HEOR Articles

Combining Social Determinants With Real-World Clinical Data for Better Mental Health Research

Michelle B. Leavy, MPH; Veena Hoffman, MPH; Jessica Paulus, PhD; Carl D. Marci, MD; OM1, Inc, Boston, MA, USA

Recent research has highlighted the extraordinary influence of social determinants of health (SDoH) on mental health, with many studies finding increased prevalence and severity of depression, anxiety, and other mental health conditions in populations that experience chronic stress and discrimination.1-3 Studies have also described the relationship between SDoH and access to care, showing that reduced access to treatment linked to SDoH results in poorer outcomes.4 The COVID-19 pandemic further heightened attention to this growing area of research, as unemployment, chronic stress, and social isolation led to increases in mental health diagnoses.5 The relationship between SDoH and mental health diagnosis, treatment, and outcomes is now an area of intense focus for both research and health policy experts.


"The relationship between SDoH and mental health diagnosis, treatment, and outcomes is now an area of intense focus for both research and health policy experts."


The increased interest in SDoH has raised new questions about how to capture and use these data in research studies for additional insights and to advance the field. Real-world data (RWD) sources are a valuable tool for facilitating mental health research, as these sources offer an efficient means of assembling large, heterogeneous patient populations and observing real-world treatment patterns and outcomes across different practice settings.6 While clinical trials typically enroll a narrow patient population, RWD sources capture information about routine clinical care from broader patient populations, including racially and ethnically diverse patients from different socioeconomic backgrounds.

Determining how to incorporate SDoH data into RWD sources is an important challenge for stakeholders interested in advancing mental health research and informing mental health policy. This article describes 2 analytic approaches to using multiple data sources linked to real-world clinical notes, outcome measures, and administrative claims to better understand the role of SDoH in mental health.


Examining Outcomes by Income and Race
One approach to understanding the impact of SDoH on mental health is to incorporate data on income and race in outcomes research. This approach was used in a retrospective, observational cohort study that described the associations between race and household income and measures of major depressive disorder (MDD) burden in a real-world cohort of Black and White patients with MDD in the United States.6

Data were drawn from the OM1 PremiOMTM MDD Dataset and linked to a SDoH dataset. The MDD dataset includes linked electronic medical records and administrative claims data from OM1’s Mental Health Network of over 3 million patients seen in more than 2000 community-based practices across all 50 states. The SDoH dataset is a patient-level data source that includes sociodemographic and behavioral attributes of adults (age ≥18) in the United States. Data elements include race, ethnicity, occupation, credit risk score, educational attainment, household income, and homeownership. The SDoH data source is considered generalizable to the broader United States population as it includes information on over 250 million people.

The analysis looked at age, race, sex, insurance type, education, household income, and Patient Health Questionnaire-9 (PHQ-9) scores (an outcome measure of depression symptom severity), as well as mental healthcare-related visits and antidepressant prescriptions. More than 123,000 Black patients and 1 million White patients were included in the analysis.

The study found that median PHQ-9 scores for Black patients were higher than for White patients at baseline (10.8 vs 8.8; P < .0001). In addition, patients with annual incomes at or below the federal poverty level of $25,000 had higher mean PHQ-9 scores than patients with incomes of at least $25,000 (9.7 vs 8.9; P <.01). In terms of access to treatment, emergency and inpatient mental healthcare use was significantly higher in Black patients, and Black patients had fewer outpatient mental health visits than White patients. Prescription fills for antidepressant therapy in the 12 months after baseline were also lower for Black patients suggesting a lack of adequate treatment. Importantly, these disparities by race and income persisted over the course of the study’s 18-month follow-up period.


Examining Outcomes by Credit Risk Score
Although similar to income, credit risk scores offer a different lens into the financial health of patients and are another important variable to help assess the impact of SDoH on mental health treatment, and outcomes. In this example, credit risk was analyzed in a retrospective, observational cohort study that described the disease burden in patients with clinical depression. Data were also drawn from the OM1 PremiOMTM MDD dataset and included information on age, race, sex, insurance type, and PHQ-9 scores as a measure of disease burden. 

This real-world study included more than 3.4 million patients with MDD. Patients with high credit risk scores had a median household income of ~$47,000, while patients with low credit risk scores had a median income of ~$80,000. Black patients made up 16% of the high credit risk group and 5% of the low credit risk group. In terms of symptom severity, patients with high credit risk had higher PHQ-9 scores than patients with low credit risk (13.3 vs. 12.4; P < .001). Similar to the previous study, patients in the high credit risk group also had more emergency and inpatient mental healthcare use, lower outpatient mental healthcare visits, and fewer prescription fills for antidepressant therapy. As in the income and race analysis, the credit risk score disparities persisted through the study follow-up period.


Implications for Real-World Research
These examples highlight the substantial independent impact of racial and financial SDoH variables on treatment patterns and outcomes in depression and emphasize the importance of incorporating these data into mental health studies. Researchers seeking to use RWD for mental health research should consider leveraging SDoH variables to better characterize their patient population and understand the complex factors that influence outcomes.

Researchers should also keep in mind the limitations and complexities of SDoH data. For example, many RWD sources are missing data on race and ethnicity. The source of the data, including whether it is patient-reported or clinician-reported, should also be considered, as should the recency of the data.

"As the crisis in mental health in the United States continues, improving outcomes will require further research and understanding of the complex interplay between social determinants and mental health variables."


Inclusion of SDoH data in mental health research may also help to identify potential policy initiatives for improving outcomes. The studies described above point to areas such as improving access to outpatient mental healthcare and identifying and addressing reasons for medication discontinuation as options for further research. When reviewing studies that focus on mental health, policy makers should consider whether SDoH variables were included, as studies that incorporate these data may provide greater insights into disparities in disease burden and offer opportunities for improving care.

Research has shown that an individual’s physical health is strongly linked to social and environmental factors both at the individual and broader community levels. There is growing evidence that this is true for mental health as well. As the crisis in mental health in the United States continues, improving outcomes will require further research and understanding of the complex interplay between social determinants and mental health variables. RWD sources can play an important role by capturing data on diverse patient populations and linking those data to SDoH datasets to further describe disparities, guide future research, and identify areas for policy-based initiatives with the goal of improving outcomes for all.


1. Alegría M, NeMoyer A, Falgàs Bagué I, Wang Y, Alvarez K. Social determinants of mental health: where we are and where we need to go. Curr Psychiatry Rep. 2018;20(11):95. doi:10.1007/s11920-018-0969-9

2. Allen J, Balfour R, Bell R, Marmot M. Social determinants of mental health. Int Rev Psychiatry. 2014 Aug;26(4):392-407. doi: 10.3109/09540261.2014.928270. PMID: 25137105.

3. Cotton NK, Shim RS. Social determinants of health, structural racism, and the impact on child and adolescent mental health. J Am Acad Child Adolesc Psychiatry. 2022;61(11): 1385–1389. https://doi.org/10.1016/j.jaac.2022.04.020.

4. Triplett NS, Luo M, Nguyen JK, Sievert K. Social determinants and treatment of mental disorders among children: analysis of data from the National Survey of Children’s Health. Psychiatr Serv. 2022;73(8):922-925. doi:10.1176/appi.ps.202100307.

5. Ettman CK, Abdalla SM, Cohen GH, Sampson L, Vivier PM, Galea S. Prevalence of depression symptoms in US adults before and during the COVID-19 pandemic. JAMA Netw Open. 2020;3(9):e2019686. doi:10.1001/jamanetworkopen.2020.19686.

6. Gliklich RE, Leavy MB, Cosgrove L, et al. Harmonized outcome measures for use in depression patient registries and clinical practice. Ann Intern Med. 2020;172(12):803-809.

7. Paulus J, Severtson G, Qian X, et al. The association between race, social determinants of health, and treatment for major depressive disorder in a real-world cohort. Presented at the 38th International Conference on Pharmacoepidemiology & Therapeutic Risk Management. August 24-28, 2022. Copenhagen, Denmark.

Your browser is out-of-date

ISPOR recommends that you update your browser for more security, speed and the best experience on ispor.org. Update my browser now