Forensic markers of physical elder abuse in medical and community contexts: implications for criminal justice interventions

PDF

Download

Open document

Flip pages

Summarize

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

FORENSIC MARKERS OF PHYSICAL ELDER ABUSE IN MEDICAL AND COMMUNITY
CONTEXTS: IMPLICATIONS FOR CRIMINAL JUSTICE INTERVENTIONS

by

Jeanine Yonashiro-Cho

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(GERONTOLOGY)

December 2018

Copyright 2018 Jeanine Yonashiro-Cho

ii

TABLE OF CONTENTS

DEDICATION iii
ACKNOWLEDGEMENTS iv
NOTIFICATION viii
LIST OF TABLES AND FIGURES ix
ABSTRACT x
CHAPTER I: INTRODUCTION 1
CHAPTER II: CHARACTERISTICS OF INTENTIONAL INJURIES AMONG
OLDER ADULTS PRESENTING TO HOSPITAL EMERGENCY
DEPARTMENTS IN 2001-2010
16
CHAPTER III: CHARACTERISTICS OF INJURIES ASSOCIATED WITH
PHYSICAL ELDER ABUSE IN COMMUNITY-DWELLING OLDER
ADULTS RECEIVING SERVICES FROM ADULT PROTECTIVE
SERVICES
46
CHAPTER IV: TOWARD A BETTER UNDERSTANDING OF THE ELDER
ABUSE FORENSIC CENTER MODEL: COMPARING AND
CONTRASTING FOUR PROGRAMS IN CALIFORNIA
83
CHAPTER V: CONCLUSION 118
REFERENCES 125

iii

DEDICATION

To Him who makes all things possible –
Whose grace is sufficient, Whose Presence sustains, and Whose plans yield hope and purpose

iv

ACKNOWLEDGEMENTS
First, I would like to thank my family for their support, encouragement, patience, and
love on this journey. I am especially grateful for the love and support of my husband, Garrett,
who encouraged me to begin this adventure and has believed in me along the way while also
managing our household, caring for our kids, listening to me process my thoughts out loud, and
editing countless paper drafts. And to Charis and Josiah, thank you for bringing me so much joy,
love, and inspiration.
I would also like to thank my parents, Damon and Irene, for their encouragement,
support, prayers, babysitting, and early investment in my life and education; my parents-by-
marriage, Walter and Diana, for encouraging and enabling me to remain committed to both
career and family through babysitting, friendship, and Sunday dinners; my sisters, Candace,
Rachel, and Rebecca, who have been my cheerleaders and ad-hoc consultants; my grandmother,
Marcela Baxa, who taught me about life, gave me with a passion for learning, and impressed
upon me the continued value and contributions of older adults; and my Aunty Diane and Uncle
William who welcomed me to Los Angeles 17 years ago and have provided me with a home
away from home for all of these years.
For almost as many years, Dr. Kathleen Wilber has provided me with an academic home.
I am thankful for the support, knowledge, guidance, and wisdom Kate has shared with me –as a
professor, teaching my undergraduate-level classes; as an advisor, introducing me to health
services research and systems evaluation; as a mentor, training me in research and academia; and
now, as my dissertation chair, guiding me toward completing this course of research. Moreover, I
appreciate and respect her commitment to conducting research with integrity and kindness. This
dissertation would not be possible without her investments of time, energy, and resources.
v

I also deeply indebted to thank the other members of my dissertation committee, Dr.
Susan Enguídanos and Dr. Laura Mosqueda. Dr. Enguídanos has been a patient and supportive
mentor to me throughout the doctoral program, spending countless hours training me in
qualitative research methods and scholarly writing. She provided me with the opportunity to
prepare my first peer-reviewed manuscript as a first-author and guided me through the writing,
submission, and revision processes. In a similar manner, Dr. Mosqueda has been a supportive
and practical mentor. Her calls to consider the translational value and relevance of research has
challenged me to rethink my assumptions and reframe my work. The feedback, critiques, and
suggestions made by my committee significantly improved the quality, readability, and practical
relevance of this dissertation. I am thankful for the gracious and generous contributions of each
committee member.
Along this journey, I have been blessed to work and train alongside a phenomenal group
of emerging scholars. These include members of my lab group, past and present – Elizabeth
Avent, Marguerite DeLiema, Gerson Galdamez, Haley Gallo, Zachary Gassoumis, Melanie
Gironda, Kylie Meyer, Adria Navarro, Laura Rath, Julia Rowan, and Yongjie Yon – for your
support, encouragement, insights, big-picture ideas, feedback, critiques, and collaboration. I
especially wish to thank Dr. Zachary Gassoumis for his guidance, support, and partnership in
project management, data analysis, and interpretation.
This research would not have been possible without the guidance, support, and
partnership of my collaborators. I wish to thank Dr. Tony Rosen and Dr. Jeffrey Hall who shared
their expertise on geriatric assault injuries, working with the National Electronic Injury
Surveillance System (NEISS), and provided feedback on an earlier draft of Chapter II.
vi

I am also thankful for the mentorship and support of Dr. Diana Homeier, who first
introduced me to the field of elder mistreatment while working as her undergraduate intern. As
the Principal Investigator for the National Institute of Justice-funded study discussed in Chapter
III, Dr. Homeier has continued to serve as one of my mentors, providing guidance in conducting
elder mistreatment research and interpreting study findings. I also wish to acknowledge the
members of the Los Angeles County Adult Protective Services Program for their dedication to
improving the safety and well-being of older and dependent adults residing in Los Angeles
County. This research would have been impossible without the partnership and tireless work of
this agency, especially Arthur Zargarian, Dave Kochen, Rich Franco, Lorenza Sanchez, Cynthia
Banks, and the countless APS workers who strived to identify willing participants. The project’s
success also relied on the dedication of study research nurses (Amanda Murray and Jacqueline
Paige) and LEAD panelists (Dr. Dierdre Anglin, Dr. Janet Arnold, and Dr. Kidest Babi). I also
wish to acknowledge the generous support of the National Institute of Justice in funding this
study and providing federal leadership to support research on topics related to elder justice,
including mistreatment.
The work to study the Elder Abuse Forensic Center model among California’s first four
program replication sites would not have been possible without the participation and support of
Dr. Diana Homeier and Allyson Young of the Los Angeles County Elder Abuse Forensic Center;
Dr. Laura Mosqueda, Dr. Kerry Burnight, and Lynn Rodriguez of the Orange County Elder
Abuse Forensic Center; Anne Rios of the San Diego HOPE Team; and Dr. Erika Falk, Talitha
Guinn, and Herschell Larrick of the San Francisco Elder Abuse Forensic Center. I also wish to
acknowledge the Archstone Foundation for its Elder Abuse and Neglect Initiative and project
funding mechanisms, which provided support for the establishment of the four California Elder
vii

Abuse Forensic Centers, funded the comparative research presented in Chapter IV, and
encouraged further research in this area.
Finally, I wish to thank my mentors and colleagues at the University of Hawai‘i: Dr.
Kathryn Braun, a true mentor who demonstrated how translational research can transform lives
and whose passion for health equity, pursuit of public health and commitment to giving voice to
others through research inspired me to pursue my doctorate; Dr. Colette Browne, Dr. Noreen
Mokuau, and Dr. Joann Tsark who shared their knowledge, expertise, and experiences in
working with Hawai‘i’s local and indigenous populations and engaging in community-based
participatory research; and Dr. Christy Nishita, who has been a source of support and
encouragement.

viii

NOTIFICATION

A portion of this presentation was supported by Award No. 2013-IJ-CX-0025 awarded by the
National Institute of Justice, Office of Justice Programs, U.S. Department of Justice. The
opinions, findings, and conclusions or recommendations expressed in this presentation are those
of the authors and do not necessarily reflect those of the Department of Justice.

I also wish to thank the Archstone Foundation for their generous support through Archstone
Grant #11-02-24.

ix

LIST OF TABLES AND FIGURES
Figure 1.1 Detection and Response to Elder Mistreatment (EM) 7
Figure 2.1 Consort diagram of sample selection and case exclusion 20
Table 2.1 Patient Characteristics of Older Adult Patients Seeking Treatment for
Non-Fatal Injuries in the Emergency Department by Injury
Intentionality, 2001-2013, N=599,792
26
Table 2.2 Characteristics of Primary Injuries Among Older Adult Patients
Seeking Treatment for Non-Fatal Injuries in the Emergency
Department by Intent of Injury Cause, N=599,792
27
Table 2.3 Prevalence of Injury Diagnosis and Location Among Older Adult
Patients Seeking Treatment for Non-Fatal Injuries in the Emergency
Department by Injury Intent, N=599,792
30
Table 2.4 Conditions in which Assault was Perpetrated Against Older Adult
Patients Seeking Treatment for Non-Fatal Injuries in the Emergency
Department
35
Figure 3.1 Recruitment protocols guiding recruitment of physically-abused APS
clients
52
Figure 3.2 CONSORT diagrams of sample selection and case exclusion of (a)
subject level data and (b) injury level data
57
Table 3.1 Demographic characteristics of study sample by recruitment arm,
N=156
63
Table 3.2 Health characteristics of study sample by recruitment arm, N=156 64
Figure 3.3 Number of days between abusive incident and study data collection
among APS-recruited subjects (N=50)
66
Table 3.3 Prevalence of Diagnoses and Locations of Injuries Reported Among
Subjects with an Injury by Recruitment Source (N = 106)
67
Table 3.4 Prevalence and Characteristics of Ecchymoses Reported Among
Community-Dwelling Older Adults by Recruitment Source (N = 152)
70
Table 3.5 Prevalence and Characteristics of Abrasions Reported Among
Community-Dwelling Older Adults by Recruitment Source (N = 92)
73
Figure 4.1 Elder Abuse Forensic Center conceptual model 93
Table 4.1 Cross-site comparison of meeting characteristics 94
Table 4.2 Cross-site comparison of multi-disciplinary team composition 96
Table 4.3 Comparison of client and case characteristics by site 101
Figure 4.2 Cross-site comparison of modified Team Effectiveness Inventory
(N=47)
104
Table 4.4 Comparison between Forensic Center and Family Justice Center
Models
109
Table 4.5 Comparison of client and case characteristics by program model 111

x

ABSTRACT
Elder mistreatment is a societal public health issue affecting millions of older adults
worldwide (Yon, Mikton, Gassoumis & Wilber, 2017), including at least 10% of older
Americans each year (Acierno, et al., 2010). Despite its initial recognition as an existing problem
over 40 years ago, it remains understudied and under-addressed. Expert estimate that the field of
elder mistreatment is developmentally 20 years behind other family violence fields, such as child
maltreatment and intimate partner violence.
Though mechanisms of elder physical abuse appear similar to those occurring in child
maltreatment and intimate partner violence, its presentation, and resulting factors may be
distinct, arising from physiological and age-related differences between older adults and younger
victims. Compared to work in other family violence fields, there has been little research
conducted on forensic markers of physical elder abuse.
Identification of such markers hold promise for improving detection of abuse and
enabling earlier health and social services responses and interventions. If conducted at the
request of victims and with their cooperation, these steps may halt abusive circumstances and
prevent future abuse. In addition to improving healthcare and social services delivery, research
on markers of abuse can also strengthen criminal justice efforts to prosecute perpetrators or seek
civil penalties on behalf of victims.
The purpose of this dissertation research is to identify forensic markers of physical elder
abuse to improve detection and inform faster service delivery processes to improve care for older
adults.
Chapter I defines elder mistreatment and physical elder abuse, including both research
and practice-oriented definitions. Because this dissertation is focused on community-based
xi

research and approaches to detecting and intervening in cases of elder mistreatment, particularly
physical elder abuse, the broader definition of abuse used by practitioners in the State of
California was used to guide analyses. This definition does not require that abuse perpetrators
have a trusted relationship with the victim, and thus includes cases of geriatric assault
perpetrated by strangers.
Chapter I also describes research conducted in other fields of family violence, including
methodological approaches used to identify markers of abuse. Existing literature on injuries
associated with physical elder abuse discussed and key findings described. Finally, Chapter I
presents the rationale for this dissertation, describing the relationship between improved abuse
detection, clearer abuse determinations, and earlier intervention. The development and use of
multidisciplinary teams (MDTs) for elder mistreatment investigation and intervention are
explored and the Elder Abuse Forensic Center (FC) model is described.
Chapter II identifies and characterizes intentional, non-fatal injuries present among older
adults seeking care from hospital emergency departments. Data collected from 2000 to 2013
through the Centers for Disease Control and Prevention’s National Electronic Injury Surveillance
System (NEISS) is analyzed, providing a nationally-representative study sample of older adults
experiencing geriatric assault, including abuse. Victim, injury, and perpetrator characteristics are
described.
Chapter III documents abuse-related injuries found among victims of physical elder
abuse receiving services from Adult Protective Services (APS) workers. Client socio-
demographic, medical history, and injury data were compared with a comparison group of non-
abused older adults receiving care from a Geriatrics Outpatient clinic. Though most victims of
abuse receive services through APS, APS clients are seldom studied.
xii

This study builds on the work of Wiglesworth and colleagues (2009) which examined
and characterized bruising found among APS clients, comparing their bruises with those found
among non-abused older adults. The research presented in Chapter III uses a similar approach to
document the location, size, color, and other characteristics of several different injury diagnoses.
Injury types included less severe physical findings such as tenderness and swelling which have
not been documented in past research on markers of abuse.
Chapter VI builds on the work of Chapter II and III by examining a promising elder
abuse intervention model. The FC model has been replicated in several different sites
nationwide, including four in California. The study conducts a cross-site analysis comparing the
policies, procedures, activities, clients, and outcomes achieved by the four California FCs to
identify the program characteristics and processes central to FC operation and management.
Finally, Chapter V provides a concluding summary of the presented dissertation research.
Key findings drawn from across the chapters are discussed as well as study limitations, policy
implications and future directions.

CHAPTER I: INTRODUCTION

Elder mistreatment, the act of intentionally causing risk of harm to an older adult through
intentional action or failure to act, is an umbrella term used to discuss several distinct types of
mistreatment including physical abuse, caregiver neglect, financial abuse, sexual abuse, and
emotional abuse. At its heart, elder mistreatment is a “wicked problem,” a complex, poorly
defined, and impossible-to-solve societal problem that requires constant re-evaluation and re-
solving to achieve some measure of resolution (Rittel & Webber, 1973).
A growing body of research suggests that exposure to elder mistreatment is detrimental to
the long-term victim well-being. Exposure to elder mistreatment has been associated with
negative psychological health outcomes, such as anger or grief (Comijs, et al., 1998),
psychological distress (Comijs, et al., 1999), depression (Dong, et al., 2009), and Post-Traumatic
Stress Disorder (PTSD) symptomology, even if abusive incidents are not recent (Acierno, et al.,
2002).
Physical injury may also result of abusive actions particularly following perpetration of
physical abuse. Potential injuries are not limited to, but may include contusions, abrasions,
lacerations, punctures, fractures, bruising, burns, pain or tenderness (Dyer, Connolly & McFeely,
2003; Wiglesworth, et al., 2009; Ziminski, Wiglesworth, Austin, Phillips & Mosqueda, 2013).
While victims sustaining injury at any age experience harm, injuries sustained among older
adults may be especially detrimental, especially given the sequelae of medical complications that
may arise as the body responds to the injury (Homeier, 2014). Such injuries may contribute to

2

recent findings that among older women, exposure to mistreatment is associated with poorer
functional health status (Cannell, Weitlauf, Garcia, Andresen, Margolis, et al., 2015).
Several studies have also found a mortality disadvantage among elder abuse victims
compared to similar, non-abused older adults. In New Haven, Lachs and colleagues (1998) found
that community-dwelling older adults reported to APS had 9% poorer survival than those who
had never been reported to APS. This disadvantage persisted despite independently controlling
for individual health characteristics, functional impairments, social and emotional support,
cognitive abilities, and depression. More recently, research among older women has found that
those exposed to physical, but not verbal abuse, had a higher risk of death attributed to “other”
causes, that is, not attributable to cancer, cardiovascular disease, accident or injury (HR=2.12,
95% CI=1.17-3.84), even after adjusting for individual health and health behavior characteristics
(Baker, et al., 2009).

Scope of the Problem
The effects of mistreatment are widespread, especially given recent estimates that 141
million older adults, or one of every six elders, worldwide experience mistreatment (Yon,
Mikton, Gassoumis & Wilber, 2017). The most recent U.S. prevalence study reported a 11.4%
one-year incidence rate among adults aged 60 and older (Acierno, et al, 2010). Though these
figures are conservative estimates as past research on mistreatment prevalence suggests that
many cases remain unreported to formal service providers (Bonnie & Wallace, 2003).
The aging of the Baby Boomer cohort is expected to increase the magnitude of affected
and at-risk individuals (Pillemer, Connolly, Breckman, Spreng & Lachs, 2015), intensifying need
for programs, services, and policies which prevent, detect, and resolve elder mistreatment and

3

provide support and justice for survivors. Development of such programs requires a clearer and
deeper understanding of mistreatment, its causes, correlates and risk factors, mechanisms, signs,
symptoms, and consequences.

Operational Definitions of Mistreatment
The term “elder mistreatment” includes both abuse committed by others and neglect by
caregivers. Because there are no federal statutes that define elder mistreatment, states have been
responsible for defining these phenomena, resulting in definitional variations and debate (Bonnie
& Wallace, 2003). Comparison of state-level statutes revealed differences in definition and level
of victim vulnerability considered; type and severity of abusive or neglectful actions; civil or
criminal jurisdiction; and remedies for and responses to perpetration of abuse, included use of
mandated reporting procedures (Bonnie & Wallace, 2003).
Recent work in the research community has yielded some consensus definitions which
define abuse as “an intentional act or failure to act by a caregiver or another person in a
relationship involving an expectation of trust that causes or creates a risk of harm to an older
adult” (Hall, Karch & Crosby, 2016). While these definitions provide a conceptual definition to
guide future research and policies and enable cross-jurisdictional meta-analysis and
generalization of findings (Bonnie & Wallace, 2003), these research definitions differ from those
often used by professionals working in the field of elder abuse, including Adult Protective
Services workers, law enforcement personnel, healthcare professionals, prosecutors, and other
service providers. Front-line workers are often guided and constrained by jurisdiction-specific
legal statutes governing their scope of practice.

4

Because the work of this dissertation and the studies included therein seek to inform and
further develop professional practices used by practitioners in the field of elder mistreatment,
definitions used in the State of California were used to guide the presented research. The
California statues are largely consistent with those presented by Hall et al. (2016). They provide
age-tested protection from abuse and neglect for older adults aged 65 and older but do not
require that perpetrators be in positions of trust, enabling the prosecution of those who are
strangers and previously unknown to the victim (California Penal Code, 2018).

Physical Abuse Definitions
Although sometimes discussed as a single phenomenon, elder mistreatment is an
umbrella term used to describe several different forms of maltreatment, including physical abuse,
emotional abuse, financial abuse, neglect, and sexual abuse (Bonnie & Wallace, 2003; Hall,
Karch & Crosby, 2016). Each type is associated with unique mechanisms, pathologies, risk
factors, signs, symptoms, and outcomes which warrant individual study and research (Jackson &
Hafemeister, 2014; Jackson & Hafemeister, 2011). Multiple abuse types may also co-occur,
yielding hybrid forms of abuse with distinct characteristics (Jackson & Hafemeister, 2012).
Due to its similarity to physical abuse sustained through other forms of family violence,
such as child maltreatment and intimate partner violence, elder physical abuse was one of the
first recognized forms of elder mistreatment and was initially characterized as “granny battering”
(Burston 1975; Kleinschmidt, 1997). The definition has since been expanded to include
inappropriate physical or chemical restraint (Hall, et al., 2016; California Welfare and
Institutions Code, 2018). Other mechanisms of physical abuse can include, but are not limited to,
hitting, kicking, pushing, squeezing, twisting, burning, poisoning, and stabbing either by hand or

5

with an object (National Committee for the Prevention of Elder Abuse, 2014). The National
Elder Mistreatment Study found that among elders abused within the past year, 16% reported
experiencing some form of physical abuse (Acierno, et al., 2010).

Lessons from Other Fields of Family Violence
Despite first gaining recognition 40 years ago, the field of elder mistreatment is
developmentally where the fields of child abuse and intimate partner violence were 20 years ago
(Bonnie & Wallace, 2003). Intervention and methodological approaches used by investigators in
these family violence fields may provide useful models for detection, intervention, and research
of elder mistreatment (NIH, 2015; Bonnie & Wallace, 2003).
Early work in child maltreatment sought to find ways to identify abuse victims in order to
respond and intervene in abusive circumstances. Fundamental work to inform detection began
through observational and descriptive studies of abuse-related injuries presenting in medical
settings. These characteristics were then compared with characteristics of accidental injuries
observed among non-abused individuals to isolate injury types and patterns highly indicative of
abuse (Murphy et al., 2013; Leventhal, Thomas, Rosenfield & Markowitz, 1993; Meservy,
Towbin, McLaurin, Meyers & Ball, 1987; Fanslow, Norton & Spinola, 1998). This included the
identification of pathognomonic injuries, injuries that solely occur as a result of abuse.
Recognizing the complexities of accurately differentiating between abused and non-
abused study subjects and the lack of an accurate and universal screening and detection
instrument, the National Academy of Sciences also recommended the use of the LEAD Standard,
as a gold-standard to assist in this work (Bonnie & Wallace, 2003). The LEAD (“Longitudinal,
Experts All Data”) methodology convenes panels of experts to review all available longitudinal

6

data and determine whether abuse has occurred. Although initially proposed for use in other
medical health contexts (Spitzer, 1983), the LEAD approach was first applied to the field of
elder mistreatment by Wiglesworth and colleagues (2009) in their seminal work characterizing
bruising associated with physical elder mistreatment and identifying forensic markers of abuse
among bruises (Wiglesworth, et al., 2009; Ziminski, et al., 2013). In the prior study, a four-
person LEAD panel of geriatricians with expertise in elder abuse met monthly to review study
cases, examining collected study data and seeking additional case data by posing questions to
APS social workers who had interacted with the clients before determining whether abuse had
occurred.

Conceptual Model of Detection and Response to Physical Elder Abuse
Figure 1.1, below, presents a conceptual model describing the activities involved in
detecting and responding to elder physical abuse. Recognition of common indicators of abuse is
a needed first step in identifying victims of abuse and providing support and assistance to halt
mistreatment, promote client well-being, and prevent future abuse. Such indicators might include
the presence of an injury as well as specific injury diagnosis, location, size, pattern, and
mechanism. While a full elder abuse investigation and evaluation can require extensive time and
worker resources, making it infeasible and inefficient to thoroughly evaluate each older adult
with injury, use of screening tools and protocols can identify a sub-set of injured older adults
with injuries consistent with abuse who might benefit from a more extensive assessment and
abuse investigation. This can include the involvement of medical, social service, and legal
professionals who work to substantiate allegations of abuse and provide services to halt ongoing

7

abuse, prevent future mistreatment, and improve and protect the health, safety, and well-being of
older adult victims.

Common Indicators of Physical Elder Abuse
As illustrated in Figure 1.1., observation of physical injury and specific injury
characteristics is an important first-step toward responding to cases of mistreatment. Recent
work on markers of physical elder abuse have begun to document and characterize injuries
present among victims of abuse.
Consensus is emerging that injuries to the head or neck, including maxillofacial and skull
brain regions are highly associated with experiencing physical abuse (Wiglesworth, Austin,
Corona, Liao, Gibbs & Mosqueda, 2009; Ziminski, Wiglesworth, Austin, Philips & Mosqueda,
2013; Gironda, Nguyen & Mosqueda, 2016; Rosen, Bloemen, LoFaso, Clark, Flomenbaum &
Lachs, 2016; Friedman, Avila, Tanouye & Joseph, 2011; Murphy et al., 2013). Abusive injuries
also appear to be commonly found in all extremities, though somewhat more commonly in upper

8

extremities (Wiglesworth, et al., 2009; Ziminski, et al., 2013, Rosen, et al., 2016, Murphy, et al.,
2013).
Bruising is a common injury associated with physical abuse with studies finding between
39-72% of subjects exhibiting bruises (Wiglesworth, et al., 2009; Rosen, et al., 2016; Gironda, et
al., 2016). Bruises sustained to the head or neck, posterior torso, lateral right arm, or the upper
extremities appear to be associated with experiencing abuse (Wiglesworth, et al., 2009; Rosen, et
al., 2016). Bruise size appears to be another significant indicator of abuse as Wiglesworth &
colleagues (2009) found abuse victims more likely than their non-abused counterparts to have
sustained at least one bruise longer than 1 cm, with the majority having a bruise at least 5 cm in
length. Bruising to specific anatomical locations was found to be associated with specific
mechanisms of injury, such as choking, punching, grabbing, and beating, revealing observable
patterns which can serve to “paint a picture” of an abusive incident (Ziminski, et al., 2013).
To a lesser extent, fractures also have been identified as a common form of injury
associated with physical abuse (Rosen, et al., 2016). Fractures to the head or face, including the
maxillofacial, dental, and neck regions have been positively associated with abuse (Gironda, et.
al., 2016; Rosen et al., 2016; Friedman, et al., 2011). There is also some evidence to suggest that
lower extremity fractures occurred frequently among physically abused older adults seeking care
from an urban emergency department (Rosen, et al., 2016).
Lacerations and abrasions are other common forms of abusive injuries that have been
noted in the literature (Dyer, et al., 2003; Rosen et al., 2016; Gironda, et al., 2016; Palmer,
Brodell & Mostow, 2013) but have not been extensively studied. Further work is needed to
describe the types of lacerations and abrasions associated with abuse and to ascertain whether

9

there are observable differences in the presentation of lacerations and abrasions among abused
and non-abused individuals.
Commonly identified mechanisms of injury include being assaulted with fists or an
object, pushing or throwing, stabbing, grabbing, and falling after being otherwise hit, kicked, or
assaulted (Rosen, et al., 2016; Ziminksi, et al., 2013; Abath, et al., 2010). Other descriptive
injury characteristics, such as injury size or pattern, have been largely understudied. Such
descriptive information may further inform detection of suspicious injuries and discernment of
abuse. In their seminal work on bruising, Wiglesworth and colleagues (2006) found that those
who had experienced abuse had, on average, at least one bruise that was 2 inches longer than
those observed among accidentally injured older adults. Additional work is needed to ascertain
whether this relationship is true among other types of abuse-related injury. Examination of injury
patterns also holds promise for improving detection of abusive injuries (Bonnie & Wallace,
2003). However, to my knowledge, empirical studies have not documented the prevalence of
specific injury patterns present among physically abusive injuries. Further work is needed to
further explore the utility of injury pattern analysis toward identifying abuse-related injuries.

Detecting Physical Elder Abuse
The majority of elder abuse incidents are never reported to medical, law enforcement, or
social services agencies (Acierno, et al., 2010; Bonnie & Wallace, 2003; Tatara, et al., 1998).
Similar findings have been observed in research conducted in other fields of family violence,
such as child maltreatment and intimate partner violence (IPV). Work in these fields has
attributed some of this underreporting to victim blaming or protectiveness of the abuser. Due to
age-related changes, elderly victims of abuse may face additional barriers to reporting such as

10

physical dependence on their abuser, susceptibility to undue influence, memory impairment,
sensory impairment, cognitive decline and incapacity.
As a result, professionals who interact with older adults must be equipped and adept at
identifying signs of abuse. However, there is little consensus about the best way to detect abuse
within a clinical setting. Many currently used approaches and instruments were adopted from
child abuse and IPV research. They may not be appropriate for use in geriatric populations with
older adults who are more susceptible to non-abuse related injuries and may have injuries that
present more severely due to reductions in their physiologic reserve (Homeier, 2014).
Recognizing these limitations, the National Academy of Sciences (Bonnie & Wallace,
2003) and National Institute of Aging (NIH, 2015) have called for research to identify indicators
of abuse that can be used to identify injuries likely to have been caused through abusive means.
Specifically, NIH cited the need to identify injuries highly suggestive of abuse and characterize
their location, distribution, multiplicity, severity, and causal mechanisms to develop specific
characterizations of abuse that will help practitioners better differentiate between accidental
injuries and those inflicted by abusive actions. Ideally, such work might also lead to the
discovery of pathognomonic injuries, injuries that are distinctly indicative of elder abuse (Bonnie
& Wallace, 2003).

Responding to Elder Mistreatment
Detection of elder abuse-related injuries is a critical step toward assisting and supporting
victims. Identifying older adults with injuries consistent with physical abuse can enable
practitioners to interact further with these individuals to conduct formal abuse screening and
assessment. Upon further evaluation, practitioners that confirm abuse has occurred or continue to

11

suspect it may have occurred may be ethically or legally compelled to report their findings or
suspicions to investigators and providers who can further investigate or take action in cases of
mistreatment.
Responding to situations involving elder mistreatment often requires the action and
continued involvement of several different organizations and providers operating through
multiple different service delivery systems. Because elder mistreatment survivors are at
increased risk for experiencing poorer physical health, psychological well-being, functional
ability and mortality (Comijs, et al., 1998; Comijs, et al., 1999; Dong, et al., 2009; Acierno, et
al., 2002; Dyer, et al., 2003; Wiglesworth, et al., 2009; Cannell, et al., 2015; Lachs, et al., 1998;
Baker, et al., 2009), medical and mental health services are needed to assess victim well-being
and health status and provide necessary treatment services. These can require long-term health
care provision and support over a prolonged period.
Social services providers can assess victim needs and eligibility for services. While APS
is the primary social services agency entrusted with investigating and intervening in cases of
mistreatment, other social services programs can play a critical role in ensuring the safety and
well-being of older adult victims. Aging Network programs are well-positioned to fulfill some of
these care needs (Yonashiro-Cho, Meyer & Wilber, 2016), though other providers are needed.
For example, supportive housing services can provide older adults with safe housing away from
abuse perpetrators, and public guardians or conservators can ensure that incapacitated older
adults are safe and cared for with their basic needs met.
Legal services interventions can operate under both criminal and civil courts. Criminal
investigations are conducted by law enforcement officers in response to reports of elder
mistreatment. While mistreatment, including elder neglect, is a prosecutable criminal offense,

12

law enforcement officers also receive and respond to other reports of crime affecting older adult
victims. Upon completion of the law enforcement investigation, officers present their cases to
prosecutors who decide whether to prosecute cases to pursue justice on behalf of the victim and
public.
Civil legal interventions occur independent of criminal investigation though they can
proceed in parallel with them. Civil litigation can be used to reverse fraudulent legal and
business transactions, including quitclaim deeds, monetary banking and investment transfers,
powers of attorney, and marriages. Civil litigation also can be used by victims to collect damages
for losses and trauma experienced. Civil courts also hold jurisdiction for deciding older adult
legal competence and establishing guardianships or conservatorships, formal appointments of an
individual, or individuals, charged with making decisions on behalf of incapacitated individuals.

Coordinated Intervention Through Multi-disciplinary Teams
Elder abuse multi-disciplinary teams (MDTs) emerged in response to the lack of
coordination and communication present among organizations and individuals involved in
investigating and responding to cases of elder mistreatment (Anetzberger, 2011; Teaster,
Nerenberg & Stansbury, 2003). Though some MDTs were strictly policy- or education-oriented,
many sought to provide case review and support for elder mistreatment professionals (Teaster, et
al., 2003). MDTs were created to address specialized community needs. For example, the
Financial Abuse Specialist Team (FAST) specializes in providing professional advice on cases of
financial exploitation; the Vulnerable Adult Specialist Team and APS-Geriatric Assessment
Team models bring together health and social services providers to improve care for victims of
mistreatment (Mosqueda, Burnight, Liao & Kemp, 2004; Dyer, et al., 1999); and the Elder

13

Abuse Forensic Center (FC), a community-based MDT comprised of medical, legal, and social
services providers specializes in complex cases of mistreatment requiring multi-systems
interventions (Wiglesworth, Mosqueda, Burnight, Younglove & Jeske, 2006).
The FC model, first developed in Orange County, CA, has been replicated in Los
Angeles, New York, San Francisco, San Diego, and other sites across the country (Schneider,
Mosqueda, Falk & Huba, 2010). Research on the Los Angeles (LA) site has documented its key
structures and processes (Navarro, Wilber, Yonashiro & Homeier, 2010) and the internal
professional considerations and decision-making processes used by members during case review
and team meetings (Navarro, Wysong, DeLiema, Schwartz, Nichol & Wilber, 2016; DeLiema,
Navarro, Moss & Wilber, 2016).
FCs convene MDTs of health, legal, and social services professionals to engage in
collaborative case review and support front-line elder mistreatment providers in their work to
investigate, resolve, and seek justice in cases of elder mistreatment (Navarro, et al., 2010).
Clinicians, including geriatricians, mental health nurses, and psychologists, play a critical role on
the team by providing their professional insight into disease processes, injury presentations, and
cognitive and psychological state and well-being. Legal and social services providers often rely
on clinicians as a source of information to guide case investigations and next-steps, which
sometimes include criminal prosecution of perpetrators, civil litigation, or conservatorship.
Compared to a propensity-score matched sample of APS cases receiving APS usual-care,
cases seen by the LAFC were found to have 8.5 times greater likelihood of achieving a guilty
plea or conviction in prosecuted cases (Navarro, Gassoumis & Wilber, 2013). FC clients also had
7 times greater odds of being referred for conservatorship through the Office of Public Guardian,
though rates of conservatorship implementation among cases received by the Office of Public

14

Guardian were similar between FC and comparison group cases (Gassoumis, Navarro & Wilber,
2015).
While the FC is a promising model for MDT elder mistreatment intervention, the extent
to which the original Orange County FC model’s policies, practices, and activities have been
implemented across replication sites remains unclear. Moreover, it is also unknown whether the
policies, activities, processes, and outcomes observed in the LAFC are replicable across sites.

The Present Study
Consistent with the model’s name, FCs have historically had strong representation and
emphasis on legal investigation, remedies, and punitive responses. As such, FCs and their team
members play a key role in identifying gaps in literature and practice which can complicate
investigation and legal resolution of abuse and neglect cases. The present dissertation research
seeks to address one of these under-studied areas of research - the identification of forensic
markers of physical elder abuse.
Chapters II and III seek to identify and characterize injuries present among older adult
victims through common service system entry points. Chapter II examines patient and injury
characteristics present among older adults seeking care for non-fatal intentional injuries from
hospital emergency departments. Chapter III utilizes a similar approach to examine client and
injury characteristics presenting to APS workers as they respond to reports of suspected physical
abuse. Findings from both studies can be used to inform the development and improvement of
abuse detection and evaluation practices to earlier identify victims of abuse and connect them
with services and supports.

15

One such intervention which brings together practitioners from several different service
delivery systems is the FC. Chapter IV evaluates the components and processes of the FC across
replication sites to identify common structures, policies, procedures, and outcomes which may be
used to characterize the FC model and promote fidelity in future replication sites.
Finally, major findings and policy and practice implications arising from these studies is
discussed in the concluding chapter to provide some guidance for future research and next-steps.

16

Chapter II: Characteristics of Intentional Injuries Among
Older Adults Presenting to Hospital Emergency Departments in 2001-2010

INTRODUCTION
Elder mistreatment is a little recognized and often misunderstood public health issue that
can have a lasting effect on victim’s mortality, physical well-being, functional ability, quality of
life, and psychological well-being (Lachs, 1998; Dong, 2009). While a good deal of research has
focused on elder mistreatment from a protective services lens, there is still relatively little
evidence to guide medical practitioners as they approach the phenomenon (Gibbs & Mosqueda,
2010). By conservative estimates, 10% of older Americans experience mistreatment each year
(Acierno, et al., 2010), meaning that most clinicians who serve older adults are likely to see
victims of elder mistreatment on a regular basis.
Clinicians may encounter victims of abusive or negligent actions or inactions that harm
older individuals, including physical abuse, neglect, financial abuse, sexual abuse, emotional or
psychological abuse, isolation, abandonment, and abduction. The mechanisms, pathologies, risk
factors, clinical presentation, and observed outcomes of the sub-types related to each of these
actions may be vastly different (Jackson & Hafemeister, 2013); therefore, investigation into each
individual sub-type is warranted.

17

Physical Abuse and Injuries in Older Adults
Elder physical abuse is perhaps the most readily observable type of mistreatment that
clinicians encounter among older adults. Abusive actions include assaultive actions such as
hitting, kicking, punching, squeezing, twisting, burning, poisoning, over- or under-medication,
and inappropriate chemical or physical restraint. The National Elder Mistreatment Study found
that 16% of elders abused within the past year reported experiencing some form of physical
abuse (Acierno, et al., 2010).
The U.S. Centers for Disease Control and Prevention defines physical abuse as, “the
intentional use of physical force that results in acute or chronic illness, bodily injury, physical
pain, functional impairment, distress, or death” (p. 31) by an individual with whom the victim
has “a relationship involving an expectation of trust” (p. 28; Hall, Karch & Crosby, 2016).
However, most practitioners adopt the legal standards used in their geographic area of practice.
For purposes of this study, the California definition of abuse has been used which broadens the
definition of physical abuse to include abusive actions perpetrated against older adults by
strangers and those who are not in a trusted relationship with the older adult (California Welfare
& Institutions Code, 2018). Further, under California statutes, the abusive actions need not result
in physical or emotional consequences, further broadening the definition of abuse.
Currently, there are few well-established forensic markers of physical abuse and little
documentation about the types and characteristics of injuries sustained through abusive actions
(Bonnie & Wallace, 2003; Friedman, Avila, Tanouye & Joseph, 2011). One project used an ad-
hoc sample to characterize the size and location of bruises in victims of physical elder abuse and
compared them with bruises seen in non-victims (Wiglesworth et al. 2009; Ziminski et al., 2013).
Bruising was more commonly found in victims than non-victims on the head and neck (20.8%

18

vs. 4.4%), posterior torso (14.6% vs. 2.9%), and lateral right arm (25.0% vs. 7.3%; Wiglesworth
et al., 2009). Another group identified nine articles that presented injury information on an
aggregate 1,247 victims of physical elder abuse, in which injuries were sustained to the upper
extremities (44.0%), maxillofacial and neck regions (22.9%), skull and brain (12.3%), lower
extremity (10.6%), and torso (10.3%; Murphy et al., 2013).
To help practitioners better differentiate between accidental injuries and those inflicted
by abusive actions, research is needed that characterizes the location, distribution, multiplicity,
severity, and causal mechanisms of accidental and inflicted injuries (Bonnie & Wallace, 2003;
Murphy et al., 2013). Research on injuries resulting from geriatric assault, including intentional
actions perpetrated by strangers, may enable practitioners to better detect abusive injuries.
Further, we aim to build on the work of these previously conducted local studies by analyzing
large-scale, nationally-representative data sources to improve the generalizability of results and
inform practice across geographically-diverse care settings.
The current study uses a nationally-representative sample of emergency departments
across the U.S. to identify patient and injury characteristics associated with physical injuries
among older adults. The purpose of this study is to identify characteristics of non-fatal injuries
sustained by older adults that may be more likely to be associated with intentional physical
assault than accidental origins.

METHODS
Data Source
Data used for this study were from the National Electronic Injury Surveillance System
(NEISS), the Centers for Disease Control and Prevention’s epidemiological system to track the

19

incidence of non-fatal injuries presenting to hospital emergency departments. The NEISS
provides a nationally representative sample of hospitals with at least six beds and a 24-hour
emergency department (U.S. Department of Health and Human Services, 2010). The dataset
excludes patients who seek care in the emergency department for injuries that later prove fatal or
those who die from their injuries prior to reaching the hospital. Trained hospital coders used
patient medical records (emergency department records, Emergency Medical Services records,
and provider notes) to extract information on the injured patient and primary injury
characteristics, which included injury diagnosis, location, cause and reported intentionality (U.S.
Department of Health and Human Services, 2010). The NEISS public-use dataset was used in
this analysis. Constructed as a stratified probability sample to account for hospital size, the
public use dataset contains data a sub-sample of NEISS facilities (n= 66; U.S. Department of
Health and Human Services, 2010).
Though data were weighted to represent the inverse of the probability of case selection to
create a nationally-representative annualized sample, of hospital emergency departments, survey
weights were not applied in the analysis because estimating national injury prevalence was not
an objective of the present study.

20

As depicted in Figure 2.1 (above), the analytical sample was drawn from the 6,823,238
emergency department visits between 2000 and 2013 that appear in the publicly-available NEISS
dataset. Of these visits, 9.0% (n = 616,417) involved older adults aged 65 or older. After

21

stratifying the sample by injury intent, injuries occurring through self-infliction (n=1,746), or
legal intervention (n=168) were excluded due to their limited relevance to the present study and
their small sample sizes. Because sexual assault is often categorized as a distinct form of elder
mistreatment with potentially distinct causes and injury manifestations, cases involving sexual
assault (n=121) were also excluded from the present analysis. Of the remaining 614,385 cases,
2.4% (n=14,593), were excluded due to data missingness on key variables of interest, including
patient sex (n=69), primary injury diagnosis (n=85), affected body part (n=3,489), and causal
mechanism (n=11,013). The resulting final analytical sample included 599,792 visits which is
reflective of the same number (N=861,862) of patients presenting to the emergency department
for treatment of a non-fatal injury. Client identifiers were not included in the dataset, so it is
unknown whether the same patient is included in the dataset at multiple time points.

Variables
Injury intent was the primary variable used throughout our analyses to stratify the sample
between abused and non-abused older adults. The NEISS captured the intentionality of the injury
causing action (variable INTENT_C) by applying a code to describe the reported or suspected
intent of the action. Injuries that were reported by patients to be purposefully caused and those
suspected of being caused with the intent to injure were coded as “Intentional” injuries, whether
or not the injured party was the intended victim of the action. The INTENT_C variable also
captures whether the injury originated through self-infliction or through interactions with law
enforcement officers. However, because the mechanisms of injury involved in self-inflicted and
legal intervention cases are likely to differ from those used in cases of geriatric assault and elder

22

abuse, and may thus result in injuries with distinct characteristics, the present analysis excluded
cases with injuries resulting from self-infliction or legal intervention.
Injuries which were reported by patients to be caused through accidental or unintentional
means were also coded into the INTENT_C variable. However, the NEISS dataset groups
together these cases involving unintentionally-caused injuries with cases in which the injury
intent was unknown and unspecified. As a result, it is impossible to identify and select cases
explicitly attributed to accidental or unintentional causes. Thus, the present analysis was
restricted to using the Unintentional/Unknown Intent group as a comparison group for the
Intentionally Injured group despite the inclusion of cases of unknown intent.
Studied patient demographic information included patient age, sex, and primary race as
recorded in the patient medical record (U.S. Department of Health and Human Services, 2010). It
is unclear whether data entered into the medical record were obtained via patient self-report or
clinician observation and documentation.
The dataset includes information on the primary injury affecting the patient for which
they are seeking care (U.S. Department of Health and Human Services, 2010). Thus, other
injuries present at the time of interaction were not recorded in the dataset. Primary injury
diagnoses were clinician-assessed and grouped by the Centers for Disease Control and
Prevention into 31 diagnosis categories (DIAG) using the International Classification of
Diseases, Ninth Revision, Clinical Modification (ICD-9_CM) taxonomy. The present analysis
recoded the diagnosis variable to reflect the most frequently occurring diagnoses present in the
study sample. The recoded diagnosis variable captures diagnosed burns; concussions; contusions
or abrasions; fractures; hematoma; internal injuries; lacerations; poisoning; punctures; strains or
sprains; and other diagnoses, which included, but were not limited to amputation, aspiration,

23

ingestion, crushing, dislocation, foreign bodies, dental injuries, nerve damage, anoxia, electric
shock, and submersion.
The 6-item NEISS clinician-assessed variable describing the primary affected body part
by anatomical region (variable BDYPTG_C) was in the analysis to describe injury location.
Affected body part categories included patients’ head/neck, arm/hand, upper trunk, lower trunk,
leg/foot, and other location. Interaction variables were constructed to create variables which were
specific to both injury diagnosis and location, capturing the occurrence of a specific injury
diagnosis at a specific anatomical area of the body. These variables were used to further describe
differences in injury prevalence between groups.
Injury mechanism, obtained per patient self-report, was defined to encompass the
immediate and precipitating causes of injury which were each documented in the NEISS dataset
(variables ICAUSE and PCAUSE, respectively). Response categories to capture reported
mechanisms of injury were the same for both ICAUSE and PCAUSE variables and included 22
possible mechanisms, including intentional falling, being struck by or against an object, being
cut or pierced, fire or burns, poisoning, transportation-related injuries, overexertion, inhalation or
suffocation, gunshot wounds, and bites or stings. The immediate cause of injury was defined as
the action directly causing the injury while the precipitating cause of injury was defined as the
action initiating the chain of events leading to the injury. For example, if an older adult cut their
hand on a rock after being shoved and falling to the ground, the shoving action would be coded
as the precipitating cause of injury while being cut or pierced would be coded as the immediate
cause of injury. In some cases, the precipitating and immediate causes of injury were the same.
In order to capture both types of injury mechanisms in a parsimonious way, the present
analysis combined the most frequently occurring immediate and precipitating injury mechanisms

24

into 8 dichotomized variables to reflect the involvement of each type of mechanism as either an
immediate or precipitating cause of injury. Thus, for each case, it was possible, though not
necessary, for two injury mechanisms to be selected. Created injury mechanism variables were
positively coded and include Bite or sting; Cut or pierced; Falls; Overexertion; Poisoning; Struck
by or against an object; Transportation-related, and Other mechanisms.
To further assess the severity of injuries and describe patterns of patient disposition
following emergency department treatment, case disposition, as recorded in hospital emergency
department records, was also analyzed. The disposition variable (DISP) coded 7 possible case
dispositions that captured whether the patient was treated and released, transferred and released,
transferred and hospitalized, hospitalized, placed under observation in the emergency
department, or left against medical advice. In the present analysis, these data categories were
recoded to reflect the patient’s eventual case disposition. The recoded variable indicates whether
the patient was treated and released; hospitalized or transferred to another facility; or otherwise
discharged. Because the NEISS is restricted to patients with non-fatal injuries, patients were
alive at the time of case disposition.
In cases in which victims reported being injured through intentional means, data on the
nature of the victim’s relationship to their perpetrator were extracted from patient records
(variable PERP). This included coding whether the patient-identified aggressor was a spouse or
partner, other relative, friend or acquaintance, unrelated caregiver, stranger, or multiple
perpetrators. Coders also recorded information on the context or reasons for the assault (variable
REASON). Codes were used to capture 8 possible reasons for the assault, which included
altercation, robbery or burglary, drug-related activities, sexual assault, gang-related activities, or
other causes. The reported reason for assault was recoded to reflect the most commonly

25

occurring reasons identified within the sample. The recoded variable included altercation,
robbery or burglary, and other reasons for the assault.

Statistical Analysis
We used descriptive statistics to identify characteristics frequently associated with elder
physical abuse. Primary injury characteristics and specific injuries were compared by injury
intent using independent samples t-tests and chi-square tests for independence. When
appropriate, Fisher’s Exact Test was used to test associations when expected sample sizes were
smaller than appropriate for use with the Independent Samples T-test.

RESULTS
The average age of patients was 77.8 years (See Table 2.1, below). Forty percent (40%)
of the sample was aged 65-74 years old and one-quarter was aged 85 years or older. Sixty-three
(63.1%) of patients was female. Primary race data were unavailable for 21.8% of patients. White
patients comprised 63.9% of the sample while patients of Black (9%), Hispanic (3.2%), Asian
(1.5%), and Other (0.7%) races accounted for the remainder of the identified sample. Patients
reporting intentionally-caused assaultive injuries accounted for 0.8% (n = 4,718) of the total
sample.

26

Table 2.1. Patient Characteristics of Older Adult Patients Seeking Treatment for Non-Fatal
Injuries in the Emergency Department by Injury Intentionality, 2001-2013, N=599,792
Total Sample
Non-Intentional
(Accidental/Unknown)

Assaultive Injuries

(N = 599,792) (n = 595,074) (n = 4,718)
Characteristic n % n % n % p
†

Age, years (M± SD)
77.8 ± 8.5 77.9 ± 8.5 72.9 ± 7.3
<0.001
65-74 years 239,492 (39.9) 236,352 (39.7) 3,140 (66.6) <0.001
75-84 years 213,168 (35.5) 212,015 (35.6) 1,153 (24.4) <0.001
85+ years 147,132 (24.5) 146,707 (24.7) 425 (9.0) <0.001
Female 378,719 (63.1) 376,801 (63.3) 1,918 (40.7) <0.001
Race

White 383,194 (63.9) 381,242 (64.1) 1,952 (41.4) <0.001
Black 53,828 (9.0) 52,542 (8.8) 1,286 (27.3) <0.001
Hispanic 19,437 (3.2) 19,048 (3.2) 389 (8.2) <0.001
Asian 9,241 (1.5) 9,130 (1.5) 111 (2.4) <0.001
Other 3,222 (0.7) 3,187 (0.7) 35 (0.7) <0.001
Unknown 130,870 (21.8) 129,925 (22.2) 945 (20.0) <0.001
† Calculated using Pearson chi-square test for independence; ‡ Calculated using Independent
samples t-test

Patients Reporting Injury Resulting From Assault
Patients reporting injuries sustained through physical assault were younger than those
reporting non-intentional (accidental or unknown) injuries (72.9 years vs. 77.9 years old;

27

p<0.001). They were also more likely to be men (59.3%) while the majority of non-intentionally-
injured victims were women (63.3%). Likewise, there also was a significant difference in the
racial background of assault victims, with a greater rate of intentional injury among racially
diverse older adults as compared with non-intentional injuries (Black: 27.3% vs 8.8%, Hispanic:
8.2% vs 3.2%, Asian: 2.4% vs. 1.5% and other race: 0.9% vs 0.7%). Race data were unavailable
for 21.8% of cases.

Table 2.2. Characteristics of Primary Injuries Among Older Adult Patients Seeking Treatment
for Non-Fatal Injuries in the Emergency Department by Intent of Injury Cause, N=599,792
Total Sample Non-Intentional
(Accidental/Unknown)
Intentional
(Assaultive)

(N=599,792) (n=595,074) (n=4,718)

Characteristic N (%) n (%) n (%) p
†

BODY PART

Head/Neck/
Maxillofacial
182,360 (30.4) 179,612 (30.2) 2,748 (58.3) <0.001
Upper Extremities 112,052 (18.7) 111,321 (18.7) 731 (15.5) <0.001
Upper Trunk 71,860 (12.0) 71,265 (12.0) 595 (12.6) <0.001
Lower Trunk 102,198 (17.0) 101,879 (17.1) 319 (6.8) <0.001
Lower Extremities 100,701 (16.8) 100,441 (16.9) 260 (5.5) <0.001
Other Site 30,621 (5.1) 30,556 (5.1) 65 (1.4) <0.001
DIAGNOSIS

Contusion/Abrasion 126,938 (21.2) 125,374 (21.1) 1,564 (33.2) <0.001

28

Laceration 87,674 (14.6) 86,765 (14.6) 909 (19.3) <0.001
Internal Injury 59,081 (9.85) 58,365 (9.81) 716 (15.2) <0.001
Fracture 152,592 (25.44) 151,943 (25.5) 649 (13.8) <0.001
Strain/Sprain 71,364 (11.9) 71,132 (12.0) 232 (4.9) <0.001
Burn 3,871 (0.7) 3,861 (0.7) 10 (0.2) <0.001
Concussion 5,485 (0.9) 5,413 (0.9) 72 (1.5) <0.001
Hematoma 7,767 (1.3) 7,709 (1.3) 58 (1.2) <0.001
Puncture 6,294 (1.1) 6,160 (1.0) 134 (2.8) <0.001
Poison 20,905 (3.5) 20,878 (3.5) 27 (0.6) <0.001
Other Injury 57,821 (9.6) 57,474 (9.7) 347 (7.4) <0.001
MECHANISM

Fall 385,867 (64.3) 385,203 (64.7) 664 (14.1) <0.001
Struck By/Against 84,597 (14.1) 80,518 (13.5) 4,079 (86.5) <0.001
Cut/Pierce 26,171 (4.4) 25,926 (4.4) 245 (5.2) <0.001
Overexertion 33,151 (5.5) 33,133 (5.6) 18 (0.4) <0.001
Poisoning 13,313 (2.2) 13,308 (2.2) 5 (0.1) <0.001
Transportation 57,435 (9.6) 57,409 (9.6) 26 (0.6) <0.001
Bite/Sting 17,485 (2.9) 17,381 (2.9) 104 (2.2) <0.001
Other Cause 28,194 (4.7) 28,031 (4.7) 163 (3.5) <0.001
DISPOSITION

Treated/Released 456,574 (76.1) 452,686 (76.1) 3,888 (0.9) <0.001
Hospitalized/
Transferred
135,935 (22.7) 135,195 (22.7) 740 (15.7) <0.001

29

Other 7,283 (1.2) 7,193 (1.20) 90 (1.9) <0.001
† Calculated using Pearson chi-square test for independence

Location of Primary Injury Associated with Assault
As depicted in Table 2.2 (above), assaultive primary injuries were most frequently
(58.3%) found on the victim’s head, neck, and maxillofacial regions. They were also commonly
found on victims’ upper extremities (15.5%) and upper trunk (12.6%). Injuries to victim’s lower
trunk (6.8%) and lower extremities (5.5%) were less common.
In contrast, injuries inflicted through non-intentional means were more evenly distributed
among anatomical sites. Though these injuries were also more frequently situated on patients’
head, neck, and maxillofacial regions (30.2%), they also commonly affected victims’ upper
extremities (18.7%), lower trunk (17.1%), lower extremities (16.9%) and upper trunk (12%).
Head, neck, and maxillofacial injuries were more likely to occur among victims reporting
assaultive injuries (58.3% vs 30.2%; p<0.001). In contrast, primary injuries occurring on
victims’ upper extremities (18.7% vs. 15.5%; p<0.001), lower trunk (17.1% vs. 6.8%; p<0.001),
lower extremities (16.9% vs. 5.5%; p<0.001), and other sites (5.1% vs. 1.4%; p<0.001) were
more likely to be attributed to non-intentional causes.
Primary Injury Diagnoses Associated with Assault
Contusions and abrasions were more likely to result from assault than unintentional or
unknown causes (33.2% vs. 21.1%; p<0.001), accounting for the largest proportion of assaultive
injuries. Lacerations (19.3% vs. 14.6%; p<0.001), internal injuries (15.2% vs. 9.8%; p<0.001),
concussions (1.5% vs. 0.9%; p<0.001), and punctures (2.8% vs. 1.0%; p<0.001) were also more
likely to result from assaults rather than non-intentional means.

30

Although accounting for a sizable proportion of intentional injuries (13.8%), fractures
accounted for one-quarter of all unintentional or unknown cases and were more likely to be the
result of a non-intentional incident (p<0.001). Strains or sprains (12.0% vs. 4.9%; p<0.001),
burns (0.7% vs. 0.2%; p<0.001); poisonings (3.5% vs. 0.6%; p<0.001) and other diagnoses
(9.7% vs. 7.4%; p<0.001) were more likely to occur non-intentionally.

Table 2.3. Prevalence of Injury Diagnosis and Location Among Older Adult Patients
Seeking Treatment for Non-Fatal Injuries in the Emergency Department by Injury Intent,
N=599,792
Total Sample
Non-Intentional
(Accidental/
Unknown)
Intentional
(Assault)
(N=599,792) (n=595,074) (n=4,718)
Characteristic n(%) n(%) n(%) p
†

Head/Neck/Maxillofacial

Contusion/Abrasion 35823 (6)
35,013 (5.9) 810 (17.2) <0.001
Laceration 47004 (7.8)
46,332 (7.8) 672
(14.2)
<0.001
Internal Injury 57338 (9.6)
56,639 (9.5) 699
(14.8)
<0.001
Fracture
9,542
(1.6)
9,305 (1.6) 237
(5)
<0.001
Strain/Sprain
13,493
(2.2)
13,424 (2.3) 69
(1.5)
<0.001
Burn
782
(0.1)
776 (0.1) 6
(0.1)
0.951
Concussion
5,485
(0.9)
5,413 (0.9) 72
(1.5)
<0.001
Hematoma
4,693
(0.8)
4,648 (0.8) 45
(1)
0.180
Puncture
426
(0.1)
403 (0.1) 23
(0.5)
<0.001
‡

Other Injury
7,774
(1.3)
7,659 (1.3) 115
(2.4)
<0.001
Upper Extremities

Contusion/Abrasion
15,526
(2.6)
15,301 (2.6) 225
(4.8)
<0.001
Laceration
30,234
(5)
30,062 (5.1) 172
(3.6)
<0.001

31

Fracture
37,668
(6.3)
37,508 (6.3) 160
(3.4)
<0.001
Strain/Sprain
6,713
(1.1)
6,676 (1.1) 37
(0.8)
0.028
Burn
1,493
(0.2)
1,492 (0.3) 1
(0)
0.002
Hematoma
931
(0.2)
925 (0.2) 6
(0.1)
0.623
Puncture
3,441
(0.6)
3,400 (0.6) 41
(0.9)
0.007
Other Injury
16,046
(2.7)
15,957 (2.7) 89
(1.9)
<0.001
Upper Trunk

Contusion/Abrasion
26,969
(4.5)
26,673 (4.5) 296
(6.3)
<0.001
Laceration
191
(0)
171 (0) 20
(0.4)
<0.001
‡

Internal Injury
1,462
(0.2)
1,453 (0.2) 9
(0.2)
0.459
Fracture
24,457
(4.1)
24,343 (4.1) 114
(2.4)
<0.001
Strain/Sprain
10,923
(1.8)
10,866 (1.8) 57
(1.2)
0.002
Burn
223
(0)
222 (0) 1
(0)
0.304
‡

Hematoma
164
(0)
164 (0) 0
(0)
0.274
‡

Puncture
146
(0)
115 (0) 31
(0.7)
<0.001
‡

Other Injury
7,325
(1.2)
7,258 (1.2) 67
(1.4)
0.212
Lower Trunk

Contusion/Abrasion
23,447
(3.9)
23,333 (3.9) 114
(2.4)
<0.001
Laceration
320
(0.1)
298 (0.1) 22
(0.5)
<0.001
‡

Internal Injury
281
(0)
273 (0) 8
(0.2)
0.002
‡

Fracture
51,880
(8.6)
51,790 (8.7) 90
(1.9)
<0.001
Strain/Sprain
18,430
(3.1)
18,395 (3.1) 35
(0.7)
<0.001
Burn
248
(0)
248 (0) 0
(0)
0.141
‡

Hematoma
596
(0.1)
595 (0.1) 1
(0)
0.043
‡

Puncture
133
(0)
113 (0) 20
(0.4)
<0.001
‡

Other Injury
6,863
(1.1)
6,834 (1.1) 29
(0.6)
<0.001

32

Lower Extremities

Contusion/Abrasion
24,605
(4.1)
24,508 (4.1) 97
(2.1)
<0.001
Laceration
9,893
(1.6)
9,871 (1.7) 22
(0.5)
<0.001
Fracture
29,003
(4.8)
28,956 (4.9) 47
(1)
<0.001
Strain/Sprain
21,758
(3.6)
21,724 (3.7) 34
(0.7)
<0.001
Burn
997
(0.2)
997 (0.2) 0
(0)
0.005
Hematoma
1,379
(0.2)
1,373 (0.2) 6
(0.1)
0.139
Puncture
2,146
(0.4)
2,127 (0.4) 19
(0.4)
0.604
Other Injury
10,920
(1.8)
10,885 (1.8) 35
(0.7)
<0.001
Other Site

Contusion/Abrasion
568
(0.1)
546 (0.1) 22
(0.5)
<0.001
‡

Laceration
32
(0)
31 (0) 1
(0)
0.197
‡

Fracture
42
(0)
41 (0) 1
(0)
0.239
‡

Strain/Sprain
47
(0)
47 (0) 0
(0)
0.690
‡

Burn
128
(0)
126 (0) 2
(0)
0.186
‡

Poisoning
20,905
(3.5)
20,878 (3.5) 27
(0.6)
<0.001
Other Injury
8,893
(1.5)
8,881 (1.5) 12
(0.3)
<0.001
† Calculated using Pearson chi-square test for independence; ‡ Calculated using Fisher's Exact
Test

Primary Injury Diagnosis by Location
Analysis of injuries by affected anatomical site and diagnoses more precisely revealed
differences in the incidence of injury characteristics caused by assault (Table 2.3, above).
Injuries to the head, neck and maxillofacial regions were more likely (p<0.001) to be caused by
assault, across most injury diagnoses. Consistent with the prior diagnosis-level analysis,
contusions and abrasions (37.7% vs 5.9%), internal injuries (35.5% vs. 9.5%), lacerations (35.0%
vs. 7.8%), concussion (12.3% vs. 0.9%), and punctures (7.0% vs. 0.1%) were more likely to

33

result from assault. In contrast, fractures (21.8% vs. 1.6%) and other injury diagnoses (15.4% vs.
1.3%) were also more likely to be associated with assault though in aggregate these diagnoses
had appeared more likely to be sustained through non-intentional means. However, consistent
with the aggregate analysis of diagnoses, muscle strain or sprain in this region was more likely to
be associated with non-intentional causes (2.3% vs. 1.5%; p<0.001).
Except for contusions and abrasions, which were more likely to be assaultive in nature
(4.8% vs. 2.6%; p<0.001), injuries present on the upper extremities, including lacerations (5.1%
vs. 3.6%; p<0.001), fractures (6.3% vs. 3.4%; p<0.001), and other injury diagnoses (2.7% vs.
1.9%; p<0.001), were more likely to be associated with non-intentional causes (p<0.001).
Injuries on the trunk varied in their association with injury intent by diagnosis and
location. Upper trunk contusions or abrasions (6.3% vs. 4.5%; p<0.001), lacerations (0.4% vs.
0.0%; p<0.001), and punctures (0.7% vs. 0.0%; p<0.001) were more likely to be the result of
assault whereas fractures (4.1% vs. 2.4%; p<0.001), strains and sprains (1.8% vs. 1.2%; p<0.01)
were more likely to be caused by actions of non-intentional intent. Lower trunk lacerations (0.5%
vs. 0.1%; p<0.001), internal injuries (0.2% vs. 0.0%; p<0.01) and punctures (0.4% vs. 0.0%;
p<0.001) were more likely to be associated with assault. In contrast, lower trunk fractures (8.7%
vs. 1.9%; p<0.001), contusions and abrasions (3.9% vs. 2.4%; p<0.001), strains and sprains
(3.1% vs. 0.7%; p<0.001), hematomas (0.1% vs. 0.0%; p<0.05), and other injury diagnoses
(1.1% vs. 0.6%; p<0.001) were more likely to be associated with non-intentional intent.
Injuries occurring on the lower extremities were more likely to be attributed to non-
intentional causes (p<0.001) across most injury diagnoses. Consistent with aggregate analyses,
fractures (4.9% vs. 1.0%), strain or sprain (3.7% vs. 0.7%), burns (0.2% vs. 0.0%), and other
injury diagnoses (1.8% vs. 0.7%) were more likely to be associated with actions caused by non-

34

intentional intent. Although in aggregate analyses, lacerations, contusions and abrasions were
more likely to result from assault, the incidence of these diagnoses on patients’ lower extremities
was more likely to be the result of actions caused by non-intentional intent (p<0.001; Laceration:
1.7% vs. 0.5%; Contusion/Abrasion: 4.1% vs. 2.1%).

Injury Mechanism Associated with Assault
Injuries resulting from being struck by or against an object or person (86.5% vs. 13.5%;
p<0.001) and those attributed to being cut or pierced by an object (5.2% vs 4.4%; p<0.001) were
more likely to be the result of assault.
Though falls accounted for 14.1% of assault injuries, they were more likely to be
attributed to unintentional or unknown injuries (64.7%; p<0.001). Non-intentional injuries were
more likely to be attributed to a broader spectrum of causes including transportation-related
incidents (9.6% vs. 0.6%; p<0.001), overexertion (5.6% vs. 0.4%; p<0.001), bites or stings
(2.9% vs. 2.2%; p<0.001), poisoning (2.2% vs. 0.1%; p<0.001) and other causes (4.7% vs. 3.5%;
p<0.001).

Case Disposition Associated with Assault
Patients with assaultive injuries were more likely to be discharged from the emergency
department (82.4% vs. 76.1%; p<0.001) and less likely to be hospitalized (22.7% vs. 15.7%;
p<0.001) than those with unintentional injuries.

35

Assaultive Incidents
Data on victim-perpetrator relationships and reasons for the assault were limited. Data on
the nature of the perpetrator’s relationship to the victim were unavailable for half of assault cases
(Table 2.4, below), and 7.9% of assaults (n=372) involved multiple assailants prohibiting the
inclusion of specific perpetrator relationships in the NEISS dataset. Data on reasons for the
assault were similarly limited with underlying reasons being unknown in 72.0% of cases.
Nevertheless, analysis of available data revealed that 15.7% of victims (n=742) reported
being assaulted by a friend or acquaintance. Romantic partners and other relatives were
implicated in 7.3% (n=345) and 1.5% (n=72) of cases, respectively. Strangers perpetrated 5.3%
of assaults. Per patient report, 17.9% of assaults (n=845) occurred during altercations and 8.4%
(n=398) were committed as part of a robbery or burglary.

Table 2.4. Conditions in which Assault was Perpetrated Against Older Adult Patients
Seeking Treatment for Non-Fatal Injuries in the Emergency Department
Intentional Injury (Assault)
(n=4718)
Characteristic n %
PERPETRATOR RELATIONSHIP
Spouse/Partner 345 (7.3)
Other Relative 72 (1.5)
Friend/Acquaintance 742 (15.7)
Multiple Perpetrators 372 (7.9)
Stranger 249 (5.3)

36

Other 72 (1.5)
Unknown/Unspecified 2360 (50.0)
REASON FOR ASSAULT
Altercation 845 (17.9)
Robbery/Burglary 398 (8.4)
Other 79 (1.7)
Unknown 3396 (72.0)

DISCUSSION
Though physical elder abuse threatens the health and well-being of a substantial number
of older adults, research to identify forensic markers of elder abuse and injury traits highly
suggestive of abuse has been limited. This study is one of few to investigate injury
characteristics that distinguish between assaultive and non-intentional injuries. To our
knowledge, it is also one of the first to analyze epidemiological population-based injury
surveillance data, yielding results with broader generalizability than those conducted in a single
community or research sites. The dataset’s large sample size enabled us to document and
compare the prevalence of specific injuries by both diagnosis and anatomical location yielding
novel findings that further differentiated assaultive and non-intentional injuries.
Unlike the non-intentionally caused injuries, there was little diversity in the diagnoses
and mechanisms associated with injuries arising from physical assault. Overall head, neck, and
maxillofacial injuries were more likely to be associated with assault rather than non-intentional
injury. Contusions and abrasions in this region were the single most common injury type found

37

among patients reporting assault followed by similarly located internal injuries and lacerations.
Though less prevalent, fractures, concussions, and punctures affecting the maxillofacial region
were also more likely to be associated with assault than non-intentional injury causes.
These findings were consistent with prior work which has documented a large prevalence
of head, neck, and maxillofacial injury among victims of assault (Wiglesworth, et al., 2009;
Rosen, Bloemen, LoFaso, Clark, Flomenbaum & Lachs, 2016; Friedman, et al., 2011; Cham &
Seow, 2000; Murphy et al., 2013). These previous studies have estimated the prevalence of head,
neck, and maxillofacial injuries to affect 12.2-52% of abuse victims. Our present analysis
documented a 58.3% prevalence of injury to this region, consistent with research conducted in an
urban emergency department by Rosen and colleagues (2016). Study site and sample
characteristics may account for some of the observed discrepancies with other studies. Because
the present analysis studies primary injuries present among older adults seeking care from
emergency departments, they represent the highest acuity injuries present among a population
that, itself, is more likely to have experienced serious injury. In contrast, some of the previous
studies were conducted in non-emergency settings (Wiglesworth, et al., 2009) or among older
adults with “mild” in nature (Murphy, et al., 2013). Others have examined injuries in
international populations which may have different injury mechanisms and patterns of healthcare
utilization (Abath, et al., 2010; Cham & Seow, 2000).
Though most injuries found on patient upper extremities were more likely to occur
through non-intentional means, contusions and abrasions to this area were positively associated
with assault (p<0.001). These findings provide further evidence of the prevalence of upper
extremity contusions or bruising resulting from assaultive action. First identified by Wiglesworth
and colleagues (2009) in a study of bruising among community-dwelling older adults who had

38

experienced abuse, bruising to the upper extremities has also been documented among physically
abused older adults seeking care from hospital emergency departments (Rosen, et al., 2016). In
particular, contusions or bruising to the lateral aspect of the right arm may result when an
assailant grabs a victim’s arm (Wiglesworth, et al., 2009).
Truncal injuries varied in their association with assault by specific injury diagnoses.
Though the overall incidence of truncal punctures and lacerations were low (n=42 and n=51,
respectively), their incidence was associated with assault (p<0.001; p<0.001). Though prior
studies have documented the presence of torso bruising among victims of abuse (Wiglesworth, et
al., 2009; Rosen, et al., 2016), our study found that injury association with assault varied by
truncal location. Specifically, contusions or abrasions on the upper trunk were more likely to be
caused by assault (p<0.001) while those to the lower trunk were more likely to be non-intentional
(p<0.001).
Because the trunk is generally protected from incidental impact, the presence of
contusions or abrasions to this area are highly suspicious. Indeed, prior research in the field of
child maltreatment has noted an association between torso injuries and assault (Christian &
Committee on Child Abuse and Neglect, 2015), and bruising to the posterior torso has been
found to be associated with elder abuse (Wiglesworth, et al., 2009). Nevertheless, the finding
that lower trunk contusions and abrasions were more likely to arise from non-intentionally-
caused events rather than assault suggests that further study of truncal injury is necessary to
better understand the region’s vulnerability and association with assault.
Some of these differences might be explained by the collection and classification of data
within the NEISS. By aggregating data on contusions and abrasions, it is not possible to
differentiate between these diagnoses. Though identified as a potential injury arising from abuse,

39

to our knowledge, the prevalence and injury distribution of abrasions have not been studied
among abused or assaulted older adults. It may be that lower trunk abrasions are common
accidental injuries and their documentation in the NEISS obscures the relationship between
lower trunk contusions and assault. Though disaggregation of these diagnoses is not possible in
the present dataset, further data collection and research on contusions and abrasions is needed to
examine the prevalence and anatomical distribution of each injury diagnosis among victims of
assault.
Our study found the overwhelming majority (86.5%) of assaults were caused by being
struck by or against something. The prevalence of injury caused by this mechanism were lower
than those reported in community contexts (99.1%; Abath, et al., 2010) but higher than those
found among severely-injured hospitalized patients in Level I trauma facilities in an urban area
(61%; Friedman, et al., 2011). Instead, study prevalence rates were like those found among
patients seeking care from an urban emergency department (90.3%; Rosen, et al., 2016),
providing further evidence of the prevalence of injury resulting from mechanisms of assault
involving mechanical energy, including blunt force trauma and assault with an object or weapon,
among patients seeking care in the Emergency Department.
The National Elder Mistreatment Study, which assessed the incidence of elder
mistreatment by conducting a telephone survey of community-dwelling older adults, reported a
lower one-year incidence of physical abuse among non-White elders (Acierno, et al., 2010). In
contrast, our study found greater ethnic diversity among patients reporting intentionally-caused
physical injuries compared to those whose injuries were attributed to non-intentional causes.
Because this study documents injury characteristics among those seeking care from hospital
emergency departments, our results may reflect the characteristics of more seriously- injured

40

physically-abused older adults. It may also be a reflection of healthcare utilization among victims
of assault. Prior research has noted disproportionately higher emergency department utilization
among non-White racial minority groups for health care services, even for conditions that might
not warrant hospital-based care (Johnson, Ghildayal, Ward, Westgard, Boland & Hokanson,
2012). Thus, our study sample may include a larger number of patients identifying with racial
minority groups than the broader community-dwelling victim population.
In the present analysis, more males than females attributed their injuries to intentional
causes. Much of the work to-date has been conflicting with some studies citing higher rates of
physical abuse among female victims (Cham & Seow, 2000; Friedman, et al., 2011; Lachs, et al.,
1997b; National Center on Elder Abuse, 1998) and others finding higher prevalence of abuse
among males (Abath, Leal, Melo Filho & Marquez, 2010; Pillemer & Finkelhor, 1988). On one
hand, older men may be at greater risk for abuse because of their higher likelihood of living with
a spouse, a documented risk factor for abuse (Pillemer & Finkelhor, 1998).
Relatively little is known about the characteristics of abuse perpetrators (Bonnie &
Wallace, 2003). The present study found the largest proportion of identified assailants were
friends or acquaintances (15.7% of the total sample, 31.4% of identified perpetrators). These
findings are similar to those reported by Abath and colleagues (2010) who found that 39.8% of
abusers were known, but unrelated, to their victims. Reported intimate partner violence among
victims accounted for a 7.3% of assault cases, suggesting that the injuries and cases included in
this analysis are more expansive than just intimate partner violence grown old. However, these
findings may also reflect study participant reticence to abuse perpetration by intimate partners or
other family members despite their willingness to report assaults committed by unrelated friends
or acquaintances.

41

Limitations
The National Electronic Injury Surveillance System (NEISS)’s heavy reliance on patient
self-report to assess the intentionality of actions causing injury may result in underreporting of
assault among patients seeking care. Some abuse victims may be reluctant to report that their
injuries were intentionally-caused out of a desire to protect their abuser; fear of future assault and
retribution from their abuser or institutional placement; misguided guilt, shame, or self-blame for
allowing or inviting abuse; or cultural norms of internal family conflict resolution and non-
reporting.
Others, including those with cognitive impairment or dementia, may have limited
physical and cognitive capacities that render them unable to comprehend and report their abusive
situation. Research has estimated that 9% of older adults aged 75-84 years of age and 20% of
those 85 years and older experience confusion or memory loss (Bernstein & Remsburg, 2007).
Given that in the present study, 35.6% (n=212,015) of those reporting non-intentional injury are
aged 75-84 years of age and 24.7% (n=146,707) are aged 85 years or older, it’s likely that some
of those with non-intentional injury may have presented to the emergency department with a
compromised ability to accurately report the intentionality of their injury.
Thus, it’s expected that some injuries attributed to non-intentional actions, those of
accidental or unknown intent, may actually have occurred through intentional assault. As such,
the present analyses are likely to be conservative in their estimates, underestimating the true
prevalence of assaultive injury within the population of emergency department patients. This
does not discount the utility of these results; instead, these findings provide evidence to better
profile and identify suspicious injury among emergency department patients.

42

Another limitation of the data arises from its sole inclusion of the patient’s primary, most
severe, injury despite the possible presence of other injuries. As a result, the dataset lacks the
detailed information necessary to identify injury patterns involving co-occurring injuries across
several locations or diagnoses. Such patterns have been identified in other studies of abuse
(Abath, et al., 2010; Clarke & Pierson, 1999) and hold promise for better identifying abusive
injuries and differentiating between injuries sustained through assaultive and accidental means.

Future Directions
Future research is needed to better capture and characterize injury patterns observed
among victims of assault. Because many victims of assault report having multiple injuries and
injury sites, data collection and analyses of all assault injury sites will likely provide a more
accurate and comprehensive picture of assaultive injuries and their characteristics (Murphy,
2013). Additionally, data on injury severity, such as the length, width, depth, or circumference of
an injury may help to better classify and categorize injury patterns associated with abuse.
This study’s use of a representative surveillance sample provided a broad perspective
about the types of injuries seen in several different sites across the country. The development of
multi-site research projects partnering hospital- and community-based medical providers with
social services and law enforcement agencies to examine, catalog, and analyze abusive injuries
in inpatient and community-based settings would also greatly advance the field and provide
substantial and generalizable results to further inform the identification of forensic markers of
abuse.

43

Practice Implications
The identification of injury characteristics commonly associated with physical elder
abuse will aid in the development of protocols and instruments to better recognize intentionally-
inflicted injuries among elders. Many current methods of abuse detection are adapted from the
fields of child abuse or intimate partner violence and may not adequately address the needs of
elderly victims of abuse. Within the older adult population age-related biological changes and
widespread medication use may alter the presentation of injuries, complicating clinician efforts
to identify abuse victims (Homeier, In Press). Elderly victims may also be physically, socially or
cognitively unable to report abuse or may experience impairments that make them poor
historians. Such limitations may also leave them functionally, socially, or financially dependent
on their abusers, making them reluctant to report abuse and vulnerable to undue influence. The
development of approaches to recognize physical elder abuse independent of victim or witness
reports promises to improve efforts to protect vulnerable and mistreated patients.
This study presents preliminary data supporting the notion that there may be differences
in the patient and injury characteristics of those presenting to the emergency department with
assaultive rather than accidental injuries.
Although further research is needed to better elucidate patterns and pathologies
associated with elder physical assault, these findings can serve as a preliminary “red flag” for
emergency department medical personnel to help them better identify patients with injuries that
may have been caused by intentional assaultive actions.
Direct observation and clinical assessment of abuse may reduce reliance on patient or
witness self-report. (Fulmer, Guadagno, Dyer & Connolly, 2004; Bonnie & Wallace, 2003).
Patients presenting with injuries sustained after being struck by or against a person or object with

44

a primary diagnosis of contusions, abrasions, or internal injuries, especially if these injuries
occur to the face, head or neck may be victims of elder physical assault. These suspicions may be
especially heightened if the patient is male.
Medical staff may decide to use protocols or procedures that are enacted in cases where
physical violence is suspected. Such steps may include interviewing the patient alone without
family members or other visitors present, directly asking the patient whether they have been
assaulted, utilizing violence screening tools, or consulting with a specialized violence
intervention team (Clarke & Pierson, 1999; McLeer & Anwar, 1989; Mosqueda, Burnight, Liao
& Kemp, 2004). Clinicians may also determine that further assessment of the individual may be
warranted. Use of cognitive assessment tools designed to assess decision making may provide
insight into an individual’s capacity to make care decisions. Additional work-up may include
screening for depression, assessing Activities of Daily Living and Instrumental Activities of
Daily Living, and gait or strength assessment.
Timely intervention in cases of abuse and thorough examination of the patient’s physical
and cognitive well-being holds promise for improving the overall health of the individual.
Halting the cycle of physical injury and risk for disability due to abuse may reduce the likelihood
of further injury. Such intervention may also theoretically reverse the observed increases in
mortality and reductions in psychological well-being and quality of life.
Findings may also contribute to criminal justice proceedings regarding cases of geriatric
assault and abuse. Clinicians are often called-upon to provide expert testimony in cases of
mistreatment. The identification of forensic markers of abuse can be used to inform clinical
assessments and provide evidence in cases of mistreatment.

45

CONCLUSION
This study provides insight into the patient and injury characteristics observed among
older adults presenting to the emergency department and seeks to identify injuries likely
sustained through assault. Such information is critical to improve elder abuse detection and
achievement of justice. As the older adult population grows in magnitude and proportion, it will
be increasingly important to establish more accurate and sensitive means of abuse surveillance
and reporting in order to promote the well-being and safety of this population. Further research,
especially using population representative samples, is necessary to inform the development of
such approaches.

46

Chapter III:
Characteristics of Injuries Associated with Physical Elder Abuse in a Population of Community-
Dwelling Older Adults Receiving Services from Adult Protective Services

INTRODUCTION
Elder physical abuse, the application of physical force with the intent to harm an older
adult, places older adults at increased risk of physical injury (Ziminski, Wiglesworth, Austin,
Phillips & Mosqueda, 2013), emotional distress (Comijs, Pennix, Knipscheer & van Tilburg,
1999), and mortality (Lachs, Williams, O’Brien, Pillemer & Charlson, 1998). Abusive actions
include both assaultive (e.g. hitting, kicking, striking with an object, pushing, pulling, biting,
choking, burning, and pinching) and restrictive (e.g. improper use of physical or chemical
restraints and over-medication) acts (CDC, 2016).
While physical abuse also occurs in other forms of family violence, such as child
maltreatment and intimate partner violence, its etiology, presentation, and consequences may
vary with age and when co-occurring with age-related conditions, complicating third-party
efforts to distinguish abuse-inflicted injury from accidental trauma (Homeier, 2014; Gibbs &
Mosqueda, 2010; Dyer, Connolly & McFeely, 2003). Difficulties distinguishing abuse from
accidental injury can hinder efforts to provide victims with opportunities for support, treatment,
and intervention. Though, research on elder abuse injuries remains relatively limited, especially

47

compared to other family violence fields (Bonnie & Wallace, 2003), recent work has begun to
address the paucity of studies.

Geriatric Injury
In a seminal study documenting the occurrence and characteristics of bruising conducted
by Mosqueda, et al. (2005), researchers provided critical insight into the location, origin, color,
and multiplicity of non-abusive bruising among older adults. This work to characterize the
spectrum of accidental bruising and healing patterns significantly informed future work
conducted by Wiglesworth and colleagues (2009) who studied the location and characteristics of
bruises found among abused older adults as compared to those found in the non-abused sample.
The study found that bruising was present in the majority (72%) of abused older adults surveyed
and that the size and location of observed bruises differed significantly from those observed
among non-abused older adults, providing evidence of measurable differences in injury
characteristic by injury cause. Bruises among abused individuals were found to be larger and
more likely to occur to the older adult’s face, lateral aspect of the right arm, and posterior torso.
Rosen and colleagues (2016) examined physical injuries present during emergency
department examination of older adults with previous interaction with Adult Protective Services
(APS), the primary community-based agency charged with investigating and intervening in cases
of suspected abuse. Bruising (39%) and fractures (29%) were common in this emergency
department population, and injuries were commonly located in the upper extremities (45%),
lower extremities (32%), maxillofacial and neck regions (26%), and skull and brain (26%). Work
in other clinical contexts have yielded similar results with regard to injury location suggesting
that abusive injuries are more commonly found in the head, maxillofacial, neck, and upper

48

extremities (Abath, Leal, Melo Filho & Marques, 2010; Friedman, Avila, Tanouye & Joseph,
2011).
Research on injuries present in successfully prosecuted, highly adjudicated physical elder
abuse cases yielded results consistent with the above clinical studies (Rosen, et al., 2015). Nearly
two-thirds (62%) of victims had multiple documented injuries, with bruising (71%), lacerations
(52%), and abrasions (33%) present within the sample. Similar to previous findings, commonly
affected anatomical regions included the upper extremities (57% of victims) and maxillofacial
and neck regions (48%). Together, these findings have laid foundational work to document and
characterize abusive injuries presenting in medical and legal contexts by victims seeking care or
legal intervention from healthcare providers or the criminal justice system.
Though the growth of research on this topic has provided insight into the types and
locations of injuries found among victims seeking medical or legal assistance, these studies may
not capture the injuries and experiences of the majority of abuse victims, many of whom never
seek assistance (Tatara, et al., 1998). As the primary agency charged with responding to
suspected mistreatment, APS proactively reaches out to those who may have experienced abuse
to provide them with an opportunity for assistance and support. Through this role, APS workers
are often privy to clients and situations which might otherwise never come to the attention of
formal health, legal, and social service providers.
However, there is a dearth of research on injuries occurring within this study population.
Little is known about the prevalence of injuries in this population and the characteristics and
typology of injuries present. It is also unclear whether specific injuries or injury characteristics
may be strongly associated, and thus, suggestive of abuse among members of this population.

49

Such data may inform the professional practice of APS workers who must assess victim injuries
and engage in immediate decision-making and case management.
The purpose of this study was to document the spectrum of injuries and injury
characteristics observed among physically-abused older adults reported to APS and compare
these findings to injuries found among non-abused older adults. The identification of
distinguishing characteristics may enable clinicians and APS workers to better detect injuries
suspicious for abuse.

METHODS
Approach
This study employs an observational, matched-comparison group design which enabled
documentation of differences in injury prevalence, location, and characteristics found between
older adult APS clients who had experienced abuse and similar non-abused older adults. The
APS and comparison groups were matched on age, gender, education, English language use,
living status, and social isolation. Study subject demographic characteristics are presented below
in Table 3.1.
In addition to being used in past studies of elder mistreatment (Wiglesworth, et al., 2009;
Ziminski, et al., 2013; Rosen, et al., 2016)., this approach has also been used in other fields of
family violence research to identify distinguishing markers of abuse perpetrated against children
(Meservy, Towbin, McLaurin, Meyers & Ball, 1987; Leventhal, Thomas, Rosenfield &
Markowitz, 1993) and intimate partners (Fanslow, Norton & Spinola, 1998).
The study protocol was reviewed and approved by the University of Southern
California’s Health Science Campus Institutional Review Board.

50

Study Recruitment
Physically abused study participants were recruited from among older adult APS clients,
aged 65 years and older, receiving services from Los Angeles County. Subjects were required to
have the capacity to assent to participate in the study or have an appropriate legal proxy available
to assent on their behalf. They also needed to speak English or Spanish, though those speaking
other languages were included if a suitable, non-abusive interpreter was available to interpret the
recruitment procedures and data collection interview.
Prospective APS participants had to have experienced a recent physical abuse incident.
Though an official APS determination of abuse was not required for participation, clients who
APS workers believed were falsely reporting abuse were excluded from recruitment efforts.
Initially, subjects were required to have experienced the abuse in the 14 days prior to data
collection. However, this timeframe was later lengthened to include participants who had
experienced an abusive incident in the 21 days prior to data collection. The longer follow-up
timeframe was adopted to better accommodate APS protocols and procedures.
An APS Supervisor was assigned to review all reports of alleged elder physical abuse
reported to APS and forward researchers the case information and corresponding APS worker
assigned to the case of all clients who aged 65 and older reported to APS with allegations of
physical elder abuse. Because study inclusion criteria allowed for the inclusion of APS clients
who spoke a language other than English or Spanish if an appropriate interpreter was present,
APS forwarded information for clients speaking any language. Similarly, because inclusion
criteria allowed for the inclusion of cognitively incapacitated clients with consent from their
conservator or other appropriate legal proxy, APS also forwarded information for clients
appearing to lack cognitive capacity who met other inclusion criteria.

51

Once client information was received by study staff, data were screened to verify that
prospective clients appeared to meet study inclusion criteria. Clients who did not appear to meet
study criteria, including those younger than 65, those who were not reported to APS for
allegations of physical abuse, and those who had experienced a physical abuse incident prior to
the required look-back period, were excluded from further follow-up. Study staff sought to
contact APS workers assigned to remaining identified APS clients who appeared to meet study
inclusion criteria in order to initiate study recruitment procedures.
Study recruitment procedures were developed with APS to adhere to the agency’s
confidentiality policies and prevent undue discomfort or danger to the client. Promotion of client,
worker, and study clinician safety and well-being were foremost in the development of
recruitment protocols. In the resulting protocol, presented in Figure 3.1 below, APS workers
were responsible for establishing contact with potential study subjects and seeking client
permission for study clinicians to contact them about the study.

52

In some cases, APS workers notified study staff about their planned visit to see a
potentially-eligible client and coordinated an in-person joint visit with the study clinician. At the
time of the visit, the APS worker approached the home alone and began their interview and
assessment. If it appeared the client would be eligible for the study and they deemed it
appropriate and feasible to discuss the study with the client, APS workers would provide a brief
description of the study and seek the client’s permission to invite the study clinician into the
home to participate in the visit. The APS worker notified the clinician via text messaging
whether the client appeared eligible for the study and whether they were amenable to including
the clinician. In all, study clinicians attempted 122 home visits using this approach, eventually
recruiting 35 in this manner.
While joint visits were preferred, it was often infeasible due to APS’ immediate response
protocols and study staffing constraints. In these situations, APS workers conducted their visit

53

independently, yet continued to provide potentially-eligible clients with a brief description of the
study and seek their permission to have study staff follow-up with them at a later date. Study
staff attempted contact clients who were amenable to follow-up to briefly assess their eligibility
and schedule a visit with study clinicians. Study clinicians attempted 37 home visits using this
approach, eventually recruiting 27 in this manner. Altogether, fewer than 5% of the cases
referred to the research team were ultimately recruited into the study.
Designed to reflect a sample of non-abused community-dwelling older adults, the
comparison group was recruited from among patients seeking usual-medical care from a County
Geriatrics Outpatient clinic. Subjects recruited to the comparison group were not necessarily
seeking care for physical injuries and were not required to have any injuries present. Prospective
comparison group members were screened for past physical abuse and excluded if they reported
experiencing elder physical abuse in the preceding 12 months. Individuals who had experienced
childhood physical abuse, intimate partner violence, or other forms of physical assault in their
early- to midlife were included in the study as injuries arising from these events would be
unlikely to emerge during physical examination.
Like the abused sample, prospective comparison group participants were required to be at
least 65 years of age and have capacity to assent to study participation or have an appropriate
legal proxy available to assent on their behalf. though incapacitated older adults were eligible to
participate with the assent of an appropriate legal proxy. They also needed to speak English or
Spanish or have a suitable interpreter, who was not identified through APS report or
investigation to be suspected of perpetrating abuse, present for the interview.
To create a demographically-similar, matched-sample, efforts were also made to match
comparison group subjects to the APS abuse sample on demographic variables, such as age,

54

gender, race, language spoken, education, and living status. During the recruitment period, APS
and comparison group subjects were compared across demographic characteristics using
bivariate tests of association, including Pearson chi-square analysis and Independent samples t-
tests. Findings were used to target comparison group recruitment to promote demographic
similarity between the APS and comparison group. Nevertheless, specific demographic
characteristics were not used to exclude potential subjects from study participation.
While waiting for their appointments, potential study subjects were approached by the
study Research Nurse and given some information about the study. If individuals were interested
in learning more about the study and potentially participating, the Research Nurse met with them
in a private room to provide them with more detailed information and screen them for potential
study inclusion. While the Research Nurse sought to discuss these topics in private, away from
other family members or caregivers accompanying the client, other individuals were permitted in
the room if they served as the appropriate legal proxy or interpreter for the potential study
subject or if the subject expressed a preference to have them present.

Data Collection
Study clinicians, which included a Registered Nurse and geriatrician met with study
subjects to obtain their assent to participate in the study and conduct the study data collection
protocol. These study clinicians conducted assessments in both the APS and clinic samples. The
protocol consisted of a research-administered survey and physical examination. The survey
collected data on subject socio-demographic characteristics, medical history, and physical and
sexual assault history. APS subjects were also asked to recount the physically abusive incident.
The interview was followed by a full-body examination in which study clinicians noted visible

55

injuries and physical findings. Patients were asked to remove all their clothing to enable study
clinicians to view body areas that are usually clothed. Examination findings were photographed
and described in written documentation.
One of the study clinicians conducted interviews with study subjects. To the extent
possible, subjects were interviewed privately without others, including the alleged abuser,
present. Study clinicians documented whether other individuals were present at the time of the
interview, taking care to note whether the alleged abuser was present.
Data collection among APS clients was primarily conducted in study subjects’ homes,
though subjects were given the option of meeting in a clinic or at another location at their
choosing. Comparison group subjects were interviewed at the clinic in a clinic office or
examination room. The same research clinicians assessed both APS and comparison subjects

LEAD Panel
To validate participant assignment to the “abused” and “non-abused” categories, all 165
subject cases were presented to a LEAD (“Longitudinal, Experts, All Data”) panel. Use of the
LEAD panel methodology has previously been applied to seminal elder abuse research
examining abuse markers in bruises present among community-dwelling older adult victims
(Wiglesworth, et al., 2009; Ziminski, et al., 2013).
The study LEAD panel was comprised of three clinicians with experience in identifying
and treating victims of family violence: A pediatrician with expertise in working with victims of
child maltreatment, an emergency room physician with expertise in working with victims of
intimate partner violence, and a geriatric nurse practitioner with expertise in working with
victims of elder mistreatment. The diversity of expertise on the panel enabled them to draw from

56

the body of work already conducted in the fields of child maltreatment and intimate partner
violence – a recommendation put forth by several federal agencies and expert panels (National
Institute of Health, 2015; Bonnie & Wallace, 2003).
During meetings, LEAD panel members reviewed all available case data. Whenever
possible, the research clinician who met with the study subject and collected the data was present
to answer any panelist questions regarding collected data and their encounter with the study
subjects. In cases where they are unable to be present, other study clinicians or research staff
reviewed the case data and answered LEAD panel members’ questions regarding the collected
study data and documentation. These included, but were not limited to, questions clarifying the
location of the injuries pictured in study photographs, obtaining additional background
information on study subjects’ physical or cognitive state at the time of the visit, and describing
the quality and nature of the relationship between the study subject and others present at the time
of the interview.
Panelists completed a data collection form in which they (1) indicated whether they
believed abuse had occurred (“Based on the discussion and the information provided, is this
incident physical abuse?”); (2) rated their level of confidence in their decision (“How confident
are you about this abuse assessment?”); and (3) provided a brief rationale for their decision
(“What are the reasons for this assessment?”). Although the abuse determination question had
the largest direct effect on study analysis procedures, panelist level of confidence in their
decision and rationale for their determination provided insight into their decision-making
processes.
After making an initial individual determination, panelists discussed their opinions and
deliberated potential differences to reach unanimous consensus on whether they believed abuse

57

had occurred. As demonstrated in the CONSORT diagram presented in Figure 3.2 (a) below,
cases in which unanimous consensus was not achieved (n = 4; 2 APS-recruited cases and 2
clinic-recruited cases) were excluded from analyses. Discordant cases (n = 5; 2 APS-recruited
cases and 3 clinic-recruited cases), those with LEAD panel determinations that were in
contradiction with study subject recruitment source, were also excluded from the study.

58

Sample
One hundred, sixty-five (n=165) older adults aged 65 years and older were recruited to
the study, 61 recruited from among clients seen by APS and 104 older adults recruited to the
comparison group. After excluding cases in which the LEAD panel was unable to reach
consensus (n=4) and cases which were discordant with recruitment source (n=5), the final
analytic sample consisted of 156 older adults: 57 older adult victims of physical abuse recruited
from APS and 99 non-abused older adults recruited from the Geriatrics outpatient clinic.
Injury-level analysis was conducted to examine differences in characteristics present
among injuries found among APS and clinic comparison group subjects. Figure 3.2 (b), above,
presents the case exclusion criteria used to obtain the analytic sample for injury-level analysis. A
total of 372 physical findings were documented by research clinicians during their total body
examination of study subjects. Of the total 382 injuries, 175 were found among APS subjects and
197 were found among clinic comparison group subjects. Findings were excluded from analysis
if they were missing a formal injury diagnosis (n = 1; 1 finding from among APS clients) or were
determined upon patient self-report or clinician assessment to be a scar (n = 40; 12 findings from
among APS clients and 28 findings from among clinic subjects) or attributable to medical
procedures or conditions (n = 6; 1 finding from among APS clients and 5 findings from among
clinic subjects). The resulting analytic sample consisted of 325 injuries: 161 injuries were
observed among APS clients and 164 injuries were observe among clinic subjects.

Variables
Data on subject socio-demographic characteristics, including their age, gender, race,
country of origin, educational attainment, living situation, and degree of social isolation were

59

collected from subjects through self-report. Social isolation was measured using a five-item
assessment modified from the Medical Outcomes Study social support module (Sherbourne &
Stewart, 1991). Modeled after work conducted by Acierno and colleagues (2017) to assess
social support among community-dwelling older adults who had experienced abuse, the total
social isolation score was dichotomized to reflect whether a subject experienced social isolation.
Participants with scores of 0-13 were positively recoded to indicate being socially isolated.
Subject medical histories were collected, including known medical conditions,
medications being taken, surgical history, and functional ability. Variables were constructed to
capture underlying medical conditions or treatments which may have an effect on injury
presentation. A dichotomously coded variable was constructed to identify the presence of a
medical condition which may affect injury presentation and characteristics. The study
geriatrician reviewed patient self-reported medical conditions and identified relevant conditions.
Subjects reporting a diagnosis of a medical conditions associated with increased risk of bruising
or easy bleeding, as well as dermatologic conditions which may mimic physical injury were
positively coded as having a medical condition which may affect injury presentation.
Similarly, a dichotomously coded variable was also created to identify the study subject’s
use of a medication which might affect the prevalence and characteristics of injuries. These
included anticoagulants, including aspirin, NSAIDs, COX2 inhibitors, coumadin, heparin,
clopidogrel, and new oral anticoagulants (NOAGs); glucocorticoids, such as prednisone, which
may cause dermatologic changes such as skin thinning; and cilostazol. Patients who reported
taking these medications were positively coded to indicate this medication use.
Two dichotomously coded variables were also constructed to capture subject medical
diagnoses or medication use which may increase risk of falls. The study geriatrician reviewed

60

patient self-reported medical conditions and identified conditions which were associated with
increased fall risk. These included but were not limited to cardiac conditions, osteoarthritis,
sciatica, scoliosis, spinal stenosis, cerebral vascular accidents, gout, Alzheimer’s Disease and
dementia, neuropathy, vertigo, cataracts, alcohol abuse, and vitamin deficiency. The process was
also repeated to examine medications associated with fall risk. Patients reporting these diagnoses
or medications used were positively coded on these variables to indicate their increased risk of
falls due to these medical contributors. Patients taking insulin, sulfonylureas, beta-blockers,
calcium channel blockers, diuretics, benzodiazepines, antipsychotics, opioids, central analgesics,
and NSAIDs were positively coded as taking a medication known to increase fall risk.
Surgical histories were available for use by LEAD panel members to identify differential
diagnoses for physical findings which may be attributed to medical causes (e.g. surgical scarring
or incision sites). Functional ability was assessed using Katz’ ADL and Lawton’s IADL scales
(Lawton & Brody, 1969).
Research staff also collected data on physical findings and injuries present at the time of
the interview, photographing and recording the location, size, color, and pain associated with
observed physical findings as well as their origin, mechanism, and timing of emergence. Injury
locations were recoded to their general anatomical regions: the head, neck, and maxillofacial
region, which also included the mouth and teeth; upper extremities, which included shoulders,
upper arm, antecubital, elbow, forearm, palms, and digits; trunk, which included both ventral and
dorsal sides; and lower extremities, including thighs, knees, ankles, feet, and toes. Injury length
was measured by assessing the length of the injury’s longest side. Injury-causing mechanisms
were recoded to reflect the four most common mechanisms and included being grabbed by
another individual, falling, contact with a person or object, and injuries arising from unknown

61

causes. Clinicians described injury patterns but did not use apriori categories to define their
observations. Characteristics were recoded to dichotomously reflect whether linear patterns or
circular or semi-circular patterns was present.
Because the study focused on the prevalence and characteristics of injuries among
physically abused APS clients, cases missing injury diagnosis data (n=1) and injury location data
(n = 1) were excluded from the study (see Figure 3.1). Physical findings which were scars or
physical findings attributable to medical intervention or procedures were documented by study
clinicians, though those which clinicians confirmed to be scars or attributable to medical
activities were excluded from the analyses.

Analysis
Data were analyzed using SAS version 9.4. Descriptive statistics were generated to
describe study subject characteristics and identified physical findings. Bivariate tests of
association, including Pearson chi-square analysis, Independent samples t-tests, and Fisher’s
Exact t-tests were used to assess the relationship between observed characteristics and exposure
to physical abuse.
These statistical approaches were also used to characterize injuries present among study
subjects and compare and contrast injury characteristics between abused APS clients and non-
abused clinic-recruited comparison group subjects.
Frequently-occurring injuries with a sufficiently large sub-sample size were further
characterized and compared between study groups. Although data collection began prior to the
dissemination of the Comprehensive Classification System for Visible Non-Accidental and

62

Accidental Acute Geriatric Injuries (Rosen, et al., 2016), whenever possible data were reported
using taxonomic categories to enable future inclusion in cross-study analyses.

RESULTS
LEAD Panel Abuse Determinations
LEAD panelists reviewed all available data for each study case. Though they were unable
to reach a consensus agreement in 4 cases (2 recruited from APS and 2 recruited from the clinic),
they unanimously agreed that abuse had occurred in 60 cases and that abuse had not occurred in
101 cases. The LEAD panel’s determinations were concordant with recruitment source in 96.9%
of cases. Of the discordant cases, 2 were recruited from APS but determined by the panel to not
involve physical abuse, and 3 were recruited from the clinic but determined by the panel to
involve physical abuse. Cases in which the LEAD panel was unable to reach consensus, and
cases in which the panel’s determination differed from the recruitment source were excluded
from the analysis.

Sample Characteristics
Across recruitment arms, study participants were similar in age, gender, use of English
language, educational attainment, living arrangements, and degree of social isolation (Table 3.1
below). Those recruited through the clinic were more racially diverse than their APS
counterparts. Clinic recruits were also less likely to be born in the United States (24.5% vs 61.8).

63

Table 3.1. Demographic Characteristics of Study Sample by Recruitment Arm, N=156
Total Sample APS Clinic

(n=156) (n=57) (n=99) p
†

Characteristic n % n % n %

Age, years (M± SD) 76.7 ± 7.7 77.7 ± 8.5 76.1 ± 7.2 0.22
‡

Female 95 60.9 37 64.9 58 58.6 0.44
Race

White 35 23.9 24 42.9 11 11.1 <0.001
Black 21 13.5 5 8.9 16 16.2 0.21
Latino 64 41.3 18 32.1 46 46.5 0.08
Asian/Pacific Islander 29 18.7 5 8.9 24 24.2 0.02
Other Race 6 3.9 4 7.1 2 2 0.11
§

Missing 1

1

U.S.-born 58 37.4 34 61.8 24 24.5 <0.001
Missing 3

2

1

English-speaking 109 70.3 44 78.6 65 65.7 0.09
Missing 1

1

Education

Less than High School 81 52.6 24 42.1 57 57.6 0.17
a

High School Graduate 38 24.7 18 31.6 20 20.2 0.17
a

College Graduate or
Higher
35 22.7 13 22.8 22 22.2 0.17
a

Missing 2 2 2

Lives Alone 37 23.9 15 26.8 22.0 22.2 0.65
With Partner and/or
Family
104 67.1 35 61.4 69 69.7 0.46
b

With Other (non-Family) 14 9 6 10.5 8 8.1 0.46
b

Missing 1

1

Socially-isolated 30 19.5 13 23.6 17 17.2 0.33
Missing 2

2

Note:
†
Calculated using Pearson chi-square test for independence;
‡
Calculated using Independent
samples t-test;
§
Calculated using Fisher's Exact T-test; Subjects were able to select more than one
race category;
a
Pearson chi-square test calculated for the table, accounting for all education
variables;
b
Pearson chi-square test calculated for the table, accounting for all living arrangement
variables

Table 3.2, below, presents the health characteristics of study participants. Older adults
recruited through APS had fewer reported medical conditions (3.7 vs. 4.7) than comparison

64

group members and a smaller proportion of APS clients reported having conditions affecting
injury presentation (8.8% vs. 29.3%). Medication usage was similar between groups with
participants taking an average of 6 medications. Sixty-percent of participants had a medical
condition or were taking a medication that was likely to affect injury presentation. Though nearly
two-thirds of the sample (62.7%) reported some IADL impairment, 76.3% were fully functional
with their ADLs. There were no statistically-significant differences, at the p<0.05-level, in
functional abilities between the APS and comparison groups.

Table 3.2. Health Characteristics of Study Sample by Recruitment Arm, N=156
Total Sample APS Clinic

(n=156) (n=57) (n=99) p
†

Characteristic n % n % N %

MEDICAL HISTORY

Num. Self-reported of Medical Conditions 4.3 ± 2.5 3.7 ± 2.3 4.7 ± 2.6 0.02
‡

Condition affecting injury presentation 34 21.8 5 8.8 29 29.3 <0.01
Condition Increasing fall risk 116 74.4 41 71.9 75 75.8 0.60
Num. Self-reported Medications Taken 6.0 ± 3.3 5.8 ± 3.5 6.1 ± 3.2 0.56
‡

Medication affecting injury presentation 74 47.4 29 52.7 45 45.9 0.42
Medication increasing fall risk 114 73.1 45 81.8 69 70.4 0.12
Missing 3

2

1

Subject has at least one self-reported
condition or medication affecting injury
presentation
93 59.6 30 52.6 63 63.6 0.18

65

Subject has at least one self-reported
condition or medication affecting fall risk
144 92.3 55 96.5 89 89.9 0.14
FUNCTIONAL ABILITY

No Identified ADL Impairment 119 76.3 45 79.0 74 74.8 0.55
Num. ADL Impairments (M± SD) 0.4 ± 0.8 0.4 ± 0.9 0.3 ± 0.7 0.74
‡

No Identified IADL Impairment 57 37.3 24 43.6 33 33.7 0.22
Num. IADL Impairments (M± SD) 2.2 ± 2.5 1.9 ± 2.3 2.3 ± 2.5 0.32
‡

Missing 3

PHYSICAL INJURIES

Injuries Present 104 66.7 44 77.2 60 60.6 0.03
Number of Injuries 2.2 ± 2.6 2.9 ± 3.0 1.8 ± 2.2 <0.01
‡

Note:
†
Calculated using Pearson chi-square test for independence;
‡
Calculated using Independent
samples t-test

Injuries Among APS Clients
Injuries were present in 77.2% of APS clients (Table 3.2) with an average of 2.9 physical
findings present at the time of assessment. Nearly a quarter (22.8%) of APS clients (n=13) had
no observable injuries at the time of examination. APS clients without injuries present were all
evaluated within 12 days of the abusive incident. Figure 3.3 presents the number of days
between the date of abusive incident to the day of study data collection and whether the recruited
abuse victims had injuries present at the time of assessment.

66

Analyses conducted among a sub-group of APS clients with injuries present upon
examination (n=43; Table 3.3), found that 69.8% had at least one ecchymosis, 46.5% had at least
one abrasion, 37.2% had at least one area of tenderness, 20.9% had at least one area of swelling,
and 11.6% had at least one erythema. Three quarters of those studied (76.7%) had at least one
injury to their upper extremity; 46.5% of APS clients had an injury to the head, neck, or
maxillofacial region; and 37.2% of clients had an injury to the lower extremities.

0
1
2
3
4
5
6
7
8
9
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Number of Subjects Recruited
Days Since Abusive Incident
Figure 3.3 - Number of Days Between Abusive Incident and Study
Data Collection Among APS-Recruited Subjects (N=50)
No Injury Present Injury Present

67

Table 3.3. Prevalence of Diagnoses and Locations of Injuries Reported Among Subjects with an
Injury by Recruitment Source (N = 106)

Total with
Injury Present
(N=106) APS (n=43) Clinic (n = 63)
N % n % n % p
†

INJURY DIANGOSIS
Abrasion 49 46.2 20 46.5 29 46 0.96
Ecchymosis 60 56.6 30 69.8 30 47.6 0.02
Erythema 14 13.2 5 11.6 9 14.3 0.69
Laceration 5 4.7 3 7 2 3.2 0.36
Swelling 16 15.1 9 20.9 7 11.1 0.17
Tenderness 24 22.6 16 37.2 8 12.7 <0.01
Other Injury 25 23.6 10 23.3 15 23.8 0.95
INJURY LOCATION
Head, Neck, Maxillofacial 32 30.2 20 46.5 12 19 <0.01
Upper Extremity 78 73.6 33 76.7 45 71.4 0.54
Trunk 15 14.2 7 16.3 8 12.7 0.60
Lower Extremity 52 49.1 16 37.2 36 57.1 0.04
DIAGNOSIS BY LOCATION
Abrasions
Head, Neck, Maxillofacial 9 8.5 3 7 6 9.5 0.64
Upper Extremity 26 24.5 16 37.2 10 15.9 0.01

68

Trunk 2 1.9 1 2.3 1 1.6 0.78
Lower Extremity 27 25.5 7 16.3 20 31.7 0.07
Ecchymosis
Head, Neck, Maxillofacial 11 10.4 9 20.9 2 3.2 <0.01
Upper Extremity 49 46.2 23 53.5 26 41.3 0.22
Trunk 4 3.8 3 7 1 1.6 0.15
Lower Extremity 12 11.3 4 9.3 8 12.7 0.59
Erythema
Head, Neck, Maxillofacial 3 2.8 1 2.3 2 3.2 0.44
§

Upper Extremity 4 3.8 2 4.7 2 3.2 0.36
§

Lower Extremity 6 5.7 2 4.7 4 6.3 0.32
§

Laceration
Head, Neck, Maxillofacial 2 1.9 2 4.7 0 0 0.16
§

Upper Extremity 3 2.8 1 2.3 2 3.2 0.44
§

Swelling
Head, Neck, Maxillofacial 4 3.8 4 9.3 0 0 0.02
§

Upper Extremity 2 1.9 2 4.7 0 0 0.16
§

Lower Extremity 8 7.5 2 4.7 6 9.5 0.20
§

Tenderness
Head, Neck, Maxillofacial 4 3.8 4 9.3 0 0 0.02
§

Upper Extremity 3 2.8 3 7 0 0 0.06
§

Trunk 1 0.9 1 2.3 0 0 0.41
§

Lower Extremity 7 6.6 1 2.3 6 9.5 0.12
§

69

Other Injury
Head, Neck, Maxillofacial 7 6.6 5 11.6 2 3.2 0.08
§

Upper Extremity 12 11.3 3 7 9 14.3 0.13
§

Trunk 1 0.9 0 0 1 1.6 0.59
§

Lower Extremity 6 5.7 3 7 3 4.8 0.29
§

Note: †Calculated using Pearson chi-square test for independence; §Calculated using Fisher's Exact
T-test

Among the injured sub-sample, examination of specific injury diagnoses by affected
location revealed frequent injury patterns present among APS clients. Commonly occurring
injuries included upper arm ecchymosis (53.5%), upper arm abrasions (37.2%), lower extremity
abrasions (16.3%), lower extremity ecchymosis (9.3%), head, neck, or maxillofacial swelling
(9.3%), and head, neck, or maxillofacial tenderness (9.3%).

Comparison of APS and Comparison Group Injuries
Analysis of all subjects revealed that injuries were more frequently found among APS
clients than clinic comparison subjects (77.2% vs. 60.6%; Table 3.2), and on average, APS
clients had more injuries present at the time of examination (2.9 vs. 1.8).
Analyses conducted among the sub-group of study subjects with injuries found that
ecchymoses (69.8% vs 47.6%, p=0.02) and tenderness without other accompanying injury
(37.2% vs. 12.7%, p<0.01) were more likely to occur among APS clients than clinic-recruited
comparison groups members (Table 3.3). Though injuries to the upper extremities were common
in both groups, there were no significant differences in prevalence between groups (p=0.54).

70

Injuries to the head, neck, and maxillofacial regions were more prevalent among APS clients
than comparison group members (46.5% vs. 19%, p<0.01), while injuries to the lower
extremities were more prevalent among comparison group members than APS clients (57.1% vs.
37.2%, p=0.04).
Examination of injury and location interactions among the sub-sample revealed
differences in injury prevalence among APS and comparison group members. Compared to
clinic comparison group members, APS clients were more likely to have at least one upper
extremity abrasion (37.2% vs. 15.9%, p=0.01) or head, neck, or maxillofacial ecchymosis
(20.9% vs. 3.2%, p<0.01), swelling (9.3% vs. 0.0%, p=0.02), or tenderness (9.3% vs. 0.0%,
p=0.02).

Ecchymoses
Comparison of ecchymoses characteristics on the injury-level, yielded some differences
in the characteristics of ecchymoses found among APS and clinic comparison group members.
Table 3.4 presents these findings. One-hundred-fifty-two (n=152) ecchymoses were found
among study participants with 84 ecchymoses documented among APS clients and 68
ecchymoses found among comparison group subjects (see Table 3.4 below).

Table 3.4. Prevalence and Characteristics of Ecchymoses Reported Among Community-
Dwelling Older Adults by Recruitment Source (N = 152)
Total Sample APS (n=84) Clinic (n=68)

n % n % N % p
†

Location

71

Head, Neck & Maxillofacial 16 10.5 14 16.7 2 2.9 0.01
Upper Extremity 111 73 60 71.4 51 75 0.62
Trunk 9 5.9 5 6 4 5.9 0.99
§

Lower Extremity 16 10.5 5 6 11 16.2 0.04
Size

Maximum Length (cm) 2.8 ± 2.5 3.4 ± 3.0 2.0 ± 1.5 <0.01
‡

Missing 8

4

4

Maximum Width (cm) 2.3 ± 1.8 2.8 ± 1.8 1.7 ± 1.2 <0.01
‡

Missing 8

4

4

Pattern

Circular Pattern 9 6.6 5 6.9 4 6.8 0.97
§

Missing 21 21 12

9

Mechanism Causing Injury

Unknown Mechanism 52 34.2 20 23.8 32 47.1 <0.01
Grabbed by Another 21 13.8 21 25 0 0 <0.00
1
Fell 6 3.9 2 2.4 4 5.9 0.27
§

Struck by Person or Object 51 33.6 42 50 9 13.2 <0.00
1
Other Mechanism 27 17.8 3 3.6 24 35.3 <0.00
1
Missing

72

Note: †Calculated using Pearson chi-square test for independence; ‡Calculated using
Independent samples t-test; §Calculated using Fisher's Exact T-test; Subjects were able to
select more than one race category

A greater proportion of bruises found among APS clients were located on the subject’s
head, neck, and maxillofacial areas (16.7%) compared to 2.7% of bruises observed among
comparison group subjects. was more commonly found among APS clients (p=0.01).
Ecchymoses observed among APS clients were 1.4 cm longer and 1.1 cm wider, on average, as
compared to those observed among clinic comparison group members. There were no significant
differences found between groups in the shape or pattern of ecchymoses.
Mechanisms causing ecchymosis varied significantly between study arms. While half of
bruises found among APS clients were inflicted through contact with an individual or object,
13.2% of bruises found among clinic comparison group members originated in this way
(p<0.001). Similarly, 25% of bruises found among APS clients were attributed, by subject self-
report, to being grabbed by another individual, though none of the bruises occurring among
clinic comparison group members originated in this way (p<0.001). Instead, nearly half (47.1%)
of bruises found among comparison group members occurred, by subject self-report, through
unknown mechanisms while only 23.8% of ecchymoses found among APS clients had unknown
origins (p<0.01).

Abrasions
Comparison of abrasion characteristics on the injury-level revealed differences in the
location and mechanisms causing injury among abrasions found on APS and clinic comparison

73

group subjects; Table 3.5, below, presents these findings. There were 92 abrasions found among
study participants with 35 present among APS clients and 57 present among clinic comparison
group members.
Over half of abrasions found among APS clients were located on the subjects’ upper
extremities (54.3%) while only 24.6% of abrasions found among comparison group members
were similarly located (p<0.01). In contrast, 59.6% of abrasions found among clinic comparison
group members were found on subjects’ lower extremities, while only 31.4% of abrasions among
APS clients were similarly located (p<0.01). There were no observed differences between groups
in abrasion length or width. There were also no differences in the prevalence of linear or
circular-shaped abrasions.

Table 3.5. Prevalence and Characteristics of Abrasions Reported Among Community-
Dwelling Older Adults by Recruitment Source (N = 92)
Total Sample APS (n=35) Clinic (n=57)

n % n % N % p
†

Location

Head, Neck & Maxillofacial 12 13 4 11.4 8 14.0 0.72
§

Upper Extremity 33 35.9 19 54.3 14 24.6 <0.01
Trunk 2 2.2 1 2.9 1 1.8 1.00
§

Lower Extremity 45 48.9 11 31.4 34 59.6 <0.01
Size

Maximum Length (cm) 1.9 ± 2.0 2.0 ± 2.4 1.8 ± 1.7 0.51
‡

Missing 4

1

3

74

Maximum Width (cm) 1.2 ± 1.5 1.4 ± 1.3 1.2 ± 1.6 0.52
‡

Missing 4

1

3

Pattern

Any Pattern 18 21.7 7 22.6 11 21.2 0.88
Linear Pattern 13 14.1 4 11.4 9 15.8 0.33
§

Circular/Semi-circular 5 6 3 9.7 2 3.8 0.21
§

Missing 9

4

5

Mechanism Causing Injury

Unknown Mechanism 34 37.8 9 27.3 25 43.9 0.12
Grabbed by Another 6 6.7 6 18.2 0 0 <0.01
Fell 7 7.8 3 9.1 4 7 0.70
§

Struck by Person or Object 19 20.7 13 37.1 6 10.5 <0.01
Other Mechanism 26 28.9 4 12.1 22 38.6 <0.01
Missing 2

Note: †Calculated using Pearson chi-square test for independence; ‡Calculated using
Independent samples t-test; §Calculated using Fisher's Exact T-test

Subject-reported mechanisms causing abrasions differed between APS and clinic
comparison group members. Over a third (37.1%) of abrasions found among APS clients were
attributed to contact with another individual or object while 10.5% of abrasions found among
comparison group subjects were caused in this way (p<0.01). Similarly, 18.2% of abrasions
found among APS clients were attributed to being grabbed by another individual while none
(0%) of abrasions found among comparison group members originated in this way (p<0.01). In

75

contrast, compared to abrasions found among APS clients, 38.6% of abrasions found among
comparison group members were attributed to other injury mechanisms, a category reflecting a
diversity of other causes.

DISCUSSION
This study documented and characterized injuries present among older adult victims of
physical abuse receiving services from APS and a comparison group of clinic patients. Our
purpose was to identify injury markers of abuse which may assist healthcare and elder abuse
professionals in distinguishing between intentionally-caused, abuse-related injuries and those
occurring through non-intentional, accidental mechanisms. Study findings identified several
commonly occurring injuries and affected areas present among the APS population as well as
characteristics which distinguished injuries seen among APS clients with those seen among non-
abused older adults.
The study was unique from others in the field because it examined injuries among the
APS population, a group that is difficult to access and study (Wiglesworth, et al., 2006).
Exploration and evaluation of injuries within this population is especially important because they
constitute a large population of victims who may never otherwise interact with medical or
criminal justice systems and are, thus, outside the sampling frame used by most prior conducted
studies. The sample also captures individuals who seek care from providers in medical settings
but do not disclose that their injuries occurred as a result of abuse. Our findings suggest that
injuries present among abuse victims at the time of APS assessment and interview may differ
from those present among victims seeking care in hospital settings. While hospital-based work
has documented the incidence of open wounds, penetrating injuries, internal injuries, fractures,

76

and to a lesser extent, bruising and lacerations in victims of abuse (Friedman, Avila, Tanouye &
Joseph, 2011; Rosen, et al., 2016; Gironda, Nguyen & Mosqueda, 2016), the present study found
and documented less severe injuries, such as ecchymoses, abrasions, tenderness in the absence of
other injury, swelling, erythema, and lacerations. It may be that those with more severe injuries
are more likely to seek emergency care.

Tenderness
To our knowledge, this study is the first to document the prevalence of tenderness
occurring in the absence of another observable injury among victims of abuse. While prior
research on elder abuse has not studied this injury type, pain and tenderness have been studied in
other areas of family violence and have been positively associated with perpetration of child
maltreatment (Kellogg & Committee on Child Abuse and Neglect, 2007; Johnson & Showers,
1985) and sexual assault (Sommers, et al., 2012).
In the present study, the significantly higher prevalence of tenderness among victims of
abuse, compared to non-abused comparison group members may suggest that area tenderness is a
documentable physical finding relevant in abuse assessment and work-up. Data suggest that
tenderness occurring in the head, neck, or maxillofacial regions may be indicative of abuse,
especially given that this particular injury-location combination occurred exclusively among
APS clients. Additional research is needed to further document this phenomenon and replicate
the present study’s findings.

Ecchymoses

77

Many of our findings were consistent with those found by Wiglesworth and colleagues
(2009), to our knowledge, the only researchers to have studied injuries among APS clients in
community settings. Although we found that abused APS clients were more likely to have
ecchymoses located on the head, neck, or maxillofacial region, we were unable to replicate their
finding of greater prevalence of abuse-related injuries to the posterior torso. These injuries
occurred infrequently in the study sample, and post-hoc power analyses using G*Power 3.1.9.2
found that the sample size of ecchymoses present on the subject’s trunk (3 APS subjects; 1 clinic
subject) did not provide enough statistical power to detect differences in prevalence among
groups. While location alone is not indicative of abusive injury origins, it may be used as a
marker to warrant further evaluation and consideration of other distinguishing characteristics,
such as injury size, color, or mechanism of origin.

Abrasions
Though recognized as a potential sign of abuse (Dyer, et al., 2003; Palmer, Brodell &
Mostow, 2013; Collins, 2006), to our knowledge, this was the first case-comparison study to
examine abrasion prevalence and characteristics occurring among abused and non-abused older
adults. Abrasions occurred in nearly half of study subjects, though prevalence did not vary
significantly between APS and comparison group members.
However, injury-level analyses of abrasion location and subject-reported injury
mechanism provide some evidence to suggest that the abrasion characteristics may differ among
groups. Findings that a larger proportion of abrasions found among APS clients were located to
the upper extremities (54.3% of abrasions found among APS clients) compared to clinic
comparison group members (24.6% of abrasions found among clinic members) are consistent

78

with findings that a higher proportion of abrasions present among APS clients were caused by
subjects to being grabbed by another individual (18.2% of abrasions found among APS clients)
or being in contact with another person or object (37.1% of abrasions found among APS
clients). Such injury-causing mechanisms may be more likely to affect the upper extremities.

Injuries by Location
Study findings analyzing the differences in the prevalence of specific injury diagnosis
and location interactions between abused APS clients and non-abused clinic comparison group
members contribute to the body of developing work on markers of physical abuse. Our findings
replicated prior studies in findings a significantly higher prevalence of ecchymoses located on
the head, neck, and maxillofacial regions among abused older adults compared to non-abused
comparison group members (Wiglesworth, et al., 2009; Ziminski, et al., 2013; Rosen, et al.,
2015).
Our findings also highlighted injury types and locations that, to our knowledge, have not
been previously described. Evidence of greater prevalence of abrasions to the upper extremities
of abused APS clients, as well as the greater prevalence of swelling and of tenderness to the
head, neck, and maxillofacial regions suggest that these injuries located in these locations may be
more likely to occur among victims of abuse. As such, they should be considered as possible
markers of abuse and their presence during interview or physical assessment should warrant
further screening or examination to rule out abuse.

79

Absence of Injury Among Abuse Victims
Because the study recruited abuse victims through APS, a community-based agency, and
validated abuse allegations through the use of a LEAD panel, researchers were able to document
instances of physical abuse which did not result in physical injury. Such individuals have been
outside the sampling frame of prior conducted hospital- or clinic-based studies on physical elder
abuse because their lack of injury preclude them from needing medical care.
Injury-absent elder abuse victims, victims who experienced abuse but had no sign of
injury, constituted 22.8% of the population of APS-recruited study subjects. Though additional
research is needed, the present study suggests that uninjured victims may constitute a sizable
proportion of the physically-abused population. These findings provide evidence that the absence
of physical injury, alone, does not negate the possibility that physical abuse may have been
perpetrated.
It’s important to acknowledge that the timing of data collection has some effect on the
types of observable injuries present at the time of examination. Work by Mosqueda and
colleagues (2005) to document and describe bruising patterns among older adults found that
bruises remained visible for varying lengths of time with the fastest resolving bruise being
indiscernible after 4 days. Eighty-one (81%) percent of bruises were resolved by Day 11 of their
study, suggesting that assessments conducted after these time periods may not have detected
abuse-related bruising. Though timing of data collection may have contributed to findings of
non-injury among the APS study sample, the extent of this issue may be limited as data
collection for most (64.3%) of the APS cases in which subjects lacked physical injury (n = 9 out
of 14) occurred prior to Day 11.

80

Understanding circumstances and protective factors contributing to the absence of
physical injury among abused APS clients is as important a contribution to the development of
forensic science informing physical abuse as identifying and describing the mechanisms that
contribute to physical injury. A clearer understanding of these topics may enable practitioners to
better evaluate and adjudicate allegations of physical abuse, which is often a key intermediating
step in obtaining needed services or pursuing justice. The identification of these mechanisms
provides criminal justice professionals with data to confirm the validity of victim and witness
statements recounting abuse. Environmental or individual protective factors or actions associated
with the lack of physical injury among abuse victims may also inform efforts to protect victims
in future physical abuse incidents.

Limitations and Challenges
This study sought to document injuries present among physically abused APS clients, a
potentially vulnerable population of older adults who had experienced a recent physical,
emotional, and social trauma. Relative to the total number of potential APS clients eligible for
the study, less than 5% of APS-identified potentially-eligible clients agreed to participate as a
result of client denials of abuse, refusals to accept APS services, and refusals of study
participation. Moreover, efforts to collect data as soon as possible after the abusive event were
complicated by APS protocols ensuring that important safeguards were adhered to and physical
abuse investigations were expedited. A central APS program administrator sent cases that
appeared to meet study criteria to the study Research Assistant who contacted case-assigned APS
workers to coordinate a joint home visit. Although agreed-upon interagency research protocols

81

dictated that the research nurse would be able to conduct joint visits with APS workers making
their initial home call, this was often impractical.
Moreover, due to the immediate threat of physical harm, many cases alleging physical
abuse were assigned to the APS After Hours Unit, a 24-hour response unit which dispatched
workers immediately upon receipt of the allegations. The immediacy of response required
workers to respond within 24 hours, often prior to research staff notification of a potential study
case.
Several important protocols delayed case review. In many cases, APS workers did not
initially seek permission from the client to enable researchers to contact them and had to re-
contact the client to request permission, often with little success. Researchers also experienced
difficulty in establishing contact with clients and scheduling a meeting within the study
eligibility period of 21 days following an abusive incident.
Among APS clients, the median number of days between abusive incident and data
collection was 12 days. Because injury healing can affect the visibility and presentation of abuse-
related injuries it is possible that some injuries were unobservable at the time of data collection
and others were reduced in scale due to healing effects.
Finally, despite efforts to systematically and comprehensively collect injury data, a lack
of taxonomy to describe basic injury characteristics for each type of diagnosis may have
contributed to missing data on some injury characteristics. Data collection in both study arms
was conducted by the same research staff increasing the likelihood that data were collected
uniformly in both arms. Thus, missing data should not have a notable effect on inter-group
comparisons. Nevertheless, a classification taxonomy for visible geriatric injuries that provides
recommended data categories for each injury diagnosis may improve data collection efforts in

82

future studies (Rosen, et al., 2016). Though their work was conducted in parallel to the present
study, many of the data points correspond with our findings, providing insight into key injury
characteristics that should be considered when collecting data on geriatric injuries.

PUBLIC HEALTH IMPLICATIONS
Study findings contribute to the growing body of research on geriatric assault and assault-
relate injuries, providing insight into less severe injuries that may be present and observable in
community-dwelling populations. Greater understanding of injury prevalence and characteristics
holds promise for improving abuse detection and adjudication by health providers and APS
workers. In turn, these steps can enable victim access to support and resources to empower them
to protect themselves from further abuse, seek assistance from health and social service
providers, and seek justice through legal recourse.
These findings make an important contribution to the detection of injuries suspicious for
abuse, a necessary first-step that can lead to initiation of protocols to engage in further interview
and evaluation. Earlier abuse detection and intervention hold promise to protect and prevent
future escalation of violence and abuse.

83

Chapter IV: Toward a Better Understanding of the Elder Abuse Forensic Center Model:
Comparing and Contrasting Four Programs in California

Forensic science, the “application of scientific or technical approaches to the recognition,
collection, analysis, and interpretation of evidence for criminal and civil law or regulatory
issues” (National Commission on Forensic Science, 2016), can be applied to strengthen and
promote elder justice. Elder Abuse Forensic Centers (FC), first developed at the University of
California, Irvine in 2003, are forensic agencies which convene a multidisciplinary team (MDT)
of professionals to assist in cases of elder abuse, neglect, and financial exploitation (Schneider,
Mosqueda, Falk & Huba, 2010).
Elder abuse, increasingly recognized as a significant social and public health problem is
estimated to affect 141 million or one in eight older adults worldwide (Yon, Mikton, Gassoumis
& Wilber, 2017). In the U.S., the most recent national prevalence study found 10% of older
Americans had experienced an abusive incident within the past year (Acierno, et al., 2010).
Elder abuse is defined as “intentional actions that cause harm or create a serious risk of
harm, whether or not intended, to a vulnerable elder by a caregiver or other person who stands in
a trust relationship to the elder,” and/or “failure by a caregiver to satisfy the elder’s basic needs
or to protect the elder from harm” (Bonnie & Wallace, 2003, pp. 1). It includes physical,
psychological and sexual abuse as well as financial exploitation and caregiver neglect. A related
phenomenon, self-neglect, is often considered a form of elder abuse because it carries high risk
of harm to a vulnerable older adult, is frequently reported to protective services, and significantly
increases risk for inter-personal abuse and emergency medical care (Dyer, Goodwin, Pickens-

84

Pace, Burnett & Kelly, 2007; Wilber & McNeilly, 2001; Dong, Simon, Mosqueda & Evans,
2012).

Responding to Abuse
Adult Protective Services (APS) takes reports of elder abuse that occur among
community-residing older and dependent adults and investigates the allegations; the Long-Term
Care Ombudsman plays a similar role for facility residents. However, because elder abuse and
self-neglect are often complex and difficult to resolve, remedies may require input from multiple
agencies and providers, including a cross-section of social, health care, and legal services
(Connolly, 2010). As these efforts are often siloed, fragmented, and difficult to coordinate, elder
abuse multidisciplinary teams (MDTs) have been used to bring specialist together to integrate
services and supports across agencies (Connolly, 2010; Elder Justice Roadmap, 2012;
Wiglesworth et al., 2006; Wilber, Navarro, & Gassoumis, 2014).

Elder Abuse Multidisciplinary Teams (MDTs)
Over the last several decades a number of different types of MDTs have been developed
to address elder abuse, including the Fiduciary Abuse Specialist Team (FAST) which facilitates
interaction between financial services experts and APS (Dyer, Heisler, Hill & Kim, 2005); the
Vulnerable Adult Specialist Team (VAST) which brings together APS and health care
professionals (Mosqueda, Burnight, Liao & Kemp, 2004; Dyer, et al., 1999; Breckman, Callahan
& Solomon, 2014); and fatality review teams which examine elder deaths suspected of being
associated with abuse (Stiegel, 2005; Teaster, Nerenberg & Stansbury, 2003). A nationwide
study of MDTs found that although most teams provide case review and expert consultation

85

there was considerable variation in MDT characteristics (e.g., size, composition, activities;
Teaster, et al., 2003).
Building on and strengthening the MDT approach, the FC model brings together
professionals from the health, social services, and justice systems to advise frontline workers,
most often APS and law enforcement. In addition to case review and advisement, FCs include
dedicated staff (project director, program coordinator) available to support and at times carry out
team recommendations (Schneider et al, 2010; Gassoumis, Navarro & Wilber, 2015). In 2006,
the Archstone Foundation initiative to “improve[e] the quality and coordination of elder abuse
and neglect services in the state of California” (Giles, Brewer, Kullman & Prevratil, 2010)
funded new centers in Los Angeles (LA), San Francisco, and San Diego in addition to the
original FC at UC Irvine (Schneider et al., 2010; Navarro, Wilber, Yonashiro & Homeier, 2010).
Expansion of FCs was further promoted through its inclusion in the Elder Justice Act of 2010.
Although this Act provides federal authorization to appropriate funds to support the
establishment and operation of FCs across the country (Elder Justice Coalition, 2016), funding
has been limited.
The LAFC has been the focus of several studies, primarily on the model’s structure,
processes, and outcomes (Navarro, Wysong, DeLiema, Schwartz, Nichol & Wilber, 2016;
Navarro, Gassoumis & Wilber, 2013; Gassoumis, et al., 2015). This FC convenes weekly to
review cases, make recommendations, and when indicated conduct home visits to assess the
client’s physical, cognitive, and mental health capacities and needs for support. The FC also
provides training to professional groups on topics related to abuse, with the goal of strengthening
community capacity to respond to elder abuse. Most workers using the FC (94%) stated they
would utilize FC services again, and 98% said they would recommend the FC to others (Navarro,

86

et al., 2010). Participating FC members evaluated the FC’s work very positively overall and with
regard to communication, empowerment, and achievement (Navarro, et al., 2010).
Outcomes studies found that financial abuse cases seen at the LAFC had a 10 times
greater likelihood of being submitted to the District Attorney’s office for review and 8.5 times
greater likelihood of a guilty plea or conviction outcome (Navarro, et al., 2013). Prosecutors
believed that enhanced prosecutorial outcomes resulted from the participation of key decision
makers around the table, availability of medical and psychological expertise, and improved
interagency cross-learning. They indicated that participation broadened their approach to cases
and increased their knowledge about the work of other agencies and service providers (DeLiema,
Navarro, Moss & Wilber, 2015). FC clients also had nearly 7 times greater odds of being
referred to the Public Guardian’s office for conservatorship, with a similar rate of successful
conservatorships among those referred (Gassoumis, et al., 2015).
Although these findings suggest that the model can achieve greater effectiveness and
efficiency in achieving programmatic outputs in areas of justice, safety, interagency
collaboration, and enhanced professional knowledge, information is needed from multiple FCs in
different areas and settings to further validate these observations. Therefore, the purpose of this
study is to identify, compare, and contrast essential program components, characteristics, and
impact across four different Forensic Centers to inform program standardization and model
fidelity.

DESIGN AND METHODS
To better describe structural and programmatic components and process outcomes, we
conducted a retrospective, systematic cross-site comparison of four FCs. Data were collected on

87

FCs’ structures and operations. This study was approved by the University of Southern
California Institutional Review Board (IRB).

Sites and Participants
Data were collected from FCs in Los Angeles, Orange County, San Diego, and San
Francisco, each of which received funding and technical assistance from the Archstone
Foundation’s initiative to develop or enhance their program. Specific site descriptions are briefly
described below, and additional site descriptions have been published elsewhere (Schneider, et
al., 2010). At the time of recruitment, each site had functioned as a self-identified FC for over a
year, serving older adults aged 65 years and older or adults aged 18 and older living with
conditions resulting in their physical, cognitive, or mental dependency on others. Analysis of
established sites ensured that programs were stable and had sufficient time to mature, develop
protocols and procedures, and address start-up issues.
The Orange County Elder Abuse Forensic Center was established by the University of
California Irvine in 2003, emerging in response to an identified lack of coordination between the
criminal justice system and medical and social service providers (Schneider, et al., 2010). The
program serves residents living in Orange County.
The LA County Elder Abuse Forensic Center was established in 2006, modeled after the
program launched in Orange County. The LA site serves residents living in the County of LA, an
ethnically and socio-economically diverse urban area.
The San Diego HOPE (Help and Outreach to Protect the Elderly) Team was first
convened in 2006 by the San Diego District Attorney’s Office. Envisioned to be an elder justice
program collaboratively engaged with existing family violence intervention programs, the HOPE

88

Team was co-located with the San Diego Family Justice Center to facilitate inter-program
interactions and activities. Central to their work, the HOPE Team sought to provide “wrap-
around” victim services, an approach applied in other fields of family violence, to “cloak”
victims with support through the development of individualized service plans.
The San Francisco Elder Abuse Forensic Center was established in 2008 as a public-
private partnership between the City of San Francisco and the Institute on Aging, a non-profit
organization serving older adults. The FC was established based on lessons learned from a prior
established elder abuse MDT - the Multidisciplinary Assessment Team (MAT), which convened
monthly meetings with APS, the Public Guardian, Long-Term Care Ombudsman, law
enforcement, and City Attorney’s Office to obtain client evaluations from a MAT
geropsychologist and geriatrician (Schneider, et al., 2010). Prior to establishing the FC, San
Francisco convened a Fiduciary Abuse Specialist Team (FAST) MDT which included several of
the same member agencies. Because of the overlap, members decided to incorporate the FAST
into FC operations.

Data Collection
Our primary source of data were quarterly Archstone Foundation grant progress reports
which were completed by each FC’s staff and submitted to study researchers between January
2010 to December 2011. Reports included information on the FC’s purpose, structure, processes,
outputs, and outcomes. When applicable, sites submitted their program’s governing documents,
collaborative agreements, and memoranda of understandings with collaborating agencies to
provide researchers with supplementary data. Centers also provided researchers with 3-months of

89

de-identified case intake data, which included victim characteristics (demographics, physical
functional status, medications, cognitive status), types of abuse alleged, and the referring agency.
If additional information or data clarification were necessary, researchers contacted FC program
managers who responded to provide the requisite information.
To complement archival data, researchers also conducted a site visit to each FC between
February and June 2011. While there, researchers observed a team meeting and conducted group
interviews with core team members to gather data on FC program planning, infrastructure, and
day-to-day implementation. A structured protocol was used to document FC attendees, types of
cases presented, and types of interaction between attendees (i.e. question, clarification,
information, professional interpretation, and recommendation).
After each Center’s observed team meeting, members were asked to anonymously
complete the Team Effectiveness Inventory (TEI), a 24-item survey assessing team mission, goal
achievement, empowerment, communication, roles/norms, and overall team effectiveness
(Navarro, et al., 2010). Members also were asked to (1) provide written feedback to the question,
“How has your involvement with the Forensic Center changed the way you practice?” and (2)
indicate, on a 4-point Likert scale ranging from strongly agree to strongly disagree, their
agreement with the statements (a) “Participation on the team has changed the way I do my job”
and (b) “Information I learn from the team discussions is information I use on other cases I
encounter.”
Because we were interested in both the program practices occurring during the grant
funding period as well as program sustainability, in 2017 we ascertained whether sites were still
in existence and self-sustaining five years after the Archstone Foundation funding was
completed.

90

Variables
The 4 FCs were compared and contrasted across four essential categories – program
structure, key activities, clients served, and outcomes.
Program structure was broadly defined to include organizational components, physical
environment, and program protocols. This included organizational components, such as Forensic
Center leadership, personnel, MDT composition, and interagency relationships; physical
environment such as Forensic Center location, meeting facilities, and remote meeting
capabilities; and program operations, such as MDT meeting frequency and duration, types of
cases reviewed, and information sharing practices.
Program activities were also reviewed to identify the specific actions undertaken by each
Center. These included regular, frequently occurring actions, such as convening MDT meetings,
as well as infrequent, though team-valued activities, such as annual team retreats or trainings.
Program impact was operationalized to include identifying the characteristics of FC
cases, the level of MDT self-assessed effectiveness, and professional impact. Case characteristics
included the number and types of abuse perpetrated, agency referring the case to the FC, and the
affected older adult’s age, gender, race, and cognitive status. MDT effectiveness was assessed
using a modified version of the Team Effectiveness Inventory previously used in research on
FCs (Navarro, et al., 2010). Professional impact was operationalized to include ways in which
team member participation in the FC affected their everyday professional practice.

91

Analysis Strategies
Qualitative data extracted from FC grant reports, operational documents, and site visit
notes were compiled into descriptive summaries that were used to compare site characteristics.
Two researchers independently identified similarities and differences across FC sites.
Free-response sections of the modified TEI were entered into NVivo version 11 for
analysis. Responses were independently content-analyzed by two researchers using inductive
approaches modeled after grounded theory methodologies. Initially, researchers read the
responses and, together, inductively identified primary codes. Responses were then
independently reviewed and coded using the agreed-upon codes. Finally, coded responses were
again discussed and interpreted by the two researchers.
De-identified FC case data and TEI scores were analyzed through descriptive statistics
and tests of association, including independent samples t-tests, and chi-square tests for
independence, using SAS version 9.4.
Study findings were shared with site representatives during cross-site in-person and
telephonic meetings, and sites were invited to provide further explanation or clarification of
results. When necessary, unclear statements or observations were clarified by contacting sites
directly through email or phone.

RESULTS
Program Structure
Leadership & Staffing

92

Each of the four sites employed a full-time program coordinator and program assistant
(Fig. 4.1, below). Program coordinators fostered interagency communication and collaboration,
managed day-to-day program operations, planned and coordinated MDT meetings, completed
case intake and preparation for presentation, and tracked client and case data and outcomes over
time. Program assistants provided instrumental support to coordinators, often managing data
systems and following up with cases to obtain information on client outcomes.

93

Sites varied in overall leadership structure and composition. Though day-today
operations and leadership were left to program coordinators, founding leaders often served as

94

program Directors. Sites A & B were founded and directed by a geriatrician; Site C was
established by a geropsychologist, and Site D was led by an attorney.

MDT Composition
Table 4.1 shows five core team members common as regular attendees across all four
sites in addition to the program manager: APS, law enforcement, prosecutors, the Public
Guardian and victim advocate. In Sites A, B, and C, a geriatrician or healthcare provider
attended regularly and though in Site D their attendance was occasional. Sites A, B, and C also
had gero- or neuropsychologists in regular attendance, though this team member was not present
in Site D. The LTC Ombudsman attended regularly in Site B and occasionally in Sites A & C.
All sites used Memoranda of Understanding/Agreement (MOU/MOA) to formalize collaborative
agreements with participating MDT member agencies.
Additional team members were used to expand services, support an existing partnership,
or support direct care provider efforts to protect clients, investigate, intervene or pursue justice.
These included County mental health services (n=2), Intimate Partner Violence
organizations(n=1), developmental disability services (n=1), community care licensing (n=1),
county counsel (n=1), and the Coroner’s office (n=1).

Table 4.1 Cross-site Comparison of Multi-Disciplinary Team Composition
Site
A B C D
FC Program Coordinator/Manager
Geriatrician or Healthcare Provider

95

Geropsychologist or Neuropsychology
Adult Protective Services
Law Enforcement Agencies
Prosecutorial Agencies
Public Guardian
Victim Advocate
LTC Ombudsman
Senior Legal Aid
Community Mental Health Services
Intimate Partner Violence Organization
Developmental Disability Services
Coroner’s Office
Community Care Licensing
Gerontologists
Legal Counsel for Participating Agencies
Key:

Structural Environment
All four sites had designated facilities that housed day-to-day operations. Each was co-
located with other service providers and agencies; Sites B and C were co-located with APS, and
Sites A and D were co-located with violence intervention programs. Each site had access to a
designated meeting space which included a large central table to facilitate meeting discussions
Regular Attendee Not on Team Occasional Attendee

96

and proceedings. Sites A, B, and C held meetings in their offices, and Site D used a separate
location in the court building to enable members to access court records during the meeting.
Geographic dispersion and provider time constraints sometimes made it difficult for team
members and case presenters to attend meetings in-person. All sites were equipped with
telecommunication technologies, (teleconferencing (Sites A, B, and C) and videoconferencing
(Site D) to enable participants to have a virtual seat at the table. While Site A relied heavily on
telecommunication approaches, teleconferencing in 1-3 presenters each meeting, Sites B and C
reported that teleconferencing was seldom used.

Team Meetings
Most FC decision-making occurred during team meetings. Although all sites established
regularly-scheduled formal meeting times, meeting frequency and duration varied (Table 4.2).
Sites A and B conducted meetings weekly and Sites C and D met bi-monthly. Meeting duration
ranged from 1-2 hours. Thus, in a given month, one MDT met for a total of 2-3 hours while
another met for 8 hours. None of the sites restricted their focus to a specific type of abuse,
though one site did not review any cases involving self-neglect.

Table 4.2: Cross-Site Comparison of Meeting Characteristics
Site
Meeting Characteristics A B C D
Number per Month 4 4 2 2
Duration (Hours) 2 1 1.5 1-1.5
Number of New Cases Reviewed 3-4 1-2 2 4-5

97

Time Spent Reviewing Case (Minutes) 20-45 20 20-30 10-15
Number of Follow-up Cases Discussed 1-3 2

Information Sharing
During FC meetings, presenters shared basic facts of the case, including confidential case
information which would otherwise be restricted from disclosure in usual-care settings. Team
members located and shared additional confidential, otherwise restricted, material about the
client, suspected abuser, or case. For example, an APS worker might present the case by
describing core facts (e.g., information about the client and suspected abuser, their relationship,
what happened over time). After the presentation of basic facts, members asked questions and
searched for relevant information: detectives might obtain records on 911 call histories, police
visits to the home and criminal history county mental health providers may find and report on the
agency’s services while prosecutors might search for any criminal history on the suspected
abuser.
Such real-time information sharing observed across all four Centers is supported by
California statues, which allow MDTs to share protected client information to facilitate the
provision of services (CA WIC section §10850.1). This enables team members to avoid more
formal information-seeking approaches such as court orders, subpoenas, search warrants, and
administrative information requests.

Key Activities
Several activities were common across all sites.

Case Review & Real-time Decision-Making

98

Observations of case review, the primary activity of FC meetings, showed similar
features across all sites. Presenters (which varied by FC, as described below) provided case facts
and circumstances, shared their professional concerns and goals for remedies, and sought
recommendations from the team. In all sites, the case review process entailed an initial
presentation, followed by MDT-presenter discussion in which team members asked the presenter
questions for clarification and to obtain further information. Then, drawing on their professional
expertise and past experiences, team members helped create an action plan by identifying case
recommendations and action steps. These included discussing strategies for working with
hesitant clients, identifying additional information needed and best practices for gathering
information from the client and other community members, and making referrals to additional
sources of support for both the client and presenter. If appropriate, members also offered their
professional services to assist the presenters and their clients (e.g. conducting cognitive
assessments, medical chart reviews, and accepting service referrals). To conclude the discussion,
the program coordinator summarized agreed upon recommendations and action items to clarify
decisions and recommendations which were then used as a basis for case tracking and updating
over time. Cases were revisited and re-discussed during later meetings if presenters found further
consultation or resources were needed.
FCs adopted two different approaches to case presentation, Sites A and C invited the
person(s) working the case to present, and Sites B and D had designated FC staff to introduce
case facts and facilitate discussions with presenters to streamline the process and mitigate
presenter burden.
FCs also varied in the number of new and ongoing cases reviewed each meeting and the
overall average of new cases seen per month (See Table 4.2, above). Average time for

99

presenting and reviewing a case ranged from 10-15 minutes to 20-45 minutes. In three sites,
cases were most often referred from APS (range 74.8-92.1% of cases), though other member
agencies were encouraged to bring cases. One site dedicated one meeting per month for non-FC
agencies to present cases to the team, expanding the research and awareness of the FC beyond
the team.

Consultation/Assessments
All four FCs provided short-term consultative activities to FC presenters, providing them
with instrumental professional assistance to support their investigation or service provision.
These activities included medical, psychological, and cognitive assessments; medical record and
legal document review; explanation of agency and system protocols; and service or resource
referral. In three sites, assessment of client decision-making capacity was identified as one of the
most important services provided. Routinely conducted by FC members or subcontracted
geriatricians or specialized psychologists (geropsychologists/neuropsychologists), these
assessments provided information on client capacity, sometimes suggesting or supporting a
particular course of action. For example, the determination of whether a client had capacity at the
time of a financially exploitive deed transfer may determine what information was needed for the
investigation and what charges might be filed against the alleged perpetrator. Client capacity also
influenced decisions related to needs for protection and whether filing for conservatorship should
be recommended if the client was at risk and vulnerable because of possible lack of capacity.
Despite the importance of medical and psychological evaluations, Centers reported that
they were among the most challenging activities to provide. Scheduling medical and
psychological assessments and following-up meetings with clients was often time consuming. In

100

addition, these assessments required preparing reports and, in some cases, extensive
documentation. Some FC team members provided assessment services in-kind as part of their
paid professional responsibilities and others were compensated based on fee-for-service billable
hours.

Documentation
All four FCs documented their activities and case outputs. This documentation served as
a central agency-independent record of case review discussions, including client and case
information and proposed team recommendations; team member activities, such as medical and
cognitive assessment notes and results, capacity declarations, consultative reports, and legal
testimony; case processing, such as status, updates and where available information on short- and
long-term outcomes. Updated client and case information was obtained by FC staff through
follow-up with case presenters and team members.
Documentation not only informed future case discussions, decision-making, and
professional activities, but also enabled the examination of program impact and identification of
best practices for replication and dissemination. Nevertheless, across the four FCs, staff reported
difficulties in longitudinally documenting case status and outcomes, citing the time-intensive
efforts required to obtain updates from case presenters and team members. Challenges were
compounded by agency staff turnover, as involved service providers who were transitioned to
other positions sometimes had to hand-off clients to co-workers who were unfamiliar with the
FC, reluctant to respond to requests for updated information, and/or limited in their knowledge of
the case history. Staff also reported difficulties with data management. Challenges included
working with several different documents, most often a spreadsheet, word processing, and/or

101

database program, and often requiring duplicative data entries. Outcome data were constrained
by data management and client tracking problems. We were unable to conduct a comparative
analysis of case outcomes using quantitative case data collected across sites FCs due to the lack
of follow-up information and inter-site differences in data categories and definitions.

Training/Outreach
Each of the four Centers provided community training and outreach. Some were ongoing,
such as hosting students and fellows from different disciplines to observe FC meetings after
obtaining prior permission and signing confidentiality agreements. Others, such as organized
community events and outreach efforts, were time-intensive. In addition to strengthening
community capacity to address abuse, FC members indicated that trainings were valued as a
means to increase community awareness of the FC and its work, build case referrals, and garner
community support.

Team Building
Because the FC’s success largely rests on the team’s ability to collaborate, centers
identified ongoing team-building activities, such as annual team retreats, to encourage
relationship-building, promote group cohesion, re-energize the team, and facilitate cross-learning
opportunities.

Table 4.3. Comparison of Client and Case Characteristics by Site
Total Sample Site A Site B Site C Site D

102

N=801 n=284 n = 144 n = 76 n=297
n % n % n % n % n %
Client Age

65+ 662 82.7 117 81.3 117 81.3 59 77.6 255 85.9
Missing 29

1

1

-

22

Overall Mean ± SD 75.9 ± 14.6 76.8 ± 16.0 75.4 ± 16.9 74.5 ± 14.3 75.7 ± 11.6
Among Elders (Mean ±
SD)
80.5 ± 8.5 82.7 ± 7.9 82.1 ± 7.1 80.3 ± 9.1 77.69 ± 8.7
Female 417 52.1 168 59.2 98 68.1 47 61.8 104 35.0
Missing 148

3

-

145

Client Race

White 452 56.4 132 46.5 113 78.5 20 26.3 187 63.0
African-American 104 13.0 57 20.1 2 1.4 12 15.8 33 11.1
Hispanic 106 13.2 47 16.6 7 4.9 4 5.3 48 16.2
Asian and Pacific
Islander
45 5.6 12 4.2 14 9.7 6 7.9 13 4.4
Other 23 2.9 11 3.9 3 2.1 4 5.3 5 1.7
Missing 41

25

5

-

11

Abuse Type

Financial 422 54.2 200 70.4 67 46.5 49 64.5 106 38.4
Emotional 249 32 58 20.4 19 13.2 23 30.3 149 54.0
Physical 194 24.9 15 5.3 22 15.3 16 21.1 141 51.1

103

Neglect 157 20.2 70 24.7 57 39.6 22 29 8 2.9
Sexual 17 2.2 6 2.1 8 5.6 - - 3 1.1
Self-Neglect 155 19.9 90 31.7 45 31.3 20 26.3 0 0
Other 141 18.1 86 30.3 18 12.5 32 42.1 5 1.8
Missing 22

1

-

-

21

Referring Agency

APS 429 53.6 213 75 107 74.3 70 92.1 39 13.1
Law Enforcement 88 11.0 56 19.7 17 11.8 - - 15 5.1
Prosecutor's Office 104 13.0 4 1.4 6 4.2 - - 94 31.7
Other Agency 45 5.6 10 3.5 12 8.3 3 4.0 18 6.1
Self-Referral 22 2.8 - - - - - - 22 7.4
Missing 112

1

4

3

0

Note: APS = Adult Protective Services

Clients Served
A majority of cases seen at FCs involved older adult victims aged 65 or older (Table 4.3,
above) with the exception of one site, Site D. This site’s clients were much less likely to be
referred through APS; instead a majority of referrals were received from the District Attorney’s
office. Clients were more likely to be younger and male compared to the other sites. This was
also the only site that received cases directly through client self-referral.

104

Process Outcomes
Similar results were observed among all FCs in three main areas: (1) Member assessed
Team Effectiveness; (2) Personal Practice, which describes the professional skills and
knowledge applied both within FC casework, and in practice not related to the FC; and (3) Inter-
agency Relationships a theme that emerged from open-ended responses, and refers to rapport
with professionals outside one’s agency or discipline.

Team Effectiveness
When surveyed using the modified Team Effectiveness Inventory (TEI), team members
in all sites reported global team effectiveness (Figure 4.2, below) average scores ranging from
3.4 to 4.0 on a 4-point Likert scale. Studied domains included team mission; goal achievement;
empowerment; communication; and roles and norms.

Open responses indicated that the FC helps members accomplish investigative and
prosecutorial goals. One member responded that FCs are “the most effective tool for
1
1.5
2
2.5
3
3.5
4
Team Mission Goal Achievement Empowerment Open, Honest
Communication
Positive Roles &
Norms
Global Score
Socre (1-4)
Aspects of Team Effectiveness
Fig. 4.2 - Cross-site Comparison of Modified Team Effectiveness Inventory
(N-47)
Site A Site B Site C Site D

105

professionals who...assist abuse victims,” and access to other team members was identified as an
essential element. Specifically, case consultation and assessments were described as “an
invaluable service,” facilitating client protection and coordination of services. According to one
prosecutor:

“We are able to prosecute more cases efficiently and quickly due to instant communication of
all members in one room at the same time, and to have quicker access on the phone or by
email when not in team session.”

Members noted that cases brought to the FC received more comprehensive review and
the multiple perspectives facilitated “the whole person approach to the problem.” This was
accomplished “with immediate feedback on possible solutions and different perspectives on
[each] case.” One member noted that “increased utilization of team contacts... [helped us] reach
desired outcomes.” Another noted, “Now that we have these services, we are less likely to feel
hopeless about complicated cases that require interagency collaboration.”

Personal Practice
Team members at all four sites reported that participation in the FC had positively
affected their professional practice with average scores across sites ranging from 3.3-3.6 out of
4.0 on a Likert scale. Open-ended written responses from members described taking a broader
perspective by utilizing knowledge gleaned from members of other disciplines.
In two sites, FC members reported an enhanced ability to pursue justice through
prosecution or civil remedies. One member wrote, “As a... [social worker] I have more

106

information about what is needed in order to get more cases prosecuted,” highlighting knowledge
outside the usual scope. Because they anticipate the documentation requirements for criminal
prosecution, members have begun to investigate and document cases differently. One team
member wrote,

“I approach cases differently, using a more forensic eye…. I document differently, more
specifically and carefully, always now considering that the case may end up in court”

Enhanced personal practice extended outside focal FC cases. When surveyed in general
about the FC, team members across sites noted that they applied information learned to non-FC
cases in their own practice (see Fig. 4.1). Open-ended written responses clarified that FC
participation is educational for observers who are then “able to investigate...[their] own case very
thoroughly.” Learning from case discussions, FC participants applied techniques to similar cases.
One member mentioned that meeting participation “assisted me in my thinking about each case,”
while another noted it “helped...[to] handle...day to day cases greatly,” and still another said FC
involvement “makes my job easier.” Heightened awareness of relevant information has enhanced
investigations, and knowledge of resources are “extremely important for the job I do every day.
It has allowed me to understand the subject matter at hand,” one respondent explained.
Similarly expanded knowledge among supervisors impacted multiple programs and staff.
One team member reported lessons learned enabled the member to “make more precise or
useful...[recommendations] to other programs that I manage/supervise.” These responses support
the idea of a “ripple effect” in which improved practice of team members extends benefits
beyond those who participate in the FC (e.g., co-workers, staff). One member noted “I take

107

information back to my staff… in order for them to do their job better,” providing techniques and
solutions outside the sphere of each member’s agency.

Inter-agency Relationships
Improved relationships among other disciplines was often mentioned as an underlying
mechanism to improve outcomes. Responses described greater awareness of other agencies,
improved ability to collaborate, and development of cross-agency respect and trust.
FC members were “more aware of the various agencies that deal with elder abuse, and
the function they serve,” gaining a comprehensive view of available solutions. With “awareness
of how APS, police, PG [Public Guardian] (etc.) function,” members outside the investigative
agencies learned “the process after mandated reporting has been completed.” Inter-agency
awareness has reciprocal benefits, described by one member who reflected that providing “input
helps others to understand what...my office...do[es].” Cross-agency knowledge enabled FC
members to understand and navigate one-another’s services and provide appropriate referrals.
Beyond inter-agency awareness, members noted increased access to direct assistance
from other members, and ability to collaborate and build trust. FC members experienced
“quicker access on the phone or by email” to information and resources, and one respondent
noted that through the FC, there is access to a “wealth of experts.” Meetings and teamwork help
“foster...more/better communication and understanding,” wrote one member; another appreciated
opportunity to “exchange ideas with the various partners.” Through sharing knowledge and
timely assistance, one member noted improved “respect [for] other agencies.” Trust is built
through time spent developing case plans during meetings, and jointly executing the plans in the
“other 38 hours.” As one team member wrote:

108

“We are confident in the other member's work and ability. We know that if we need to
reach out to them, we will receive the help and support we need to reach our goals.”

DISCUSSION
Building on prior work describing individual FCs, we systematically compared four
existing self-identified Elder Abuse Forensic Centers. Our aim was to support future program
replication efforts by better understanding program characteristics, including commonalities and
differences in structure, processes process outcomes, and sustainability. Identifying key model
components can inform the establishment of a standard framework to ensure that replication
efforts adhere to the model’s core components with fidelity. Moreover, once a common model
has been established, further evaluative research can be conducted to assess the short and long-
term outcomes of the intervention including its impact on clients.
Comparing the four FCs, we found that three programs had similar practices and
approaches, MDT composition, protocols, procedures, and process outcomes. The fourth, was
distinct in scope of practice, approach, team composition, and outcomes.

Model Clarity and Fidelity: A Tale of Two Approaches
While all sites were forensic-focused and funded under the Archstone Foundation’s Elder
Abuse and Neglect Initiative as “forensic centers,” there appeared to be two different models.
Three of the sites (A, B, C) had similar programmatic philosophies, structures and operational
approaches (Table 4.4, below); the fourth (D) was different in several respects (Tables 4.4 and
4.5, below).

109

Table 4.4. Comparison Between the Forensic Center and Family Justice Center Models

Forensic Center (FC) Family Justice Center (FJC)
Service Delivery Model Provider-supportive Client-empowerment
Direct Service Recipient Service Providers Older adult victims
Client Consent Required
to Discuss Case
No Yes
Client Preferences
Dictate Service Plan
No
Although client preferences are
taken into account, the final action
plan may include steps to protect the
client or community despite being
contrary to client wishes
Yes
Client opinions and preferences
guide service plan development
because without client consent
the team cannot implement any
actions
Client Decisional
Capacity Required
No
Clients may have cognitive
impairment or suspected incapacity
Yes
Clients must be able to consent
in order to receive services
Capacity Assessments
Provided
Yes
Capacity assessments are essential
services provided by team members
for clients who might need.
No
Capacity assessments are not
conducted since all clients are
assumed to have capacity.

110

111

Table 4.5. Comparison of Client and Case Characteristics by Program Model

Total Sample
Forensic
Center
Family Justice
Center

N=801 n=504 n=297

n % n % n % P-value
Client Age

65+ 662 82.65 407 80.75 255 85.86 <0.001
a

Missing 29

7

22

Overall Mean ± SD 75.93 ± 14.55 76.05 ± 15.97 75.72 ± 11.56 0.762
Among Elders (Mean ± SD) 80.45 ± 8.49 82.19 ± 7.86 77.69 ± 8.73 <0.001
a

Female 417 52.10 313 62.10 104 35.02 <0.001
a

Missing 148

3

145

Client Race

White 452 56.43 265 52.58 187 62.96 0.013
African-American 104 12.98 71 14.09 33 11.11 0.154
Hispanic 106 13.23 58 11.51 48 16.16 0.081
Asian and Pacific Islander 45 5.62 32 6.35 13 4.38 0.173
Other 23 2.87 18 3.57 5 1.68 0.105
Missing 41

30

11

Abuse Type

Financial 422 54.2 316 62.8 106 38.4 <0.001
a

Emotional 249 32 100 19.9 149 54.0 <0.001
a

Physical 194 24.9 53 10.5 141 51.1 <0.001
a

112

Neglect 157 20.2 149 29.6 8 2.9 <0.001
a

Sexual 17 2.2 14 2.8 3 1.1 0.121
Self-Neglect 155 19.9 155 30.8 0 0 <0.001
a

Other 141 18.1 136 27 5 1.8 <0.001
a

Missing 22

Referring Agency

APS 429 53.56 390 77.38 39 13.13 <0.001
a

Law Enforcement 88 10.99 73 14.48 15 5.05 <0.001
a

Prosecutor's Office 104 12.98 10 1.98 94 31.65 <0.001
a

Other Agency 45 5.62 25 4.96 18 6.06 <0.001
a

Self-Referral 22 2.75 0 0.00 22 7.41 <0.001
a

Missing 112

8

0

Note: APS = Adult Protective Services;
a
Calculated using Independent Samples T-test;
b
Calculated
using Pearson Chi-square Test for Independence

Three of the four FCs targeted client interventions by offering expertise and resources to
assist case presenters with case investigation, intervention and implementation of a multi-
disciplinary action plan. Although this approach seeks to account for reported client preferences,
clients themselves do not participate. This enabled practitioners to seek FC support for a variety
of clients, including those with diminished cognitive capacity who would be unable to consent to
services.
In contrast, one program was built on the client-empowerment model of its parent
organization, a based on a Family Justice Center (FJC) approach. Whereas a primary foundation

113

of the FC model was to offer expertise and support to enhance the work of protective service
professionals (e.g., APS, LE; Gassoumis, et al., 2015), the client-empowerment program
centered on providing “wrap-around” direct services to victims of abuse. In this model, the focus
is on giving clients the power to decide which MDT services or recommendations to accept and
pursue. Because program staff were required to obtain client consent to discuss cases with MDT
members, victims had to have decisional capacity and consent to receive services. In contrast, in
the FC model, a professional could present a case based on concerns that the person appeared to
lack capacity and needed neuropsychological testing to determine if a petition for
conservatorship was appropriate. This allowed the FC model to serve incapacitated older adults,
a client population distinct from the FJC site. As such, client-empowerment site’s core team did
not include a health professional or a gero- or neuropsychologist, unlike the other three
programs.
Longitudinal comparison of model characteristics and operations revealed further
distinctions. Archstone grant funding ended in July 2013. By 2017, nearly five years after the
end of the data collected for this study, three of the sites were continuing with the FC mission;
the FJC-like program no longer operated as a FC.

The Ripple Effect
FCs appear to impact communities and professionals beyond those at the table for case
review. Many FC team members work with other professional colleagues providing another
avenue for dissemination of lessons learned and transmission of promising practices. These
interactions widen the FC scope of influence as members implement acquired knowledge and
practices in their own professional practice. By facilitating exchange of knowledge, improved

114

access to cross-agency assistance, and enhanced relationships, the FC has the potential to
strengthen responses to specific cases of elder abuse and potentially to more broadly improve
members’ organizations.

Replicability of Outcomes
Prior work has focused on evaluating the structure, processes, and outcomes of specific
Centers. The replication of these findings in other centers strengthens confidence in the
generalizability of previously documented results. The observation of common approaches and
processes, and the subsequent achievement of common process outcomes across confirmed FC
sites in the present study provides preliminary evidence to suggest that replicated programs
adhering to the observed essential FC programmatic components might achieve similar results,
though future research is needed. The finding that deviations from the pure FC model, such as
those applied by the FJC-like model, resulted in process outcomes distinct from those observed
among the true FCs further strengthens this argument. Replication and evaluation of FCs in other
states or geographic areas may provide further insights on the expected outcomes resulting from
FC implementation.

Fidelity Markers
While further research is needed to refine the establishment of program fidelity markers,
we propose the following key programmatic components for consideration in future efforts to
establish such markers. The primary service modality of FCs should be to partner with already-
involved investigators and providers to collaborate with them in their efforts to investigate,
intervene, and seek justice in cases of suspected abuse or neglect. Though FCs may adopt direct-

115

service approaches, interacting directly with clients, the provider-focused approach should be the
default employed at all FCs. FC MDTs should be led by a dedicated program manager and
include at least one representative from APS, law enforcement, prosecutors, healthcare
clinicians, neuro- or geropsychologist, public guardian, and victim advocacy programs. MDT
members should be willing and able to not only provide their professional opinion on cases
presented to the FC, but also have the capacity to actively engage in case activities and work-up
during and after MDT meetings if appropriate. At minimum, FC services should include MDT
case review, medical and legal record review, and physical and cognitive assessment. FCs should
engage in continuing case monitoring and follow-up after MDT case presentation to provide
ongoing support and technical assistance, which may include re-presentation at a future FC
meeting. Though it may be possible for replicative programs to implement some variation in
program components to better serve their clients and communities, we recommend that these
characteristics be implemented at-minimum to promote the establishment of programs that yield
similar successful programmatic outcomes.

Multi-site Data Collection
Future work should also focus on the development of standardized data collection and
management systems. Our efforts to compare client characteristics, abuser characteristics, and
program outcomes across sites were limited by a lack of consistency in data across sites,
including differences in the information collected, operationalization of concepts, and data
elements. More uniformity across sites will enable more comprehensive evaluation of FC
programs and contribute to the development of evidence-based program guidelines.

116

Further, future work should focus on documenting program outputs and outcomes across
sites. While past studies have sought to track prosecutions and guardianships among FC cases,
research on client well-being and quality-of-life following FC involvement has not been
conducted, to our knowledge. Multi-site studies employing standardized data collection
strategies may provide opportunities to explore these topics and inform the development,
replication, and success of future FCs.

Limitations
This study was conducted in conjunction with a grant funding mechanism designed to
promote the replication and development of FC sites. Though the research team sought to remain
objective in their assessment, the primary data collection tools used in the study were directly
associated with project funding mechanisms. As a result, sites may have been more likely to
report project activities and resulting impacts more positively than warranted. To address these
concerns, the research team conducted in-person observational visits to validate report details
and information. Upon visitation, sites operated consistent with reported practices.
Though self-assessment approaches have been used in past FC evaluations and in
evaluation of other MDTs, it’s critical to note that reliance on the Team Effectiveness Inventory
as the sole assessment of team effectiveness and interaction quality may yield misleading results.
Some positivity bias may exist as MDT members seek to portray their team’s activities more
positively than actually the case. Moreover, selection bias may be present among the sample as
some MDT members may volunteer to attend meetings on behalf of their agencies and be more
willing than mandatorily assigned providers to rate the MDT as effective. Again, we sought to
counter these limitations by ensuring that participant responses would be anonymous, freeing

117

team members to answer honestly without fear of retribution. The research team also observed
the FC MDT during its meeting, verifying, in part, their ability to work cohesively and
collaboratively.

CONCLUSION
While elder abuse MDTs have been around for over a decade, the FC model is unique in
its ability to incorporate an array of professionals needed to process the most complex elder
abuse cases in real-time with a “forensic lens,” both during the MDT meeting and in the “other
38 hours.” Although past work has sought to describe FCs and their outcomes, more work is
needed to clarify the model’s essential program characteristics and fidelity markers. In
conducting a systematic evaluation and comparison of California FC programs, their models and
operational approaches, this article identifies essential program components needed to establish a
FC and presents examples of how programs can be tailored to address specific community
contexts, resources and needs. This information can guide FC replication efforts as communities
seek to implement this program with fidelity in aim to increase likelihood of MDT effectiveness,
achievement of positive client outcomes, and accomplish prosecution of abusers and appropriate
conservatorship for vulnerable incapacitated victims.

118

CHAPTER V: CONCLUSION

The presented dissertation research sought to improve existing elder mistreatment
detection-intervention service pathways and support replication of promising interventions.
Chapters II and III sought to identify forensic markers of physical elder abuse by
documenting and characterizing injuries present among victims of physical elder abuse and
comparing these injuries to those found among non-abused, non-intentionally injured older
adults. Though similar in intent and analytical approach, the chapters examined distinct service
populations at their point-of-entry into the elder abuse service delivery system. Chapter II
provided insight into injuries found among older adult patients seeking care from hospital
emergency departments who self-identified as being victims of an intentional assault. In contrast,
Chapter III examined injuries present among older adults reported to APS for allegations of
physical elder abuse who confirmed abuse had occurred but who may not have otherwise sought
outside assistance.
Injury types and characteristics reported among geriatric patients seeking care from
hospital emergency departments in Chapter II are consistent with those reported in previous
studies. Because the study employs a nationally-representative geographic sample, findings are
generalizable to geographic locations nationwide. The sample, drawn from patients seeking care
from hospital emergency departments for non-fatal injuries, is likely to include older adults with
injuries severe enough to warrant medical attention though not severe enough to result in death.
In contrast, physically-abused APS subjects studied in Chapter III reported less severe
injuries than those found among patients in Chapter II and other studies conducted in medical
settings (Friedman, et al., 2011; Rosen, et al., 2016; Gironda, et al., 2016). Given that a majority

119

of study subjects did not seek medical care for injuries sustained, it is probable that most of these
victims would not have been included in a hospital-based study and have been absent from past
research study frames. They represent a unique population of elder abuse victims who have been
seldom otherwise studied (Wiglesworth, et al., 2009) and provide novel insight into abuse-
related injuries and markers of abuse.
Despite these differences, similar findings in injury location suggest that some markers of
physical elder abuse may be common throughout the spectrum of injury diagnoses and severity.
Injuries to the head, neck, and maxillofacial regions were strongly associated with abuse while
those located in the lower extremities, including the leg and foot, were more likely to be found
among patients who did not report abuse. Being struck by or against a person or object was an
injury-causing mechanism strongly associated with abuse in both study populations.
Though further work is needed to continue identifying forensic markers of physical abuse
and verifying their applicability to unique abuse populations, the present findings continue to
build the body of literature on this topic and provide insight into the types and characteristics of
injuries occurring among APS clients. These findings can inform the work of clinical
practitioners and social service providers as they seek to investigate and respond to cases of elder
mistreatment. They may be especially applicable to the work of Elder Abuse Forensic Center
(FC) teams which convene healthcare providers, social services providers, and legal justice
professionals into a collaborative working group of elder mistreatment professionals.
Chapter IV seeks to explore the program model, policies, activities, and outcomes found
among 4 FC teams in California. Cross-site comparison of site policies, protocols, clientele
characteristics, and services revealed that three of the sites were very similar to one another. The
third employed a unique program model which was more similar to the Family Justice Center

120

model of its parent organization than to the other three centers studied. Though neither model
was superior to the other and both sought and achieved positive case and client outcomes, the
programmatic approach and philosophy were distinct enough to distinguish one from the other.
The Family Justice Center model operated with a client-empowerment model which
placed the older adult victim in a decision-making role, requiring them to have capacity to
consent to services. In contrast, Forensic Center programs were provider-supportive and
provider-focused, working collaboratively with the provider to investigate and respond to cases
of mistreatment. Because the FC sought to work through or collaborate with already-involved
providers client consent to have cases reviewed by the FC was never sought nor required. This
enabled the FCs to provide services and supports for clients with cognitive impairment and
suspected incapacity, filling a critical service provision gap within their communities.
Positive programmatic, service delivery, and sustainability outcomes were achieved
through the FCs. Though future work is needed to continue studying FC outcomes across sites,
the present study identified key model characteristics found across sites which may be
appropriate for use as fidelity markers to inform future model replication efforts.

Research Limitations
The three presented studies seek to improve physical abuse detection and elder
mistreatment intervention programs. Though efforts were made to examine understudied
populations of victims through recruitment analyses of a nationally-representative sample and a
sample recruited through APS, older adult victims of abuse who are not reported to APS and are
uninjured or do not seek care for their injuries remain outside the sampling frame of these
studies. Research is needed to better identify and provide support for these individuals. The

121

standardization of screening protocols has been implemented in gynecological healthcare
practices to provide otherwise unidentifiable victims an opportunity to seek assistance. The
development of such practices may enable us to better address this population in future research.
Study generalizability may also be limited by study selection effects. Individuals willing
to participate in research on elder mistreatment and confirm the occurrence of an abusive event
may differ from those who would abstain from research or deny the occurrence of abuse. It is
unclear whether these differences in participant characteristics would have an effect on the
injuries present in the sample and available for analyses. The magnitude and directionality of
such effects are difficult to assess. It is also possible that mechanisms of injuries used to
perpetrate mistreatment may vary across culture or socioeconomic status. These research
questions may need to be addressed by continuing to study these phenomena in different
populations and through more informal data collection and sampling methods.

Practice Implications
Though existing research on forensic markers of physical elder abuse have not yet
identified a pathognomonic elder abuse injury, these characteristics can be used to identify
injuries which may be suspicious for abuse. Upon identifying suspicious injuries, practitioners
may be better poised to begin further discussion and investigation of cases, taking care to respect
the older adult client’s right to self-determination and self-dignity. The identification of
suspicious injuries might initiate protocols to create opportunities to interview the older adult
away from their care recipient or to obtain a history of the purported injury mechanism. This
information may be types of information needed for criminal or civil cases should the victim
decide to seek legal remedies.

122

Earlier abuse detection can also enable earlier intervention and provision of victim
supportive services. Many of the negative physical, functional, and mental health outcomes of
abuse may be assessed and addressed through the collaborative work and services of healthcare
clinicians and social services providers. Criminal justice professionals may be able to assist in
creating safer living situations through the enactment of protective orders, perpetrator
incarceration, or mandated perpetrator counseling. Though some front-line elder abuse service
providers, such as APS workers or law enforcement officers may be unsure of how to access
such remedies for their clients, FCs and other MDTs can provide practitioners with the resources,
support, and training to support victims of mistreatment. Replication of this model with fidelity
to the original program model may help communities provide these services.

Future Directions
Future work is needed to continue documenting and characterizing the scope of injuries
associated with elder physical abuse. Further work is needed to continue data collection from
community-dwelling older adults, particularly those who experience abuse but are never in
contact with formal health or social services professionals. Partnering with community agencies
and well-established and trusted community leaders may enable researchers to recruit previously
unreached victims.
Continuing research with APS clients is also needed to further describe abuse-related
injuries occurring within these populations. Research is needed to assess whether less severe
forms of injury, such as tenderness and swelling, may serve as markers of abuse. Exploration of
this population may also lead to an examination of protective factors or responses which may
prevent injury or mitigate harm resulting from abuse. Moreover, it may also provide evidence to

123

support victim claims of abuse absent observable physical injury, and conversely be used to
refute false claims of mistreatment.
There is also great value in documenting and characterizing accidental, non-intentionally-
inflicted injuries occurring among geriatric patients. This work to identify what abuse is not,
similar to that approach used by Mosqueda, et al. (2005), can also inform clinical practice and
serve as a comparison from which abuse-related injuries can be compared. Documentation of
accidental, non-intentionally-caused tenderness, swelling, abrasions, lacerations, and fractures is
needed.
Because these studies sought to conduct clinically-relevant and practice-oriented
research, California statutes defining elder mistreatment detection and dictating appropriate
responses were used to guide analyses. However, further research is needed to explore whether
use of practice-oriented definitions of abuse, namely the inclusion of cases and injuries that arise
from assaultive or physically forceful actions perpetrated by strangers and those not well-known
to the client are fundamentally or characteristically different from those resulting from abuse by
a trusted individual. Further analysis of the NEISS dataset, which may have a large enough
sample to support sub-analyses of injuries between injuries arising from geriatric assault by non-
trusted others and abusive actions by trusted other, may provide greater insight into these
questions.
Finally, collaborative research approaches may enable researchers to recruit larger, more
diverse samples for analysis. Use of standardized injury collection instruments and data
categories may enable data sharing across sites and projects to increase the sample size and
resulting statistical analytic power of studies. Research collaborations may prove essential in

124

continuing efforts to study these phenomena which are vastly underreported and affect
vulnerable populations.

125

REFERENCES

Abath, M. D. B., Leal, M. C. C., Melo Filho, D. A. D., & Marques, A. P. D. O. (2010). Physical
abuse of older people reported at the Institute of Forensic Medicine in Recife,
Pernambuco State, Brazil. Cadernos de Saúde Pública, 26(9), 1797-1806.
Acierno, R., Brady, K., Gray, M., Kilpatrick, D. G., Resnick, H., & Best, C. L. (2002).
Psychopathology following interpersonal violence: A comparison of risk factors in older
and younger adults. Journal of Clinical Geropsychology, 8(1), 13-23.
Acierno, R.A., Hernandez, M. A., Amstadter, A. B., Resnick, H. S., Steve, K., Muzzy, W. &
Kilpatrick, D. G. (2010). Prevalence and correlates of emotional, physical, sexual and
financial abuse and potential neglect in the United States: The National Elder
Mistreatment Study. American Journal of Public Health, 100: 292-297.
Acierno, R., Hernandez-Tejada, M. A., Anetzberger, G. J., Loew, D., & Muzzy, W. (2017). The
National Elder Mistreatment Study: An 8-year longitudinal study of outcomes. Journal
of elder abuse & neglect, 29(4), 254-269.
Anetzberger, G. (2011). The evolution of a multidisciplinary response to Elder Abuse. Marquette
Elder’s Advisor, 13(1), Article 1.
Bonnie, R. J., & Wallace, R. B. (Eds.). (2003). Elder Mistreatment: Abuse, Neglect, and
Exploitation in an Aging America. National Academies Press.
Breckman, Risa; Callahan, Jean; Solomon, J. (2014). Planning for the Future. Elder Abuse
Multi-disciplinary Teams: Planning for the Future. New York, NY.

126

Connolly, M.-T. (2010). Where Elder Abuse and the Justice System Collide: Police Power,
Parens Patriae, and 12 Recommendations. Journal of Elder Abuse & Neglect (Vol. 22).
http://doi.org/10.1080/08946560903436338.
Bernstein, A. B., & Remsburg, R. E. (2007). Estimated prevalence of people with cognitive
impairment: results from nationally representative community and institutional
surveys. The Gerontologist, 47(3), 350-354.
Burston, G. R. (1975). Letter: Granny-battering. British Medical Journal, 3(5983), 592.
California Welfare & Institutions Code, California Civil Code § 10850.1.
Cannell, M. B., Weitlauf, J. C., Garcia, L., Andresen, E. M., Margolis, K. L., & Manini, T. M.
(2015). Cross-sectional and longitudinal risk of physical impairment in a cohort of
postmenopausal women who experience physical and verbal abuse. BMC women's
health, 15(1), 98.
Cham, G. W., & Seow, E. (2000). The pattern of elderly abuse presenting to an emergency
department. Singapore medical journal, 41(12), 571-574.
Christian, C. W., & Committee on Child Abuse and Neglect. (2015). The evaluation of suspected
child physical abuse. Pediatrics, peds-2015.
Clarke, M. E., & Pierson, W. (1999). Management of elder abuse in the emergency department.
Emergency medicine clinics of North America, 17(3), 631-644.
Collins, K. A. (2006). Elder maltreatment: a review. Archives of pathology & laboratory
medicine, 130(9), 1290-1296.
Comijs, H. C., Pot, A. M., Smit, J. H., Bouter, L. M., & Jonker, C. (1998). Elder abuse in the
community: prevalence and consequences. Journal of the American geriatrics
Society, 46(7), 885-888.

127

Comijs, H. C., Penninx, B. W., Knipscheer, K. P., & van Tilburg, W. (1999). Psychological
distress in victims of elder mistreatment: The effects of social support and coping. The
Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 54(4),
P240-P245.
Crimes Against Elders, Dependent Adults, and Persons with Disabilities, Cal. Penal Code §368.
DeLiema, M., Navarro, A. E., Moss, M., & Wilber, K. H. (2016). Prosecutors’ Perspectives on
Elder Justice Using an Elder Abuse Forensic Center. American Journal of Criminal
Justice, 1–16. http://doi.org/10.1007/s12103-015-9321-7.
Dong, X., Simon, M., de Leon, C. M., Fulmer, T., Beck, T., Hebert, L., Dyer, C., Pavez, G. &
Evans, D. (2009). Elder self-neglect and abuse and mortality risk in a community-
dwelling population. Journal of the American Medical Association, 302(5): 517-526.
Dong, X., Simon, M. A., Mosqueda, L., & Evans, D. A. (2012). The Prevalence of Elder Self-
Neglect in a Community-Dwelling Population. Journal of Aging and Health, 24(3),
507–524.
Dyer, C B.; Gleason, Mary S.; Murphy, Kathleen Pace; Pavlik, Valory N.; Portal, Betty; Regev,
Tziona; Hyman, D. J. (1999). Treating elder neglect: Collaboration between a geriatrics
assessment team and adult protective services. Southern Medical Journal, 92(2), 242–
244.
Dyer, C. B., Goodwin, J. S., Pickens-Pace, S., Burnett, J., & Kelly, P. A. (2007). Self-neglect
among the elderly: A model based on more than 500 patients seen by a geriatric
medicine team. American Journal of Public Health, 97(9), 1671–1676.
Dyer, C. B., Heisler, C. J., Hill, C. A., & Kim, L. C. (2005). Community approaches to elder
abuse. Clinics in geriatric medicine, 21(2), 429-447.

128

Elder Justice Coalition. (2016). Elder Justice Act Summary (Fact Sheet). Retrieved online from:
http://elderjusticecoalition.com/sites/default/files/documents/Elder%20Justice%20Act%
20Facts%20Funding%20Resources.pdf on April 11, 2016.
Fanslow, J. L., Norton, R. N., & Spinola, C. G. (1998). Indicators of assault-related injuries
among women presenting to the emergency department. Annals of Emergency Medicine,
32(3), 341-348.
Friedman, L. S., Avila, S., Tanouye, K., & Joseph, K. (2011). A case–control study of severe
physical abuse of older adults. Journal of the American Geriatrics Society, 59(3), 417-
422.
Gassoumis, Z. D., Navarro, A. E., & Wilber, K. H. (2015). Protecting victims of elder financial
exploitation: the role of an Elder Abuse Forensic Center in referring victims for
conservatorship. Aging & Mental Health, 19(9), 790-798.
Gibbs, L., & Mosqueda, L. (2010). Elder abuse: a medical perspective. Aging health, 6(6), 739-
747.
Giles, L., Brewer, E. T., Kullman, M. E., & Prevratil, J. F. (2010). Advancing the field: The
design of the Archstone Foundation Elder Abuse and Neglect Initiative. Journal of Elder
Abuse & Neglect, 22(3-4), 225-230.
Gironda, M. W., Nguyen, A. L., & Mosqueda, L. M. (2016). Is this broken bone because of
abuse? Characteristics and comorbid diagnoses in older adults with fractures. Journal of
the American Geriatrics Society, 64(8), 1651-1655.
Hall, J. E., Karch, D. L. & Crosby, A. E. (2016). Elder abuse surveillance: Uniform definitions
and recommended core data elements for use in elder abuse surveillance, version 1.0.

129

Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease
Control and Prevention.
Homeier, D. (2014). Aging: Physiology disease and abuse. Clinics in Geriatric Medicine, 30(4):
671-686.
Jackson, S. L., & Hafemeister, T. L. (2014). How case characteristics differ across four types of
elder maltreatment: Implications for tailoring interventions to increase victim
safety. Journal of applied gerontology, 33(8), 982-997.
Jackson, S. L., & Hafemeister, T. L. (2012). Pure financial exploitation vs. Hybrid financial
exploitation co-occurring with physical abuse and/or neglect of elderly
persons. Psychology of Violence, 2(3), 285.
Jackson, S. L., & Hafemeister, T. L. (2011). Risk factors associated with elder abuse: The
importance of differentiating by type of elder maltreatment. Violence and Victims, 26(6),
738.
Johnson, P. J., Ghildayal, N., Ward, A. C., Westgard, B. C., Boland, Hokanson, J. S. (2012).
Disparities in potentially avoidable emergency department (ED) care. Medical Care, 50:
1020-1028.
Johnson, C. F., & Showers, J. (1985). Injury variables in child abuse. Child abuse &
neglect, 9(2), 207-215.
Kellogg, N. D. & Committee on Child Abuse and Neglect. (2007). Evaluation of suspected child
physical abuse. Pediatrics, 119(6), 1232-1241.
Kleinschmidt, K. C. (1997). Elder abuse: A review. Annals of Emergency Medicine, 30: 463-472.

130

Lachs, M. S., Williams, C., O’Brien, S. O., Hurst, L. & Horowitz, R. (1997a). Risk factors for
reported elder abuse and neglect: A nine-year observational cohort study. The
Gerontologist, 37(4): 469474.
Lachs, M. S., Williams, C. S., O'Brien, S., Hurst, L., & Tinetti, M. E. (1997b). ED use by older
victims of family violence. Annals of Emergency Medicine, 30(4), 448-454.
Lachs, M.S., Williams, C. S., O’Brien, S., Pillemer, K.A., & Charlson, M.E. (1998). The
mortality of elder mistreatment. Journal of the American Medical Association, 280(5):
428-432.
Leventhal, J. M., Thomas, S. A., Rosenfield, N. S., & Markowitz, R. I. (1993). Fractures in
young children: distinguishing child abuse from unintentional injuries. Archives of
Pediatrics & Adolescent Medicine, 147(1), 87.
McLeer, S. V. & Anwar, R. (1989). A study of battered women presenting in an emergency
department. American Journal of Public Health, 79(1): 65-66.
Meservy, C. J., Towbin, R., McLaurin, R. L., Myers, P. A., & Ball, W. (1987). Radiographic
characteristics of skull fractures resulting from child abuse. American Journal of
Roentgenology, 149(1), 173-175.
Mosqueda, L., Burnight, K., & Liao, S. (2005). The life cycle of bruises in older adults. Journal
of the American Geriatrics Society, 53(8), 1339-1343.
Mosqueda, L., Burnight, K., Liao, S., & Kemp, B. (2004). Advancing the field of elder
mistreatment: A new model for integration of social and medical services. The
Gerontologist, 44(5), 703-708.
Murphy, K., Waa, S., Jaffer, H., Sauter, A. & Chan, A. (2013). A literature review of findings in
physical elder abuse. Canadian Association of Radiologists Journal, 64(1): 10-14.

131

National Committee for the Prevention of Elder Abuse. Physical abuse. Available at
http://www.preventelderabuse.org/elderabuse/physical.html.
National Institute of Health. 2015. Report - NIH Workshop: Multiple Approaches to
Understanding and Preventing Elder Abuse and Mistreatment. Bethesda, MD.
Navarro, A. E., Gassoumis, Z. D., & Wilber, K. H. (2013). Holding abusers accountable: An
elder abuse forensic center increases criminal prosecution of financial exploitation. The
Gerontologist, 53(2), 303-312.
Navarro A, Wilber KH, Yonashiro J, Homeier DC. (2010). Do we really need another meeting?
Lessons from the Los Angeles County Elder Abuse Forensic Center. Gerontologist;
50(5): 702-711.
Navarro, A. E., Wysong, J., DeLiema, M., Schwartz, E. L., Nichol, M. B., & Wilber, K. H.
(2016). Inside the black box: The case review process of an Elder Abuse Forensic
Center. The Gerontologist, 56(4): 772-781.
Palmer, M., Brodell, R. T., & Mostow, E. N. (2013). Elder abuse: dermatologic clues and critical
solutions. Journal of the American Academy of Dermatology, 68(2), e37-e42.
Pillemer, K., Connolly, M. T., Breckman, R., Spreng, N., & Lachs, M. S. (2015). Elder
mistreatment: Priorities for consideration by the White House Conference on Aging. The
Gerontologist, 55(2), 320-327.
Pillemer, K., & Finkelhor, D. (1988). The prevalence of elder abuse: A random sample survey.
Gerontologist, 28(1), 51-57.
Rosen, T., Bloemen, E. M., LoFaso, V. M., Clark, S., Flomenbaum, N. E., & Lachs, M. S.
(2016). Emergency department presentations for injuries in older adults independently
known to be victims of elder abuse. The Journal of emergency medicine, 50(3), 518-526.

132

Rosen, T., Lofaso, V. M., Bloemen, E. M., Reisig, C., Clark, S., Mulcare, M. R., ... & Lachs, M.
S. (2015). Injury Patterns in Physical Elder Abuse: Preliminary Findings from a Pilot
Sample of Highly Adjudicated Cases. Academic Emergency Medicine, 22, S268.
Rosen, T., Reisig, C., LoFaso, V. M., Bloemen, E. M., Clark, S., McCarthy, T. J., ... & Lachs, M.
S. (2016). Describing visible acute injuries: development of a comprehensive taxonomy
for research and practice. Injury prevention, injuryprev-2016.
Rittel, H. W., & Webber, M. M. (1973). Dilemmas in a general theory of planning. Policy
sciences, 4(2), 155-169.
Schneider, D. C., Mosqueda, L., Falk, E., & Huba, G. J. (2010). Elder abuse forensic
centers. Journal of Elder Abuse & Neglect, 22(3-4), 255-274.
Sherbourne, C. D., & Stewart, A. L. (1991). The MOS social support survey. Social science &
medicine, 32(6), 705-714.
Sommers, M. S., Brown, K. M., Buschur, C., Everett, J. S., Fargo, J. D., Fisher, B. S., ... & Zink,
T. M. (2012). Injuries from intimate partner and sexual violence: Significance and
classification systems. Journal of forensic and legal medicine, 19(5), 250-263.
Spitzer, R. L. (1983). Psychiatric diagnosis: are clinicians still necessary?. Comprehensive
psychiatry, 24(5), 399-411.
Stiegel, L. (2005). Elder Abuse Fatality Review Teams: A Replication Manual. American Bar
Association: Washington, D.C.
Tatara, T., Kuzmeskus, L. B., Duckhorn, E., Bivens, L., Thomas, C., & Gertig, J. (1998).
National elder abuse incidence study: Final report. Washington, DC: National Center on
Elder Abuse, American Public Human Services Association.

133

Teaster, P. B., Nerenberg, L., & Stansbury, K. L. (2003). A national look at elder abuse
multidisciplinary teams. Journal of Elder Abuse & Neglect, 15(3-4), 91-107.
U.S. Department of Health and Human Services, Centers for Disease Control and Prevention,
National Center for Injury Prevention and Control and United States Consumer Product
Safety Commission, National Electronic Injury Surveillance System All Injury Program.
(2010). Codebook. Ann Arbor, MI: Inter-university Consortium for Political and Social
Research.
U.S. Department of Justice and Department of Health and Human Services. (2014). The elder
justice roadmap: A stakeholder initiative to respond to an emerging health, justice,
financial and social crisis. Washington, D.C.: U.S. Department of Health and Human
Services.
Wiglesworth, A., Austin, R., Corona, M., Schneider, D., Liao, S., Gibbs, L., & Mosqueda, L.
(2009). Bruising as a marker of physical elder abuse. Journal of the American Geriatrics
Society, 57(7), 1191-1196.
Wiglesworth, A., Mosqueda, L., Burnight, K., Younglove, T., & Jeske, D. (2006). Findings from
an elder abuse forensic center. The Gerontologist, 46(2), 277-283.
Wilber, K. H., & McNeilly, D. P. (2001). Elder abuse and victimization. In J. E. Birren & K. W.
Schaie (Eds.), Handbook of the psychology of aging (pp. 569-591). San Diego, CA:
Academic Press.
Wilber, K. H., Navarro, A. E., & Gassoumis, Z. D. (2014). Evaluating the Elder Abuse Forensic
Center Model. Submitted to the National Institute of Justice.

134

Yon, Y., Mikton, C. R., Gassoumis, Z. D., & Wilber, K. H. (2017). Elder abuse prevalence in
community settings: a systematic review and meta-analysis. The Lancet Global Health,
5(2), e147–e156.
Yonashiro-Cho, J., Meyer, K., Wilber, K. H. (2017). The Aging Network. In X. Dong (Ed.),
Elder abuse: Research, practice, and policy. Springer Publishing.
Ziminski, C.E., Wiglesworth, A., Austin, R., Phillips, L.R. & Mosqueda, L. (2013). Injury
patterns and causal mechanisms of bruising in physical elder abuse. Journal of Forensic
Nursing, 9(2): 84-91.

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Asset Metadata

Creator Yonashiro-Cho, Jeanine (author)

Core Title Forensic markers of physical elder abuse in medical and community contexts: implications for criminal justice interventions

School Leonard Davis School of Gerontology

Degree Doctor of Philosophy

Degree Program Gerontology

Publication Date 12/03/2018

Defense Date 05/01/2018

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag abuse detection,abuse indicators,accidental injury,adult guardianship,Adult Protective Services,aging,assault,criminology,elder,elder abuse,elder justice,elder mistreatment,elder physical abuse,emergency department,emergency medicine,emergency room,Family violence,forensic center,forensic exam,forensic markers,forensic nurse,geriatric assault,geriatric trauma,geriatrician,Gerontology,Health and Environmental Sciences,home visit assessment,injuries,injury assessment,injury characteristics,injury intent,intentional injury,intervention,Law,legal services,multidisciplinary team,National Electronic Injury Surveillance System,OAI-PMH Harvest,older adult,operations management,organizational development,perpetrator,physical abuse,promising practice,Public Health,social science,social services intervention,stressful life events,unintentional injury,victim,Violence

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Wilber, Kathleen H. (committee chair), Enguidanos, Susan (committee member), Mosqueda, Laura (committee member)

Creator Email jyonashirocho@gmail.com,yonashir@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-110963

Unique identifier UC11676662

Identifier etd-YonashiroC-6994.pdf (filename),usctheses-c89-110963 (legacy record id)

Legacy Identifier etd-YonashiroC-6994.pdf

Dmrecord 110963

Document Type Dissertation

Format application/pdf (imt)

Rights Yonashiro-Cho, Jeanine

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA