Melanie Frances Molina
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Melanie Frances Molina has a total of 78 co-authors. Here are the top co-authors by number of shared publications:
Cynthia Fenton
Division of Clinical Informatics and Digital Transformation, Department of Medicine, University of California, San Francisco, CA
3 shared publicationsRobert M. Rodriguez
Department of Internal Medicine, University of California, Riverside
3 shared publicationsSamuel D. Pimentel
Department of Statistics, University of California, Berkeley, CA
3 shared publicationsAaron E. Kornblith
Department of Emergency Medicine, University of California, San Francisco, CA; Department of Pediatrics, University of California, San Francisco, CA
2 shared publicationsAnita N. Chary
Center for Innovations in Quality Effectiveness and Safety, Michael E. DeBakey United States Department of Veterans Affairs Medical Center, Houston, Texas, USA
2 shared publications- Anita Nandkumar Chary
Department of Emergency Medicine Baylor College of Medicine Houston Texas USA
2 shared publications
Latest Publications
- 10 September 2026
- Published by Massachusetts Medical Society in New England Journal of Medicine
- Vol. 395 (10) , 939-941
AI-Based Translation of Discharge Instructions — Aligning Policy with Practice
- 10 September 2026
- Published by Massachusetts Medical Society in New England Journal of Medicine
- Vol. 395 (10) , 939-941
AI-Based Translation of Discharge Instructions — Aligning Policy with Practice
- Open Access
- 9 July 2026
- Published by Public Library of Science (PLoS) in PLOS Digital Health
- Vol. 5 (7) , e0001491
Large language model-generated clinical summaries in emergency departments: A blinded comparison study
Emergency department (ED) clinicians routinely construct “one-liner” summaries, distilling a patient’s history and presentation into one high-yield sentence that supports rapid decision-making. Producing these summaries is cognitively demanding. Large language models (LLMs) may assist by synthesizing longitudinal electronic health record (EHR) data. In this blinded study of 99 ED encounters, emergency physicians evaluated paired LLM- and physician-authored summaries on accuracy, completeness, and clinical utility, indicating their overall preference with free-text explanation. LLM-generated one-liner summaries were produced using k-nearest-neighbor few-shot prompting. We used linear mixed-effects to compare ratings. We also examined the LLM’s selective note inclusion and used rapid content analysis to summarize free-text explanations. We found that across all dimensions, LLM-generated summaries received higher ratings than physician-authored summaries. Mean (SE) estimated marginal means for accuracy were 4.18 (0.09) vs 3.40 (0.11) (β = 0.78; 95% CI 0.50–1.07), for completeness 3.69 (0.10) vs 3.25 (0.12) (β = 0.44; 95% CI 0.14–0.74), and for clinical utility 3.88 (0.10) vs 3.21 (0.12) (β = 0.67; 95% CI 0.35–0.99). The LLM utilized 53.2% of available notes, with History & Physical (H&P) notes requested in 97.0% of cases and imaging reports in 81.8%, demonstrating context-sensitive document selection that varied significantly by chief complaint. Qualitative analysis indicated that LLM summaries were often more inclusive and neutrally phrased, whereas physician summaries exhibited greater contextual nuance but occasionally omitted key details. These findings represent an important first step in evaluating whether LLMs can produce clinically acceptable one-liner summaries from longitudinal EHR data; prospective validation is required before deployment in real-world ED workflows.
- 9 July 2026
- Published by Public Library of Science (PLoS) in PLOS Digital Health
- Vol. 5 (7) , e0001491
Open AccessLarge language model-generated clinical summaries in emergency departments: A blinded comparison study
Emergency department (ED) clinicians routinely construct “one-liner” summaries, distilling a patient’s history and presentation into one high-yield sentence that supports rapid decision-making. Producing these summaries is cognitively demanding. Large language models (LLMs) may assist by synthesizing longitudinal electronic health record (EHR) data. In this blinded study of 99 ED encounters, emergency physicians evaluated paired LLM- and physician-authored summaries on accuracy, completeness, and clinical utility, indicating their overall preference with free-text explanation. LLM-generated one-liner summaries were produced using k-nearest-neighbor few-shot prompting. We used linear mixed-effects to compare ratings. We also examined the LLM’s selective note inclusion and used rapid content analysis to summarize free-text explanations. We found that across all dimensions, LLM-generated summaries received higher ratings than physician-authored summaries. Mean (SE) estimated marginal means for accuracy were 4.18 (0.09) vs 3.40 (0.11) (β = 0.78; 95% CI 0.50–1.07), for completeness 3.69 (0.10) vs 3.25 (0.12) (β = 0.44; 95% CI 0.14–0.74), and for clinical utility 3.88 (0.10) vs 3.21 (0.12) (β = 0.67; 95% CI 0.35–0.99). The LLM utilized 53.2% of available notes, with History & Physical (H&P) notes requested in 97.0% of cases and imaging reports in 81.8%, demonstrating context-sensitive document selection that varied significantly by chief complaint. Qualitative analysis indicated that LLM summaries were often more inclusive and neutrally phrased, whereas physician summaries exhibited greater contextual nuance but occasionally omitted key details. These findings represent an important first step in evaluating whether LLMs can produce clinically acceptable one-liner summaries from longitudinal EHR data; prospective validation is required before deployment in real-world ED workflows.
- Open Access
- 13 June 2026
- Published by Elsevier in Annals of Emergency Medicine
- Vol. 88 (5) , 562-569
Computable Structured Phenotype Versus Large Language Model Identification of Opioid Use Disorder Using Electronic Health Record Data
- 13 June 2026
- Published by Elsevier in Annals of Emergency Medicine
- Vol. 88 (5) , 562-569
Open AccessComputable Structured Phenotype Versus Large Language Model Identification of Opioid Use Disorder Using Electronic Health Record Data
- Open Access
Comparing computable structured phenotype- versus large language model-identification of opioid use disorder using electronic health record data
Study Objective: To compare a computable structured opioid use disorder (OUD) phenotype currently used to trigger emergency department (ED) clinical decision support (CDS) with a large language model (LLM) for OUD identification, using expert physician review as the reference standard. Methods: We conducted a retrospective study of randomly sampled adult ED encounters (January 1, 2023-October 17, 2024) at a single academic health system. Encounters were stratified by structured phenotype status and weighted to reflect population prevalence. The structured phenotype, implemented operationally to activate CDS, incorporated diagnosis codes, medications for OUD, urine toxicology results, addiction consultations, and keyword recognition. An LLM (ChatGPT 4.1) analyzed ED notes from the index visit using a zero-shot prompt. Two board-certified emergency physicians independently determined OUD status by full chart review; discrepancies were adjudicated by a third reviewer. We calculated weighted sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Results: Among 302 encounters, weighted OUD prevalence was 5.6% (95% CI 4.0-7.0%). The structured phenotype demonstrated sensitivity 0.84 (95% CI 0.42-0.97) and specificity 0.964 (95% CI 0.96-0.97) (PPV 0.58; NPV 0.99). The LLM demonstrated sensitivity 0.81 (95% CI 0.70-0.88) and specificity 0.996 (95% CI 0.993-0.998) (PPV 0.92; NPV 0.99). Specificity was significantly higher for the LLM (p<0.0001). Conclusion: Both approaches demonstrated strong diagnostic performance. Although the structured phenotype showed slightly higher sensitivity, the LLM achieved higher specificity and PPV, suggesting potential to reduce false-positive alerts in ED workflows. Prospective validation in larger, representative populations is needed to guide clinical implementation.
Open AccessComparing computable structured phenotype- versus large language model-identification of opioid use disorder using electronic health record data
Study Objective: To compare a computable structured opioid use disorder (OUD) phenotype currently used to trigger emergency department (ED) clinical decision support (CDS) with a large language model (LLM) for OUD identification, using expert physician review as the reference standard. Methods: We conducted a retrospective study of randomly sampled adult ED encounters (January 1, 2023-October 17, 2024) at a single academic health system. Encounters were stratified by structured phenotype status and weighted to reflect population prevalence. The structured phenotype, implemented operationally to activate CDS, incorporated diagnosis codes, medications for OUD, urine toxicology results, addiction consultations, and keyword recognition. An LLM (ChatGPT 4.1) analyzed ED notes from the index visit using a zero-shot prompt. Two board-certified emergency physicians independently determined OUD status by full chart review; discrepancies were adjudicated by a third reviewer. We calculated weighted sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Results: Among 302 encounters, weighted OUD prevalence was 5.6% (95% CI 4.0-7.0%). The structured phenotype demonstrated sensitivity 0.84 (95% CI 0.42-0.97) and specificity 0.964 (95% CI 0.96-0.97) (PPV 0.58; NPV 0.99). The LLM demonstrated sensitivity 0.81 (95% CI 0.70-0.88) and specificity 0.996 (95% CI 0.993-0.998) (PPV 0.92; NPV 0.99). Specificity was significantly higher for the LLM (p<0.0001). Conclusion: Both approaches demonstrated strong diagnostic performance. Although the structured phenotype showed slightly higher sensitivity, the LLM achieved higher specificity and PPV, suggesting potential to reduce false-positive alerts in ED workflows. Prospective validation in larger, representative populations is needed to guide clinical implementation.
- Open Access
Characterizing emergency clinician engagement with social drivers of health data among patients with opioid use disorder
Objectives: To characterize emergency department (ED) clinician engagement with electronic health record (EHR)-based social drivers of health (SDOH) data; test whether engagement differs in encounters with opioid use disorder (OUD); and, among OUD encounters, assess whether engagement is associated with medications for OUD (MOUD). Materials and Methods: We conducted a cross-sectional study of adult ED encounters (January 2023–October 2024). OUD encounters, identified with a structured phenotype, were matched (1:2) to non-OUD encounters. Audit logs captured clinician engagement with structured SDOH questions (“SDOH Wheel”), ICD-10 Z codes in the Problem List, Social History free text, and social work notes. Engagement was any SDOH documentation or review of preexisting SDOH data during the encounter. Logistic regression estimated associations. Results: Among 17,103 encounters (5,701 OUD; 11,402 non-OUD), clinician SDOH documentation was rare (<1%). Clinicians most often reviewed Z codes (610/620; 98.4%), followed by the SDOH Wheel (1,103/3,953; 27.9%), social work notes (1,711/10,670; 16.0%), and Social History free text (232/6,942; 3.3%). Engagement occurred in 19.5% of encounters and was higher with OUD (26.6% vs 16.0%; OR 1.91, 95% CI 1.77–2.07). Among OUD encounters, engagement was not associated with MOUD (OR 1.11, 95% CI 0.84–1.47); however, MOUD treatment varied by race and ethnicity, reflecting persistent disparities. Discussion: ED clinicians infrequently documented but did review structured, accessible SDOH data. Engagement increased in OUD encounters yet neither predicted MOUD nor mitigated racial and ethnic treatment disparities. EHR interfaces that surface SDOH data coupled with targeted decision supports might influence equitable, time-sensitive ED care.
Open AccessCharacterizing emergency clinician engagement with social drivers of health data among patients with opioid use disorder
Objectives: To characterize emergency department (ED) clinician engagement with electronic health record (EHR)-based social drivers of health (SDOH) data; test whether engagement differs in encounters with opioid use disorder (OUD); and, among OUD encounters, assess whether engagement is associated with medications for OUD (MOUD). Materials and Methods: We conducted a cross-sectional study of adult ED encounters (January 2023–October 2024). OUD encounters, identified with a structured phenotype, were matched (1:2) to non-OUD encounters. Audit logs captured clinician engagement with structured SDOH questions (“SDOH Wheel”), ICD-10 Z codes in the Problem List, Social History free text, and social work notes. Engagement was any SDOH documentation or review of preexisting SDOH data during the encounter. Logistic regression estimated associations. Results: Among 17,103 encounters (5,701 OUD; 11,402 non-OUD), clinician SDOH documentation was rare (<1%). Clinicians most often reviewed Z codes (610/620; 98.4%), followed by the SDOH Wheel (1,103/3,953; 27.9%), social work notes (1,711/10,670; 16.0%), and Social History free text (232/6,942; 3.3%). Engagement occurred in 19.5% of encounters and was higher with OUD (26.6% vs 16.0%; OR 1.91, 95% CI 1.77–2.07). Among OUD encounters, engagement was not associated with MOUD (OR 1.11, 95% CI 0.84–1.47); however, MOUD treatment varied by race and ethnicity, reflecting persistent disparities. Discussion: ED clinicians infrequently documented but did review structured, accessible SDOH data. Engagement increased in OUD encounters yet neither predicted MOUD nor mitigated racial and ethnic treatment disparities. EHR interfaces that surface SDOH data coupled with targeted decision supports might influence equitable, time-sensitive ED care.
- 17 November 2025
- Published by Wiley in Academic Emergency Medicine
- Vol. 33 (3) , e70195
Financial Compensation of Academic Emergency Medicine Physicians in the United States
- 17 November 2025
- Published by Wiley in Academic Emergency Medicine
- Vol. 33 (3) , e70195
Financial Compensation of Academic Emergency Medicine Physicians in the United States
- Open Access
- 29 October 2025
- Published by Elsevier in JACEP Open
- Vol. 6 (6) , 100272
Documentation of Adverse Social Drivers of Health in US Emergency Departments
- 29 October 2025
- Published by Elsevier in JACEP Open
- Vol. 6 (6) , 100272
Open AccessDocumentation of Adverse Social Drivers of Health in US Emergency Departments
- Open Access
- 1 October 2025
- Published by American Medical Association (AMA) in JAMA Network Open
- Vol. 8 (10) , e2537655
COVID-19 Booster Vaccine Messaging in Emergency Departments
Importance: Patient uptake of updated COVID-19 vaccines is crucial for reducing severe outcomes, yet national uptake remains low.Objective: To determine if tailored messaging or simple inquiry about vaccine acceptance increases 30-day uptake of updated COVID-19 vaccines among emergency department (ED) patients.Design, Setting, and Participants: This 3-arm, cluster randomized clinical trial conducted from January 29 to June 18, 2024, enrolled adult ED patients who had not received a COVID-19 vaccine in the prior 6 months across 6 EDs in San Francisco, California; Philadelphia, Pennsylvania; Houston, Texas; and Durham, North Carolina.Interventions: Intervention M involved tailored messaging about updated COVID-19 vaccines and inquiry about vaccine acceptance. Intervention Q involved only inquiry about vaccine acceptance. Usual care (no messaging or vaccine acceptance questions) served as the control.Main Outcomes and Measures: The primary outcome was updated COVID-19 vaccine receipt within 30 days of the ED visit, assessed via electronic health record review and follow-up telephone calls. Secondary outcomes included vaccine acceptance and vaccination during the ED visit. Outcomes were also compared between study sites that had the updated COVID-19 vaccine available and those that did not.Results: Of 852 participants (median age, 47 years [IQR, 33-63 years]; 464 [54.5%] women), 247 (29.0%) were in the intervention M group, 273 (32.0%) in the intervention Q group, and 332 (39.0%) in the control group. Vaccine uptake at 30 days was not significantly higher in either the intervention M group compared with control (14 participants [5.7%] vs 10 [3.0%]; absolute difference, 2.7 percentage points [pp] [95% CI, −0.8 to 6.3 pp]) or the intervention Q group compared with control (11 [4.0%] vs 10 [3.0%]; absolute difference, 1.0 pp [95% CI, −2.0 to 4.2 pp]). However, at sites where vaccines were available, the intervention M group had higher uptake compared with control (13 of 132 [9.8%] vs 5 of 150 [3.3%]; absolute difference, 6.5 pp [95% CI, 0.5-12.5 pp]).Conclusions and Relevance: In this cluster randomized clinical trial of adult ED patients, tailored messaging and simple inquiry alone did not significantly increase 30-day updated COVID-19 vaccine uptake among patients. The slight increase in uptake among participants in the EDs that had access to the updated COVID-19 vaccines suggested that vaccine availability was an effect modifier, underscoring the importance of opportunity and convenience in vaccine delivery.Trial Registration: ClinicalTrials.gov Identifier: NCT06156215
- 1 October 2025
- Published by American Medical Association (AMA) in JAMA Network Open
- Vol. 8 (10) , e2537655
Open AccessCOVID-19 Booster Vaccine Messaging in Emergency Departments
Importance: Patient uptake of updated COVID-19 vaccines is crucial for reducing severe outcomes, yet national uptake remains low.Objective: To determine if tailored messaging or simple inquiry about vaccine acceptance increases 30-day uptake of updated COVID-19 vaccines among emergency department (ED) patients.Design, Setting, and Participants: This 3-arm, cluster randomized clinical trial conducted from January 29 to June 18, 2024, enrolled adult ED patients who had not received a COVID-19 vaccine in the prior 6 months across 6 EDs in San Francisco, California; Philadelphia, Pennsylvania; Houston, Texas; and Durham, North Carolina.Interventions: Intervention M involved tailored messaging about updated COVID-19 vaccines and inquiry about vaccine acceptance. Intervention Q involved only inquiry about vaccine acceptance. Usual care (no messaging or vaccine acceptance questions) served as the control.Main Outcomes and Measures: The primary outcome was updated COVID-19 vaccine receipt within 30 days of the ED visit, assessed via electronic health record review and follow-up telephone calls. Secondary outcomes included vaccine acceptance and vaccination during the ED visit. Outcomes were also compared between study sites that had the updated COVID-19 vaccine available and those that did not.Results: Of 852 participants (median age, 47 years [IQR, 33-63 years]; 464 [54.5%] women), 247 (29.0%) were in the intervention M group, 273 (32.0%) in the intervention Q group, and 332 (39.0%) in the control group. Vaccine uptake at 30 days was not significantly higher in either the intervention M group compared with control (14 participants [5.7%] vs 10 [3.0%]; absolute difference, 2.7 percentage points [pp] [95% CI, −0.8 to 6.3 pp]) or the intervention Q group compared with control (11 [4.0%] vs 10 [3.0%]; absolute difference, 1.0 pp [95% CI, −2.0 to 4.2 pp]). However, at sites where vaccines were available, the intervention M group had higher uptake compared with control (13 of 132 [9.8%] vs 5 of 150 [3.3%]; absolute difference, 6.5 pp [95% CI, 0.5-12.5 pp]).Conclusions and Relevance: In this cluster randomized clinical trial of adult ED patients, tailored messaging and simple inquiry alone did not significantly increase 30-day updated COVID-19 vaccine uptake among patients. The slight increase in uptake among participants in the EDs that had access to the updated COVID-19 vaccines suggested that vaccine availability was an effect modifier, underscoring the importance of opportunity and convenience in vaccine delivery.Trial Registration: ClinicalTrials.gov Identifier: NCT06156215
- Open Access
Physician- versus Large Language Model-Generated Clinical Summaries in the Emergency Department
High-quality one-liner summaries are essential in the emergency department (ED) to support rapid decision-making, but generating them is cognitively demanding and adds to documentation burden; large language models (LLMs) may help by synthesizing longitudinal electronic health record (EHR) data into concise, clinically useful summaries. In this blinded, within-subject study of 99 ED encounters from adult patients at the University of California, San Francisco with prior inpatient admissions (March 2022–March 2024), 26 ED physicians (14 attendings, 12 residents) evaluated paired LLM- and physician-generated summaries in randomized order, rating each on accuracy, completeness, and clinical utility using 5-point Likert scales and indicating their preferred summary with optional free-text explanations. LLM-generated summaries were preferred in 50.5% of encounters, physician summaries in 38.4%, and 11.1% were ties; compared with physician summaries, LLM summaries had higher mean (SD) scores for accuracy (4.27 [0.98] vs 3.49 [1.43]; P < .001), completeness (3.72 [1.08] vs 3.28 [1.25]; P = .006), and clinical utility (3.95 [1.20] vs 3.28 [1.51]; P < .001). Qualitative feedback suggested LLMs tended to produce more inclusive and neutrally phrased summaries, while physicians offered richer nuance but sometimes omitted key details. These findings suggest that domain-adapted LLM-generated one-liners can outperform physician-authored summaries on multiple quality dimensions and, with clinician oversight, may aid rapid synthesis of complex EHR data in high-stakes settings.
Open AccessPhysician- versus Large Language Model-Generated Clinical Summaries in the Emergency Department
High-quality one-liner summaries are essential in the emergency department (ED) to support rapid decision-making, but generating them is cognitively demanding and adds to documentation burden; large language models (LLMs) may help by synthesizing longitudinal electronic health record (EHR) data into concise, clinically useful summaries. In this blinded, within-subject study of 99 ED encounters from adult patients at the University of California, San Francisco with prior inpatient admissions (March 2022–March 2024), 26 ED physicians (14 attendings, 12 residents) evaluated paired LLM- and physician-generated summaries in randomized order, rating each on accuracy, completeness, and clinical utility using 5-point Likert scales and indicating their preferred summary with optional free-text explanations. LLM-generated summaries were preferred in 50.5% of encounters, physician summaries in 38.4%, and 11.1% were ties; compared with physician summaries, LLM summaries had higher mean (SD) scores for accuracy (4.27 [0.98] vs 3.49 [1.43]; P < .001), completeness (3.72 [1.08] vs 3.28 [1.25]; P = .006), and clinical utility (3.95 [1.20] vs 3.28 [1.51]; P < .001). Qualitative feedback suggested LLMs tended to produce more inclusive and neutrally phrased summaries, while physicians offered richer nuance but sometimes omitted key details. These findings suggest that domain-adapted LLM-generated one-liners can outperform physician-authored summaries on multiple quality dimensions and, with clinician oversight, may aid rapid synthesis of complex EHR data in high-stakes settings.
- Open Access
- 1 April 2025
- Published by American Medical Association (AMA) in JAMA Network Open
- Vol. 8 (4) , e257951
Screening and Response for Adverse Social Determinants of Health in US Emergency Departments
Importance: Regulatory agencies have begun incentivizing screening for adverse social determinants of health (SDOH) and responses in inpatient settings, missing a crucial safety net: the emergency department (ED). Little is known about the prevalence of ED-based adverse SDOH screening and response practices nationally.Objective: To describe the prevalence of ED-based adverse SDOH screening and response policies and to identify associated hospital characteristics.Design, Setting, and Participants: This survey study utilized a 5% random sample from the National Emergency Department Inventory–USA, including EDs stratified by geography, urbanicity, and practice setting (academic vs community). Data regarding 2022 policies were collected in 2023.Exposures: Practice setting, urbanicity, visit volume, and availability of social work.Main Outcomes and Measures: The presence of written policies for any adverse SDOH (housing, food, transportation, and utility payment difficulties) screening and responses, as well as other requirement-driven screening for SDOH risk factors (intimate partner violence, substance use, and mental health conditions). Responses were categorized as consultations (eg, social work), standardized information sheets, individualized resource information, or other.Results: Of a total of 280 EDs, 232 responded (83% response rate). Among 232 EDs, 28.4% (survey-weighted proportion; 95% CI, 21.0%-37.2%) had screening policies for at least 1 adverse SDOH domain, and 93.1% (95% CI, 89.2%-95.7%) performed at least 1 other requirement-driven screening (eg, intimate partner violence). Of EDs performing any screening (adverse SDOH or other), 81.6% (95% CI, 73.4%-87.7%) had response policies, primarily involving consultations (78.2%; 95% CI, 67.2%-86.3%), standardized information sheets (43.0%; 95% CI, 32.5%-54.3%), and individualized resource information (12.9%; 95% CI, 7.2%-21.8%). Among all responding EDs, only 23.4% (95% CI, 17.1%-31.2%) had around-the-clock social work availability, and 20.5% (95% CI, 14.2%-28.6%) had an ED-based social worker. There was no association between practice setting, urbanicity, visit volume, or around-the-clock social work with adverse SDOH screening or response policies.Conclusions and Relevance: Despite the high prevalence of adverse SDOH in ED populations, in this survey study of 232 EDs, less than one-third performed screening, and one-fifth did not have policies requiring a response to positive screens. Bridging this gap may require expanding adverse SDOH screening practices while also ensuring that EDs have the resources and infrastructure to respond appropriately to identified social needs. Future research might explore advanced technological solutions to enhance screening and responses in these resource-constrained settings.
- 1 April 2025
- Published by American Medical Association (AMA) in JAMA Network Open
- Vol. 8 (4) , e257951
Open AccessScreening and Response for Adverse Social Determinants of Health in US Emergency Departments
Importance: Regulatory agencies have begun incentivizing screening for adverse social determinants of health (SDOH) and responses in inpatient settings, missing a crucial safety net: the emergency department (ED). Little is known about the prevalence of ED-based adverse SDOH screening and response practices nationally.Objective: To describe the prevalence of ED-based adverse SDOH screening and response policies and to identify associated hospital characteristics.Design, Setting, and Participants: This survey study utilized a 5% random sample from the National Emergency Department Inventory–USA, including EDs stratified by geography, urbanicity, and practice setting (academic vs community). Data regarding 2022 policies were collected in 2023.Exposures: Practice setting, urbanicity, visit volume, and availability of social work.Main Outcomes and Measures: The presence of written policies for any adverse SDOH (housing, food, transportation, and utility payment difficulties) screening and responses, as well as other requirement-driven screening for SDOH risk factors (intimate partner violence, substance use, and mental health conditions). Responses were categorized as consultations (eg, social work), standardized information sheets, individualized resource information, or other.Results: Of a total of 280 EDs, 232 responded (83% response rate). Among 232 EDs, 28.4% (survey-weighted proportion; 95% CI, 21.0%-37.2%) had screening policies for at least 1 adverse SDOH domain, and 93.1% (95% CI, 89.2%-95.7%) performed at least 1 other requirement-driven screening (eg, intimate partner violence). Of EDs performing any screening (adverse SDOH or other), 81.6% (95% CI, 73.4%-87.7%) had response policies, primarily involving consultations (78.2%; 95% CI, 67.2%-86.3%), standardized information sheets (43.0%; 95% CI, 32.5%-54.3%), and individualized resource information (12.9%; 95% CI, 7.2%-21.8%). Among all responding EDs, only 23.4% (95% CI, 17.1%-31.2%) had around-the-clock social work availability, and 20.5% (95% CI, 14.2%-28.6%) had an ED-based social worker. There was no association between practice setting, urbanicity, visit volume, or around-the-clock social work with adverse SDOH screening or response policies.Conclusions and Relevance: Despite the high prevalence of adverse SDOH in ED populations, in this survey study of 232 EDs, less than one-third performed screening, and one-fifth did not have policies requiring a response to positive screens. Bridging this gap may require expanding adverse SDOH screening practices while also ensuring that EDs have the resources and infrastructure to respond appropriately to identified social needs. Future research might explore advanced technological solutions to enhance screening and responses in these resource-constrained settings.