Running head: CHILD PHYSICAL ABUSE SCREENING PROGRAM

Implementation of a Child Physical Abuse Screening Program in the Emergency Department
Sheri M. Carson
Arizona State University

1

CHILD PHYSICAL ABUSE SCREENING PROGRAM

2

Abstract
Children often present to the emergency department (ED) for treatment of abuse-related injuries.
ED healthcare providers (HCPs) do not consistently screen children for physical abuse, which
may allow abuse to go undetected and increases the risk for re-injury and death. ED HCPs
frequently cite lack of knowledge or confidence in screening for and detecting child physical
abuse. The purpose of this evidence-based quality improvement project was to implement a
comprehensive screening program that included ED HCP education on child physical abuse, a
systematic screening protocol, and use of the validated Escape Instrument. After a 20-minute
educational session, there was a significant increase in ED HCP knowledge and confidence
scores for child physical abuse screening and recognition (p < .001). There was no difference in
diagnostic coding of child physical abuse by ED HCPs when evaluating a 30-day period before
and after implementation of the screening protocol. In a follow-up survey, the Escape Instrument
and educational session were the most reported screening facilitators, while transition to a new
electronic health system was the most reported barrier. The results of this project support
comprehensive ED screening programs as a method of improving HCP knowledge and
confidence in screening for and recognizing child physical abuse. Future research should focus
on the impact of screening on the diagnosis and treatment of child physical abuse. Efforts should
also be made to standardize child abuse screening programs throughout all EDs, with the
potential for spread to other settings.
Keywords: child abuse, physical abuse, non-accidental trauma, screening, detection,
recognition, systematic protocol, emergency department, healthcare provider, staff education

CHILD PHYSICAL ABUSE SCREENING PROGRAM

3

Implementation of a Child Physical Abuse Screening Program in the Emergency Department
Not only can a child exposed to violence suffer physical harm, but the emotional and
psychological damage secondary to abuse can also predispose the child to post-traumatic
distress, a future of criminal activity, and a host of other mental and behavioral health issues
(Felitti et al., 1998; Teeuw, Derkx, Koster, & van Rijn, 2012). To prevent or reduce these
adverse effects, healthcare providers (HCPs) should routinely screen children for signs of abuse.
This is especially important in the emergency department (ED). Abused children frequently
present to the ED for care, and if HCPs do not consistently screen children for non-accidental
trauma (NAT), a child can be released unintentionally back into an abusive home, which may
lead to extremely detrimentalor even fatalresults.
Overview of the Problem
The United States (U.S.) has one of the worst child abuse records among industrialized
nations (U.S. Department of Health and Human Services [USDHHS], 2017). A child abuse
report is made every 10 seconds, and four to seven children die from abuse and neglect every day
(Childhelp, 2017). The economic burden of child abuse is also staggering. The lifetime estimated
financial costs for just one year of newly confirmed cases of child abuse and neglect are
approximately $124 billion (Fang, Brown, Florence, & Mercy, 2012). Data from the USDHHS
(2017) reveals that more than 683,000 children (9.2 children per 1,000) were the victims of abuse
or neglect in 2015—an increase of 3.8% from 2011. Of those cases, 117,560 (17.2%) were
victims of physical abuse, and 1670 of them died (2.25 per 100,000 children). However, because
only reports received and investigated by Child Protective Services (CPS) are included in these
statistics, officials estimate the actual number of children who are victims of abuse and neglect is
much higher—potentially by 50% or more (Childhelp, 2017). Additionally, many cases of child

CHILD PHYSICAL ABUSE SCREENING PROGRAM

4

abuse go unrecognized and unreported due to insufficient HCP knowledge and skill in
recognizing, diagnosing, and reporting suspicious childhood injuries (Crichton et al., 2016).
Most states have defined physical abuse as any non-accidental physical injury to the
child, which can include kicking, biting, striking, burning, or any action that causes physical
impairment of the child (Child Welfare Information Gateway [CWIG], 2016). The federal
government provided a similar, albeit more expansive, definition of child abuse and neglect in
the Child Abuse Protection and Treatment Act (CAPTA) of 1974. Notably, although the first
documented case of child abuse in the U.S. was in 1874, it was not until the enactment of
CAPTA in 1974 that the federal government provided funding for state-based programs directed
toward the prevention, identification, and treatment of child abuse and neglect (CWIG, 2011).
This financial assistance has continued with the reauthorizations of CAPTA in 1996 and 2010.
The CAPTA Reauthorization Act of 2010 also introduced the statutory requirement that all states
mandate child abuse reporting, although mandatory screening was not included. In addition to
CAPTA, there have been widespread initiatives, advocacy, and research funding for child abuse
prevention, recognition, and treatment through the USDHHS, the National Children’s Alliance
(NCA), CPS, the American Academy of Pediatrics (AAP), the National Association of Pediatric
Nurse Practitioners (NAPNAP), and many others (AAP, 2017; NAPNAP, 2016; NCA, 2014).
Background and Significance
In the U.S. alone, nearly 1,000,000 children are victims of NAT each year, with an
estimated 1.3% to 15% of ED childhood injury visits resulting from physical abuse (Allareddy et
al., 2014; Escobar et al., 2016; Teeuw et al., 2012). Moreover, 25% to 30% of abused infants
have already had a previously noted sentinel injury, such as bruising or an intraoral injury, at the
time they receive a child abuse diagnosis (Glick, Lorand, & Bilka, 2016; Petska & Sheets, 2014;

CHILD PHYSICAL ABUSE SCREENING PROGRAM

5

Sheets et al., 2013). Despite these sobering statistics, there is insufficient and inconsistent
screening of children by ED HCPs for NAT. Abused children have a higher rate of ED usage
than nonabused children, but the abuse often remains unrecognized (Acehan et al., 2016;
Crichton et al., 2016; Guenther, Knight, Olson, Dean, & Keenan, 2009; Jordan & Moore-Nadler,
2014; Louwers et al., 2011). In fact, the early detection rate of child abuse in the ED is a mere
10%, with the estimated percentage of missed abuse cases ranging from 11% to 64% (Allareddy
et al., 2014; Sittig et al., 2016). This lack of effective screening contributes to child physical
abuse being an underreported problem. Additionally, if physical abuse goes undetected at the
initial presentation, the abused child has a 35–50% chance of experiencing recurrent abuse and a
10–30% chance of eventual death from that abuse (Acehan et al., 2016; Escobar et al., 2016;
Teeuw et al., 2012).
Barriers to Child Abuse Screening in the ED
Abused children often present to the ED with various injuries and chief complaints, and
ED HCPs may be their first and only medical contact (Bair-Merritt & Lane, 2011; King, Kiesel,
& Simon, 2006; Tiyyagura, Gawel, Koziel, Asnes, & Bechtel, 2015). It is imperative that ED
HCPs consistently screen children for NAT to ensure early identification, intervention, and
prevention of continued or worsening abuse. Yet standardized screening rarely occurs due to a
variety of identified barriers, including insufficient HCP knowledge of injuries consistent with
abuse; lack of a validated ED child abuse screening tool; limited time to conduct screening or
develop screening policies; and HCP desire to believe caregivers and prevent false CPS reports
(Bair-Merritt & Lane, 2011; Crichton et al., 2016; Louwers, Korfage, Affourtit, De Koning, &
Moll, 2012a; Jordan & Moore-Nadler, 2014; Jordan & Steelman, 2015; Tiyyagura et al., 2015).
As a result, a multifaceted approach is needed to improve HCP confidence in and performance of

CHILD PHYSICAL ABUSE SCREENING PROGRAM

6

child physical abuse screening, as well as reduce the number of children who are not
appropriately identified as being victims of abuse.
ED staff worldwide utilize a variety of methods to screen for child abuse, including
screening checklists, structured clinical examination of the undressed child (“top-toe”
inspection), and specialized training of HCPs. Unfortunately, the majority of these screening
methods are not substantiated by empirical evidence (Bailhache, Leroy, Pillet, & Salmi, 2013;
Hoytema van Konijnenburg, Teeuw, Zwaard, van der Lee, & van Rijn, 2014; Louwers, Affourtit,
Moll, de Koning, & Korfage, 2010; Teeuw et al., 2012). A systematic review by Woodman et al.
(2008) found the quality of screening tests at the time was poor, with no evidence that any test
was highly predictive of physical abuse. Woodman et al. (2009) also noted that screening
markers—such as a child’s age, type of injury, and rate of ED usage—do not reliably identify
abused versus nonabused children. ED HCPs in the Netherlands often administer the
SPUTOVAMO checklist to screen for abuse, but its high false positive rate, even in the revised
version, warrants careful consideration prior to use (Sittig et al., 2014). This prompted Louwers
et al. (2014) to develop the “Escape Form” as an alternative child abuse screening instrument. A
study by Dinpanah, Pasha, and Sanji (2017) showed excellent diagnostic accuracy of the Escape
tool (99.2%). Yet regardless of the specific tool used, assessment by a child abuse pediatrician or
clinician properly trained in the assessment and treatment of child abuse injuries should always
follow any positive screening test (Sittig et al., 2016; Teeuw et al., 2016). This will reduce the
number of false positive reports while also limiting the risk of failing to diagnose child abuse.
Methods to Improve HCP Screening
Recommendations in the literature appear to focus more on adequate training for HCPs
and the implementation of systematic guidelines than on the use of a specific screening tool.

CHILD PHYSICAL ABUSE SCREENING PROGRAM

7

Researchers agree a validated child abuse screening tool is imperative, but they emphasize the
importance of HCP education to improve screening results and reduce false positive reports
(Acehan et al., 2016; Hoft & Haddad, 2017; Jordan & Moore-Nadler, 2014; Louwers et al.,
2012b; Teeuw et al., 2016). Standardized educational programs and well-defined screening
protocols not only improve detection of child physical abuse in the ED, but they also decrease
bias and increase self-efficacy in the HCPs performing the screenings (Higginbotham et al.,
2014; Milani, Vianello, Cantoni, Agostoni, & Fossali, 2016; Smeekens et al., 2011). In addition,
hospital policymakers and ED administrators need to support child physical abuse screening by
embedding it into the routine structure of the hospital, integrating it into electronic systems, and
forming multidisciplinary teams to properly assess and treat children with positive screening
results (Benger & Pearce, 2002; Escobar et al., 2016; Louwers et al., 2012a). Through the
integration of administrative support, a validated screening tool, systematic guidelines, and HCP
education, child abuse screening programs have the potential to be widely successful.
Internal Evidence
In a pediatric ED that is part of a large medical center in southern Arizona, there is no
formal method of screening for or tracking cases of child physical abuse. A retrospective chart
review completed by their social work team revealed that not all childhood injuries secondary to
physical abuse were properly identified and coded by ED staff. The ED HCPs do not use a child
abuse screening tool to identify injuries consistent with NAT, nor do they receive specialized
training on how to recognize injuries that are concerning for abuse. Rather, most of the identified
child abuse cases are the result of injuries leading to hospital admissions, social work consults,
child abuse team (CAT) consults, and/or CPS referrals, with the remaining unidentified abuse
cases being discharged from the ED without further intervention or follow-up (D. Woolridge,

CHILD PHYSICAL ABUSE SCREENING PROGRAM

8

personal communication, November 11, 2016). Finally, although the ED has an on-call
multidisciplinary CAT, many of the ED HCPsincluding physicians and nurseshave
verbalized concern over their own inabilities to recognize injuries suggestive of abuse.
Problem Statement
Even though all HCPs are mandated reporters of child abuse and neglect, no standardized
process currently exists for screening children for abuse in the ED. As a result, children
presenting to the ED with NAT might “slip through the cracks” and be discharged without the
appropriate follow-up or referral. Sadly, research has shown that children who die from abuse
often have been seen by a medical provider at least once prior to their death (Acehan et al., 2016;
Bair-Merritt & Lane, 2011; King et al., 2006; Teeuw et al., 2012). It is imperative that ED HCPs
identify abusive injuries early in order to reduce the risk of re-injury or death secondary to
recurrent and escalating abuse.
Search Process
The initial inquiry and exploration of the literature led to the clinically relevant PICO
question: In ED HCPs, how does a systematic protocol compared to no systematic protocol
affect screening for child physical abuse?
A systematic and exhaustive literature search was conducted to address the above PICO
question and gather the best evidence to support a practice change. Scholarly databases searched
included PubMed, Cumulative Index of Nursing and Allied Health Literature (CINAHL), Web
of Science, Academic Search Premier, PsycINFO, JSTOR, ProQuest Dissertations and Theses,
the Cochrane Library, and the National Guideline Clearinghouse. The initial search strategy
consisted of the following keywords, combinations, and Boolean phrases: (“child physical
abuse” OR “child abuse” OR “child maltreatment” OR “nonaccidental trauma” OR “non-

CHILD PHYSICAL ABUSE SCREENING PROGRAM

9

accidental trauma” OR “intentional injury”) AND (screening OR detection OR identification
OR assessment OR evaluation OR detect OR identify) AND (protocol OR guideline OR tool OR
instrument OR questionnaire OR algorithm OR process OR standardized OR systematic) AND
(“emergency department” OR “emergency room” OR “emergency” OR “accident & emergency
department”). Due to the wide variety of terms used to refer to the same concept across studies
and the paucity of studies on child abuse screening, these word combinations allowed for more
robust results. Database searches were initially conducted as “all field” searches and then refined
by “abstract” or “title/abstract” searches. However, applying these limits significantly reduced
the number of studies retrieved, resulting in the need to return to a broader, all-field search focus.
Initial Search Results
The above search strategy yielded a combined total of 724 results: 278 from PubMed;
101 from CINAHL; 138 from Web of Science; 126 from Academic Search Premier; and 81 from
PsycINFO. The same search strategy performed in JSTOR, ProQuest Dissertations and Theses,
the Cochrane Library, and the National Guideline Clearinghouse did not produce any additional
results, so they were excluded. After the initial keyword search, a second search was performed
in PubMed using a combination of MeSH terms and keywords to ensure no pertinent studies
were missed. The MeSH terms “child abuse” and “emergency service, hospital” were combined
with the following keywords and Boolean connectors: (screening OR detection OR identification
OR assessment OR evaluation OR detect OR identify) AND (protocol OR guideline OR tool OR
instrument OR questionnaire OR algorithm OR process OR standardized OR systematic). This
search produced only 91 results, all of which were included in the initial yield of 278. Therefore,
to guarantee the exhaustive quality of this systematic literature search, all 278 studies from the
initial PubMed search were included in the second stage of the search process.

CHILD PHYSICAL ABUSE SCREENING PROGRAM

10

Review and Refinement of Retrieved Citations
The complete list of retrieved database citations was reviewed and duplicates were
removed. The remaining 430 individual articles were filtered for language (English only) and
publication dates (2011–2017 or “past 5 years”), resulting in 193 articles eligible for further
review (Appendix A). To ensure inclusion of all relevant studies, supplemental ancestry searches
of study reference lists, hand searches of child abuse journals, and an electronic grey literature
search were performed. These searches led to the retrieval of eight additional studies. Each
article was reviewed by title and abstract to ensure relevance to the current study. In addition to
considering the article titles and abstracts, specific inclusion and exclusion criteria were applied.
The primary inclusion criterion was that the study directly addressed screening for or detection
of child physical abuse in the ED, including discussion of barriers and facilitators to screening
and/or interventions to improve HCP detection of child physical abuse. The primary exclusion
criterion was any study conducted in a non-ED setting. Additionally, studies were excluded if
their main focus was sexual or emotional abuse or neglect; interpersonal or domestic violence;
treatment of an already established child abuse diagnosis; detection of child abuse based on
parental characteristics (the Hague protocol); pre-hospital child abuse assessment; injury-specific
evaluation (e.g., burns, bruises, fractures); child protection (CPS) decision-making; child abuse
prevention; and/or accidental trauma.
Critical Appraisal and Synthesis
A total of 41 studies met the criteria for critical appraisal (Appendix A). Once critical
appraisal was complete, 10 studies were retained for in-depth evaluation (Appendix B) and
synthesis (Appendix C). Although several appraised studies contributed to the background and
significance of this research topic, many were excluded from further evaluation because they

CHILD PHYSICAL ABUSE SCREENING PROGRAM

11

were of poor quality, addressed parental characteristics instead of the child’s characteristics, or
focused more on child abuse prevention and/or treatment than on the actual screening process.
The majority of studies were level four evidence, with the remaining studies comprising
levels two, three, and six evidence (Melnyk & Fineout-Overholt, 2015). The studies included one
randomized controlled trial, one quasi-experimental study, four prospective cohort studies, two
diagnostic accuracy studies, and two qualitative studies. Considering the ethical issues
surrounding child abuse screening, these are the best levels of evidence to answer the PICO
question. Six of the studies were conducted in the Netherlands, three in the United States, and
one in Iran. This is not surprising given the fact that the Netherlands has made child abuse
screening compulsory in the ED and, as a result, has created the only validated child abuse
screening tool to date: the Escape Instrument (Dinpanah et al., 2017; Louwers et al., 2014).
All 10 studies were conducted in the pediatric and/or general ED, and each included up to
seven sites. The participants in six of the studies were children (one study limited participants to
12 months of age or younger), with mean ages ranging from 6.4 months to 7.95 years. These
studies all had large sample sizes, and all but one had slightly more males than females. Aside
from the one study that directly addressed socioeconomic status (SES) and racial bias, no other
significant demographic variations were noted. Participants in the remaining four studies were
HCPs. Interestingly, all four of these studies had very small sample sizes, with one having a high
attrition rate (36%) due to participants’ work shift patterns. Finally, a few of the studies raised
concern about potential sample bias due to the temporal nature of the data, site selection, and
small sample sizes, but no other bias was evident (Appendix C).
There were a variety of measurement tools and independent variables used across studies
(Appendix C). The eight quantitative studies used a combination of child abuse screening

CHILD PHYSICAL ABUSE SCREENING PROGRAM

12

instruments (SI), physical exam, CAT or expert panel evaluation, and/or medical chart review,
while the two qualitative studies used tape-recorded, transcribed semi-structured interviews. Two
studies evaluated the diagnostic accuracy of the Escape Instrument, which demonstrated validity
and reliability in screening for child physical abuse in the ED. Two other studies used the
SPUTOVAMO checklist to screen for child abuse in the ED, and the authors of both studies
cautioned against its widespread use due to its high false-positive rate.
The dependent variables also exhibited wide variation across studies (Appendix C). The
majority of studies evaluated child abuse screening rates and/or detection of child abuse risk to
some degree. Individual studies also examined barriers and facilitators to child abuse detection,
HCP knowledge and self-efficacy in recognizing child abuse, racial and SES bias, and factors
influencing completion of child abuse SIs and screening protocols in the ED. Despite this
heterogeneity, clear relationships and themes emerged. Screening protocols increased screening
consistency and screening rates in four studies and decreased SES bias in one study. Child abuse
education programs improved HCP knowledge and self-efficacy in two studies, while
simultaneously increasing child abuse screening and detection rates in three studies and use of
screening protocols in a forth. Finally, the two qualitative studies revealed similar barriers and
facilitators to child abuse detection. Barriers included lack of child abuse knowledge, practical
problems, personal beliefs, and fast ED staff turnover, while facilitators included educational
training, administrative support, and presence of a CAT (Appendix C).
Purpose Statement
The evidence clearly shows that child abuse education improves ED HCP knowledge and
self-efficacy, which subsequently increases screening for and recognition of injuries consistent
with abuse. Additionally, systematic screening protocols—especially when combined with a

CHILD PHYSICAL ABUSE SCREENING PROGRAM

13

validated SI—promote screening consistency, decrease SES bias, and increase rates of child
abuse detection. Therefore, the purpose of this evidence-based quality improvement project was
to implement a comprehensive child physical abuse screening program that incorporated HCP
education on child physical abuse, a systematic screening protocol, and use of the validated
Escape Instrument to support best practice for child physical abuse screening in the ED.
Logic Model
To assist with project planning, a logic model was created prior to development of the
comprehensive child physical abuse screening program (Appendix D). The logic model was
utilized to plan the project and provide a tool for program evaluation. The model also clearly
depicted the relationships between the inputs, outputs, outcomes, and impact of the screening
program, which helped gain buy-in from the project site’s ED HCPs and administrators.
Evidence Based Practice Model
The model chosen to guide implementation of this project was Rosswurm and Larrabee’s
(1999) revised Model for Evidence-Based Practice Change. This model utilizes research and
change theory to support evidence-based practice (EBP) change in the acute care environment. It
integrates teamwork, quality improvement (QI), and the translation of evidence to promote and
sustain practice change (Melnyk & Fineout-Overholt, 2015). The child physical abuse screening
program developed for this project aligns well with this model for several reasons: the ED is an
acute care environment; the QI project needs to be integrated into the standard of care to change
practice throughout the ED; and the components of teamwork are vital to the sustainability of
this EBP project, as it is a practice change that requires adoption by all ED HCPs and staff.
The Model for EBP Change is comprised of six steps that include assessing the need for
change, locating the best evidence, critically analyzing the evidence, designing a practice change,

CHILD PHYSICAL ABUSE SCREENING PROGRAM

14

implementing and evaluating the change in practice, and integrating and maintaining the change
in practice. Although the steps are progressive, they are not explicitly linear (Larrabee, 2009).
This allows for reappraisal and monitoring throughout the process, which is crucial for the
practice change in this project.
Development of the child physical abuse screening program directly followed the steps of
the EBP model. Key stakeholders, including ED HCPs and administrators, identified the need for
improved child abuse screening in the ED. The proposed interventionimplementation of a
comprehensive child physical abuse screening programwas linked to the desired outcome of
improved child physical abuse recognition, and an exhaustive literature review and synthesis of
the evidence supported this practice change. A screening protocol and ED HCP badge cards
containing the Escape Instrument and child physical abuse diagnostic codes were distributed
based on identified needs, and ED HCPs and administrators were educated on the specifics of the
new protocol and Escape screening tool. Once implemented, process evaluation occurred
throughout the duration of the project, and outcomes were evaluated at project conclusion.
Finally, recommended changes were communicated to stakeholders, and regular monitoring and
evaluation of the screening process and desired outcomes will continue in order to fully integrate
and sustain the new child physical abuse screening program as a standard of practice in the ED.
Theoretical Framework
Bandura’s Self-Efficacy Theory guided development of the educational module and
systematic screening protocol for this project. Self-efficacy is the belief in one’s ability to
successfully achieve a task or reach a desired goal and is influenced by one’s past mastery
experiences, vicarious experiences, verbal persuasion, and physiological/emotional state
(Bandura, 1977, 1986, 1997). Self-efficacy beliefs drive a person’s actions, thoughts, feelings,

CHILD PHYSICAL ABUSE SCREENING PROGRAM

15

and motivations. High self-efficacy results in a person believing that difficult tasks are
challenges to be mastered, not threats to be avoided. Conversely, low self-efficacy results in low
aspirations, feelings of inadequacy, and poor commitment to new goals or tasks (Bandura, 1994).
Self-efficacy theory informed the content and structure of this project through its premise
that improving one’s self-efficacy will support positive practice change (Bandura, 2009).
Increasing the HCPs’ child abuse knowledge and self-efficacy through education and positive
feedback will increase their motivation to screen, overcome barriers to detection, improve
confidence, and produce positive experiences. The ED HCPs have already recognized the need
for improved child abuse education and screening practices, and the retrospective chart review
supported the need for improved diagnostic coding of child physical abuse in the ED. Therefore,
educating the HCPs about child physical abuse and providing them with a systematic screening
protocol should remove sociostructural barriers, clearly delineate outcome expectations, and
improve self-efficacy, thereby producing the desired screening behavior.
Methods
Ethical Considerations
This QI project was approved and deemed exempt from full review by the Arizona State
University (ASU) Institutional Review Board (IRB) (Appendix E). After receiving ASU IRB
approval, the project was approved by the project site’s Director of Professional Practice and the
Research and Innovation Council (Appendix F).
Participants 18 years of age or older were recruited by the project team leader during
regularly scheduled staff meetings in July and August 2017 (for the Child Abuse Awareness
pretest/posttest) and again in December 2017 (for the Project Evaluation Survey). Participants
provided their consent for each phase of data collection through completion of the anonymous

CHILD PHYSICAL ABUSE SCREENING PROGRAM

16

project surveys. The Child Abuse Awareness pretest and posttest were linked by random, prenumbered codes that were unique to each participant; identities were not linked to the codes.
Random participant codes were also assigned for the Project Evaluation Survey. No personally
identifiable information was collected during either of the project surveys, and participants were
not asked questions intended to target or exclude special populations.
Setting and Organizational Culture
This QI project took place in the pediatric ED (PED) of a children’s hospital that is
housed within a large, urban, academic level 1 trauma center in southern Arizona. The PED has
18 beds and provides care to approximately 20,000 children each year. One attending physician
staffs each shift, along with several emergency medicine (EM) residents and registered nurses
(RNs). The PED does not currently utilize nurse practitioners or physician assistants. All patients
are initially evaluated in the ED’s triage area, with the exception of trauma patients, who are
taken directly to the trauma bay. Once triaged, pediatric patients are admitted to the PED.
During the course of this project, there were changes in ED management, as well as
turnover of several RNs and transition of EM residents from one rotation to the next. The
medical center also transitioned to a new electronic health record (EHR) six weeks after
implementation of the child abuse screening project. These changes resulted in some project
challenges that will be addressed during the results and discussion sections of this paper.
Participants
All ED HCPs (attending/resident EM physicians and RNs) were invited to participate in
the project during regularly scheduled staff meetings. Participation was voluntary. Consent was
obtained prior to the Child Abuse Awareness pretest/posttest and again prior to the Project
Evaluation Survey. All participants who completed the Child Abuse Awareness posttest had

CHILD PHYSICAL ABUSE SCREENING PROGRAM

17

attended the educational session on child physical abuse. Completion of the Project Evaluation
Survey was independent of the Child Abuse Awareness pretest/posttest and educational session.
Participants only needed to have worked in the PED during the project period and be aware of
the child physical abuse screening program to complete the Project Evaluation Survey.
Intervention
The project intervention was two-phased. First, a 20-minute educational session on child
physical abuse was offered to all ED HCPs during two regularly scheduled staff meetings (one
for EM physicians and one for PED RNs). The educational sessions were delivered in-person by
the project team leader and included an evidence-based overview of child physical abuse; types
of injuries that raise the index of suspicion for child abuse; use of the validated Escape
Instrument; the specific steps in the systematic screening protocol; and documentation of
findings, including diagnostic coding of suspected and confirmed child physical abuse.
After the educational sessions were completed, the systematic child abuse screening
protocol was implemented in the PED. The project team leader placed laminated copies of the
screening protocol (Appendix G), Escape Instrument (Appendix H), and child physical abuse
diagnostic codes (Appendix I) in the PED physician’s room and at the nurses’ station. All ED
HCPs were given laminated badge cards containing the Escape Instrument and child abuse
diagnostic codes. Laminated reminders with the words “Did you remember to ESCAPE?” were
placed on each computer screen in the PED patient rooms and at the nurses’ station to remind the
RNs to complete the Escape Instrument during patient intake. An email was sent to all PED
HCPs and staff announcing the official launch of the child physical abuse screening program.
After implementation, the project team leader visited the PED weekly to bi-weekly to answer
questions, educate new HCPs on the screening protocol, and offer support to the staff.

CHILD PHYSICAL ABUSE SCREENING PROGRAM

18

Outcome Measures
The impact of the comprehensive screening program on ED HCP recognition and
documentation of child physical abuse was the primary outcome evaluated for this project. The
secondary outcomes evaluated were (a) ED HCP knowledge in screening for and recognizing
child physical abuse, (b) ED HCP confidence in screening for and recognizing child physical
abuse, and (c) the utility of the child physical abuse screening program.
Instruments
The impact of the screening program on HCP recognition and documentation of child
physical abuse was measured by analyzing the number of International Classification of
Diseases, 9th and 10th edition (ICD-9 and ICD-10) diagnostic codes entered by ED HCPs for
both suspected and confirmed child physical abuse. ICD codes can be reliably used to identify
cases of child abuse. The specificity of ICD codes for child physical abuse ranges from 92% to
100%, but sensitivity is lower (74% to 98.2%) due to errors in HCP documentation and coding
(Hooft et al., 2015; Hooft et al., 2013; McKenzie, Scott, Waller, & Campbell, 2011).
ED HCP knowledge and confidence in screening for and recognizing child physical
abuse were both measured with the 7-item Child Abuse Awareness pretest (Appendix J) and
posttest (Appendix K). Items one and two used a 5-point Likert scale to assess HCP confidence,
and items three through seven used multiple choice and multi-select responses to evaluate HCP
knowledge. The tests were similar, except for items five and six, which were changed to evaluate
accurate application of the Escape Instrument on the posttest. Since there was no valid
instrument to measure the desired outcomes, test questions were developed to align with the
educational session objectives (Appendix L). The test questions were evaluated by three child
abuse experts to establish face and content validity (McDonald, 2014), and one psychometric

CHILD PHYSICAL ABUSE SCREENING PROGRAM

19

expert. Content validity was assessed using a content validity index (CVI) (Polit & Beck, 2006).
Only items with a CVI score of 1.00 and agreed upon face validity by all three experts were
included in the final pretest and posttest.
The utility of the child physical abuse screening program was measured with the 8-item
Project Evaluation Survey (Appendix M). The purpose of this final survey was to evaluate the
implementation process from the HCP perspective; ascertain the level of HCP support of systemwide implementation of the program throughout all EDs; and determine the effect of the
evidence-based screening program on HCP confidence and self-efficacy in screening for and
recognizing child physical abuse. Questions regarding attendance at previous child abuse
educational seminars were also included to provide information on additional training received
by the HCPs who completed the final project survey.
In addition to the instruments used for data collection, RNs completed the validated
Escape Instrument (Appendix H) for any child ≤ 18 years old who was admitted to the PED. This
6-item child abuse screening tool requires yes/no responses for each item. Item three also
includes a non-applicable (n/a) response for children presenting to the ED with a non-injury
complaint. One or more aberrant answers indicates a positive screenand a heightened concern
for child physical abusewhile no aberrant answers constitutes a negative screen and a
decreased risk of abuse (Louwers et al., 2012b). As previously noted, the Escape Instrument has
demonstrated high reliability (99.2%) when used to screen children for physical abuse in the ED
(Dinpanah et al., 2017). Additionally, Louwers et al. (2014) found the specificity and negative
predictive value of the Escape tool to be high (98% and 99%, respectively), indicating that child
physical abuse is unlikely with a negative screen.
Data Collection and Analysis

CHILD PHYSICAL ABUSE SCREENING PROGRAM

20

ICD-10 codes entered by ED HCPs for suspected and confirmed child physical abuse
were collected for 30 consecutive days prior to implementation of the screening program (June 1,
2017 – June 30, 2017) and 30 consecutive days after implementation (August 20, 2017 
September 18, 2017). ICD-10 sub-classification codes were included in the data sets to ensure all
child physical abuse diagnostic codes were captured (Appendix I). To control for seasonal
factors, data sets for the same ICD-10 codes were collected for the exact two sets of 30
consecutive days during the previous year (2016). The ICD-9 code for child physical abuse
(995.54) was also included in the 2016 data sets due to the transition from ICD-9 to ICD-10 in
October 2016. All ICD coding data was collected by the Clinical Data Warehouse (CDW) staff
and presented to the project team leader in aggregate form. Coding data was intended to include
four months of data both before and after project implementation. However, due to the project
site’s change to a new EHR shortly after implementation of the child physical abuse screening
program, the CDW staff were not able to collect any ICD coding data after September 18, 2017.
The Child Abuse Awareness pretest and posttest were administered at two individual staff
meetings in July and August 2017. The pretest was administered immediately before the 20minute educational session, and the posttest was administered immediately after the session. The
Project Evaluation Survey was subsequently administered at two individual staff meetings in
December 2017four months after implementation of the child physical abuse screening
program (see Appendix N for the complete project flow diagram).
Data analysis was performed using Statistical Package for the Social Sciences (SPSS)
version 24, with a statistical significance level set at ≤ .05. Descriptive statistics were used to
describe the sample and outcome variables. A paired-samples t test was run on data from the
Child Abuse Awareness pretest and posttest to measure the difference in HCP knowledge (items

CHILD PHYSICAL ABUSE SCREENING PROGRAM

21

3, 4, 7) and confidence (items 1, 2) before and after the educational session. Items five and six
were not included in the total score or paired-samples t test because they differed between tests.
Qualitative data from the Project Evaluation Survey was entered into SPSS to allow
quantification of common responses. Text responses and comments were summarized to fully
capture project feedback from all survey participants.
Budget
There was no formal financial budget for this project. No external funding or grant
monies were received. The only material costs incurred were printing and laminating costs for
the HCP badge cards and the PED project-related signage, as well as refreshments served during
the child abuse educational sessions and the occasional PED project-related visits. Total material
costs for this project were less than $200 and were at the expense of the project team leader.
Budgeting of personnel time was vital to this project. Approximately 50 hours were
required for the project team leader to develop and deliver the child abuse educational session to
ED HCPs. Time spent by ED HCPs who participated in the educational session and Child Abuse
Awareness pretest/posttest was 30 minutes, and time spent by ED HCPs who participated in the
Project Evaluation Survey was 10 minutes. The educational session and data collection occurred
during non-mandatory staff meetings that were scheduled during normal work hours. Finally, the
running and compilation of ICD coding data by the CDW staff required 60 minutes and only
involved one staff member.
Results
Demographic Data
To protect the privacy of the ED HCPs participating in the project, limited demographic
data about HCPs was collected. Fifty-two ED HCPs completed both the Child Abuse Awareness

CHILD PHYSICAL ABUSE SCREENING PROGRAM

22

pretest and posttest, including 39 (75%) EM physicians and 13 (25%) PED RNs. Fourteen ED
HCPs completed the final Project Evaluation Survey, including 10 EM physicians (71.4%) and
four PED RNs (28.6%). Of these 14 survey participants, 12 (85.7%) attended the 20-minute child
physical abuse educational session at the start of the project, and two (14.3%) did not. The
number of previous child abuse educational sessions attended by the participants ranged from 0–
11, with an average of 4.43 (SD = 3.92).
No demographic data was collected on patients. ICD data searches were limited to only
include child physical abuse codes entered by ED HCPs on patients ≤ 18 years of age. All ICD
data was reported by the CDW staff to the project team leader in aggregate form.
Outcomes
Impact of the screening program on ED HCP recognition and documentation of
child physical abuse. There were no child physical abuse diagnostic codes entered by ED HCPs
in the 30-day period prior to implementation of the screening program, which was similar to the
same 30-day period in the preceding year. In the 30-day period immediately following project
implementation, there were three child physical abuse diagnostic codes entered by ED HCPs,
compared to two in the same 30-day period the preceding year. Due to the small sample sizes, no
statistical analysis was conducted.
ED HCP knowledge in screening for and recognizing child physical abuse. The
average total knowledge score on the Child Abuse Awareness pretest was 21.4% (SD = 21.71).
Post educational session, the average score was 73.2% (SD = 25.64). A significant increase in
knowledge was found from the pretest to posttest (t(51) = -13.831, p < .001) (Table O1). Items
five and six on the pretestwhich assessed knowledge of when to screen for child physical
abusewere both answered correctly by 86.5% (n = 45) of participants. On the posttestwhich

CHILD PHYSICAL ABUSE SCREENING PROGRAM

23

evaluated accurate use of the Escape Instrument88.5% (n = 46) of participants answered item
five correctly, and 92.3% (n = 48) answered item six correctly.
ED HCP confidence in screening for and recognizing child physical abuse. Prior to
the educational session, the average confidence score for recognizing child physical abuse was
3.48 (SD = .70), and the average confidence score for screening for child physical abuse was
3.31 (SD = .67). Post educational session, the average confidence score for child physical abuse
recognition was 3.87 (SD = .56), and the average confidence score for child physical abuse
screening was 3.77 (SD = .61). A significant increase from pretest to posttest was found for both
confidence in recognizing child physical abuse (t(51) = -5.236, p < .001) and confidence in
screening for child physical abuse (t(51) = -5.778, p < .001) (Table O2).
The utility of the child physical abuse screening program. Only 14 ED HCPs
completed the Project Evaluation Survey. The average confidence score for recognizing child
physical abuse was 3.93 (SD = .48), while the average confidence score for screening for child
physical abuse was 3.79 (SD = .70). The Escape Instrument and the educational session were
each noted as facilitators to screening and detection of child physical abuse by 35.7% (n = 5) of
participants. The systematic screening protocol was noted as a facilitator by 14.3% (n = 2) of
participants, as was nursing involvement (RN-driven screening process, with positive results
communicated to EM physicians). Five participants (35.7%) did not note any facilitators.
Fifty percent (n = 7) of survey participants noted the EHR transition to be the most
significant barrier to child physical abuse screening in the PED. Two participants (14.2%)
identified the busy ED setting and time restrictions as another barrier, and two participants
(14.2%) did not note any barriers to screening. Finally, the following barriers were identified by
a total of one participant (7.1%) each: delays in obtaining a CAT consult; uncertainty regarding

CHILD PHYSICAL ABUSE SCREENING PROGRAM

24

the details of the screening program; RN-driven screening process (participant reported RNs are
not appropriately trained to screen for child physical abuse); positive Escape screens that did not
have significant concern for abuse; no dedicated place to document the Escape screening results
in the EHR; and ED triage staff not being as familiar with the Escape Instrument as PED RNs.
Building the Escape Instrument into the EHR was recommended by 57.1% (n = 8) of
survey participants, with two participants (14.3%) specifically recommending that completion of
the screening tool be required before the ED HCP could proceed to other sections of the EHR.
Additional changes that were recommended by a total of one participant (7.1%) each included:
increased and more widespread training; having the EM physicians administer the Escape
Instrument instead of the RNs; having a more thorough discussion with EM physicians prior to
implementing the screening protocol; increasing screening awareness and knowledge of next
steps; and having occasional lectures on child physical abuse at future staff meetings. Three
survey participants (21.4%) did not recommend any changes to the child physical abuse
screening protocol or educational session.
Of the 14 ED HCPs who completed the Project Evaluation Survey, 11 (78.6%) supported
system-wide implementation of the ED child physical abuse screening program, and three
(21.4%) did not. Survey participants provided a variety of rationales in support of system-wide
implementation. Two participants (14.3%) noted increased knowledge and awareness of child
physical abuse; two (14.3%) noted increased identification of abuse cases that would otherwise
be missed; and two (14.3%) noted the program is a system that is designed to protect children.
Additional supporting factors noted by one participant (7.1%) each included: the Escape
Instrument is practical and easy to use; the systematic screening protocol facilitates RN to
physician communication; and the program allows collection of data on the actual prevalence of

CHILD PHYSICAL ABUSE SCREENING PROGRAM

25

child abuse. Of the three participants not supporting system-wide implementation, one noted the
short implementation time and unclear results at the project site; one stated that RNs do not have
adequate training to screen for child physical abuse; and one noted the program should first be
widely validated and peer reviewed.
Discussion
Despite some unanticipated system changes that occurred during implementation of this
QI project, the comprehensive child physical abuse screening program demonstrated statistically
significant results that were consistent with the evidence. Both ED HCP knowledge and
confidence in screening for and recognizing child physical abuse improved after the 20-minute
educational session. The Escape Instrument, screening protocol, and educational session were all
noted as facilitators to child physical abuse screening and detection in the ED. Participants also
noted that integration of the Escape Instrument into the EHR would greatly facilitate the
screening process, as it would provide a specific place to document and view screening results.
Finally, at the end of the project, the majority of ED HCPs surveyed supported full
implementation of the screening program throughout all EDs in the health system.
The results of the final Project Evaluation Survey also pointed to some interesting power
dynamics within the ED. One participant commented that RNs do not have the proper training to
screen for child physical abuse and should not be entrusted to administer the Escape Instrument.
Rather, it was suggested that the EM physicians be responsible for completing the Escape
Instrument and interpreting the results. Due to the limited participant demographics collected on
the survey, it is uncertain whether the respondent who questioned the appropriateness of the RNdriven screening process directly works with the PED RNs.
Limitations

CHILD PHYSICAL ABUSE SCREENING PROGRAM

26

The biggest limitation to this project was the loss of access to all ICD coding data on
child physical abuse after September 18, 2017. This loss of access resulted in only 30 days of
coding data being available for collection after official implementation of the newly developed
PED child physical abuse screening protocol. The number of child physical abuse cases that
present to the ED can vary significantly from month to month. Having only one month of postimplementation data available for comparison to pre-implementation data prevented analysis that
would demonstrate the true impact of the screening programspecifically, whether or not ED
HCP detection and documentation of child physical abuse increased after implementation of the
educational session and systematic screening protocol.
Another limitation was the lack of a previously validated tool to measure ED HCP
knowledge and confidence in screening for and recognizing child physical abuse. However, once
developed by the project team leader, three child abuse experts reviewed the Child Abuse
Awareness pretest and posttest to establish face and content validity prior to participant
administration. This expert review helped reduce the effects of this limitation on the project.
Although the project team leader desired to include nurse practitioners (NPs) as
participants in this project, the PED at the project site does not currently employee NPs.
Additionally, the PED RNs and EM physicians were not required to attend the staff meetings at
which the child physical abuse educational sessions were offered. As a result, not all ED HCPs
participated in the educational session that preceded implementation of the systematic screening
protocol, nor did all ED HCPs receive training in the use of the Escape Instrument. Finally, due
to the constant change in EM residents that rotate through the PED, as well as RN staffing and
management changes that occurred during the tenure of this project, not all the ED HCPs were
aware of the specifics of the screening program and what their particular roles in screening were.

CHILD PHYSICAL ABUSE SCREENING PROGRAM

27

Implications for Practice
Child abuse and neglect costs the U.S. $220 million each day. This equates to more than
$80 billion every year (Gelles & Perlman, 2012). In addition to a validated screening tool and
systematic screening guidelines, provider-based education is needed to increase HCP knowledge
of child abuse injuries (Jordan & Moore-Nadler, 2014; Louwers et al., 2012b; Louwers et al.,
2014). Comprehensive child physical abuse screening programssuch as the one developed for
this projectcan help ED HCPs make significant strides in the effort to identify child physical
abuse early and prevent its immediate and long-term effects. Not only will these types of
programs increase ED HCP knowledge, confidence, and self-efficacy in child physical abuse
screening and recognition, but they can also result in improved detection of non-accidental
childhood injuries. Ultimately, these implications can be far-reaching, positioning ED HCPs to
be at the frontlines of early detection and treatment of child physical abuse.
Conclusion
This evidence-based QI project evaluated the implementation of a comprehensive child
physical abuse screening program in the PED that included delivery of a 20-minute educational
session to ED HCPs, creation of a systematic child physical abuse screening protocol, and
completion of the validated Escape Instrument by PED RNs for all patients ≤ 18 years of age.
Overall, the project appeared to be successful. ED HCPs had statistically significant increases in
their knowledge and confidence scores for child physical abuse screening and recognition after
the educational session. The Escape Instrument, educational session, and systematic protocol
facilitated screening in the PED, while transition to a new EHR presented significant barriers to
screening for child physical abuse. However, no significant difference in ED HCP diagnostic
coding of child physical abuse was appreciated after implementation of the screening program.

CHILD PHYSICAL ABUSE SCREENING PROGRAM

28

The results of this project are consistent with past research supporting comprehensive
programs to improve ED HCP knowledge and confidence in screening for and recognizing child
physical abuse. They also highlight the importance of integrating the child physical abuse SI into
the EHR to facilitate screening. Future research should focus on the impact of screening
programs on HCP diagnosis and treatment of child physical abuse. Longitudinal studies
examining the relationship between screening results, child physical abuse diagnosis, and final
case adjudication are also needed. Finally, efforts should be made to support best practice by
standardizing child physical abuse screening programs throughout all EDs, with the potential for
spread to other settings, including primary care clinics and urgent care centers.

CHILD PHYSICAL ABUSE SCREENING PROGRAM

29

References
Acehan, S., Avci, A., Gulen, M., Segmen, M. S., Akoglu, H., Kozaci, N., . . . Satar, S. (2016).
Detection of the awareness rate of abuse in pediatric patients admitted to emergency
medicine department with injury. Turkish Journal of Emergency Medicine, 16(3), 102106. doi: 10.1016/j.tjem.2016.06.001
Allareddy, V., Asad, R., Lee, M. K., Nalliah, R. P., Rampa, S., Speicher, D. G., . . . Allareddy,
V. (2014). Hospital based emergency department visits attributed to child physical abuse
in United States: Predictors of in-hospital mortality. PLoSOne, 9(2), e100110.
doi:10.1371/journal.pone.0100110
American Academy of Pediatrics. (2017). The section on child abuse and neglect. Retrieved
from https://www2.aap.org/sections/childabuseneglect
Bailhache, M., Leroy, V., Pillet, P., & Salmi, L. (2013). Is early detection of abused children
possible? A systematic review of the diagnostic accuracy of the identification of abused
children. BMC Pediatrics, 13(1), 202-213. doi: 10.1186/1471-2431-13-202
Bair-Merritt, M. H., & Lane, W. G. (2011). An evidence-based approach to the evaluation and
treatment of child physical abuse. Pediatric Emergency Medicine Practice, 8(5), 1-20.
Retrieved from https://www.ebmedicine.net/topics.php?paction=showTopic&topic_id=
260
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change.
Psychological Review, 84(2), 191-215. Retrieved from https://www.uky.edu/~eushe2/
Bandura/Bandura1977PR.pdf
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice-Hall.

CHILD PHYSICAL ABUSE SCREENING PROGRAM

30

Bandura, A. (1994). Self-efficacy. In V. S. Ramachaudran (Ed.), Encyclopedia of human
behavior (Vol. 4) (pp. 71-81). New York, NY: Academic Press.
Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY: W.H. Freeman.
Bandura, A. (2009). Cultivate self-efficacy for personal and organizational effectiveness. In E.
A. Locke (Ed.), Handbook of principles of organizational behavior (2nd ed.) (pp. 179200). New York, NY: Wiley.
Benger, J. R., & Pearce, A. V. (2002). Quality improvement report: Simple intervention to
improve detection of child abuse in emergency departments. BMJ, 324(7340), 780-782.
doi: 10.1136/bmj.324.7340.780
Childhelp. (2017). Child abuse statistics & facts. Retrieved from https://www.childhelp.org/
child-abuse-statistics
Child Welfare Information Gateway. (2011). About CAPTA: A legislative history. Washington,
DC: U.S. Department of Health and Human Services, Children's Bureau. Retrieved from
https://www.childwelfare.gov/pubPDFs/about.pdf
Child Welfare Information Gateway. (2016). Definitions of child abuse and neglect. Washington,
DC: U.S. Department of Health and Human Services, Children's Bureau. Retrieved from
https://www.childwelfare.gov/topics/systemwide/laws-policies/statutes/define
Crichton, K. G., Cooper, J. N., Minneci, P. C., Groner, J. I., Thackeray, J. D., & Deans, K. J.
(2016). A national survey on the use of screening tools to detect physical child abuse.
Pediatric Surgery International, 32(8), 815-818. doi: 10.1007/s00383-016-3916-z
Dinpanah, H., Pasha, A. A., & Sanji, M. (2017). Potential child abuse screening in emergency
department: A diagnostic accuracy study. Emergency, 5(1), e8. doi:
10.22037/emergency.v5i1.12396

CHILD PHYSICAL ABUSE SCREENING PROGRAM

31

Escobar, M. A., Pflugeisen, B. M., Duralde, Y., Morris, C. J., Haferbecker, D., Amoroso, P. J., . .
. Pohlson, E. C. (2016). Development of a systematic protocol to identify victims of nonaccidental trauma. Pediatric Surgery International, 32(4), 377-386. doi: 10.1007/s00383016-3863
Fang, X., Brown, D. S., Florence, C. S., & Mercy, J. A. (2012). The economic burden of child
maltreatment in the United States and implications for prevention. Child Abuse &
Neglect, 36(2), 156-165. doi: 10.1016/j.chiabu.2011.10.006
Felitti, V. J., Anada, R. F., Nordenberg, D., Williamson, D. F., Spitz, A. M., Edwards, V., . . .
Marks, J. S. (1998). Relationship of childhood abuse and household dysfunction to many
of the leading causes of death in adults: The adverse childhood experiences (ACE) study.
American Journal of Preventative Medicine, 14(4), 245-258. doi: 10.1016/S07493797(98)00017-8
Gelles, R. J., & Perlman, S. (2012). Estimated annual cost of child abuse and neglect. Chicago,
IL: Prevent Child Abuse America. Retrieved from http://preventchildabuse.org/images/
research/PCA_COM2012.pdf
Glick, J. C., Lorand, M. A., & Bilka, K. R. (2016). Physical abuse of children. Pediatrics in
Review, 37(4), 146-156. doi: 10.1542/pir.2015-0012
Guenther, E., Knight, S., Olson, L. M., Dean, M., & Keenan, H. T. (2009). Prediction of child
abuse risk from emergency department use. The Journal of Pediatrics, 154(2), 272-277.
doi: 10.1016/j.jpeds.2008.07.047
Higginbotham, N., Lawson, K. A., Gettig, K., Roth, J., Hopper, E., Higginbotham, E., . . .
Garcia, N. M. (2014). Utility of a child abuse screening guideline in an urban pediatric

CHILD PHYSICAL ABUSE SCREENING PROGRAM

32

emergency department. Journal of Trauma and Acute Care Surgery, 76(3), 878-883. doi:
10.1097/TA.0000000000000135
Hoft, M., & Haddad, L. (2017). Screening children for abuse and neglect: A review of the
literature. Journal of Forensic Nursing, 13(1), 26-33. doi:
10.1097/JFN.0000000000000136
Hooft, A. M., Asnes, A. G., Livingston, N., Deutsch, S., Cahill, L., Wood, J. N., & Leventhal, J.
M. (2015). The accuracy of ICD codes: Identifying physical abuse in 4 children’s
hospitals. Academic Pediatrics, 15(4), 444-450. doi: 10.1016/j.acap.2015.01.008
Hooft, A., Ronda, J., Schaeffer, P., Asnes, A. G., & Leventhal, J. M. (2013). Identification of
physical abuse cases in hospitalized children: Accuracy of international classification of
diseases codes. The Journal of Pediatrics, 162(1), 80-85. doi:
10.1016/j.jpeds.2012.06.037
Hoytema van Konijnenburg, E. M., Teeuw, A. H., Zwaard, S. A., van der Lee, J. H., & van Rijn,
R. R. (2014). Screening methods to detect child maltreatment: High variability in Dutch
emergency departments. Emergency Medicine Journal, 31(3), 196-200. doi:
10.1136/emermed-2013-202733
Jordan, K. S., & Moore-Nadler, M. (2014). Children at risk of maltreatment: Identification and
intervention in the emergency department. Advanced Emergency Nursing Journal, 36(1),
97-106. doi: 10.1097/TME.0000000000000011
Jordan, K. S., & Steelman, S. H. (2015). Child maltreatment: Interventions to improve
recognition and reporting. Journal of Forensic Nursing, 11(2), 107-113. doi:
10.1097/JFN.0000000000000068

CHILD PHYSICAL ABUSE SCREENING PROGRAM

33

King, W. K., Kiesel, E. L., & Simon, H. K. (2006). Child abuse fatalities: Are we missing
opportunities for intervention? Pediatric Emergency Care, 22(4), 211-214. doi:
10.1097/01.pec.0000208180.94166.dd
Larrabee, J. H. (2009). Nurse to nurse: Evidence-based practice. New York, NY: McGraw-Hill.
Louwers, E. C., Affourtit, M. J., Moll, H. A., de Koning, H. J., & Korfage, I. J. (2010). Screening
for child abuse at emergency departments: A systematic review. Archives of Disease in
Childhood, 95(3), 214-218. doi: 10.1136/adc.2008.151654
Louwers, E. C., Korfage, I. J., Affourtit, M. J., De Koning, H. J., & Moll, H. A. (2012a).
Facilitators and barriers to screening for child abuse in the emergency department. BMC
Pediatrics, 12(1), 167-172. doi: 10.1186/1471-2431-12-167
Louwers, E. C., Korfage, I. J., Affourtit, M. J., Ruige, M., van den Elzen, A. P., de Koning, H. J.,
& Moll, H. A. (2014). Accuracy of a screening instrument to identify potential child
abuse in emergency departments. Child Abuse & Neglect, 38(7), 1275-1281. doi:
10.1016/j.chiabu.2013.11.005
Louwers, E. C., Korfage, I. J., Affourtit, M. J., Scheewe, D. J., van de Merwe, M. H., VooijsMoulaert, A. F., . . . de Koning, H. J. (2011). Detection of child abuse in emergency
departments: A multi-centre study. Archives of Disease in Childhood, 96(5), 422-425.
doi: 10.1136/adc.2010.202358
Louwers, E. C., Korfage, I. J., Affourtit, M. J., Scheewe, D. J., van de Merwe, M. H., VooijsMoulaert, A. F., . . . de Koning, H. J. (2012b). Effects of systematic screening and
detection of child abuse in emergency departments. Pediatrics, 130(3), 457-464. doi:
10.1542/peds.2011-3527

CHILD PHYSICAL ABUSE SCREENING PROGRAM

34

McDonald, M. E. (2014). The nurse educator’s guide to assessing learning outcomes (3rd ed.).
Burlington, MA: Jones & Bartlett Learning.
McKenzie, K., Scott, D. A., Waller, G. S., & Campbell, M. (2011). Reliability of routinely
collected hospital data for child maltreatment surveillance. BMC Public Health, 11(1), 815. doi: 10.1186/1471-2458-11-8
Melnyk, B. M., & Fineout-Overholt, E. (2015). Evidence-based practice in nursing &
healthcare: A guide to best practice (3rd ed.). Philadelphia, PA: Wolters Kluwer.
Milani, G. P., Vianello, F. A., Cantoni, B., Agostoni, C., & Fossali, E. F. (2016). New program
for identification of child maltreatment in emergency department: Preliminary data.
Italian Journal of Pediatrics, 42(1), 66-67. doi: 10.1186/s13052-016-0275-2
National Association of Pediatric Nurse Practitioners. (2016). NAPNAP position statement on
child maltreatment. Journal of Pediatric Health Care, 30(5), A15-A17. doi:
10.1016/j.pedhc.2016.05.006
National Children’s Alliance. (2014). National initiatives. Retrieved from
http://www.nationalchildrensalliance.org/national-initiatives
Petska, H.W., & Sheets, L. K. (2014). Sentinel injuries: Subtle findings of physical abuse.
Pediatric Clinics of North America, 61(5), 923-935. doi: 10.1016/j.pcl.2014.06.007
Polit, D. F., & Beck, C. T. (2006). The content validity index: Are you sure you know what’s
being reported? Critique and recommendations. Research in Nursing & Health, 29(5),
489-497. doi: 10.1002/nur.20147
Rosswurm, M. A., & Larrabee, J. H. (1999). A model for change to evidence-based practice.
Image: Journal of Nursing Scholarship, 31(4), 317-322. doi: 10.1111/j.15475069.1999.tb00510.x

CHILD PHYSICAL ABUSE SCREENING PROGRAM

35

Sheets, L. K., Leach, M. E., Koszewski, I. J., Lessmeier, A. M., Nugent, M., & Simpson, P.
(2013). Sentinel injuries in infants evaluated for child physical abuse. Pediatrics, 131(4),
701-707. doi: 10.1542/peds.2012-2780
Sittig, J. S., Post, E. D., Russel, I. M., van Dijk, I. A., Nieuwenhuis, E. E., & van de Putte, E. M.
(2014). Evaluation of suspected child abuse at the ED: Implementation of American
academy of pediatrics guidelines in the Netherlands. The American Journal of Emergency
Medicine, 32(1), 64-66. doi: 10.1016/j.ajem.2013.08.038
Sittig, J. S., Uiterwaal, C. S., Moons, K. G., Russel, I. M., Nievelsteain, R. A., Nieuwenhuis, E.
E., & van de Putte, E. M. (2016). Value of systematic detection of physical abuse at
emergency rooms: A cross-sectional diagnostic accuracy study. BMJ Open, 6(3),
e010788. doi: 10.1136/bmjopen-2015-010788
Smeekens, A. E., Boekhuijsen-van Henten, D. M., Sittig, J. S., Russel, I. M., ten Cate, O. T.,
Turner, N. M., & van de Putte, E. M. (2011). Successful e-learning programme on the
detection of child abuse in emergency departments: A randomised controlled trial.
Archives of Disease in Childhood, 96(4), 330-334. doi: 10.1136/adc.2010.190801
Teeuw, A. H., Derkx, B. H., Koster, W. A., & van Rijn, R. R. (2012). Educational paper:
Detection of child abuse and neglect at the emergency room. European Journal of
Pediatrics, 171(6), 877-885. doi: 10.1007/s00431-011-1551-1
Teeuw, A. H., Sieswerda-Hoogendoorn, T., Sangers, E. J., Heymans, H. S., & van Rijn, R. R.
(2016). Results of the implementation of a new screening protocol for child maltreatment
at the emergency department of the academic medical center in Amsterdam. International
Emergency Nursing, 24, 9-15. doi: 10.1016/j.ienj.2015.05.002

CHILD PHYSICAL ABUSE SCREENING PROGRAM

36

Tiyyagura, G., Gawel, M., Koziel, J. R., Asnes, A., & Bechtel, K. (2015). Barriers and
facilitators to detecting child abuse and neglect in general emergency departments.
Annals of Emergency Medicine, 66(5), 447-454. doi: 10.1016/j.annemergmed.2015.
06.020
U.S. Department of Health and Human Services, Administration for Children, Youth, and
Families, Children’s Bureau. (2017). Child maltreatment 2015. Retrieved from
https://www.acf.hhs.gov/cb/resource/child-maltreatment-2015
Woodman, J., Lecky, F., Hodes, D., Pitt, M., Taylor, B., & Gilbert, R. (2009). Screening injured
children for physical abuse or neglect in emergency departments: A systematic review.
Child: Care, Health and Development, 36(2), 153-164. doi: 10.1111/j.13652214.2009.01025x
Woodman, J., Pitt, M., Wentz, R., Taylor, B., Hodes, D., & Gilbert, R. E. (2008). Performance of
screening tests for child physical abuse in accident and emergency departments. Health
Technology Assessment, 12(33), iii, xi-xiii, 1-95. doi: 10.3310/hta12330

CHILD PHYSICAL ABUSE SCREENING PROGRAM

37

Appendix A
Search Strategy Flow Diagram

Stage 1: Initial
Search Results

Identified citations from
database search after
duplicate citations removed
(n = 430)

Stage 2: Filters
Applied

Included according
to filters
(n = 193)

Stage 3: Title
Review

Included by title
(n = 59)

Stage 4:
Abstract
Review

Included by abstract
(n = 33)

Excluded by abstract
(n = 26)

Stage 5: Critical
Appraisal

Full text articles
critically appraised
(n = 41)

Ancestry, hand, and
grey literature search
(n = 8)

Stage 6: Final
Search Results

Included studies
(n = 10)

Articles excluded
based on article content
(n = 31)

Duplicate
citations excluded
(n = 294)

Excluded according
to filters
(n = 237)

Excluded by title
(n = 134)

CHILD PHYSICAL ABUSE SCREENING PROGRAM

38
Appendix B
Evaluation Table

Citation

Dinpanah, H., et al.
(2017).
Potential child
abuse screening in
emergency
department: A
diagnostic accuracy
study
Funding: None
Conflicts/Bias:
None
Country: Iran

Conceptual/
Theoretical
Framework
Cochrane
protocol for
systematic
review of
interventions

Design/
Method
Design:
Prospective
DAS
Purpose:
Evaluate
accuracy of
Escape tool in
screening
children at risk
for CA

Sample/Setting

N = 6120
Demographics:
F = 52%
Reside in city:
71.5%
MA = 2.19 +/- 1.12
years
1-4 years: 2244
(36.6)
4-8 years: 1548
(25.3)
8-12 years: 1192
(19.5)
12-16 years: 1136
(18.6)
AR: 0
Setting: 2 EDs
Eligibility:
Children < 16 y.o.
presenting to ED
from 2011 – 2014

Major Themes,
Variables, &
Definitions
IV: Escape tool
(SI)
DV: Detection
of children at
risk for CA

Measurement/
Instrumentation

Data Analysis

Findings/
Themes

Escape SI for
potential CA
 ≥ 1 aberrant
answer =
positive screen
 V&R per
Louwers et al
(2014): Sn
0.80; Sp 0.98;
PPV 0.10;
NPV 0.99

Sn, Sp, PLR,
NLR, PPV, and
NPV, and area
under the ROC
were calculated
with 95% CI to
evaluate
accuracy of SI

+ Escape tool:
n = 137
1 +SII: 120 (2)
2 +SII: 4 (0.1)
3 +SII: 1 (0.01)
4 +SII: 1 (0.01)

Physical exam
by ED specialist
+ CA diagnosis
by ED specialist
→ reevaluation
of child by CAT
to confirm CA
diagnosis (gold
standard)

STATA 11.0
used for data
analysis

CAT opinion:
+ CA = 35 (0.5)
DV:
Escape tool
accuracy:
Sn = 100 (87.6100)
Sp = 98.3 (97.9
– 98.6
PLR = 25.5
(18.6-33.8)
NLR = 100
(99.9-100)
PPV = 0.34
(0.25-0.46)
NPV = 0 (0NAN)

Decision for Use/
Application to
Practice
LOE: IV
Strengths:
Large N (minimum
required = 2696); 2
study sites; ED and
CAT specialists
blind to SI results (↓
bias); all pediatric
age groups included;
final CA diagnosis
known
Limitations:
Convenience
sampling – ↑ risk of
bias; not controlled
Conclusions:
Escape is a suitable
SI for detecting
potential CA cases
in ED (99.2%
accuracy in this
study).

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

39
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

Data Analysis

Area under
ROC curve =
99.2 (98.9-99.4)

Inclusion criteria:
Consent for
participation;
cooperation in
filling out SI;
hemodynamically/
clinically stable

Higginbotham, N.,
et al. (2014).
Utility of a child
abuse screening
guideline in an
urban emergency
department.
Funding: Dell
Children’s Medical
Center of Central
Texas/Seton
Healthcare
Conflicts/Bias:
None noted

Guideline for
the evaluation
of suspected
NAT

Design: QES;
Data collected
18 months
before and 18
months after SP
implementation
Purpose:
Implement SP
with goal of
limiting bias in
screening for
suspected CA
Hypothesis:
Implementing a
SP for all

Exclusion criteria:
Cases of suicide
injury, poisoning,
peer injury, or
disclosed CA
N = 332
PREG: n = 111
PSTG: n = 221
Demographics:
PREG
MA (SD), months
= 6.7 (3.1)
F = 53 (47.8)
M = 58 (52.3)
W = 46 (41.4)
H = 49 (44.1)
B/A/O = 16 (14.4)
PI = 38 (34.2)
GSNI = 73 (65.8)

Findings/
Themes

IV: CA SP
DV1: Racial
bias/referral
patterns
DV2: SES bias/
referral patterns
DV3:
Performance of
SP-mandated
tests

Hospital
databases and
medical chart
review (PREG
vs. PSTG)

χ2 analysis to
compare patient
characteristics,
SP diagnostic
test use, CAT/
CPS referral
UVR to
examine
association
between race
and insurance
status with CA
screening
likelihood
p < 0.05

No SSD among
groups for age,
race, sex, SF
type, insurance
DV1:
PREG
SS =
H/B/A/O > W
(OR, 1.31; 95%
CI, 0.61-2.83;
p= 0.49)
CAT referral =
GSNI > PI
(OR, 2.6; 95%
CI, 1.11-5.95; p
= 0.028

Decision for Use/
Application to
Practice
Feasibility:
This is a brief, 6item SI that can
easily and quickly
be administered by
ED nurses or
providers to screen
children for CA.
Minimal impact on
work flows if filled
out during triage.
No cost to use SI.
LOE: III
Strengths:
Good LOE; no SSD
between PREG and
PSTG; evidencebased SP
Limitations:
N only included
infants; temporal
nature of data –
possible bias due to
changes in patient
population during
tenure of study
(authors suggest this

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Country: U.S.A.

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

patients
younger than 12
months who
present to the
hospital with a
SF not caused
by a MVC will
attenuate racial
and SES bias on
the part of the
clinician.

PSTG
MA (SD), months
= 6.1 (3.4)
F = 94 (42.5)
M = 137 (57.5)
W = 108 (48.9)
H = 85 (38.5)
B/A/O = 28 (12.7)
PI = 90 (40.7)
GSNI = 131 (59.3)
AR: 0
Setting: PED
trauma center
Inclusion criteria:
age < 12 months;
any type of SF
(non MVC-related)
Exclusion criteria:
None listed

40
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

Data Analysis

Findings/
Themes
CPS referral =
GSNI > PI (OR,
3.85; 95% CI,
1.66-8.91;
p=0.002)
PSTG
SS = No SSD
between W and
H/B/A/O (OR,
1.05; 95% CI,
0.51-2.91; p =
0.89)
CAT referral =
GSNI > PI (OR,
2.5; 95% CI,
1.34-4.50; p =
0.004)
CPS referral =
GSNI > PI (OR,
3.0; 95% CI,
1.71-5.26; p <
0.001)
DV2:
PREG
SS = GSNI > PI
(OR, 2.69; 95%
CI; p = 0.017)

Decision for Use/
Application to
Practice
is limited due to
similarities between
PREG and PSTG);
study lacked data on
final case
adjudication by CPS
and law
enforcement
Conclusions:
Implementation of a
SP can help
attenuate SES bias,
but not racial bias,
when screening for
CA; SP allows CA
screening to become
more consistent and
evidence-based; SP
results in ↑
frequency of
appropriate and
↓frequency of
unnecessary CA
screening tests/
referrals
Feasibility:
Recommend SP
implementation.

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

41
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

Data Analysis

Findings/
Themes
PSTG
SS = Insurance
status no longer
a factor (OR,
1.18; 95% CI,
0.56-2.46; p =
0.66)
DV1 & DV2:
ORs for UA &
TA screening
showed similar
relationships
with race/SES,
but no SSD.

Decision for Use/
Application to
Practice
Low risk
intervention.
Hospital currently
has a CAT, but no
SP. Management is
supportive of
implementing SP
that will attenuate
bias, ↑ screening
consistency, and ↓
unnecessary testing/
referrals.

DV3:
PSTG > PREG
for UA, TA, &
SS (p < 0.001)
UA = 56.6% vs.
13.6%)
TA = 62.0% vs.
10.9%
SS = 84.6% vs.
60.4%
OC = PSTG <
PREG (20.8%
vs. 40.4%)
p < 0.001
A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

42
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

Data Analysis

Findings/
Themes

Decision for Use/
Application to
Practice

CAT referral
PSTG < PREG
(32.6% vs.
44.6%), p=0.05

Jordan, K. S., &
Moore-Nadler, M.
(2014).

Cognitive
learning
theory

Design: RCR/
PCS; pre- and
post-test design

N = 31 nurses
(completed CA
education program)

Children at risk of
maltreatment:
Identification and
intervention in the
emergency
department.

combined
with

Purpose:
Implement a
series of
strategies in a
PED to identify
children at risk
of CA and
ensure their
safety and
protection

N = 2309 children
presented to PED

Funding: Not
disclosed
Conflicts/Bias:
Authors deny any
conflicts/bias

Theory of
reasoned
action

IV1:
Comprehensive
CA nursing
education
program
IV2: CA SI

Children who
should have been
screened for CA
according to RCR:
n = 42

DV1:
Knowledge and
skill set of PED
nursing staff

Children actually
screened for CA:
n = 37 (88)

DV2: CA SR
for PED
patients with

Likert scale: 3part pretest data
collection tool on
nurses’ CA
confidence,
attitudes, selfefficacy, and
clinical practice
behaviors
Validated risk of
CA was assessed
by reviewing and
matching CA SI
data to nurse’s
documentation in
patient’s EMR

Only
percentiles
reported; no
other statistical
tests/data
analysis noted

After SP
implemented,
more injuries
adjudicated as
AI (62% vs
74%) than NAT
(38% vs. 26%);
p = 0.03
DV1:
Pretest: 76% of
questions
answered
correctly
Posttest: 90%
of questions
answered
correctly
Questions most
missed: age
when child
most vulnerable
to CA; injuries
specific to CA

LOE: IV
Strengths:
All nurses
completed the
educational
program; 94% of
nurses had > 6 years
experience; pre-/
post-test design
allows evaluation of
knowledge/practice
changes
Limitations:
Small convenience
sample limited to

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Country: U.S.A.

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

Demographics:
Nurses:
>6 years nursing
experience: n = 29
(94);
Prior CA training:
n = 16 (52)
AR = 0
Setting: PED
Inclusion criteria:
All nurses working
in PED; all
children presenting
to PED over a 1month period with
high-risk diagnoses
for CA were to be
screened with SI
High-risk
diagnoses for CA
classified as SF in
child <36 months;
head injury in child
<12 months; burnrelated injuries;
alleged sexual or

43
Major Themes,
Variables, &
Definitions
high-risk
diagnoses

Measurement/
Instrumentation
Patient census
log reviewed and
compared with
completed SI or
EMR daily to
assess for
documentation
of CA risk

Data Analysis

Findings/
Themes
100% of nurses
rated CA skill
set; knowledge
level; and
confidence in
ability to
identify &
document on
CA higher after
educational
program
DV2:
37 out of 42
high-risk
children
screened; IR
88%
35 of the 37
children
(94.6%) had
evidence of CA
(1.5% of total
PED
population)

Decision for Use/
Application to
Practice
only 1 PED and 5
preselected high-risk
CA diagnoses – ↑
risk of sample bias;
tools designed by
study coordinator –
no V&R data; study
coordinator solely
conducted the RCR
– limits IR
Conclusions:
Nurses need ↑ CA
education to ↑ SE,
knowledge, and skill
set; value in ↑ CA
education for nurses,
implementing a CA
SP in ED, and
working with a
collaborative and
cooperative
approach to achieve
best practice for CA
screening and
management
Feasibility:
Recommend CA SP
and staff education.

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

44
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

IV: Audiotaped SSIs with
HPs focused on
CA detection in
EDs and related
training and
policy

Tape recorded
transcribed
interviews

Data Analysis

Findings/
Themes

Qualitative
content analysis

DV1:
Facilitators to
CA detection:
Support of
hospital board;
presence of CA
attendant;
presence of
CAT; intensive
training of ED
staff; financial
support

physical abuse; and
factitious disorder
by proxy

Louwers, E.C., et
al. (2012a).
Facilitators and
barriers to
screening for child
abuse in the
emergency
department.
Funding: The
Netherlands
Institute for Health
Research and
Development
Conflicts/Bias:
Authors deny any
conflicts/bias

Child Abuse
Framework
of the Dutch
Health Care
Inspectorate

Design: 2phased QS;
audio-recorded,
SSIs
Purpose:
Phase 1: Define
facilitators/
barriers to CA
screening
Phase 2: Make
suggestions on
methods to
optimize ED
CA screenings

Exclusion criteria:
Children not
meeting the
inclusion criteria;
no other exclusion
criteria noted
N = 33
n = 27 (phase 1)
n = 6 (phase 2)
Demographics:
Phase 1 interviews:
n = 7 pediatricians
n = 2 surgeons
n = 6 ED nurses
n = 6 ED managers
n = 6 hospital
board members
n=6
Phase 2 interviews:
n = 5 CA experts
n = 1 expert in
implementation
AR: 0

DV1:
Facilitators to
CA detection
DV2: Barriers
to CA detection
DV3: Methods
to overcome
barriers to CA
screening

DV2:
Barriers to CA
detection:
Lack of CA
knowledge;
practical
problems (i.e.,

Decision for Use/
Application to
Practice
Low risk
intervention.
Possible cost to train
staff, but costs
should ↓ once all
current ED staff
trained.

LOE: VI
Strengths:
HPs had mixed
disciplines and
backgrounds; used
national CA
framework to order
study results; most
hospital Boards
were supportive of
CA screening; most
hospitals had CA
attendant, protocol
for suspected CA,
and/or a suitable SI;
all Dutch hospitals
that see children
were represented in
N

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Country: The
Netherlands

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

Setting: 7 EDs
who treat children
Inclusion criteria:
HP must be
associated with at
least 1 of the 7
participating
hospitals
Exclusion criteria:
None listed

45
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

Data Analysis

Findings/
Themes
lack of suitable
location/limited
time); personal
barriers (i.e.,
fear of
unjustified
suspicion);
insufficient
communication
skills (talking
with family);
lack of support
from hospital
board; fast ED
staff turnover

Decision for Use/
Application to
Practice
Limitations:
Low LOE; small N
and disciplinespecific n’s – risk of
bias; HP interviews
conducted before
the Dutch Health
Care Inspectorate
published its official
CA framework, so
registry and
information section
of framework was
not addressed in this
study

DV3:
Methods to
overcome
barriers to CA
screening:
↑ training of
staff to better
recognize CA;
improving
communication
skills; coding
injuries as CA
to receive
funding; full SP

Conclusions:
For a successful SP,
providers must be
properly educated
on CA detection and
effective
interviewing
techniques; CA
attendants and
CATs must be
available to ED
staff; and hospital
management must
support the required

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Louwers, E.C., et
al. (2014).
Accuracy of a
screening
instrument to
identify potential
child abuse in
emergency
departments.
Funding: The
Netherlands
Institute for Health

Conceptual/
Theoretical
Framework

Cochrane
protocol for
systematic
review of
interventions
combined
with
Standards for
the Reporting
of Diagnostic
Accuracy

Design/
Method

Sample/Setting

Design: PCS

N = 38,136

Purpose:
1. Measure the
accuracy of a
newly
developed SI
for CA in EDs
(Escape tool)

IG: n = 18,275
(48); M = 10,322
(56)
CG: n = 19,861
(52); M = 11,389
(57)

2. Examine the
possibility to
minimize
burden of

Demographics:
MA = 5.5 years
0-4 years:
IG = 10,035 (55)
CG = 9759 (49)

46
Major Themes,
Variables, &
Definitions

IV: Escape tool
(6-item SI;
yes/no answer
options)
DV: Actual risk
of/potential for
CA
CA definition:
“Any form of
threatening or
violent
physical, mental

Measurement/
Instrumentation

CA risk
measured by
Escape tool
 SI developed
by team of
pediatricians
and screening
experts (study
authors)
 ≥ 1 aberrant
answer =
positive screen

Data Analysis

χ2 to compare
categorical
variables
LRs to validate
Escape tool’s
Sn, Sp, PPV
and NPV
UVR and MVR
to determine PV
of each SII
95% CI

Findings/
Themes
support by
management;
implementing
CATs to ensure
good interdisciplinary
collaboration;
appointing a
CA attendant;
and
implementing
an SP and
validated SI to
ensure uniform
CA screening
process
IV:
+ SI screen
n = 420 (2.3)
- SI screen
n = 17,855
DV:
CA DR 5 times
↑ in IG than CG
Referred to
CAT: n = 89

Decision for Use/
Application to
Practice
training and SP
implementation.
Feasibility:
Recommend CA SP
for ED. Low risk.
Management
supportive of CA SP
development/staff
training. CAT team
already in place.
Possible cost to train
staff, but cost should
↓ once all current
ED staff trained.
LOE: IV
Strengths:
Good LOE; multicenter setting; large
IG; all pediatric age
groups included;
several potential CA
cases identified; SI
detected most
children at risk for
CA; extremely high
correlations between
the SIIs.

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Research and
Development and
Foundation
Physico
Therapeutic
Institute
Conflicts/Bias:
None noted
Country: The
Netherlands

Conceptual/
Theoretical
Framework
Studies
(STARD)

Design/
Method
completing the
tool while
maintaining Sn
and Sp.

Sample/Setting

5-8 years:
IG = 3537 (19)
CG = 3457 (17)
9-12 years:
IG = 2614 (14)
CG = 2908 (15)
13-18 years:
IG = 2089 (12)
CG = 3737 (19)
AR: 0
Setting: 3 EDs
Eligibility: All
children aged ≤ 18
years, regardless of
ED visit reason

47
Major Themes,
Variables, &
Definitions
or sexual
interaction with
a minor which
is perpetrated
actively or
passively by
parents or other
persons on
whom the
minor is
dependent and
causes or will
probably cause
physical or
mental injury
and serious
harm to the
minor” (p.
1277).

Measurement/
Instrumentation
SI data
independently
evaluated by EP
of 4 physicians
with extensive
experience in CA

Data Analysis

p < 0.05
SPSS 17.0 used
for analysis

Findings/
Themes
Potential CA
case: n = 55
(56); + screen:
n = 44 (80);
- screen: n = 11
SI ≥ 1 item +
OR, 189.8 (97.3
-370.4); p <
0.001; Sn 0.80;
Sp 0.98; PLR
40; NLR 0.20;
PPV 0.10; NPV
0.99

Cases included for
analysis if ≥2 CAT
professionals
classified case as a
“potential case”

Individual SII
analysis
Sp and NPV of
each SII = 0.99;
p < 0.001
SII 1 – OR, 50
(23.6-106.2);
Sn 0.71; PPV
0.11

“Potential case”:
≥ 1 inclusion
criteria checked by
CAT professional

SII 2 – OR,
17.4 (7.3-41.3);
Sn 0.12; PPV
0.04

Decision for Use/
Application to
Practice
Limitations:
Rate of confirmed
CA unavailable;
using data of
potential CA cases
could result in
overestimation of
rate of true cases of
CA.
Conclusions:
Escape tool may
miss some cases of
CA, but CA not
likely when SI result
is negative.
Excluding some SIIs
from the SI is not
recommended due
to decreased Sn/Sp.
Combining the
Escape SI with
training of ED staff
on CA risk factors is
an effective way to
↑ CA SR and DR.
Feasibility:
This is a brief, 6item SI that can

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting
“No case”:
≥ 1 exclusion
criteria checked by
CAT professional
(even if ≥ 1
inclusion criteria
also checked)
Inclusion criteria:
Injury caused by
person on whom
child is dependent;
from caregiver
neglect; withheld
from medical care;
psychological harm
from actions or
failure of person on
whom child is
dependent; child
witnessed domestic
violence; child
witnessed or was
victim of sexual act
Exclusion criteria:
Suspicion of CA
known prior to ED
visit; alcohol
intoxication;

48
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

Data Analysis

Findings/
Themes
SII 3 – OR,
137.0 (72.7258.5); Sn 0.34;
PPV 0.21
SII 4 – OR,
65.3 (32.3131.9); Sn 0.21;
PPV 0.13
SII 5 – OR,
82.1 (37.9178.2); Sn 0.17;
PPV 0.17

Decision for Use/
Application to
Practice
easily and quickly
be administered by
ED nurses or
providers to screen
children for CA.
Low risk
intervention. No
cost to use SI.
↓ staff burden if
filled out during
triage.

SII 6 – OR,
182.9 (102.3327.4); Sn 0.59;
PPV 0.18
↓ SI burden
analysis:
Exclude SII 2 –
Sn ↓ to 0.73
Exclude SII 1,
4, or 5 – Sn ↓ to
0.78
Exclude SII 1, 4
and 5 – Sn ↓
0.75; Sp 0.98

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Louwers, E.C., et
al. (2012b).
Effects of
systematic
screening and
detection of child
abuse in
emergency
departments.
Funding: The
Netherlands
Institute for Health
Research and
Development and
Foundation
Physico
Therapeutic
Institute
Conflicts/Bias:
Authors deny any
conflicts/bias
Country: The
Netherlands

Conceptual/
Theoretical
Framework

Cochrane
protocol for
systematic
review of
interventions
combined
with
Cognitive
learning
theory

Design/
Method

Design:
Interventional
PCS
Purpose:
Determine if
introducing a SI
and training of
ED nurses
increases DR of
CA

Sample/Setting

suicide attempt;
injury caused by a
stranger or peers
N = 104,028; MA
= 7.2 years; M =
58,445 (56)
n = 37,404 (36)
(screened for CA)
Demographics:
Suspected CA:
n = 243 (0.2)
M = 123 (51)
Age =
0-4 years: 150 (62)
5-8 years: 37 (15)
9-12 years: 25 (10)
13-18 years:31 (13)
Not CA:
n = 103,785 (99.8)
M = 58,322 (56)
Age =
0-4 years: 41,942
(40)
5-8 years: 17,865
(17)
9-12 years: 17,220
(17)

49
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

IV1: “Escape
Form” (6-item
SI; yes/no
answer options)

Data collection:
EMRs and ED
triage systems

χ2 to compare
categorical
variables

CA risk
measured by
Escape Form
 SI developed
by the authors
of this study
 ≥ 1 aberrant
answer =
positive screen

Interrupted
time-series
analyses to
calculate effect
of interventions
on CA SR

IV2: Training
(interactive
workshop in
interview
techniques) for
ED nurses
DV1: CA SR
DV2: CA DR
CA definition:
“Any form of
threatening or
violent
physical, mental
or sexual
interaction with
a minor which
is perpetrated
actively or
passively by

Suspected CA
cases were
independently
scored by 4 HPs
with variables of
age, gender,
signs at ED
presentation,
history and
findings at ED,
conclusion of the
SI, and physician
diagnosis

Data Analysis

Pooled ORs for
detection of CA
in screened
children

Findings/
Themes

DV1:
Overall CA SR
↑ from 20% in
February 2008
to 67% in
December 2009
SR in the 5
intervention
hospitals ↑
twice as much
(14% to 69%)
as in the 2
control
hospitals (35%
to 63%)

p < 0.05
SPSS 17.0 and
R 2.7.1 were
used for the
analysis

DV2:
DR 5 times
higher in
children
screened than
not screened for
CA (0.5% vs.
0.1%, p <
0.001)

Decision for Use/
Application to
Practice

LOE: III
Strengths:
Controlled; good
LOE; large N and
screened n;
inclusion of all
pediatric age
groups; universal
implementation of
Escape form; long
study period; 6month baseline
monitoring period
clearly shows SR
and DR changes
after interventions
implemented; results
represent variety of
ED settings – ↑
generalizability of
findings
Limitations:
Hospitals were not
randomized to the
intervention and

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

13-18 years:
26,748 (26)
Setting: 7 EDs in
the province of
South Holland –
5 intervention
hospitals; 2 control
hospitals
Eligibility:
All children aged 0
to 18 years who
visited the ED from
February 2008 –
December 2009
Cases included for
analysis if at least 2
CAT professionals
confirmed case as a
“potential case”
“Potential case”:
≥ 1 inclusion
criteria checked by
CAT professional
“No case”:
≥ 1 exclusion
criteria checked by

50
Major Themes,
Variables, &
Definitions
parents or other
persons on
whom the
minor is
dependent and
causes or will
probably cause
physical or
mental injury
and serious
harm to the
minor” (p. 459).

Measurement/
Instrumentation

Data Analysis

Findings/
Themes
Suspected CA
cases were
significantly
younger than
total PED
population (4.7
vs. 7.2 years; p
< 0.001) and
were more often
screened by ED
staff than
children in the
total PED
population
(75% vs. 36%,
p < 0.001)
Pooled ORs for
DR in children
screened at all 7
ED sites:
4.88 (95% CI,
3.58-6.68)

Decision for Use/
Application to
Practice
control arms
(authors note
impossibility due to
logistical and ethical
reasons); possible
over-estimation of
“actual” CA cases
since only potential
CA cases were
presented; number
of false-negative CA
cases not known;
possibility of
inconsistent SI use;
Dutch Health Care
Inspectorate
mandated CA
screening in all EDs
during middle of
this study, which
could partially
explain ↑ in SR
Conclusions:
Systematic CA
screening in EDs is
effective in ↑ the
DR of CA. Training
ED staff and
mandating CA

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

CAT professional
(even if ≥ 1
inclusion criteria
also checked)
Inclusion criteria:
Injury caused by
person on whom
child is dependent;
from caregiver
neglect; withheld
from medical care;
psychological harm
from actions or
failure of person on
whom child is
dependent; child
witnessed domestic
violence; child
witnessed or was
victim of sexual act
Exclusion criteria:
Suspicion of CA
was known prior to
ED visit; alcohol
intoxication;
suicide attempt;
injury caused by a
stranger or peers

51
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

Data Analysis

Findings/
Themes

Decision for Use/
Application to
Practice
screening in EDs
significantly ↑ the
extent of CA
screening.
Feasibility:
Recommend
implementing a
systematic SP for
CA screening in the
ED that includes
both an SI and an
educational module
for ED staff. The
Escape tool is a
feasible SI for this
purpose. No cost to
use tool. Low risk
intervention.
Possible cost to train
staff, but costs
should ↓ once all
current ED staff
trained.

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Sittig, J.S., et al.
(2016).
Value of
systematic
detection of
physical abuse at
emergency rooms:
a cross-sectional
diagnostic accuracy
study.
Funding: The
Netherlands
Institution for
Health Research
and Development
Conflicts/Bias:
Authors deny any
conflicts/bias
Country: The
Netherlands

Conceptual/
Theoretical
Framework
Quality
Assessment
of Diagnostic
Accuracy
Studies
(QUADAS)
combined
with
Standards for
the Reporting
of Diagnostic
Accuracy
Studies
(STARD)

Design/
Method

Sample/Setting

Design: Cross
Sectional DAS
with a 6-month
follow-up

N = 4290 (screened
with SI); Age 0-7
years

Purpose:
Establish
whether the
SPUTOVAMOR checklist
accurately
detects or
excludes CA
among children
presenting to
the ED with
physical injury.

n = 720 (RSP and
EP Assessment)
Positive screen: n =
112 (2.6)
15% random
sample of negative
screens: n = 645
Demographics:
M = 417 (58)
North European =
498 (69)
Low SES = 170
(24)
Age < 1 year = 43
(6)
MA in years (SD)
AI: 3.8 (1.8)
II: 1.03 (0.8)
Neglect: 2.1 (1.3)
AR: 0.9% (37
parents refused to
participate in RSP)

52
Major Themes,
Variables, &
Definitions
IV:
SPUTAVAMOR checklist (6
questions with
yes/no answer
options)
DV1: Injury
due to CA
(defined as “use
of physical
force or
implements
against the child
that has resulted
in physical
injury” [p. 2]).
DV2: Injury
due to neglect
(defined as
“failure to meet
a child’s basic
physical needs
or failure to
ensure a child’s
safety” [p. 2]).
DV3: Need for
help from social

Measurement/
Instrumentation

Data Analysis

Findings/
Themes

SPUTOVAMOR checklist (SI)
 Revised
version of
original 9-item
checklist
 ≥ 1 aberrant
answer =
positive screen

PVs, Sn, and Sp
with 95% exact
CIs to establish
diagnostic
accuracy of SI

DV1:
Based on RSP,
II in 3 children
– prevalence
0.07% (95% CI,
0.01-0.2)

RSP: The
SPUTOVAMOR checklist (SI)
was tested
against the
majority opinion
of a 3-member
CA EP
 ED visit
medical file
 Detailed injury
history by CA
pediatrician
 Risk factors
derived from
questionnaires
(HPs, CPS)
 6 months
follow-up
information

RR for being
reported to CPS
with + SI result
as compared to
– SI result was
calculated with
approximate
95% CI
IR between EP
members
assessed by 2way ICC:
< 0.20 = poor
0.21-0.40 = fair
0.41-0.60 =
moderate
0.61-0.80 =
good
0.81-1.00 =
very good
agreement

PPV 0.03; NPV
1.00; Sn 1.000;
Sp 0.865; falsepositive rate =
0.97 (95% CI,
0.915-0.994);
false-negative
rate = 0.0 (95%
CI, 0.0-0.006)
DV2:
Based on RSP,
injury caused
by neglect in 6
children –
prevalence of
0.27% (95% CI,
0.15-0.49)
PPV 0.050;
NPV 0.998; Sn
0.833; Sp 0.867

Decision for Use/
Application to
Practice
LOE: IV
Strengths:
Large N; met all
QUADAS criteria
for diagnostic
accuracy studies;
good agreement (IR)
between EP
members; all truenegative cases
correctly identified;
panel members
blind to SI results; ↓
incorporation bias
Limitations:
Possible
misclassification of
outcome as reason
for low CA
prevalence
compared to prior
studies; risk of
implicit bias; unable
to unequivocally
diagnose injury due
to neglect and need
for help from social
services

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

Setting: 4 EDs in
region of Utrecht
Inclusion criteria:
All children aged
0-7 years admitted
to ED between
June 2009 and
December 2010 for
any physical injury
Exclusion criteria:
Obvious victims of
CA (admitted by
perpetrator at
presentation);
victims of
witnessed MVCs;
children who died
before ED arrival

53
Major Themes,
Variables, &
Definitions
services
(defined as “any
concern about
the situation of
the child that
requires
consultation of
social services”
[p. 2]).

Measurement/
Instrumentation

Data Analysis

Findings/
Themes

PASW statistics
V.20.0 and
STATA/SE
V.11.0 were
used for the
analysis

DV3:
Based on RSP,
help from social
services was
needed in 102
children –
prevalence of
11.6% (95% CI,
10.6-12.6)
PPV 0.330;
NPV 0.889; Sn
0.326; Sp 0.892
CPS reports for
physical CA =
1.6%: + SI =
70; - SI = 63
(RR 4.51; 95%
CI, 2.14-9.95)
CPS reports for
general CA =
4.8%: + SI =
15; - SI = 188
(RR 3.31; 95%
CI, 2.03-5.39)
Panel IR for II
= 0.82 (95% CI,

Decision for Use/
Application to
Practice
Conclusions:
SPUTOVAMO-R
has a very high
false-positive rate
for physical CA and
did not reliably
diagnose injury due
to neglect and need
for help. Results in
↑ workload for CA
pediatricians and
CPS workers.
Should consider use
of this SI carefully
before de novo
implementation.
Feasibility:
Would NOT
recommend use of
SPUTOVAMO-R as
CA SI in EDs due to
high false-positive
rate. ↑ risk of
unsubstantiated CPS
reports could result
in ↑ costs as well as
negatively impact
children and their
families.

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Smeekens, A.E., et
al. (2011).
Successful elearning
programme on the
detection of child
abuse in
emergency
departments: A
randomised
controlled trial.
Funding: Augeo
Foundation
Conflicts/Bias:
Authors deny any
conflicts/bias
Country: The
Netherlands

Conceptual/
Theoretical
Framework

Bandura’s
social
learning
theory
combined
with
Cognitive
learning
theory

Design/
Method

Sample/Setting

Design:
Blinded RCT;
pre- and postintervention
design

N = 38 (included)
n = 25 (analyzed)

Purpose:
Evaluate the
effectiveness of
an electronic
learning (elearning)
program on
nurses’
recognition of
CA in a
simulated ED
case

Demographics:
F = 18 (72)
MA (years) = 42
(SD 9)
Mean work
experience (years)
= 9 (SD 7)

IG: n = 13
CG: n = 12

AR: 36% (due to
participants’ work/
shift patterns)
Setting: University
Medical Center ED
Inclusion criteria:
All nurses with

54
Major Themes,
Variables, &
Definitions

IV: 2-hour elearning
program
focused on CA
recognition
DV1:
Performance
during case
simulation to
detect CA
DV2: Selfreported SE for
the detection of
CA

Measurement/
Instrumentation

Simulation
performance:
assessed by EP
using a
standardized
assessment form
– scored quantity
and quality of
questions posed
by the nurse
(minimum score
0; maximum
score 114)
SE in detection
of CA: measured
using visual
analogue scale:
consisting of 8
statements (total
score minimum 0
mm; maximum
800 mm)

Data Analysis

Student t-test
used for
apparently
normally
distributed data
Mann-Whitney
U to assess
group
differences
IR assessed
using ICC to
determine
consistency
among EP
members
p ≤ 0.05
SPSS 17.0 used
for analysis

Findings/
Themes
0.80-0.84);
neglect = 0.07
(95% CI, 0.020.11); social
service help =
0.40 (95% CI,
0.35-0.44)
No SSD in
characteristics
of IG and CG
DV1:
Total
performance of
IG significantly
better than CG
during post-test
(89 vs. 71; 95%
CI 2.9-33.3; p <
0.022; ES = 1)
+ correlation
between length
of simulation
test and # of
adequate ?’s
(Spearman
correlation
0.586, p =
0.002)

Decision for Use/
Application to
Practice

LOE: II
Strengths:
High LOE; large ES
for simulation
performance;
medium ES for SE;
EP blind to IG/CG
allocation; objective
measurement to
determine effect of
educational program
on CA detection; no
baseline SSD when
comparing nurses
lost to follow-up
with those analyzed;
simulations based
on real CA cases
Limitations:
Small N (↑ risk of
bias); high AR

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

permanent contract
in the ED during
the study period
Exclusion criteria:
None listed

55
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

Data Analysis

Findings/
Themes
Separate IG/CG
post-/pretest
performance
comparison:
IG: 89 vs. 74
(p = 0.053)
CG: 71 vs. 69
(p = 0.728)
IR of EP: 0.70
(95% CI, 0.510.84, p = 0.000)
DV2:
Mean post-test
SE: IG = 502;
CG = 447 (95%
CI -25.4 to
134.7, p =
0.171; ES =
0.56)
+ correlation
between total
SE score and
performance on
simulation test
(Spearman
correlation
0.387, p=0.056)

Decision for Use/
Application to
Practice
(36%) might ↓
generalizability of
findings; used
simulation rather
than observation of
clinical practice;
time for CA
assessment may be
restrictive factor in
ED, whereas it was
not in simulated
cases
Conclusions:
E-learning improved
the performance in
case simulations and
the SE of nurses in
detection of CA in
the ED.
Feasibility:
2-hour e-learning
CA module feasible
for ED practice.
Small time
commitment for
providers; training
can be done at work
or home. Cost of

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Teeuw, A.H., et al.
(2016).
Results of the
implementation of
a new screening
protocol for child
maltreatment at the
emergency
department of the
academic medical
center in
Amsterdam.
Funding: Stichting
Kinderpostzegels
Nederland
Conflicts/Bias:
Authors deny any
conflicts/bias

Conceptual/
Theoretical
Framework

Protocol for
the evaluation
of suspected
NAT
combined
with
Cognitive
learning
theory

Design/
Method

Sample/Setting

Design: PCS

N = 1089

Purpose:
Evaluate the
implementation
of a new CA SP
directly after
introduction
(February 2010)
and 9 months
later
(November
2010)

February cohort:
n = 560
November cohort:
n = 529
Demographics:
February cohort:
F = 265 (47.3)
M = 295 (52.7)
MA = 7.95; median
age 6.4
November cohort:
F = 212 (40.1)
M = 317 (59.9)
MA = 7.95; median
age 7.5

56
Major Themes,
Variables, &
Definitions

IV: SP –
SPUTOVAMO
9-item checklist
plus (TTI)
DV1:
Completion of
SPUTOVAMO
DV2:
Completion of
TTI
DV3: Possible
factors
influencing
completion of
SPUTOVAMO
and/or TTI

Measurement/
Instrumentation

DVs were
measured by
comparison of
data collected
directly after
introduction of
the TTI to the SP
(February 2010)
and 9 months
later (November
2010)

Data Analysis

Median and
IQR to describe
non-normally
distributed data
Pearson Chi
square used to
calculate
statistical
difference in
performing
SPUTOVAMO
and TTI
between
February and
November
Factors
influencing
chances for
performing
SPUTOVAMO

Findings/
Themes
Significant ↑ in
SE score in IG
and CG with
post-/pretest
comparison:
IG: 502 vs. 402
(p = 0.000)
CG: 447 vs. 364
(p = 0.045)
DV1:
SPUTOVAMO
completed in
416 of 560
February cases
(74%); 223 of
529 November
cases (41.2%);
Completion 1.8
times ↓ in
November than
February (95%
CI 1.6-2.0; p =
0.00)
ULR: older
children less
likely to have
SPUTOVAMO.
February: odds
↓ 7.4% for

Decision for Use/
Application to
Practice
implementation
unknown, but
benefits of ↑ in CA
detection and SE
likely outweigh any
cost. No associated
risks.

LOE: IV
Strengths:
Large overall N and
cohort n’s suitable
for analysis; SP was
implemented into
daily routine; all ED
HPs attended
training session
prior to addition of
TTI to SP
Limitations:
Only able to analyze
documented
SPUTOVAMO and
TTI reports; ED
changed to EMR
during study,
resulting in ↓

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Country: The
Netherlands

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

AR: Per authors,
estimated to be
15% due to staff
turnover. However,
all new personnel
received short
introductory course
to the SP process
Setting: ED of a
large teaching
hospital
Inclusion criteria:
All patients
between 0-8 years
who presented to
the ED
Exclusion criteria:
None noted

57
Major Themes,
Variables, &
Definitions
SPUTOVAMO
Checklist
Definition: 9
risk factors to
detect CA; to be
completed by
ED triage nurse
TTI
Definition: Full
physical
examination
(top-toe
inspection) of
the undressed
patient; to be
completed by
pediatrician, ED
physician, or
ED nurse

Measurement/
Instrumentation

Data Analysis

and TTI were
assessed with
Pearson Chi
square in case
of 2 variables
and with ULR
in case of > 2
variables or
numeric
variables
p < 0.05
PASW statistics
18.0 was used
for statistical
analysis

Findings/
Themes
every year ↑ in
age (p = 0.00);
November:
odds ↓ 5.2% for
every year ↑ in
age (p = 0.00)
DV2:
TTI completed
in 217 of 477
February cases
(45.5%); 89 of
433 November
cases (20.6%)
RR for having
TTI 2.4 times ↓
in November
than February
(95% CI 1.92.9; p = 0.00)
UVR: older
children less
likely to have a
TTI completed.
February: odds
↓ 4.9% for
every year ↑ in
age (p = 0.001);

Decision for Use/
Application to
Practice
visibility of the SP;
assumptions for
Pearson Chi square
test may not be
totally true in ED
(patients differ, ED
staff change);
reasons for not
performing TTI
were seldom
documented and
could not be
analyzed; some
patients referred to
general practitioner,
so SP not completed
Conclusions:
Implementation of
the new CA SP,
where the TTI was
added to the
SPUTOVAMO, was
only mildly
successful and ↓
with time. SP
completion was
highest right after
implementation,
which was

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

58
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

Data Analysis

Findings/
Themes
November:
odds ↓ 5.6% for
every year ↑ in
age (p = 0.005)
DV3:
Completion of
SPUTOVAMO:
Diagnosis “skin
& subcutaneous
tissue”: OR 1.9
(p = 0.003)
Diagnosis
“gastrointestinal
infection”: OR
0.08 (p = 0.001)
Known chronic
illness (31.5%)
versus no
chronic illness
(46.6%) (RR
0.68; 95% CI
0.52-0.88; p =
0.00)
SPUTOVAMO
completion ↑
chance of TTI
completion; RR

Decision for Use/
Application to
Practice
immediately
preceded by
extensive training
sessions. Negative
correlation between
↑ child age and
having a chronic
illness and
completion of the
SP.
Feasibility:
If new SP for CA is
implemented,
regular training
sessions should be
held to ↑ adherence
and sustain HP
motivation to
complete the SP.
Most practical for
ED nurses to
complete the
SPUTOVAMO
during triage and the
ED physicians to
perform the TTI.

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

59
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

Data Analysis

Findings/
Themes

Decision for Use/
Application to
Practice

5.4 (95% CI 3.8
-7.5; p = 0.000)
Completion of
TTI:
No correlation
between child’s
diagnosis and
TTI completion

Tiyyagura, G., et
al. (2015).
Barriers and
facilitators to
detecting child
abuse and neglect
in general
emergency
departments.
Funding: Specific
funding source not
stated

Grounded
theory

Design: QS;
one-on-one
SSIs; consistent
with grounded
theory

N = 29
Physicians: n = 9
Nurses: n = 16
Physician
Assistants: n = 4

Purpose:
1. Explore
general ED
providers’
experiences
with screening
and reporting of
CAN

Recruited
participants
through purposeful
sampling and
snowball sampling

IV: Face-toface, audiotaped, SSI with
open-ended ?’s
and prompts to
encourage
detailed
discussion
DV1: Barriers
to recognizing
CAN

AR: 0
Setting: 3 general
EDs in different

DV2: Barriers
to reporting
CAN

Tape recorded
transcribed
interviews

Qualitative
content analysis
Constant
comparative
method of data
collection and
analysis
continued past
the point of
saturation

Known chronic
illness: RR 0.61
(95% CI 0.381.00; p = 0.05)
DV1:
Barriers to
recognizing
CAN:
Desire to
believe
caregiver;
failure to
recognize signs/
symptoms of
CAN; personal
biases about
caregivers;
challenges
innate to
working in ED

LOE: VI
Strengths:
Sample included 3
sites with different
pediatric models of
care; providers had
variety of
experience and ED
roles; data
composed of rich,
candid narratives
Limitations:
Low LOE; small N
from only 3 sites

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conflicts/Bias:
Authors deny any
conflicts/bias
Country: U.S.A.

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

2. Identify
barriers and
facilitators to
recognizing and
reporting CAN
in ED setting

regions of
Connecticut and
with different
models of pediatric
care; none had fulltime, on-site social
work or CA expert
support
Inclusion criteria:
Recently evaluated
and treated patients
with CAN; various
experience levels
and roles in ED
setting; willing to
discuss CAN
experiences with
research team
Exclusion criteria:
None noted

60
Major Themes,
Variables, &
Definitions
DV3:
Facilitators to
CAN detection
DV4: Provider
preferences
regarding CAN
education and
support

Measurement/
Instrumentation

Data Analysis

Findings/
Themes
DV2:
Barriers to
reporting CAN:
the reporting
process (time
and effort);
negative
consequences
of reporting
DV3:
Facilitators to
CAN detection:
immediate case
discussion with
peers; belief
that it was
better to report
with any
suspicion; clear
understanding
of role of CPS
and mandated
reporter law
DV4:
Provider
requests for
CAN education
and support:

Decision for Use/
Application to
Practice
within 1 state may
limit generalizability
of results; did not
include direct
observations of
provider’s clinical
practices; focus
groups may have
stimulated greater
discussion among
providers than SSIs
Conclusions:
Detection of CAN
by general ED
providers can be
improved by
providing education
through case review,
improving follow-up
by CPS agencies,
and increasing realtime assistance with
patient care
decisions.
Feasibility:
Institution of a
CAT, CA SI,
educational case-

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Citation

Conceptual/
Theoretical
Framework

Design/
Method

Sample/Setting

61
Major Themes,
Variables, &
Definitions

Measurement/
Instrumentation

Data Analysis

Findings/
Themes
immediate
consultation for
unclear cases of
CAN; feedback
about case
disposition;
case-based
education;
standardized
process of CPS
reporting

Decision for Use/
Application to
Practice
based interventions
for providers, and a
CPS liaison are
feasible, low-risk
ways to improve CA
detection and
reporting in the ED.

A – Asian race, AI – accidental injury, AR – attrition rate, B – Black race, C = compliant, CA – child abuse, CAN – child abuse and neglect, CAT – child abuse team, CG – control group, CI –
confidence interval, CPS – child protective services, DAS – diagnostic accuracy study, DR – detection rate, DV – dependent variable, ED – emergency department, EMR – electronic medical record,
EP – expert panel, ES – effect size, F – female, GSNI – government subsidized/ no insurance, H – Hispanic race, HP – health professionals, ICC - intraclass correlation coefficient, IG – intervention
group, II – inflicted injury, IQR – interquartile range, IR – interrater reliability, IV – independent variable, LOE – level of evidence, LR – likelihood ratio, M – male, MA – mean age, MVC – motor
vehicle collision, MVR – multivariate regression, N – number of participants (study), n – number of participants (group), NAT – non-accidental trauma, NC – non-compliant, NLR – negative
likelihood ratio, NPV – negative predictive value, O – Other race, OC – ophthalmology consult, OR – odds ratio, p – statistical significance, PCS – prospective cohort study, PED – pediatric
emergency department, PI – private insurance, PLR – positive likelihood ratio, PPV – positive predictive value, PREG – preguideline group, PSTG – postguideline group, PV – predictive value,
QES – quasi-experimental study, QS – qualitative study, R – reliability, RCR – retrospective chart review, RCT – randomized controlled trial, RR – risk ratio, RSP – reference standard procedure,
SD – standard deviation, SE – self-efficacy, SES – socioeconomic status, SF – skeletal fracture, SI – screening instrument, SII – screening instrument item, Sn – sensitivity, Sp – specificity, SP –
screening protocol/guideline, SPUTOVAMO – 9-item screening checklist, SPUTOVAMO-R – revised 6-item checklist, SR – screening rate, SS – skeletal survey, SSD – statistically significant
difference, SSI – semi-structured interview, TA – transaminase levels, TTI – “top-toe” inspection, UA – urinalysis, UVR – univariate regression, V – validity, W – White race, χ2 – Chi-squared

CHILD PHYSICAL ABUSE SCREENING PROGRAM

62
Appendix C
Synthesis Table

Level of Evidence
Design
Pre/post design
Study Demographics
Setting
GED
PED
Number of sites
Country
Iran
Netherlands
USA
Sample
Total participants
Attrition rate
Population
Children
HPs
Mean Age
Months (mos) / Years (yrs)
Gender
% Male
Measurement/Instrumentation
CA SI
Physical exam/TTI
CAT and/or EP evaluation
CA simulation performance
Hospital databases/MCR
Likert scale – CA SE/detection
Visual analogue scale – CA SE
Tape-recorded, transcribed SSIs

Dinpanah
(2017)
IV
PDAS

Higginbotham
(2014)
III
QES
x

Jordan
(2014)
IV
PCS/RCR
x

x
1

x
1

x

x

6120
0%

332
0%

31
0%

33
0%

x

x
x

x

x
2

Louwers
(2012a)
VI
QS

Louwers
(2014)
IV
PCS

Louwers
(2012b)
III
PCS

Sittig
(2016)
IV
CSDAS

Smeekens
(2011)
II
RCT
x

Teeuw
(2016)
IV
PCS

Tiyyagura
(2015)
VI
QS

x
x
7

x
x
3

x
x
7

x
x
4

x
1

x

x

1

3

x

x

x

x

x

x

x
x
38,136
0%

104,028
n/a

720
0.9%

x

x

x

38
36%

1089
15%

29
0%

x
x

x

2.2 yrs

6.4 mos

ND

ND

5.5 yrs

7.2 yrs

2.3 yrs

42 yrs

7.95 yrs

ND

48

56

ND

ND

57

56

58

28

56

ND

x

x

x

x
x

x

x
x
x

x

x
x

x
x

x
x

x
x

x

C – consistency, CA – child abuse, CAT – child abuse team, CSDAS – cross sectional diagnostic accuracy study, DR – detection rate, Escape – 6-item child abuse screening tool, EP – expert panel,
GED – general emergency department, HP – healthcare provider, K – knowledge, MCR – medical chart review, n/a – not applicable, ND – no data, NR – not recommended, PCS – prospective cohort
study, PDAS – prospective diagnostic accuracy study, PED – pediatric emergency department, QES – quasi-experimental study, QS – qualitative study, RCR – retrospective chart review, RCT –
randomized controlled trial, SE – self-efficacy, SES – socioeconomic status, SI – screening instrument, Sn – sensitivity, Sp – specificity, SP – screening protocol, SPUTOVAMO – 9-item child abuse
screening checklist, SPUTOVAMO-R – revised 6-item screening checklist, SSI – semi-structured interviews, SR – screening rate, ST – screening test, TTI – “top-toe inspection”, U – unnecessary

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Dinpanah
(2017)
Independent Variables
Audiotaped SSI
CA educational program
CA SP
Escape Form
Independent CA SI
SPUTOVAMO-R checklist
Dependent Variables
CA SR
CA risk/DR
Barriers to CA detection
Barriers to CA reporting
Facilitators to CA detection
HP CA education/support needs
HP CA knowledge/skill set
SE in CA detection
Racial bias/referral patterns
SES bias/referral patterns
Completion of SPUTOVAMO/TTI
Factors influencing SP completion
Findings/Outcomes
Escape SI diagnostic accuracy
SPUTOVAMO-R diagnostic
accuracy
SP effect on bias
SP effect on CA screening
Effect of CA educational program
Barriers to CA detection/reporting
Lack of CA knowledge
Practical problems (time, space)
Desire to believe caregiver
Personal barriers/beliefs
Deficient communication skills
Lack of administrative support

Higginbotham
(2014)

Jordan
(2014)

x

x
x

63
Louwers
(2012a)

Louwers
(2014)

Louwers
(2012b)

Sittig
(2016)

Smeekens
(2011)

Teeuw
(2016)

x

Tiyyagura
(2015)
x

x

x
x

x

x

x

x
x
x
x

x

x
x

x

x

x

x
x
x
x

x
x
x

x

x
x
x
x
Sn 100%;
Sp 98.3%

Sn 80%;
Sp 98%
↑ false +
rate: 97%
↓ SES bias
↑C

↑C
↑K, ↑SE

↑SR, ↑DR
x
x
x
x
x

↑SR
↑SR, ↑DR

↑DR, ↑SE

↑C , ↑SR
↑SP use
x
x
x
x

C – consistency, CA – child abuse, CAT – child abuse team, CSDAS – cross sectional diagnostic accuracy study, DR – detection rate, Escape – 6-item child abuse screening tool, EP – expert panel,
GED – general emergency department, HP – healthcare provider, K – knowledge, MCR – medical chart review, n/a – not applicable, ND – no data, NR – not recommended, PCS – prospective cohort
study, PDAS – prospective diagnostic accuracy study, PED – pediatric emergency department, QES – quasi-experimental study, QS – qualitative study, RCR – retrospective chart review, RCT –
randomized controlled trial, SE – self-efficacy, SES – socioeconomic status, SI – screening instrument, Sn – sensitivity, Sp – specificity, SP – screening protocol, SPUTOVAMO – 9-item child abuse
screening checklist, SPUTOVAMO-R – revised 6-item screening checklist, SSI – semi-structured interviews, SR – screening rate, ST – screening test, TTI – “top-toe inspection”, U – unnecessary

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Dinpanah
(2017)
Fast ED staff turnover
Negative results of reporting
Facilitators to CA detection
Administrative support
Presence of CA attendant
Presence of CAT
Immediate discussion with peers
Intensive educational training
Financial support
Standardized reporting process
Recommendations for ED Practice
Escape Tool SI
SPUTOVAMO-R
CA SP
CA educational program for HPs

Higginbotham
(2014)

Jordan
(2014)

64
Louwers
(2012a)
x

Louwers
(2014)

Louwers
(2012b)

Sittig
(2016)

Smeekens
(2011)

Teeuw
(2016)

x
x
x

Tiyyagura
(2015)
x
x

x
x
x

x
x

x
x

x

x

x

x
x

NR
x

x
x

x

x

x
x

x

C – consistency, CA – child abuse, CAT – child abuse team, CSDAS – cross sectional diagnostic accuracy study, DR – detection rate, Escape – 6-item child abuse screening tool, EP – expert panel,
GED – general emergency department, HP – healthcare provider, K – knowledge, MCR – medical chart review, n/a – not applicable, ND – no data, NR – not recommended, PCS – prospective cohort
study, PDAS – prospective diagnostic accuracy study, PED – pediatric emergency department, QES – quasi-experimental study, QS – qualitative study, RCR – retrospective chart review, RCT –
randomized controlled trial, SE – self-efficacy, SES – socioeconomic status, SI – screening instrument, Sn – sensitivity, Sp – specificity, SP – screening protocol, SPUTOVAMO – 9-item child abuse
screening checklist, SPUTOVAMO-R – revised 6-item screening checklist, SSI – semi-structured interviews, SR – screening rate, ST – screening test, TTI – “top-toe inspection”, U – unnecessary

CHILD PHYSICAL ABUSE SCREENING PROGRAM

65
Appendix D
Logic Model

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Appendix E
Institutional Review Board (IRB) Approval Letter

EXEMPTION GRANTED
Debra Hagler
CONHI Academic Innovation
602/496-0802
DEBRA.HAGLER@asu.edu Dear
Debra Hagler:
On 7/7/2017 the ASU IRB reviewed the following protocol:
Type of Review: Initial Study
Title: A quality improvement project to increase emergency
department healthcare provider recognition of child
physical abuse
Investigator: Debra Hagler
IRB ID: STUDY00006473
Funding: None
Grant Title: None
Grant ID: None
Documents Reviewed: • Consent - Child Abuse Awareness Pre-test/Post-test,
Category: Consent Form;
• Banner Health HIPAA Patient Signature Form,
Category: Other (to reflect anything not captured
above);
• Letter of Support from Banner UMC Site, Category:
Off-site authorizations (school permission, other IRB
approvals, Tribal permission etc);
• Child Abuse Educational Session Objectives,
Category: Other (to reflect anything not captured
above);
• Recruitment Script 2 - Program Evaluation Survey,
Category: Recruitment Materials;
• References - IRB application , Category: Other (to
reflect anything not captured above);
• Recruitment Script 1 - Child Abuse Awareness
Pre/Post-test, Category: Recruitment Materials;
• Consent - Project Evaluation Survey, Category:

66

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Consent Form;
• Project Evaluation Survey, Category: Measures
(Survey questions/Interview questions /interview
guides/focus group questions);
• Notice of Privacy Practices for Banner Health,
Category: Other (to reflect anything not captured
above);
• Social Behavioral Protocol - HCP Recognition of
Child Abuse, Category: IRB Protocol;
• Escape Instrument - Screening Tool, Category:
Screening forms;
• Project Flow Diagram, Category: Technical
materials/diagrams;
• Child Abuse Awareness Pre-test/Post-test, Category:
Measures (Survey questions/Interview questions
/interview guides/focus group questions);

The IRB determined that the protocol is considered exempt pursuant to Federal Regulations
45CFR46 (4) Data, documents, or specimens, (2) Tests, surveys, interviews, or observation
on 7/7/2017.
In conducting this protocol you are required to follow the requirements listed in the
INVESTIGATOR MANUAL (HRP-103).
Sincerely,
IRB Administrator
cc:
Sheri Carson

67

CHILD PHYSICAL ABUSE SCREENING PROGRAM

68

Appendix F
Site Approval Letter

Date: Wednesday July 12, 2017
To: Sheri Carson, MSN, RN, CPN, CPNP
Cc: Debbie Hagler, PhD, RN, ACNS-BC, CNE, CHSE, ANEF, FAAN
Clinical Professor and Coordinator, Scholarship of Teaching and Learning
From: Jill Arzouman, DNP,RN,ACNS,BC,CMSRN
Re: A quality improvement project to increase emergency department healthcare provider recognition
of child physical abuse.

Thank you for submitting the required documentation from Arizona State IRB as well as your project
template. As per our previous discussion, we have assessed your project proposal for implementation
potential and appropriateness of the project within BUMCT. From our final review we have determined
that the project is feasible and congruent with Banner Health initiatives.
I will forward your proposal and IRB approval to our Nursing Research Director at the Banner Corporate
Research Offices. You may now begin your project. Please do not hesitate to contact me for any
questions during the process. The Research and Innovation Council will look forward to you presenting
your results when the project is complete.
Best wishes on the successful completion of your project.
Sincerely,
Jill Arzouman
Director of Professional Practice
BUMCT/S

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Appendix G
Systematic Child Physical Abuse Screening Protocol

69

CHILD PHYSICAL ABUSE SCREENING PROGRAM

70

Appendix H
Escape Instrument

The Escape Instrument was published in an open-access article distributed under the terms of the
Creative Commons Attribution-NonCommercial-ShareAlike License, which permits noncommercial use, distribution, and reproduction in any medium, provided the original author and
source are credited. The above tool was created for the ED HCPs and adapted from “Accuracy of
a Screening Instrument to Identify Potential Child Abuse in Emergency Departments,” by E. C.
Louwers, I. J. Korfage, M. J. Affourtit, M. Ruige, A. P. van den Elzen, H. J. de Koning, and H.
A. Moll, 2014, Child Abuse & Neglect, 38(7), p. 1276.

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Appendix I
Child Physical Abuse Diagnostic Codes

71

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Appendix J
Child Abuse Awareness Pretest

72

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Appendix K
Child Abuse Awareness Posttest

73

CHILD PHYSICAL ABUSE SCREENING PROGRAM

74

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Appendix L
Child Abuse Educational Session Learning Objectives
Upon completion of the child abuse educational session, participants will be able to:
1. Identify common “red flags” for child physical abuse
2. Demonstrate accurate use of the Escape Instrument based on a case study
3. Describe the process of screening for child physical abuse
4. State essential elements that must be documented in cases of suspected child physical
abuse

75

CHILD PHYSICAL ABUSE SCREENING PROGRAM
Appendix M
Project Evaluation Survey

76

CHILD PHYSICAL ABUSE SCREENING PROGRAM

77
Appendix N

Project Flow Diagram for Data Collection

CHILD PHYSICAL ABUSE SCREENING PROGRAM

78

Appendix O
Paired Samples Statistical Analysis
Table 1
Paired Differences between Pre- and Post-Educational Session ED HCP Knowledge
Paired Variable

M (SD)

SEM

95% CI

t(51)

p

PRE CA Reporting –
POST CA Reporting

-.83 (.38)

.05

[-.93, -.72]

-15.61

<.001

PRE CA Red Flags –
POST CA Red Flags

-.44 (.50)

.07

[-.58, -.30]

-6.36

<.001

PRE CA Documentation –
POST CA Documentation

-27 (.45)

.06

[-.39, -.15]

-4.34

<.001

PRE Total % Score –
POST Total % Score

-51.75
(26.98)

3.74

[-59.26,
-44.24]

-13.83

<.001

Note. CI = confidence level. CA = child abuse.

Table 2
Paired Differences between Pre- and Post-Educational Session ED HCP Confidence
Paired Variable

M (SD)

SEM

95% CI

t(51)

p

PRE Confidence to
Recognize CA –
POST Confidence to
Recognize CA

-.39 (.53)

.07

[-.53, -.24]

-5.24

<.001

PRE Confidence to
Screen for CA –
POST Confidence to
Screen for CA

-.46 (.58)

.08

[-.62, -.30]

-5.78

<.001

Note. CI = confidence level. CA = child abuse.