetics. 09 March 2022. 2100056. https://doi.org/10.1002/ggn2.202100056
3. Phuong J, Hong S, Palchuk MB, Espinoza J, Meeker D, Dorr DA, Lozinski G, Madlock-Brown C, Adams WG. Advancing Interoperability of Patient-level Social Determinants of Health Data to Support COVID-19 Research. In AMIA Joint Informatics Summit Proceedings 2022 (Vol. 2022). American Medical Informatics Association. Accepted 2021 Dec 9.
4. Phuong J, Zampino E, Dobbins N, Espinoza J, Meeker D, Spratt H, Madlock-Brown C, Weiskopf NG, Wilcox A. Extracting Patient-level Social Determinants of Health into the OMOP Common Data Model. InAMIA Annual Symposium Proceedings 2021 (Vol. 2021, p. 989). American Medical Informatics Association. Published 2022 Feb 21. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8861735/