Child Adversity and Trauma-Informed Care Teaching Interventions: A Systematic Review

Methods for this systematic review are specified in the protocol registered on January 3, 2020, with PROSPERO (no. CRD42020148077),³¹  the International Prospective Register of Systematic Reviews. The review followed Preferred Reporting Items for Systematic Reviews and Meta-analyses³²  guidelines.

A medical librarian (J.R.) searched PubMed, Embase, Web of Science, CINAHL, Cochrane Central Register of Controlled Trials, PsychInfo, and MedEdPORTAL from inception through January 2021. A comprehensive search strategy for each database used relevant keywords and controlled vocabulary to identify publications addressing educational curricula for medical, nursing, and allied health professionals and students that focused on TIC or ACE concepts (Table 1).

To ensure completeness, we conducted an additional author-based search of PubMed, Web of Science, and Google Scholar to find resources by specific authors recommended by 4 national content experts in TIC education. A hand search was also conducted.

We used EndNote X9 to identify and remove duplicates. The remaining abstracts were exported to the review software Rayyan,³³  where they were screened independently by 2 authors (S.H.J. and M.S.), 1 of whom is a research-trained TIC expert. Discrepancies were resolved in consultation with the medical librarian (J.R.). A study was eligible for the review if it (1) was a full-text article with a full abstract and not a case report; (2) was published in a peer-reviewed English-language journal; (3) described an appropriate learning intervention or curriculum for a health professional study population that addressed ACEs, TIC, child abuse, and child maltreatment; and (4) reported and assessed teaching and evaluation methods. The flowchart in Fig 1 summarizes the search and selection process.

For each study that met the eligibility criteria, we completed a standardized data extraction form to record title, authors, year of publication, geographic location, study design, study population (type of health professional or learner), curriculum details (number and type of trainings) and curriculum evaluation. We used descriptive statistics to assess changes in teaching trends over time on the basis of content and learner type and organized studies by year of publication for an informal trend analysis.

One article was excluded because it could not be obtained, despite repeated interlibrary loan attempts.

The κ coefficient for this systematic review was 0.8567 (95% confidence interval, 0.7964–0.9169), suggesting a strong level of agreement between the 2 independent reviewers.

To create focus and efficiency in an in-depth curriculum analysis, we selected from the 51 studies a smaller subset that (1) addressed the broad topics of ACEs, TIC, and child maltreatment and (2) were published after the first Felitti et al¹  study (1998), when ACEs were first defined. This selection process yielded 30 studies,^{20–30,34–52}  which were then analyzed using the criteria described below (see Supplemental Fig 2). From among the 30 studies, we selected for deeper analysis 10 exemplar studies,^21–30  chosen for 2 reasons: (1) They met standards of high educational quality, applying accepted adult learning strategies,⁵³  and 2) they collectively revealed diversity in learner types and formats and exemplified different educational strengths. The purpose of the detailed analysis of exemplars was to help other programs to select from well-designed curricular models in the published literature to build on the work of those who have preceded them.

The framework used for evaluation of curriculum quality is based on GNOME,⁵⁴  a mnemonic for curriculum planning published in 1994 that aligns well with the 6 steps of curriculum development later developed by Kern et al.⁵⁵  The GNOME acronym stands for 5 curricular elements: goals, needs, objectives, methods, and evaluation of learners and curriculum (details in Table 2 legend).

Our criteria for evaluating each of the 5 GNOME elements (see Table 2 column headings) were adapted from the Toolbox for Evaluating Educators,⁵⁶  a resource developed by a task force of expert educators organized by the Association of American Medical Colleges. The toolbox contains evaluation criteria for 5 domains of educational activities, including curriculum development. The criteria for our analysis are derived from key components of curriculum development that are organized in the toolbox around Glassick’s criteria for educational excellence.⁵⁶  For our purposes, 4 of Glassick’s 6 criteria were relevant to this study: clear goals, adequate preparation, appropriate methods, and significant results. In the area of significant results, we evaluated learner and curriculum evaluation methods using the Kirkpatrick hierarchical model,⁵⁷  which categorizes evaluation at 4 levels: reaction, learning, behavior, and results (details in Table 2 legend). Although all 4 levels are important, achieving the higher levels of evaluation provides stronger evidence of curriculum quality.

The database searches yielded 2234 potential studies for inclusion in the review. Thirty additional studies^{20–30,34–52}  were identified through a search of specific authors recommended by national experts in TIC education and a hand search (N = 2264). After elimination of duplicates, 1537 studies underwent title and abstract screening. Information from 79 studies was recorded on data abstraction forms. Of the 79 studies, 28 were excluded after full-text analysis, and 51 were included for final, qualitative synthesis (Fig 1).

Tables 3 and 4 summarize the reviewed studies, including authors and publication year, geographic location, primary focus, learner type, and a brief intervention description. From the English-language studies in the final review, 38 (74.5%) of the 51 studies originated from the United States, so cross-national comparisons were not attempted.

We divided learner types targeted by the curricula into 2 sets. Those for practicing medical professionals (34 studies, Table 3) included physicians, residents, nurse practitioners, registered nurses, social workers, and allied health professionals from behavioral health, child life, and occupational therapy. These teaching interventions were typically offered as optional single-day sessions, ranging from 1 to 8 hours. Curricula for students (17 studies, Table 4) included medical, nursing, and social work students. Student teaching interventions often included multisession curricula embedded in semester-long courses.

Among the 51 publications included, we found 27 studies on ACEs/TIC,^{20,22,23,25–30,35–48,50–52,58}  14 on child abuse,^59–72  6 on domestic violence (DV)/intimate partner violence (IPV),^73–78  and 4 on child maltreatment.^21,24,34,49  Terminology about child trauma has changed over the years. As shown in Tables 3 and 4, which are chronologically sorted, teaching about child trauma in earlier years appeared to be focused mainly on specific kinds of abuse rather than on the broader phenomenon of trauma (ACEs) and its care (TIC). The first publication defining ACEs was published in 1998,¹  and since 2011, increasing numbers of studies focusing on children with ACEs have been published. TIC has also come into increasing use as a collective approach to care for children who have experienced abuse. The first study of a curriculum about TIC, ACEs, and/or child maltreatment in our review was published in 2011.²¹ 

Figure 2 shows that overall, the number of publications about childhood trauma increased dramatically between 2010 and 2021, with the largest changes in curricula focused on ACEs and TIC. Before 2010, most studies focused on ≥1 narrow topics of child maltreatment (eg, child abuse, DV/IPV, parental physical punishment). In contrast, of the 43 studies published between 2010 and 2021, 31 were focused on teaching the broader topics of ACEs, TIC, and child maltreatment.^{20–30,34–52,58}  These curricula for practicing health professionals became common after 2011 (28 of 34 studies),^{20–23,25,26,28–30,35–38,42,43,46,49,51,52,58,62–67,76,77}  whereas those for students became frequent after 2017 (14 of 17 studies).^{24,27,34,39–41,44,45,47,48,50,70–72} 

The subset of 30 curricula^{20–30,34–52}  selected for analysis varied from a 30-minute online module, to 1- to 4-hour didactic/interactive sessions, to full-day workshops with simulation experiences, to multiple sessions (usually included in student courses). A tabular summary of this analysis can be found in Supplemental Fig 2. We reviewed these articles for potential focus on emerging topics related to child adversity and found 7 curricula that addressed resilience,^{29,40,44–47,50}  4 on secondary traumatic stress/compassion fatigue,^20,29,45,48  and only 1 on broader social determinants of health.⁴⁰ 

We selected 10 curricular exemplars^21–30  (see Table 2, studies numbered C1–C10 for cross reference) and evaluated them in more detail, with a focus on the quality of teaching and evaluation methods. Curriculum planning, as outlined in GNOME, was often not described in specific terms, but nearly all curricula authors proposed goals focused on patient-centered communications and screening, with some addressing patient follow-up for positive screens. Authors of 6 studies collected individual needs assessment data from learners through surveys or pilot studies. Objectives were generally omitted, except in studies published in MedEdPORTAL (which requires them). A particularly well-planned study (C8) used the methods of implementation science to develop both a TIC program and supportive ACE/TIC training activities.

A program’s selection of teaching and evaluation methods depended on not only context, particularly the time available to learners, but also availability of expert teachers, key resources, and stakeholder buy-in. Curricula with stronger teaching methods allowed enough learner time for experiential activities with simulated patients (SPs) (C3, C10). The strongest evaluation methods used the randomized controlled trial (RTC) design (C1–C3, C5), which usually required that the educators had sufficient access to learners to assess their skills in real or simulated practice settings, ideally over time. Student curricula seldom offered these opportunities.

Robust curricula used a combination of didactic, interactive, and experiential activities, in keeping with the principles of adult learning. Didactic methods were typically interspersed with small group discussions, role plays, or practice exercises.

Slide presentations in didactic sessions were activated in many ways, such as progressive case studies to illustrate patient histories of trauma and typical behaviors (C5) or cases for practice with assessment tools (C1, C2). One excellent workshop (C8) taught with unfolding vignettes to illustrate patients with emotional stress indicators. Four curricula offered practice with tools to guide application of learning to patient care: a parent screening tool (C1, C2), an ACE screening tool and resilience questionnaire (C8), and the self-assessment/evaluation tools discussed below (C1–C4).

Small-group discussions were a standard component of all the curricula and provided opportunities for learner feedback. More interactive learning elicited strong positive learner response (C3, C6–C8).

Three curricula used SP care activities as a powerful tool to augment instruction (C3, C7, C10). Patient vignettes provided an alternative to SPs (C1, C2). Use of SPs for evaluation is discussed below; evaluation experiences at their best can be powerful teaching/learning experiences as well.

Learner and curriculum evaluation methods for the 10 exemplar studies varied in type and quality. These methods are rated and described in Table 2 in relation to Kirkpatrick’s hierarchical framework (K1–K4).

At level K1 (reaction), the most informative evaluation of learner reaction (C6) was focus groups with qualitative analysis. Four other studies used quantitative and/or qualitative survey data (C6–C9).

At level K2 (learning), 6 of the studies used self-report survey data (C1, C2, C4, C6–C9). Two curricula (C1, C2, which used the same teaching model) used a carefully designed, structured pre- and postquestionnaire with which learners self-assessed their knowledge, skills, and attitudes with reference to 5 vignettes linked to patient problems targeted by the curriculum. Four studies reported no K2 data (C3, C5, C6, C10), but 3 of these reported data at the higher K3 level.

At the important level K3 (behavior or application of learning), the exemplar studies used varied and often creative methods, including an objective structured clinical examination using SPs, rubric evaluation by faculty, and feedback (C7) and videotaped SP encounters using a standardized behavioral coding tool (C10). This ambitious study also attempted to compare learners at year 1 and year 4 of a longitudinal curriculum, but a poor response rate led to equivocal results. In studies C3 and C4, videotapes of SP encounters were analyzed by the rigorously developed Roter Interaction Analysis System^23,25  to create a patient-centeredness score (C3, C5). Moving beyond simulation, 2 curricula evaluated observations or audiotapes of real patient encounters (C1, C3), and 2 also used chart reviews to track learners’ screening and follow-up for ACEs in practice (C1, C2). The most rigorous tool used for K3 curricular evaluation was the RCT. An intervention versus control group design in 3 studies was used to compare measurements longitudinally, demonstrating increases in practice-based competence (C1), screening frequency (C2), and patient-centeredness scores (C3). A fourth study used an RCT with immediate versus delayed intervention design and revealed pre- and postintervention increases in patient-centeredness scores both between and within groups (C5).

Level K4 (results) was addressed in only 2 studies. C2 used a parent-doctor interaction scale to measure parent satisfaction with care. C6 used interviews at 6 months after training to document delivery of a train-the-trainer intervention at participants’ home settings.

This systematic review includes 51 studies published between 1979 and 2021 that described curricula on ACEs, TIC, and child maltreatment taught to health care professionals and trainees. The curricula varied widely in breadth, duration, and quality. Three quarters of these studies were conducted in the United States. Our original hypothesis that few teaching interventions would be published was confirmed only for the years before 2011. Three studies matched criteria before 2001,^59,68,69  2 studies matched criteria between 2002 and 2005,^60,73  and 3 studies matched criteria between 2006 and 2009.^61,74,78  However, eligible publications accelerated in the next decade. Twenty curricula about TIC/ACEs were taught to medical professionals after 2011,^{20–23,25,26,28–30,35–38,42,43,46,49,51,52,58}  and 10 curricula were taught to students after 2017.^{27,34,39–41,44,45,47,48,50} 

Terminology for ACEs changed considerably over time and may partly explain our failure to find many articles before 2000 that met our criteria. We found 20 studies focused on teaching about specific kinds of child trauma (eg, child abuse, DV/IPV, parental physical punishment).^59–78  Only after 2011 did studies focus on teaching broader principles of ACEs and TIC, reflecting publication of the ACE studies by Felitti and colleagues from 1998 through 2006.^1–6 

Initially, studies of ACEs fell under the purview of adult medicine, and ACEs were described as a “chronic health morbidity”¹  because Felitti’s early ACE studies were based on adult recall of experiences in childhood and health consequences later in life. These publications were groundbreaking because they led the field to understand that “trauma” in children extends beyond physical injury to psychological stress²⁰  and that these stressors have lifelong effects. More inclusive perspectives on child trauma have developed along with growing recognition of the importance of social determinants of health to child and adult well-being.^79,80 

Teaching interventions for health professionals were more likely to be optional workshops of 2 to 8 hours compared with multisession learning opportunities for students, often embedded within semester-long courses. This difference reflects the real-world challenge of finding time for continuing education for health professionals in busy clinical practice. The findings reinforce our previous work showing the challenges of teaching practitioners about TIC in clinical settings.²⁰ 

We identified 10 curricular exemplars^21–30  that represent a variety of instructional designs targeting students, residents, and practicing health professionals. These are potentially useful models that demonstrate excellence in curriculum planning (C5, C6, C8, C9), robust teaching methods (C1, C2, C4, C5, C7, C8), and/or development of strong evaluation tools (C1–C3, C5, C8, C10). Embedded in Table 2, readers will find several resources worth studying in-depth from the original publications.

In this rapidly changing educational field, some important new topics for teaching about child adversity were not fully covered in the curricula we reviewed. Looking forward, educators teaching about TIC are now beginning to focus not only on treating pathology but also on building resilience. Among the 51 studies, only 7 addressed this important topic.⁵⁰  The AAP Pediatric Approach to Trauma, Treatment, and Resilience curriculum⁸¹  provides continuing medical education in multisession case-based interventions that show “how kids and caregivers can be supported to promote resilience through attachment, regulation, and efficacy.”¹⁶  Innovative programs by the National Child Traumatic Stress Network,⁸²  as well as by the AAP,⁸¹  are teaching primary care providers to reconceive relational approaches to pediatric care to promote resilience and respond to trauma. Another perspective that is emerging in educational practice and in the literature is the concept of secondary traumatic stress, or compassion fatigue, which is reported by medical professionals who care for traumatized children.²⁰  Two nursing curricula in our study address this issue in detail,^45,48  along with a curriculum for pediatric residents by Jee et al²⁰  and a curriculum for pediatricians by Schmitz et al.²⁹ 

The ACE studies of Felitti and colleagues heralded a new era in conceptualizing child adversity. Built on this foundation, newer interpretations of ACEs relate trauma to more broadly defined social determinants of health, such as food insecurity, community violence, poverty, housing instability, structural racism, environmental blight, and climate change. We found only 1 article that addressed social determinants of health,⁴⁰  but more will follow. These newer approaches to framing the social circumstances that shape children’s health and drive health disparities are rapidly advancing the field of child adversity.⁸³  We recommend that medical educators who teach about child trauma take a broad view, including the origins of trauma across a child’s life, and follow the impact of child trauma on health across the lifespan.

This systematic review reveals a steady growth in curricula that teach health professionals to understand and care for children who have experienced trauma. The identification of ACEs has deepened and broadened our understanding about how trauma affects humans in early development, when their vulnerability is greatest. It has also led us to focus our attention on the effects of social determinants of health on child development. These new insights are rapidly being incorporated into health professional training.

Emergence of the discipline of TIC has also been revolutionary, establishing best practices for child screening and treatment and identifying secondary traumatic stress in providers who treat child trauma. Effective teaching about TIC should be both patient and provider centered. Health professionals need to be mindful of how their own history and reactions to trauma affect the care they provide.⁸⁴  Moreover, pediatric providers need to learn to exercise caution in screening children for ACEs, to avoid creating a self-fulfilling “expectancy effect” that labels a high-risk child as likely to have poor outcomes.⁸⁵  Screening for ACEs, which was the focus of most of the curricula in our review, is critical,^85,86  but curricula need to go farther to emphasize resilience training as well.⁸⁷ 

A strength of this study was that we conducted this systematic review with the guidance of a medical librarian and followed Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines. The use of searches unrestricted by publication date allowed us to track major changes between 1979 and 2021 in the way professionals are taught to deal with child trauma. We examined changes over time in the content of the curricula identified and evaluated their quality using well-established educational criteria.

Changes in the terminology for child trauma complicated our evaluation of trends in educational content over time and may have led to missed studies. Another limitation of this study is that we included only published English-language studies, most conducted in the United States, so generalizability of the review to other parts of the world is limited. Moreover, the explosion of academic publications in recent decades has hampered interpretation of frequency estimates as a marker of evolving interest in child trauma.

Finally, evaluation of educational curricula requires judgment in the application of best practice criteria; our quality assessments, although conducted by an educational veteran (C.D.B.), are intended to serve as a well-informed guide rather than as a definitive test of excellence. In addition, although we included MedEdPORTAL as a key resource for peer-reviewed publication of medical education curricula, many teaching interventions are not published. We found but omitted from the review many non–peer-reviewed and unpublished curricula on child trauma (eg, workshops at national meetings, conference proceedings, webinars, Web sites, and toolkits). Such curricula would provide a more complete picture of how child trauma is taught to medical professionals, but they are difficult to replicate without evaluation and peer review.

In a literature search over 5 decades, we found an emerging and, later, rapidly expanding interest in child adversity and since 1998, a broader emphasis on ACEs and TIC. Application of a critical framework to evaluate these published resources may encourage the development of new medical curricula on child trauma. Access to sound models should help educators to select well-designed, peer-reviewed curricula to build on the work of their predecessors. Good teaching, in turn, will give the next generation of health care professionals foundational knowledge to address ACEs proactively and implement effective TIC to improve the lives of children. In the future, exemplary curricula should address the broad health impacts of childhood trauma and teach proven strategies to proactively improve health, particularly by building children’s resilience. We need to develop providers who have the knowledge, attitudes, and skills required to identify and treat ACEs early in life, when further trauma may be prevented.

We gratefully acknowledge the advice of our content experts Moira Szilagyi, MD, PhD, Heather Forkey, MD, Jody Todd Manly, PhD, and Sheree Toth, PhD, as well as the content experts who they recommended and whose work we reviewed. We appreciate the helpful comments on our manuscript from Dr Todd Manly.

ACE

adverse childhood experience

DV

domestic violence

GNOME

goals, needs, objectives, methods, and evaluation of learners and curriculum

IPV

intimate partner violence

RCT

randomized controlled trial

SP

simulated patient

TIC

trauma-informed care