Leveraging VA geriatric emergency department accreditation to improve elder abuse detection in older Veterans using a standardized tool

Abstract

Elder abuse (EA) is common and has devastating health impacts, yet most cases go undetected limiting opportunities to intervene. Older Veterans receiving care in the Veterans Health Administration (VHA) represent a high-risk population for EA. VHA emergency department (ED) visits provide a unique opportunity to identify EA, as assessment for acute injury or illness may be the only time isolated older Veterans leave their home, but most VHA EDs do not have standardized EA assessment protocols. To address this, we assembled an interdisciplinary team of VHA social workers, physicians, nurses, intermediate care technicians (ICTs; former military medics and corpsmen who often conduct screenings in VHA EDs) and both VHA and non-VHA EA subject matter experts to adapt the Elder Mistreatment Screening and Response Tool (EM-SART) to pilot in the Louis Stokes Cleveland VA Medical Center geriatric ED (GED) program. The cornerstone of their approach is an interdisciplinary GED consultation led by ICTs and nurses who screen high-risk older Veterans for geriatric syndromes and unmet needs. The adapted EM-SART was integrated into the electronic health record and GED workflow in December 2020. By July 2022, a total of 251 Veterans were screened with nine (3.6%) positive on the prescreen and five (2%) positive on the comprehensive screen. Based on the first-year pilot experience, the interdisciplinary team was expanded and convened regularly to further adapt the EM-SART for wider use in VHA, including embedding flexibility for both licensed and nonlicensed clinicians to complete the screening tool and tailoring response options to be specific to VHA policy and resources. The national momentum for VHA EDs to improve care for older Veterans and secure GED accreditation offers unique opportunities to embed this evidence-based approach to EA assessment in the largest integrated health system in the United States.

INTRODUCTION

An estimated 10% of community-dwelling older adults experience abuse, neglect, and/or exploitation at the hands of a trusted person annually in the United States.^1,2 This elder abuse (EA) leads to myriad negative health impacts, including early mortality,^3–5 depression,⁶ emergency department (ED) visits,⁷ hospitalization,⁸ and nursing home placement.⁹ Despite the high prevalence and devastating health consequences of EA, studies suggest that <5% of cases are identified and referred to the appropriate authorities,¹⁰ limiting opportunities to intervene and improve health, safety, and well-being for affected older adults. Thus, improving detection and response to EA is an important public health priority.

While there have been no formal EA prevalence studies in the Veterans Health Administration (VHA), older Veterans receiving care in VHA have a high prevalence of health-related social needs and a higher prevalence than the similarly aged civilian population of most known EA risk factors, including cognitive impairment, functional impairment, and social isolation.¹¹ A VHA national directive originally implemented in 2012 and subsequently updated in 2017¹² stipulates that VHA employees (who, as federal employees, are typically not bound by state laws when performing official work duties) must follow their state laws for reporting cases of suspected EA and promptly document in the Veteran’s medical record the report, examination, and subsequent treatment offered. This directive also requires that a referral be made to a local VHA social worker following a mandated report. Additionally, the VHA is uniquely well-equipped to respond to EA detected in its health system. A highly trained and integrated social work workforce provides comprehensive support to Veterans focused on evaluation of key social determinants of health domains, including access to care, finances, housing, psychological status, cognitive status, and social support. VHA has high-quality geriatric programs across the spectrum of care, robust mental health programs, a national caregiver support program,¹³ and experience with developing a successful national program to address intimate partner violence, which also serves to respond to EA cases between intimate partners in later life.¹⁴ As the largest integrated health care system in the United States, the VHA, which serves a high-risk population for EA, is an optimal environment to develop, test, and disseminate innovative programs for EA.

VHA ED visits provide an important opportunity to identify adults suspected of experiencing EA and to initiate interventions that may mitigate and prevent future harm. Medical assessment in a VHA ED for acute injury or illness may be the only time isolated older Veterans leave their home.^7,15 Currently, however, most VHA EDs do not have standardized protocols to assess for EA and many do not have 24/7 immediate access to a social worker.

In 2018, the American College of Emergency Physicians created the geriatric ED (GED) accreditation program to recognize hospitals for higher standards of geriatric care.¹⁶ GED accreditation is granted at three different levels; to achieve the highest two levels, EDs must demonstrate integration of either 10 (Level 2) or 20 (Level 1) of 27 possible policies, guidelines, or procedures from the ED model of care for older adult patients,¹⁷ of which one is a “protocol for identification of elder abuse with appropriate follow-up.” With over half the nation’s 19 million Veterans aged ≥ 60 years¹⁸ and this population receiving high rates of ED care,¹⁹ VHA has long recognized the importance of achieving excellence in ED care for older Veterans. As such, VHA has been a committed early adopter of the GED accreditation program, focusing on a nationwide strategy to ensure that their EDs meet guidelines.²⁰ VHA now has 43 medical centers with some level of GED accreditation. The ability to integrate EA detection and intervention practices into the ED workflow to help achieve GED accreditation has provided a compelling opportunity for VHA EDs to identify and adopt evidence-based practices.

The Elder Mistreatment Screening and Response Tool (EM-SART)²¹ was developed in 2019 by the National Collaboratory to Address Elder Mistreatment (NCAEM), a group of leading national EA experts convened by the nonprofit Education Development Center (EDC) to design, test, and implement a model for more effective EA detection and response in the health care setting. EM-SART is based on an EA screening tool validated in the ED setting²² but modified to increase time efficiency and feasibility for use in busy EDs.²¹ It includes a brief prescreen intended for all patients followed by a longer comprehensive screen for those with positive responses on the prescreen. It also includes prompted older adult and caregiver observations derived from the DETECT screening tool, validated to detect EA in the field by emergency responders,²³ as well as physical examination components and suggested response pathways following assessment. In a multicenter validation study, the original EA screening tool on which EM-SART is based, the ED Senior AID Tool, had a sensitivity of 94.1% and specificity of 84.3% for identifying EA cases confirmed by multidisciplinary expert panel review and the approach was found to be acceptable to ED providers.²²

As a complex biopsychosocial problem, EA detection may best be addressed by an interprofessional stakeholder team of frontline clinicians and researchers. High-quality ED care is also by its nature interdisciplinary. Our objective was to bring together a team to adapt and pilot an evidence-based EA screening tool for use in VHA. In this article, we describe: (1) the experience of conducting a single-site pilot quality improvement initiative to integrate EA screening into the VHA ED workflow using the EM-SART and (2) the subsequent convening of a highly interdisciplinary stakeholder team to adapt the EM-SART to the VHA system for broader dissemination.

PILOTING EM-SART AT A HIGH-FUNCTIONING VHA GERIATRIC ED

The Louis Stokes Cleveland VA Medical Center (LSCVAMC) has received Level 1 GED accreditation, recognizing their dedication to leadership in providing optimal care to older adults. The cornerstone of their approach is a comprehensive GED consultation, which screens high-risk older Veterans for geriatric syndromes (e.g., cognitive impairment, polypharmacy, falls) and unmet social needs (e.g., food insecurity, lacking transportation).^19,24 Patients are identified as high risk if they are age 65 years or older and have an Identifying Seniors at Risk (ISAR)²⁵ score of >2 or a positive delirium triage screen²⁶ or for whom staff have a concern for unmet needs or caregiver stress.¹⁹ These patients receive a flag on the ED tracking board next to their status so that all staff are aware of their elevated risk. The interdisciplinary GED consultation is led by an intermediate care technician (ICT; former military medics and corpsmen who often conduct screenings in VHA EDs)²⁷ or nurse screener and may include a comprehensive evaluation by social work. The consult team provides care coordination, connections to VHA, and community services and assists with health care navigation. GED consultation at LSCVAMC is available weekdays from 8 a.m. to 5 p.m. as well as occasional evening hours.

Early in 2020, the LSCVAMC GED leadership team recognized the need to identify an EA screening tool developed specifically for use in the ED as staff reported that the current tool did not meet their needs, citing lack of a standardized way to document observational EA red flags as one major concern (Figure 1). Together with a VHA geriatrician/EA researcher (LKM) and a VHA social work leader involved with VHA’s national GED efforts and National Social Work Program EA response (JJH), the LSCVAMC GED leadership team (JKB and CMM) partnered with the NCAEM to adapt and pilot the EM-SART tool as part of an already ongoing multisite pragmatic implementation study occurring at five non-VHA ED sites. We met monthly over 6 months to adapt the EM-SART tool to the VHA electronic health record (EHR) and receive regulatory approvals to pilot the tool in the LSCVAMC GED program (Figure 1). The modified EM-SART was integrated into the GED workflow under a quality improvement initiative and added to the local EHR in December 2020 with a plan for a 9-month pilot period. After the completion of the initial pilot period, LSCVAMC operational quality review procedures recommended transition to permanent incorporation of the tool into the comprehensive GED process. This decision was based on end-user feedback, ability of the screen to identify safety concerns related to unmet care needs, and improved communication between the screeners and social workers and ED providers on identified EA concerns. Therefore, we present here data from the initial pilot period plus the subsequent year of experience with the EM-SART at the LSCVAMC ED. The reporting of this quality improvement project follows SQUIRE guidelines.²⁸

FIGURE 1.

Interdisciplinary team process for EM-SART pilot, adaptation, and planned dissemination. EA, elder abuse; EM-SART, Elder Mistreatment Screening and Response Tool; GED, geriatric emergency department; ICT, intermediate care technician; VHA, Veterans Health Administration

GERIATRIC EMERGENCY PROGRAM EXPERIENCE AND RESULTS

From December 2020 through July 2022, a total of 8431 unique patients age ≥ 65 made a total of 18,542 visits to the LSCVAMC ED. Of these encounters, GED consultations were ordered for 2779 (15% of encounters, 1994 unique patients) and completed for 923 (33% of consults ordered, 836 unique patients). The observed difference between GED consults ordered and completed was due in part to staffing and workflow changes related to the COVID-19 pandemic, leading the GED screening program to be intermittently suspended or abbreviated.

During the 21-month period analyzed, 251 unique Veterans were screened with the EM-SART in 261 encounters (28% of encounters with a completed GED consultation). Of the 836 patients who had completed GED consultations, 459 (54.9%) were also seen by an ED social worker during their encounter. When a social worker was present on duty in the ED, he or she might comprehensively assess the patient prior to the GED team performing their consult, as the high-risk older adult flag was visible to the entire ED provider team. The fact that many Veterans received GED consultations but not screening with the EM-SART is partially explained by the workflow established. ED social work evaluations with older adults included an assessment for EA, so the GED team would not perform the EM-SART if the Veteran had already seen ED social work that day.

Among Veterans screened, the average age was 76, with 95% men, 39% Black, 2% Hispanic, and 18% with abnormal cognitive screens (Table 1). Nine (3.6%) had positive responses on the prescreen. Of the nine Veterans with a positive prescreen, five (2% of total, 56% of positive prescreens) screened positive on the comprehensive screen. These positive screening rates are very similar to those found in a multicenter validation study of the ED Senior AID tool upon which the EM-SART is partially based.²² Details of the nine Veterans who had a positive prescreen are presented in Table 2. Presenting chief concerns varied and included falls, shortness of breath, pain in various locations, and nonspecific complaints. EA concerns most commonly involved financial exploitation and physical abuse, but also included neglect and emotional abuse. Of the five Veterans who screened positive on the comprehensive screen, all were admitted to the hospital except one who eloped. In Ohio, Adult Protective Services (APS) will not open a case on a community living older adult when the plan is to discharge him or her to a care facility if the change in living situation will address the source of abuse. APS may therefore request that reports for hospitalized adults are made closer to the time of discharge. For this reason, not all cases that were admitted to the hospital from the ED were reported to APS directly from the ED. In these cases, mistreatment concerns are communicated to the inpatient providers, and the GED team monitors the patient’s hospital course and disposition plan to ensure that a report is made, if appropriate. Nearly all older Veterans with a positive prescreen also screened positive for other geriatric syndromes and unmet needs (Table 2).

TABLE 1.

Characteristics of Veterans screened for EA using the EM-SART tool from December 2020 through July 2022

	Overall (N = 251)
Age (years)	76.25 (±7.60)
Male gender	239 (95.2)
Race
Black	99 (39.4)
White	134 (53.4)
Other race	18 (7.2)
Ethnicity
Hispanic or Latino	6 (2.4)
Not Hispanic or Latino	228 (90.8)
Missing	17 (6.8)
ISAR score
0	2 (0.8)
1	16 (6.4)
2	23 (9.2)
3	109 (43.4)
4	52 (20.7)
5	9 (3.6)
6	1 (0.4)
Missing	39 (15.5)
ISAR score	3.06 (±0.98)
AMT-4 screen positive	46 (18.3)
EM-SART prescreen positive	9 (3.6)
EM-SART comprehensive screen positive	5 (2.0)
Disposition
Admitted to ICU	2 (0.8)
Admitted to telemetry	22 (8.8)
Admitted to ward	88 (35.1)
Left against medical advice	1 (0.4)
Home	134 (53.4)
Psychiatric observation	1 (0.4)
Sent to specialty clinic	2 (0.8)
Transferred to non-VA facility	1 (0.4)

Note: Data are reported as mean (±SD) or n (%).

Abbreviations: AMT-4, Abbreviated Mental Test 4; EA, elder abuse; EM-SART, Elder Mistreatment Screening and Response Tool; ICU, intensive care unit; ISAR, Identifying Seniors at Risk.

TABLE 2.

Characteristics of patients who screened positive on EM-SART prescreen (n = 9)

Age (years)	Presenting chief concern	Comprehensive screen positive	Report made to APS	Seen by social work while in ED	Type of potential abuse	Recommended interventions	ED disposition	Other positive ED geriatric screens
70	Chest pain, shortness of breath	Yes	No	Yes	Financial and physical	Capacity evaluation	Admitted	None
64	Shoulder pain	Yes	Yes	No	Neglect	Home safety evaluation, social work	Eloped	Falls
87	Shortness of breath	Yes	Yes	No	Financial	Capacity evaluation	Admitted	ADLs, IADLs, falls, polypharmacy
68	Foot color change	Yes	No	Yes	Physical	Physical and occupational therapy, cognitive testing, home meal delivery	Admitted	ADLs, IADLs, falls and polypharmacy
77	Social needs	Yes	No	No	Financial	None	Admitted	Impaired cognition, polypharmacy, ADLs, IADLs, food insecurity
64	Wound check	No	No	Yes	Neglect	Home safety evaluation, primary care follow-up, social work, skilled wound care	Home	Polypharmacy, falls, ADLs, food insecurity
79	Fall	No	No	Yes	Physical	Geriatrics comprehensive evaluation, physical/occupational therapy, extended care facility placement	Admitted	Polypharmacy, impaired cognition, falls, ADLs, transportation
74	Memory concern	No	No	No	Physical	Social work follow-up	Admitted	ADLs, IADLs, falls
68	Pain	No	No	Yes	Emotional	None	Home	Polypharmacy

Abbreviations: ADLs, activities of daily living; EM-SART, Elder Mistreatment Screening and Response Tool; IADLs, instrumental activities of daily living.

Over the study period, 10 Veterans received the screen twice; in three cases the results changed, with two becoming negative and one becoming positive. In one case where the screen was initially positive and then subsequently negative, 9 months had elapsed between screens during which time the Veteran had been enrolled in an intensive transitional care program. In the other case, the EM-SART screen was initially positive due to concerns for financial exploitation in the setting of cognitive impairment. When the Veteran was screened again 8 months later, he had obtained a fiduciary to aid in management of his finances and was no longer felt to be at risk. In the one case that changed from negative to positive, the Veteran was a nonverbal patient with significant cognitive impairment who was unable to answer screening questions. During the first ED evaluation, the caregiver was not present and so no observational red flags were triggered to indicate a positive screen. Five months later, the Veteran re-presented with his caregiver and several concerning observations of the caregiver by the EM-SART screening triggered observational red flags that indicated a positive screen.

As part of the Level 1 GED program, all processes related to accreditation, including EA screening, are required to be in a continuous quality review cycle. As part of this quality review, ED team members were queried on their experience with the EM-SART. The GED screeners (n = 4) reported that the care model and EM-SART were easy to use and valuable to identify older Veterans who would benefit from services and intervention. ED providers appreciated the support with EA identification and noted that this structured approach was very helpful, particularly when a social worker was not in the ED to assist with assessment. Even when an ED social worker was on duty, in cases where the GED team performed the EM-SART prior to social work evaluation, positive prescreens allowed the team to communicate their elevated concerns to the ED social worker. This helped the social worker prioritize evaluating these patients in depth. The ED staff and GED team also noted they were able to identify cases of EA without having to perform the EM-SART secondary to receiving additional education during the screen implementation, especially in cases of financial mistreatment. This suggests that the process of implementing the EM-SART, which involves staff and provider training, may increase the detection of EA beyond those formally screened. The team reported that the EM-SART was helpful in cases where obvious abuse was not occurring but was unnecessary when abuse was apparent/obvious based on initial physical examination or history taken during the ED course.

As the NCAEM EM-SART multisite implementation study closed, the VHA and NCAEM convened an interdisciplinary team with the goal to review the pilot experience and develop a pathway for widespread adoption of the EM-SART within the VHA.

EXPANDING THE TEAM AND PREPARING FOR BROADER DISSEMINATION

Real-world use of the EM-SART tool during the pilot at the LSCVAMC ED revealed several issues with the structure, language, and flow of the tool that required optimization to (1) enhance useability for the unique providers, clinical environment, and policy structure of the VHA health system and (2) embed flexibility in the tool to accommodate the variability in ED staffing and workflow of different VHA sites. Recognizing that it was critical to integrate the perspective of each type of frontline clinical provider that might be utilizing the tool, the pilot team (Team A, Figure 1) was expanded to include ICTs, nurses, and social workers (Team B, Figure 1). This interprofessional team of frontline VHA health care providers collaborated with VHA and non-VHA EA experts and researchers biweekly over 4 months first to identify key issues and then to iteratively modify the EM-SART template to address them. Table 3 outlines both VHA-specific challenges and other challenges that were identified and key adaptations that were made by the team. For example, to allow for VHA health care professionals who do not perform certain assessments (e.g., social workers and psychologists do not perform physical examinations) to complete the EM-SART while still ensuring collection of as complete data as possible, flexibilities were built in to allow skipping certain sections while still being prompted to complete all sections possible within the confines of the screener’s scope of practice. The response section of the EM-SART prompting specific follow-up for EA was also adapted to include many of the health-related services available within VHA that may not be available in other health care settings (e.g., intimate partner violence assistance program referral, referral for capacity assessment by mental health).

TABLE 3.

Key adaptations made to EM-SART for VHA implementation

Challenges	Key adaptations
VHA-specific challenges
Embed flexibility for different types of licensed and nonlicensed clinicians to complete	• Provided examples and changed language to allow ICTs to complete most portions of the EM-SART without needing to make clinical judgments • Option to opt-out of physical examination portion • Changed language prompting final determination of EA suspicion to: “Based on the totality of information gathered in completed portions of this screen” • Added fourth final determination option that allows getting other interdisciplinary providers involved: “Concern is present, additional information/evaluation needed prior to determining suspicion of EA”
Response tailored to VHA policy and resources	• Prompt to make an EA report to appropriate authorities accompanied by VHA specific policies to guide user • Prompt to document mandatory EA report in appropriate VHA specific note template for standardization • VHA specific resource referral list
Remove redundancies to align with VHA-ED workflow	• Removed initial cognitive assessment (AMT4) prior to triggered screen as this is a separate assessment in most VHA GED programs • Removed repeated screener observations from triggered screen
Other challenges
Completion of observation portions of EM-SART for patients who cannot answer questions	• Added question to beginning of screener: “Is the patient able to answer questions?” with answer options: ∘ Yes [If selected, goes to “prescreen”] ∘ No (e.g., nonverbal, too ill, despite attempts unable to get patient alone) [If selected, skip to “screener observations”]
Documentation of relevant physical examination components for all patients getting triggered screen	• Prompt for EA-specific physical examination component for all providers that can perform physical examinations and have not already done so, regardless of confidence in patient’s ability to report
Increase ease of completion and reliability of data	• Removed “no” option for each physical exam component and changed to encourage affirmative responses only for those items present/observed

Abbreviations: AMT-4, Abbreviated Mental Test 4; EM-SART, Elder Mistreatment Screening and Response Tool; GED, geriatric emergency department; ICT, intermediate care technician; VHA, Veterans Health Administration.

Other challenges were described by the Team B frontline providers when using the EM-SART tool that were not specific to the VHA system, as might be anticipated for a recently developed tool without a long track record of real-world clinical ED use. For example, even though the EM-SART comprises prompted observations and physical examination components that may provide important information in addition to the question responses, the original EM-SART did not include a prompt to remind clinicians to complete these portions for patients who cannot answer questions. Team B felt that this missed an opportunity for GED team members to gather important information on potential EA that did not rely on patient responses. To address this, we added questions and guidance to the adapted EM-SART that explicitly prompt completion of observation and exam components for older Veterans who cannot answer the screening questions.

At each step in the template review process, all members of the interprofessional team were given opportunities to provide feedback and propose changes. The team met to discuss changes together and did not proceed until consensus on a path forward for each issue was reached. After the entirety of the EM-SART content was reviewed and needed modifications were made, the new version was circulated to all members of Team B for final review and approval.

NEXT STEPS

Based on the LSCVAMC experience and access to the EHR template, six additional early adopter VHA EDs around the country began participation in a pilot program that involves using the EM-SART²⁹ within their GED programs between May and October 2021. These programs are located in Salt Lake City, UT; San Diego, CA; Palo Alto, CA; Durham, NC; Dallas, TX; and Grand Junction, CO, with each serving wide catchment areas of both urban and rural patients. These sites vary in their level of complexity, resources, personnel, and workflows, highlighting the adaptability and flexibility of the EM-SART model, which was designed for use in any ED setting, not only those with GED accreditation. Presence of the EM-SART template in the VHA EHR allows integration of this tool into VHA ED workflows regardless of GED accreditation. Screening for EA using the EM-SART can be modified to meet the needs of each local ED setting; how the cases detected are addressed and who the handoff goes to can be tailored based on individual need and staffing. In addition, updated versions of the EM-SART tool along with relevant trainings and other resources are publicly available in a toolkit on the GED collaborative website for any site interested in implementing this tool locally.³⁰

While the LSCVAMC and these six early adopter VHA medical centers continue to use the original EM-SART in their GED programs, our team has solicited broad-based stakeholder feedback from national VHA program leaders on the adapted EM-SART tool and has recently completed collaboratively integrating this feedback (Figure 1). While the EM-SART was originally created for use in the ED and the process we undertook was to integrate it into the VHA ED workflow, our team recognized that a tool found to be time efficient and useful in the ED would likely have value in other VHA clinical settings as well. In the absence of uniform recommendations for EA screening,³¹ wide variation exists in practice in other VHA clinical locations. As a result, our team also recognized the potential of integrating EM-SART into standard clinical assessment in other VHA clinical settings in the future. Therefore, in addition to leaders from the national VHA emergency medicine office, we solicited national stakeholder feedback from leaders in primary care, hospital medicine, social work, nursing, home and community-based services, mental health, and geriatrics.

Following finalization and approval, this VHA-adapted version of EM-SART will be built into a national note template for the VHA EHR that will allow data collection and will be tested in a multisite pilot to assess feasibility and impact. It will also be made available to VHA GED programs nationally planning to integrate EA screening into their programs and may be particularly useful for sites seeking Level 1 or 2 GED accreditation. As the EM-SART is disseminated in VHA to address the pressing current need to assess older Veterans for EA, it will also be important to track meaningful downstream outcomes to examine the impact of the screening program among Veterans who are screened compared to those who are not. These efforts will allow VHA to contribute to the growing body of knowledge on optimizing EA assessment in the ED setting more broadly.

CONCLUSIONS

In conclusion, as more EDs both within and external to Veterans Health Administration seek to improve the care they provide to older adult patients and apply for geriatric ED accreditation, they will need to integrate new processes and procedures to best serve the needs of older ED patients. This multistep implementation process involving a large and invested interdisciplinary team may serve as a model for adapting evidence-based practices to local resources, personnel, and workflows to standardize sociomedical care in Veterans Health Administration EDs and other clinical settings. For the important and understudied problem of elder abuse, the geriatric ED accreditation movement presents critical opportunities for health systems to implement detection and response strategies, collect data, evaluate experiences, and learn from the field to continually optimize the approach to providing high-quality care for older ED patients.