From Intake to Intervention: How Technology Is Transforming SDOH Screening in the Emergency Department

Foster Goss, DO, MMSc, FACEP, FAMIA
Alex Lupi, MD
Emergency Medicine Informatics Section

The emergency department is, for millions of Americans, the most reliable point of contact with the healthcare system. On any given shift, emergency physicians encounter patients whose chief complaints (shortness of breath, chest pain, a minor laceration) mask deeper, persistent challenges: food insecurity, unstable housing, lack of transportation, social isolation, or violence in the home. These social determinants of health (SDOH) are not incidental to the clinical picture. They are often its cause.

Yet for years, SDOH screening in the ED has been inconsistent, burdensome, and poorly connected to actionable resources. Clinicians know the need is there, but the tools to address it systemically have lagged behind. That is changing. A new generation of digital tools, integrated platforms, and AI-assisted workflows is giving emergency physicians practical, scalable ways to screen for social needs, connect patients with community resources, and close the loop on referrals, all within the constraints of a high-volume shift.

This article reviews the current landscape of technology-enabled SDOH screening and resource navigation in emergency medicine, with practical guidance for ED physicians looking to engage with these tools.

The ED: An Underutilized SDOH Touchpoint

Emergency departments disproportionately serve patients experiencing poverty, housing instability, substance use disorders, and limited access to primary care. Research consistently shows that patients with unmet social needs are higher utilizers of emergency services and have worse outcomes across nearly every disease category.^1,2. Collecting Structured SDoH data, when paired with referral actions, has been shown to reduce hospitalizations and costs.³ Additionally, there is evidence that timely social needs referrals can improve outcomes for vulnerable populations.⁴  The social and structural drivers of these disparities don’t present on labs or imaging. They surface in conversation, in repeated visits, in the address on the registration form.

Despite this, systematic SDOH screening in the ED has faced real barriers: time pressure, workflow fragmentation, clinician uncertainty about what to do with a positive screen, lack of integration with community resources, and documentation burden. Surveys of emergency physicians reveal that most recognize the importance of SDOH, but fewer than half report regularly screening for them, and even fewer feel equipped to act on what they find.¹

Technology does not solve these problems by itself. But the right tools, thoughtfully implemented, can dramatically reduce the friction of screening and dramatically increase the likelihood that a positive screen leads to a meaningful connection to resources.

The Technology Landscape

EHR-Integrated SDOH Screening Tools

The most sustainable SDOH workflows are those embedded directly into the EHR. Major platforms now support structured SDOH data capture, often aligned with the Gravity Project’s standardized terminology and ICD-10-CM Z-codes for social determinants.^5,6. Epic’s Social Determinants tools, for example, allow for structured documentation of housing, food security, transportation, and safety concerns, with patient-facing questionnaires that can be completed on a tablet or MyChart app during triage.

For ED physicians, the practical value is in the workflow: a screen completed by the patient or a nurse, triggered automatically at check-in or triage, flags positive responses before the physician walks in the room. Rather than a separate conversation starting from zero, the physician can engage with specific needs already identified, validate them briefly, and initiate a referral pathway.

Validated tools commonly used in ED settings include the PRAPARE (Protocol for Responding to and Assessing Patients’ Assets, Risks, and Experiences),⁷ the AHC Health-Related Social Needs (HRSN) Screening Tool, and the Accountable Health Communities tool. Many ED systems are moving toward a shorter, CMS-aligned set of questions that can be completed in under three minutes.

Closing the Referral Gap: Community Resource Platforms

Identifying a social need is only half the challenge. The other half, connecting the patient to a resource that actually exists, is available, and is accessible to them, has historically required institutional memory, printed handouts, and a lot of phone calls. Technology platforms have emerged specifically to address this gap.

Key platforms now deployed in ED settings include:

• Unite Us: Unite Us maintains a curated, regularly updated network of social service organizations by geography, with eligibility filters, and supports closed-loop referrals, meaning the ED team can track whether a patient actually connected with a referred agency.
• Findhelp (formerly Aunt Bertha): Integrated into several major EHR systems, Findhelp allows clinicians or social workers to search for local resources by zip code and domain (food, housing, utilities, mental health) and send referrals or resource lists directly to the patient via text or email.
• 2-1-1 Integration: The 2-1-1 national social services helpline now has API integrations that allow EHRs to pull real-time resource data into the clinical workflow, with some systems enabling direct referral handoffs.
• WellSky Social Care Coordination (formerly Healthify): A platform designed specifically for care coordination and population health management, with SDOH dashboards that track unmet need at the panel and community level.

For emergency physicians, the key question is what role they play in the referral handoff. In many high-functioning ED SDOH programs, physicians validate the need and warm-hand off to a social worker or community health worker (CHW) who manages the resource navigation. The technology infrastructure supports this by providing real-time, location-specific resource information to whoever is handling the handoff.

AI-Assisted Screening and Risk Stratification

A growing number of health systems are deploying machine learning tools that predict social risk from EHR data, including diagnosis codes, visit patterns, pharmacy records, and claims.⁸ At NorthShore-Edward-Elmhurst Health in Illinois, an NLP-based system now extracts SDOH from unstructured clinical notes for ED social workers, enabling them to redirect time from chart review to direct patient care. In a 2022–2023 pilot, 21% of the 1,465 patients assessed screened positive for at least one social need. Medical Home Network, using the ClosedLoop.ai platform, demonstrated that integrating SDOH data into risk models identified 41% more high-risk patients than traditional approaches.⁹. At Mass General Brigham, researchers fine-tuned clinical language models to extract SDOH from visit notes with 93.8% sensitivity, compared to just 2% capture through diagnostic codes alone.¹⁰

The regulatory picture is evolving. CMS introduced SDOH screening measures (SDOH-1 and SDOH-2) for hospital quality reporting but has since finalized their removal beginning FY 2026, making screening voluntary rather than mandatory. Still, health systems participating in accountable care arrangements retain strong financial incentives to identify and address SDOH upstream of costly ED utilization, and the broader push toward health equity in quality reporting continues to drive investment in these tools.

For the practicing emergency physician, AI risk stratification is largely a background function: it shapes which patients get a social work consult, which get a CHW follow-up call, which get a care management flag. The clinical decision, ultimately, remains human. But these tools make that decision more informed and more efficient.

Patient-Facing Digital Tools

Patient-facing screening tools, including tablets, kiosks, and patient portal questionnaires, offer a promising avenue for SDOH data collection, particularly for sensitive domains like housing instability, intimate partner violence, and substance use, where patients may be more candid without a clinician present. ED waiting rooms represent a natural opportunity for this approach: patients have time, and many health systems have deployed tablet lending programs or kiosk setups to support the workflow. However, the evidence base remains early-stage, and concerns about selection bias are real. Research has shown that patients with prior financial insecurity may be less likely to complete self-administered screens, underscoring the need for multiple modalities rather than a single digital-only approach.¹¹

Emerging tools also include LLM-powered chatbots designed for conversational SDoH data collection, which may combine the scalability of digital screening with a more adaptive, patient-centered interaction.¹² Meanwhile, text-based follow-up via SMS is being explored as a way to check in with patients after an ED visit, confirm whether they connected with referred resources, and provide an easy pathway to re-engage if they did not. This is particularly relevant for addressing the “last mile” problem: positive SDOH screens that never translate into completed connections. Published data on referral completion rates after ED-based screening remain limited, highlighting a critical gap for future research.

What EM Physicians Can Do Right Now

Not every ED has an integrated SDOH platform or a dedicated social work team available around the clock. But regardless of institutional resources, there are steps emergency physicians can take today.

• Know what your EHR already supports. Many Epic and Cerner instances have SDOH screening modules that are built but not activated or not promoted in the clinical workflow. Talk to your informatics team.
• Bookmark a resource navigation tool. Even without institutional integration, free tools like Findhelp.org, 211.org, and SAMHSA’s treatment locator can be pulled up quickly on any workstation to give patients a real, specific referral rather than a generic handout.
• Document SDOH using structured fields and Z-codes. Using Z55–Z65 ICD-10 codes for social determinants creates a codified, searchable record of social need that supports population health tracking, quality reporting, and future risk stratification. This documentation also matters for reimbursement as payers increasingly recognize social complexity.
• Advocate for team-based SDOH response. Emergency physicians don’t need to be the primary SDOH responder, but they need to be the champions who ensure the team has the tools and workflow support to do this work. Community health workers, social workers, and care managers are more effective when technology enables seamless handoffs and closed-loop follow-up.
• Engage with national efforts. ACEP’s Social Emergency Medicine and Population Health Section, the American College of Emergency Physicians’ Social Emergency Medicine resources, and the Gravity Project are all building the standards and tools that will shape this work for the next decade. EM voices matter in these forums.

Equity and Implementation

Technology-enabled SDOH screening is not a neutral intervention. How and where these tools are implemented, and how results are used, carries real equity implications. Screening patients without the infrastructure to respond risks raising expectations that cannot be met and eroding trust. AI-driven risk tools built on biased historical data can replicate and amplify existing disparities if not carefully validated across demographic subgroups. Patient-facing digital tools may exclude patients with limited English proficiency, low digital literacy, or no smartphone access.

Effective implementation requires not just technical infrastructure but community partnership, culturally responsive communication, language access, and sustained commitment from institutional leadership. The tools described here are enablers, not solutions. The solution is a healthcare system that takes social need seriously as a medical issue, and emergency medicine is well positioned to lead that shift.

The Bottom Line

The emergency department’s role in SDOH screening is not a distraction from clinical medicine; it is an extension of it. Technology cannot substitute for the upstream policy changes and social investments that would make SDOH screening less necessary. But it can make the ED’s daily practice of care more responsive to the full reality of patients’ lives: today, on each shift, with the tools that exist now.

References

1. Loo S, Molina M, Ahmad NJ, et al. Implementing social determinants of health screening in US emergency departments. JAMA Netw Open. 2025;8(3):e250137.
2. Wallace AS, Luther BL, Sisler SM, Wong B, Guo JW. Integrating social determinants of health screening and referral during routine emergency department care: evaluation of reach and implementation challenges. Implement Sci Commun. 2021;2(1):114.
3. Yan AF. Collecting and Using Social Needs Data in Health Settings: A Systematic Review of Health Service Utilization and Costs. PMC/Academic Press; 2025.
4. Gottlieb LM, Wing H, Adler NE. A systematic review of interventions on patients’ social and economic needs. Am J Prev Med. 2017;53(5):719-729.
5. Gravity project. Gravity Project. June 23, 2022. Accessed April 27, 2026. https://thegravityproject.net/.
6. Chishtie JA, Tea J, Ester M, et al. Integration of screening and referral tools for social determinants of health and modifiable lifestyle factors in the Epic electronic health record system: Scoping review. J Med Internet Res. 2025;27:e73615.
7. PRAPARE. Accessed April 27, 2026. https://prapare.org/.
8. Holcomb J, Oliveira LC, Highfield L, Hwang KO, Giancardo L, Bernstam EV. Predicting health-related social needs in Medicaid and Medicare populations using machine learning. Sci Rep. 2022;12(1):4554.
9. Carroll NW, Jones A, Burkard T, Lulias C, Severson K, Posa T. Improving risk stratification using AI and social determinants of health. Am J Manag Care. 2022;28(11):582-587.
10. Guevara M, Chen S, Thomas S, et al. Large language models to identify social determinants of health in electronic health records. NPJ Digit Med. 2024;7(1):6.
11. Vest JR, Mazurenko O. Non-response bias in social risk factor screening among adult emergency department patients. J Med Syst. 2023;47(1):78.
12. Ong JCL, Seng BJJ, Law JZF, et al. Artificial intelligence, ChatGPT, and other large language models for social determinants of health: Current state and future directions. Cell Rep Med. 2024;5(1):101356.