Population Health Analytics
Social Needs and Population Health: What the Research Shows
Social determinants of health, such as housing, food access, economic stability and transportation, shape health outcomes in ways that show up consistently in clinical data: higher rates of hospitalization, emergency department use, chronic disease burden and cost.
Collecting and addressing SDOH presents complex challenges for health care organizations. Success requires more than identifying patients’ social needs — it depends on documenting those needs in a meaningful way, making the information accessible to the right care teams, connecting patients with appropriate interventions, delivering support that is culturally and linguistically appropriate, and measuring outcomes to improve and sustain results.
Social Need Markers and GeoHealth in the ACG System
Within the ACG System, there are multiple tools to identify needs and risk. The System’s person-centered approach generates outputs at the individual patient level, giving care teams the specificity they need to act. Social Need Markers (SNMs) and ACG GeoHealth are two tools that allow users to assess and address social needs at the patient level.
Social Need Markers identify patient-level social needs documented in the clinical record, primarily through ICD-10 Z-codes. SNMs are organized across five domains — Social, Education, Health Care System, Economic and Physical Environment — and give care teams a structured way to identify individuals with known social needs, understand prevalence across a population, and track the results of intervention programs over time.
ACG GeoHealth works at the community level, using a patient’s address to surface geography-based social determinants of health. This is particularly relevant when there is a limitation of Z-code data, as documentation of Z-codes remains a challenge in many health care organizations. A patient with no coded social needs may still face meaningful social risk. GeoHealth helps fill that gap, providing context that individual clinical records may not capture.
Used together, SNMs and GeoHealth offer both a patient-level and community-level picture of social risk. For a deeper look at how these features work, see our SDoH Handbook.
Social Needs Research
Peer-reviewed literature documents the relationship between social needs and health outcomes. The findings are consistent across populations and settings: social needs are associated with higher rates of hospitalization, emergency department use and readmission, and they tend to compound one another when multiple needs are present.
A 2021 analysis of more than 13 million patients in the Nationwide Readmissions Database found that 30-day readmission rates rose sharply with the number of documented social need domains: from 11.5% for patients with no documented needs to 63.5% for those with needs across all five domains. Housing instability and employment were among the strongest predictors. A 2024 study of patients at a federally qualified health center found that 68% reported at least one unmet social need, and having multiple needs was associated with meaningfully higher rates of ED visits and hospitalization in the following year.
Researchers at the Johns Hopkins Center for Population Health IT (CPHIT) have contributed directly to this area of study. One study of 1.3 million Johns Hopkins Health System patients found that individual-level social needs and community-level SDOH provided complementary information — patients living in more disadvantaged neighborhoods who also had documented social needs showed higher ED utilization than neighborhood context alone would have predicted. A separate CPHIT study developed and validated an EHR-based social risk score using ACG Social Need Markers and Resource Utilization Bands to prospectively identify patients likely to experience future social needs, with consistent performance across older adults, racial and ethnic minority groups, and patients in high-deprivation neighborhoods. Additional CPHIT research applied latent class analysis to Medicaid populations and identified four clinically meaningful groupings of social need — stress-predominant, safety-predominant, access-to-care-predominant, and socioeconomic — each with distinct morbidity profiles characterized using ACG Aggregated Diagnosis Groups (ADG) counts.
Selected Research on Social Needs and Health Outcomes
The studies below represent a selected sample of key literature on social needs and health outcomes, including research authored or co-authored by CPHIT. This work reflects an active and ongoing area of inquiry, one that continues to inform how health systems identify, measure, and respond to social risk.
- Bensken WP, Alberti PM, Koroukian SM. Health-Related Social Needs and Increased Readmission Rates: Findings from the Nationwide Readmissions Database. J Gen Intern Med. 2021.
A retrospective analysis of 13.2 million patients found a dose-response relationship between the number of documented social need domains and 30-day readmission rates, with housing and employment needs among the strongest individual predictors.
https://pubmed.ncbi.nlm.nih.gov/33634384/
- Zeng C et al. Patterns of Social Needs Predict Quality-of-Life and Healthcare Utilization Outcomes in Patients from a Large Hospital System. J Gen Intern Med. 2024.
Latent class analysis identified six distinct patterns of co-occurring social needs, each associated with poorer health-related quality of life and higher 90-day healthcare utilization, with findings consistent across racial and age subgroups.
https://pubmed.ncbi.nlm.nih.gov/38710869/
- Drake C et al. Association of Patient-Reported Social Needs with Emergency Department Visits and Hospitalizations Among Federally Qualified Health Center Patients. J Gen Intern Med. 2024.
In a safety-net primary care population where 68% of patients reported at least one unmet need, multiple social needs were associated with higher probability of ED visits and hospitalization in the following 12 months, even after adjusting for comorbidity burden.
https://pubmed.ncbi.nlm.nih.gov/38717665/
- Sidebottom AC et al. Health-Related Social Needs and Health Care Utilization in the Accountable Health Communities Model. JAMA Netw Open. 2025.
Drawing on 166,682 Medicare and Medicaid recipients screened under the CMS Accountable Health Communities program, this study found housing instability and transportation needs showed the strongest associations with ED visits and inpatient admissions.
https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2833960
- Brandt EJ et al. Assessing and Addressing Social Determinants of Cardiovascular Health: JACC State-of-the-Art Review. J Am Coll Cardiol. 2023.
A clinical practice review recommending structured screening, ICD-10 Z-code documentation, and multidisciplinary care team workflows for addressing social determinants in cardiovascular populations.
https://pubmed.ncbi.nlm.nih.gov/37019584/
- Hatef E et al. The Impact of Social Determinants of Health on Hospitalization in the Veterans Health Administration. Am J Prev Med. 2019. CPHIT
A national study of 6.6 million Veterans found that neighborhood socioeconomic context was independently associated with hospitalization risk after adjusting for individual clinical characteristics, with spatial analysis identifying high-risk clinic sites across the country.
https://pubmed.ncbi.nlm.nih.gov/31003812/
- Hatef E, Kitchen C, Pandya C, Kharrazi H. The Synergistic Effect of Social Needs and Social Determinants of Health on Healthcare Utilization at a Multilevel Academic Healthcare System. J Med Syst. 2023. CPHIT
Using data from 1.3 million Johns Hopkins Health System patients, individual-level social needs and community-level SDOH provided complementary information about ED utilization risk, with the combination producing a stronger signal than either data source alone.
https://pubmed.ncbi.nlm.nih.gov/37656284/
- Hatef E et al. Development of a Social Risk Score in the Electronic Health Record to Identify Social Needs Among Underserved Populations. JMIR Form Res. 2024. CPHIT
CPHIT researchers developed and validated an EHR-based social risk score incorporating ACG Social Need Markers and Resource Utilization Bands to prospectively identify patients likely to experience future social needs, with consistent performance across key subpopulations.
https://pubmed.ncbi.nlm.nih.gov/38470477/
- Pandya CJ, Wu J, Hatef E, Kharrazi H. Latent Class Analysis of Social Needs in Medicaid Population and Its Impact on Risk Adjustment Models. Med Care. 2024. CPHIT
Latent class analysis of EHR-documented social needs in a Medicaid population identified four clinically meaningful groupings with distinct morbidity profiles, characterized using ACG ADG counts, offering a more structured view of social risk than single-domain documentation.
https://pubmed.ncbi.nlm.nih.gov/38085115/
Sign up for blog alerts and other insights from the ACG System team
Follow Us
