Predictive Modeling for Social Needs in Emergency Department Settings

Protocol for Evaluating the Effectiveness of a Clinical Decision Support System With Prediction Modeling to Identify Patients With Health-related Social Needs in the Emergency Department

The overall objective of this study is to support emergency department management of patients' health-related social needs. This study will measure the impact of a decision support system that informs clinicians about which patients are likely to screen positive for a health-related social need. The system uses statistical models to create a health-related social need risk score for each patient. The main questions, the study aims to answer are:

• Does providing emergency department clinicians with risk scores on health-related social needs increase screening and referral activities?
• Does providing emergency department clinicians with risk scores on health-related social needs change patients' use of healthcare services?

The decision support system with health-related social needs risk scores will be introduced for all adult patients at one emergency department. Screening rates, referrals, and subsequent healthcare encounters will be compared with emergency departments that did not have access to the decision support system.

Study Overview

• Status: Completed
• Conditions: Social Determinants of Health; Emergency Service, Hospital
• Intervention / Treatment: Other: Health-related social needs decision support system

• Study Type: Interventional
• Enrollment (Actual): 518512
• Phase: Not Applicable

Contacts and Locations

Study Locations

• United States
  • Indiana
    • Indianapolis, Indiana, United States, 46202
      • Indiana University Health

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: Yes

Description

Inclusion Criteria:

• Adults (>18 years old)
• Seeking care at Indianapolis, Indiana area emergency departments (EDs).

Exclusion Criteria:

• Children
• Encounters by patients that present with a critical illness/injury (e.g. severe trauma patients or those with Emergency Severity Index (ESI) classification level 1)
• Encounters by patients who have been transferred from another inpatient facility
• Patients that die during the ED encounter
• Encounters among patients who were ultimately admitted during their ED visits from our analysis

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Health Services Research
• Allocation: Non-Randomized
• Interventional Model: Parallel Assignment
• Masking: None (Open Label)

Number of Arms

2

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
Experimental: Decision support intervention group; Adult ED patients seeking care the ED site with the health-related social needs decision support system live.	Other: Health-related social needs decision support system; The clinical decision support intervention will present emergency department clinicians at an Indianapolis, IN ED with a likelihood score for an adult patient screening positive for the following health-related social needs (HRSNs): housing instability, food insecurity, transportation barriers, financial strain, and history of legal involvement. For each HRSN, the likelihood of screening positive is reported as "high", "medium", or "low". These categorizations are the product of logistic regression models. The clinical decision support intervention will be delivered through an existing FHIR (Fast Healthcare Interoperability Resources) standards-based clinical decision support platform.
No Intervention: Comparison group; Adult ED patients created using statistical matching from ED sites in the same metropolitan area.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Percent of emergency department encounters screened for health-related social needs (HRSNs)	The numerator will be an emergency department encounter with any indication of HRSN screening using any tool or questionnaire, regardless of patient completion or results. The denominator will be all eligible ED encounters.	At time of emergency department encounter (or within 24 hours)
Percent of emergency department encounters that were referred for health-related social needs (HRSNs) services	The numerator will be emergency department encounters with a referral to social worker, case management, community health workers, or related services within 24 hours of the ED encounter. The denominator will be all eligible ED encounters	At time of emergency department encounter (or within 24 hours)

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Percent of encounters with an emergency department revisit measured at 3 days	The numerator will be an emergency department encounter at any facility included in the Indiana for Network Care database within 3 days of an ED encounter at an intervention or comparator site. ED revisits may serve as the index visit for subsequent revisits. The denominator will be all eligible ED encounters. Encounters resulting in an inpatient admission will be excluded from the numerator and denominator.	within 3 days of emergency department encounter
Percent of encounters with an emergency department revisit measured at 7 days	The numerator will be an emergency department encounter at any facility included in the Indiana for Network Care database within 7 days of an ED encounter at an intervention or comparator site. ED revisits may serve as the index visit for subsequent revisits. The denominator will be all eligible ED encounters. Encounters resulting in an inpatient admission will be excluded from the numerator and denominator.	within 7 days of emergency department encounter
Percent of encounters with an emergency department revisit measured at 30 days	The numerator will be an emergency department encounter at any facility included in the Indiana for Network Care database within 30 days of an ED encounter at an intervention or comparator site. ED revisits may serve as the index visit for subsequent revisits. The denominator will be all eligible ED encounters. Encounters resulting in an inpatient admission will be excluded from the numerator and denominator	within 30 days of emergency department encounter
Percent of emergency department encounters with primary care visit within 7 days of an ED encounter	The numerator will include all emergency department encounters with a completed family medicine, internal medicine, OBGYN, or geriatrician visit64 within 7 days of the ED visit. The denominator will be all eligible ED encounters. Encounters resulting in an inpatient admission will be excluded from the numerator and denominator.	within 7 days of emergency department visit

Other Outcome Measures

Outcome Measure	Measure Description	Time Frame
Percent of emergency department encounters where the health-related social needs (HRSN) decision support system intervention was accessed.	The numerator will include encounters with access of the social needs section (containing the risk prediction scores) of the decision support system intervention during the study visit (defined as within 24 hours). Access will be defined as record of the end user visiting the HRSN page in the system user logs. The denominator will be all eligible ED encounters. Limited to the intervention site only.	At time of emergency department encounter (or within 24 hours)
Percent of emergency department encounters where the clinical decision support platform was accessed	The numerator will include encounters with access of clinical decision support platform during the study visit (defined as within 24 hours). Access will be defined as record of the end user visiting initiating a request to the decision support application from the EHR. Any portion of the decision support platform (not just the health-related social needs section) is included. The denominator will be all eligible ED encounters (see Inclusion criteria, above). Limited to the intervention site only	At time of emergency department encounter(or within 24 hours)

Collaborators and Investigators

• Sponsor: Indiana University
• Investigators: Principal Investigator: Joshua R Vest, PhD,MPH, Indiana University

Publications and helpful links

General Publications

• Mazurenko O, Hirsh AT, Harle CA, McNamee C, Vest JR. Acceptance of Automated Social Risk Scoring in the Emergency Department: Clinician, Staff, and Patient Perspectives. West J Emerg Med. 2024 Jul;25(4):614-623. doi: 10.5811/westjem.18577.
• Mazurenko O, Harle CA, Blackburn J, Menachemi N, Hirsh A, Grannis S, Boustani M, Musey PI Jr, Schleyer TK, Sanner LM, Vest JR. Effectiveness of a clinical decision support system with prediction modeling to identify patients with health-related social needs in the emergency department: Study protocol. PLoS One. 2025 May 12;20(5):e0323094. doi: 10.1371/journal.pone.0323094. eCollection 2025.

Study record dates

Study Major Dates

• Study Start (Actual): March 10, 2025
• Primary Completion (Actual): December 1, 2025
• Study Completion (Actual): December 1, 2025

Study Registration Dates

• First Submitted: August 23, 2024
• First Submitted That Met QC Criteria: October 22, 2024
• First Posted (Actual): October 24, 2024

Study Record Updates

• Last Update Posted (Actual): March 31, 2026
• Last Update Submitted That Met QC Criteria: March 26, 2026
• Last Verified: March 1, 2026

More Information

Terms related to this study

• Keywords: informatics
• Additional Relevant MeSH Terms: Pathologic Processes; Disease Attributes; Pathological Conditions, Signs and Symptoms; Emergencies
• Other Study ID Numbers: 2011558232

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO
• IPD Plan Description: All data produced during the project will be preserved, but raw and derived data (at the patient-level) will not be posted publicly because of our use of secondary data from privately held electronic health record and health information exchange systems. Due to the data use restrictions put in place by consortium agreements among the health system partners that contribute EHR data to the Indiana Network for Patient Care, patient-level data cannot be shared or disseminated outside this project. However, de-identified derived data (at the patient-level) used in this study may be shared with investigators whose formal request is approved by the data owners. Requests can be sent to askRDS@regenstrief.org. Access to these data requires investigator support for use and a signed data access agreement between the Regenstrief Institute and the investigator's institution.

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No