History of the CASTER™
History of the Development of the CASTER™ (Child and Adolescent Screener for Traumatic Events and Response)
Given the high prevalence of trauma in children referred for mental health services, it has been recommended that mental health professionals ensure that they are regularly screening for exposure to traumatic events and indicators of posttraumatic distress so that the appropriate, evidence-based interventions can be employed (e.g., Alisic et al., 2014 ; Briggs et al., 2013; Chemtob et al., 2016; Rolon-Arroyo et al., 2020). Some of the available instruments for assessing child and adolescent trauma require detailed clinical interviews with the child and/or the child’s parent/caregiver (e.g. Children’s PTSD Inventory, Saigh, 2004; K-SADS, Kaufman et al., 1997). The mental health Community Clinic at the George Hull Centre for Children and Families provides services to approximately 500 clients each year, so it is not feasible to utilize these resource-intensive measures as a screening tool for each client. As an alternative to clinical interviews, a number of self- and parent-report questionnaires have been developed to assess aspects of trauma. However, some questionnaires have been designed to inquire solely about potential trauma symptoms—without assessing exposure to traumatic events (e.g., Child Report of Post-Traumatic Symptoms, Greenwald & Rubin, 1999; Trauma Symptom Checklist for Children, Briere, 1997), whereas other measures have been developed to inquire about trauma exposure—without assessing potential traumatic distress (e.g., Childhood Trauma Questionnaire, Bernstein & Fink, 1998). Other available measures have been designed to provide information regarding only a single traumatic event (e.g., Acute Stress Index for Children, Kassam-Adams, 2006; UCLA PTSD Index for Children/Adolescents-DSM5, Pynoos & Steinberg, 2013), so information regarding multiple traumatic events is not captured. Furthermore, many of the available measures are focused on a somewhat narrow range of potentially traumatic events, and they miss potentially traumatic events such as bullying, financial distress in the family, parental substance abuse, and removal from the home by child welfare authorities, for example. Our goal, therefore, was to develop a more comprehensive and contemporary screening tool.
2017: Initial Development
Development of the Child and Adolescent Screener for Traumatic Events and Response (CASTER) began in 2017 by Dr. James Worling, consulting clinical and forensic psychologist at the George Hull Centre. At the outset, there were two versions of this two-page tool: a self-report version for youth aged 12-18 and a parent-report version for children from birth to age 18. Page 1 of each version contained a list of 29 potentially traumatic events from several categories, including loss (e.g., “Death of a parent”), community violence (e.g., “Witnessed a death or attempted suicide”), family violence/neglect (e.g., “Physically hurt by parent/caregiver”), serious accident/injury (e.g., “Serious building fire”), separation from parent/caregiver (e.g., “Removed from the care of parents/caregivers by child welfare authorities”), sexual abuse (e.g., “Made to engage in sexual activity by an adult”), and war (e.g., “Resident of a refugee camp”). The list of potentially traumatic events was informed by Dr. Worling’s clinical experiences working with child and adolescent trauma and by several of the available trauma screening measures (e.g., Childhood Trauma Questionnaire, Bernstein & Fink, 1997; UCLA PTSD Index for Children/Adolescents-DSM5, Pynoos & Steinberg, 2013).
On page 1 of the CASTER, respondents were asked to indicate which of the 29 potentially traumatic events had occurred by indicating Yes or No for each event (there was also space for a 30th item, “other”, so that respondents could list a traumatic event that was not otherwise included in the list). If respondents indicated that, Yes, an event had occurred, they were also asked to indicate whether the event in question occurred either less than a year ago or more than a year ago. The goal here was to collect information regarding the recency of the traumatic event(s), as this is not typically included on the available trauma screening measures. At the time of initial development, it was thought that recency information may assist with intervention planning (e.g., Matte-Landry et al., 2023). Note below, however, that this was eventually removed from the CASTER following feedback from parents/caregivers and clinicians.
Page 2 of the CASTER had 2 sections: one small section contained 4 questions regarding potentially significant changes/attachment disruptions that the child/adolescent may have experienced (e.g., “How many changes have there been in parenting figures for the child/adolescent? 0, 1 or 2, 3 or 4, 5-10, or >10”). These questions were included on the tool, as information regarding stability and attachment is not often included in such screening measures. The second section of page 2 contained a list of 47 possible responses to child/adolescent trauma (e.g., “impulsive behaviour”, “avoids social activities”, and “nightmares”). The response choices for these items was Never, Sometimes, or Often. As was the case for the traumatic events listed on page 1, these potential trauma responses were informed by Dr. Worling’s clinical experiences and by existing measures of traumatic distress (e.g., Child Report of Post-Traumatic Symptoms, Greenwald & Rubin, 1999; Trauma Symptom Checklist for Children, Briere, 1997). Several wording suggestions were provided by Leticia Gracia (former Director of the Community Clinic at the George Hull Centre) and Diane Bartlett (Director of Research and Evaluation, George Hull Centre) in 2017 to enhance the clarity of the CASTER prior to piloting the measure at the George Hull Centre.
2017-2019: Pilot Implementation
In the late spring of 2017, the CASTER was introduced to clinicians in the George Hull Centre Community Clinic. A number of additional wording changes were made to the self- and parent/caregiver-report versions of the CASTER following helpful input from the clinicians. Additionally, after just a few months of piloting the measure, one of the four items designed to understand instability (i.e., number of changes in parenting figures) was removed, as it was deemed too cumbersome for clients to complete. With this revised version of the CASTER, clinicians were asked to implement the self- and parent/caregiver-report measures routinely with new clients; typically within the first few sessions. This process began towards the end of 2017 and continued throughout 2018 and 2019. This initial implementation aimed to … (just wondering if we want to make note here that it was done in association with research and/or what we were looking at/for).
In March of 2019, the CASTER research team (which initially included Diane Bartlett, Leticia Gracia, and James Worling) collated and examined CASTER self-report data from 101 youth and CASTER parent/caregiver report data from 196 parents/caregivers. A number of interesting findings emerged from this initial analyses regarding the incidence of traumatic events and the frequency of potential trauma responses. For example, youth in the initial sample reported an average of 4 adverse life events, with at least 30% of the youth reporting a bully-victimization history, death of a family member/close friend, parental divorce/separation, or emotional abuse. With respect to possible responses to trauma, at least 30% of the youth indicated that they often felt anxious, depressed, or dissociative, or that they experienced frequent difficulties with their sleep. Most importantly, it was also found that the correlation between the total number of traumatic experiences (Total score) and the responses to trauma (Total score) was significant (r=0.60, p<.001). During a discussion of these initial results with clinicians at the Community Clinic, additional suggestions were made regarding the content and wording of the CASTER. For example, it was suggested that the CASTER include an item to address experiences of discrimination based on factors such as race, sexual orientation, gender, and place of birth. Discussions with clinicians at the Community Clinic also led to the refinement of the instructions for the administration of the CASTER.
2019: Content Validity
A research proposal was developed in late 2019 to begin a study of the psychometric properties (i.e., reliability and validity) of the CASTER. Prior to submission to the Research Ethics Board at the George Hull Centre, however, the CASTER research team invited Dr. Kristin Cleverly from the Centre for Addiction and Mental Health in Toronto to provide input on the proposed research design. Dr. Cleverly suggested that, instead of looking at the reliability and validity of the current versions of the CASTER, we instead consider having the CASTER reviewed by youth, parents/caregivers, and trauma experts external to the George Hull Centre to ensure that the item wording was clear and that the tool contained the most pertinent items. As a result, a new proposal was developed in 2020 focused on the content validity of the CASTER. This resulted in an extensive investigation that included collecting detailed feedback on the CASTER from (i) a focus group with emerging adults (ages 18-25), (ii) parents/caregivers of children and youth aged 0-18, and (iii) trauma experts from Canada and the U.S. with specialized knowledge and experience in the area of child and adolescent trauma.
(i) Young Adult Focus Group
Four young adults participated in a focus group in 2020 to review the content of the CASTER. These participants all had personal experiences with trauma and with mental health services in Ontario, and they included members from the Black, Muslim, and LGBTQ+ community. Detailed feedback from these participants resulted in several important changes to the wording of CASTER items. In particular, some items were perceived by participants as being overly “formal” or “clinical”, so they were reworded. For example, “Lived in a high-crime neighbourhood” was changed to “Lived somewhere that felt dangerous, stressful, or unsafe” to capture instances where youth may have felt unsafe—regardless of the incidence rate of crime in their neighbourhood. The young adults in the focus group also felt that some of the items were too ‘blaming’. As such, a number of changes were made, such as rewording, “Refuses to attend school”, to “Not wanting to go to school/day program”. Another important change based on the feedback from the young adults was to separate, “Treated badly because of race, gender, sexual orientation, religion, or place of birth” into separate items, as it was perceived as dismissive or minimizing to have the categories lumped into a single item. Finally, the young adults in the focus group stressed the need for clinicians to ensure that individuals completing the CASTER were well aware of the limits of confidentiality, to safeguard the information collected from the CASTER, and to have sufficient training to assist individuals who have disclosed traumatic experiences.
(ii) Parent/Caregiver Interviews
A total of six parents (3 parents/caregivers of former clients at the George Hull Centre and 3 parents/caregivers from other agencies in the Toronto area) were interviewed individually in 2020 and 2021 regarding their feedback on the CASTER. Two parents identified as Black, and one identified as Hispanic (and spoke English as a second language). Three parents had children under the age of 12, and four had children aged 12 to 18. Two parents had children with relatively little trauma exposure, three parents had children with significant trauma exposure, and one parent had a child diagnosed with developmental trauma.
Parent/caregiver feedback highlighted some wording on the CASTER that was perceived as awkward or unclear. For example, “Unable to gain weight” was changed to “Difficulty eating (e.g., refusing to eat, eating too much or too little)”. Concerns were also raised by some parents/caregivers regarding how some items may lead to respondents feeling personally blamed for things that may have been out of their control. For example, the item, “Lack of food, affection, appropriate clothing from parent/caregiver” was revised to “Not enough food, appropriate or clean clothing, or other basic needs”. Several parents/caregivers also noted that some of the language on the CASTER felt quite subjective and, therefore, that it would be helpful to add examples and/or behavioural anchors to many of the items (both traumatic events and responses). The research team did make some changes in wording to address this concern (e.g., “ ‘Jumpy’, or easily startled” was revised to “Easily startled [e.g., “jumpy” in response to touch, sound, or being approached”]); however, it was felt that adding more specific behavioral anchors to many of the items would reduce the utility of the instrument by too narrowly restricting the meaning.
(iii) Trauma Experts Review
Detailed feedback on the CASTER was obtained in the fall of 2020 from 14 individuals with expertise with respect to childhood trauma. The individuals invited to participate in this process were clinical and academic professionals from medicine, psychology, and social work in Canada and the U.S. The experts were selected because of their knowledge and experience in pertinent areas, including trauma, attachment, and scale development. Furthermore, within this group, there was expertise regarding preschoolers, school-age children, adolescents, and adults, and expertise regarding diverse service-delivery systems, including community mental health, residential treatment, youth justice, medical research, and child welfare, for example.
A great deal of quantitative and qualitative feedback was received during this process. Quantitative feedback regarding the importance of each item on the CASTER was collected from the experts and analyzed using Lawshe’s (1975) Content Validity Ratio (CVR) rating scale. More specifically, participants were asked to rate how essential each item (for both traumatic events and responses) was on a 3-point scale: Essential, Useful but not essential, or Not necessary. Using formula and critical-value tables from Lawshe (1975), it was determined that, ideally, 10 of the experts would need to rate an item as Essential for the item to be retained. Most of the traumatic events on page 1 of the CASTER were seen as Essential by at least 10 of the expert reviewers. One item (“Criticized or put down by close family member NOT including parent/caregiver”) was removed because it did not meet the threshold; however, there were a few items that were deemed important to retain by the research team (e.g., “Death of a beloved pet”, “Frightening or painful medical procedure”) despite not quite meeting the threshold amongst expert reviewers. Similarly, several potential trauma responses were removed (e.g., “Difficulty making decisions”, “Unable to gain weight”, “Reacting aggressively to situations that feel threatening”) through this process, as few experts viewed these as Essential to include on a trauma screener. As in the case of traumatic events, however, there were several items that did not quite meet the threshold of 10 experts but were nonetheless judged by the research team as critical to retain (and, in some instances, reworded) by the research team (e.g., “Risky or reckless behaviour”).
A great deal of qualitative feedback was also provided by the trauma experts, and all comments and editorial suggestions were carefully reviewed by the research team. Qualitative feedback was additionally obtained from 2 experts in the area of diversity and inclusion in mental health services.
The CASTER-Speech-Language Pathology (“CASTER-SLP”)
During 2020, an additional and related project was underway that significantly impacted the development of the CASTER. More specifically, Anna Rupert (Program Manager, Preschool Speech and Language Services, George Hull Centre) and Diane Bartlett were conducting an investigation regarding speech-language pathologists’ awareness of how trauma may be impacting their clients and, subsequently, speech-and-language treatment goals/approaches. The results of their investigation revealed that speech-language pathologists were very interested in understanding the impact of childhood trauma and that there was an opportunity for early identification of children exposed to traumatic stress through routine screening by speech-and-language professionals (Rupert & Bartlett, 2022). At this juncture, Anna Rupert and Kim Curran (Director, Prevention and Early Intervention Services, George Hull Centre) joined the CASTER research team, and this began work to develop a version of the CASTER that would be appropriate for children seen in Preschool Speech and Language Services at the George Hull Centre (i.e., children aged 0 to 4 years). During the initial phases of development, this brief version of the tool was referred to by the research team as the “CASTER-SLP”.
As part of the development process of the CASTER-SLP, three speech-language pathologist focus groups were conducted (those at the George Hull Centre, SLPs working from other Ontario Preschool Speech and Language System programs, and SLPs working in other settings/systems with young children) to provide detailed feedback regarding CASTER items and screening implementation. Several parents of children attending the Preschool Speech and Language Services at the George Hull Centre were also asked for their input regarding the pilot version of the CASTER-SLP. Feedback from speech-language pathologists and parents resulted in a number of important changes to address the wording of several items. For example, “Sensitive to, or bothered easily by, sound, smell, light, taste, or texture” was flagged as somewhat awkwardly worded, and it was changed to “Easily bothered by certain sounds, smells, sights, tastes or textures”. A number of items were also reworded to address concerns regarding the literacy level of the CASTER-SLP. For example, “Loss of previously mastered skills (e.g., toileting, dressing, feeding, communication)” was changed to “Is not able to do things that they used to be able to do (e.g., toileting, dressing, feeding, talking”). Feedback from speech-language pathologists also led to re-ordering several items so that related items would be adjacent to one another, thus facilitating ease of completion. Input from parents and speech-language pathologists during this phase also resulted in modifications to the instructions to parents/caregivers for completing the screener. Many of the suggested modifications to the CASTER during this process ultimately contributed to the development of a third version of the CASTER, the parent/caregiver version for children aged 0-5.
2020 & 2021 CASTER Revisions
During late 2020 and early 2021, the expanded CASTER research team (which also included several student volunteers) made a number of key revisions to the self- and parent/caregiver-report versions of the CASTER based on the extensive and detailed feedback from the youth focus group, parent interviews (from both the content validity study and the CASTER-SLP development process), expert review, and feedback from Community Clinic clinicians, speech-language pathologists, and diversity specialists. More specifically, many items were edited for greater clarity, and a number of items were added to address issues raised through the various consultation processes. It was also decided at this time to change the recency rating for traumatic events (i.e., less than a year ago vs. over a year ago) to a measure of how much a traumatic event currently impacts the individual (Not at all, Somewhat, or Very Much). Feedback from clients and clinicians indicated that it was often challenging to pinpoint the timing of specific, past events. Based on feedback from the trauma experts, parents, and clinicians, the time period regarding the presence of potential trauma responses was changed from “During the past 3 months” to “During the past 6 months”, as it was felt by many that the 3 month interval was too brief.
In the late fall of 2020, clinicians at the Community Clinic were asked to complete a survey regarding CASTER items and the CASTER screening process, and this resulted in several additional modifications to the wording of several items to enhance clarity. It should also be noted here that the research assistants carefully compared the wording of the final CASTER items to the wording on available measures to ensure that there were no copyright-infringement issues.
The most significant change to the CASTER throughout this process was the development of two separate parent/caregiver versions based on the age of the child. Although it was initially thought that a single version of the tool would be sufficient for capturing trauma exposure and responses for children and youth, the feedback received from experts and parents/caregivers during the content-validity study strongly suggested the need for separate versions for the two age groups (i.e., 0-5 and 6-18), as some potential trauma exposures (e.g., being offered gifts or money to do “sexual things”) and trauma responses (e.g., suicide threats) were deemed most applicable only for older children and youth. The development of a shortened version of the CASTER for parents/caregivers of children aged 0-5 years dovetailed very nicely with the work that the research team had already been doing with respect to the development of a brief version of the tool for speech-language pathologists to employ with parents of young children (i.e., the “CASTER-SLP”). The end result was a shortened and slightly modified version of the CASTER for parents/caregivers of children aged 0-5 and a more comprehensive version for parents/caregivers of children and youth aged 6 to 18—together with a more comprehensive self-report version for children and youth aged 12-18. The final parent/caregiver-report version of the CASTER for individuals aged 6-18 (and the self-report version for youth aged 12-18) contains 44 potentially traumatic events and 65 possible responses to trauma. The parent/caregiver version of the CASTER for children aged 0-5 contains 21 potential traumatic events and 41 possible trauma responses (and this is the version that is currently being used to screen children at the Preschool Speech and Language Service at the George Hull Centre). All versions of the CASTER also contain 3 questions regarding possible changes in parenting/caregiving figures, schools/day programs, and out-of-home placements. The final versions of the CASTER were completed in 2021, and they were subsequently translated into French and Spanish. A detailed user guide for the CASTER was also developed in 2021, and the CASTER materials were subsequently made available for free download through the George Hull Centre’s website.
Future Research with the CASTER
We are planning to conduct research at the George Hull Centre regarding the psychometric properties of the CASTER and, importantly, the nature of child and adolescent trauma. With respect to the psychometric properties of the CASTER, we will be examining test-retest (2 weeks) reliability with a sample of clients to establish the temporal stability of the CASTER. We are also planning to compare client responses on the CASTER to client responses on established measures of potential trauma responses to investigate the concurrent validity of the CASTER.
Another area of research will be to determine what information is gained through the use of the CASTER that is over and above the information that is already collected via (i) the standard telephone intake process and (ii) the single-session interview with the child/adolescent and parent (called the “consultation interview”). For this investigation, we hope to complete the CASTER using file documents from the intake and interview processes and then compare the information from these sources to information from the client-completed CASTER. As part of this study, we will also have a second rater code approximately 30% of the files to permit an examination of the interrater reliability of the CASTER.
De-identified CASTER data from a large sample will permit us to look at the prevalence of specific traumatic events and responses that clients are presenting with at the George Hull Centre. This information will help to inform future training and service-delivery needs. The factor structure of the CASTER will also be examined with these data, as it will be informative to determine if there are meaningful clusters of traumatic events and traumatic responses. Statistically reducing the number of items in this manner would facilitate further research into the possible associations between certain classes of traumatic events (e.g., family violence vs. community violence) and classes of traumatic responses (e.g., mood vs. self-regulation).
Another line of investigation will be to examine the linkages between attachment disruptions/instability and possible traumatic responses. Additionally, we are hoping to look at the level of agreement between the youth- and parent/caregiver-report versions of the CASTER and outline if there are certain traumatic events and responses that are under- or over-reported by youth relative to their parents/caregivers. We hope that the CASTER will prove to be a reliable and valid screening tool that provides critical information regarding the impact of traumatic childhood experiences. We also hope that the information that we glean from the CASTER will add to the scientific literature regarding child and adolescent trauma and attachment.
References
Alisic, E., Zalta, A. K., Van Wesel, F., Larsen, S. E., Hafstad, G. S., Hassanpour, K., & Smid, G. E. (2014). Rates of post-traumatic stress disorder in trauma-exposed children and adolescents: meta-analysis. The British Journal of Psychiatry, 204(5), 335-340. https://doi.org/10.1192/bjp.bp.113.131227
Bernstein, D.P, & Fink, L. (1998). Childhood Trauma Questionnaire: A retrospective self-report manual. Psychological Corporation.
Briere, J. (1996). Trauma symptom checklist for children. Psychological Assessment Resources, 00253-8.
Briggs, E. C., Fairbank, J. A., Greeson, J. K. P., Layne, C. M., Steinberg, A. M., Amaya-Jackson, L. M., Ostrowski, S. A.,
Gerrity, E. T., Elmore, D. L., Belcher, H. M. E., & Pynoos, R. S. (2013). Links between child and adolescent trauma exposure and service use histories in a national clinic-referred sample. Psychological Trauma: Theory, Research, Practice, and Policy, 5(2), 101–109. https://doi.org/10.1037/a0027312
Chemtob, C. M., Gudiño, O. G., Luthra, R., Yehuda, R., Schmeidler, J., Auslander, B., … & Panzer, P. G. (2016). Child Trauma Exposure and Posttraumatic Stress Disorder: Identification in Community Mental Health Clinics. Evidence-Based Practice in Child and Adolescent Mental Health, 1(2-3), 103-115. https://doi.org/10.1080/23794925.2016.1227949
Greenwald, R., & Rubin, A. (1999). Assessment of posttraumatic symptoms in children: Development and preliminary validation of parent and child scales. Research on Social Work Practice, 9(1), 61-75. https://doi.org/10.1177/104973159900900105
Kassam‐Adams, N. (2006), The Acute Stress Checklist for Children (ASC‐Kids): Development of a child self‐report measure. Journal of Traumatic Stress, 19, 129-139. https://doi.org/10.1002/jts.20090
Kaufman, J., Birmaher, B., Brent, D., Rao, U. M. A., Flynn, C., Moreci, P., … & Ryan, N. (1997). Schedule for affective disorders and schizophrenia for school-age children-present and lifetime version (K-SADS-PL): initial reliability and validity data. Journal of the American Academy of Child & Adolescent Psychiatry, 36(7), 980-988. https://psycnet.apa.org/doi/10.1097/00004583-199707000-00021
Lawshe, C. H. (1975). A quantitative approach to content validity.” Personnel Psychology, 28(4), 563-575. https://doi.org/10.1111/j.1744-6570.1975.tb01393.x
Matte-Landry, A., Grisé Bolduc, M.-È., Tanguay-Garneau, L., Collin-Vézina, D., & Ouellet-Morin, I. (2023). Cognitive outcomes of children with complex trauma: A systematic review and meta-analyses of longitudinal studies. Trauma, Violence, & Abuse, 24(4), 2743-2757. https://doi.org/10.1177/15248380221111484
Pynoos, R. S. & Steinberg, A. M. (2013). UCLA PTSD Reaction Index for Children/Adolescents-DSM 5. Author. Available from www.reactionindex.com.
Rolon-Arroyo, B., Oosterhoff, B., Layne, C. M., Steinberg, A. M., Pynoos, R. S., & Kaplow, J. B. (2020). The UCLA PTSD Reaction Index for DSM-5 brief form: A screening tool for trauma-exposed youths. Journal of the American Academy of Child & Adolescent Psychiatry, 59(3), 434–443. https://doi.org/10.1016/j.jaac.2019.06.015
Rupert, A. C., & Bartlett, D. E. (2022). The Childhood trauma and attachment gap in speech-language pathology: Practitioner’s knowledge, practice, and needs. American Journal of Speech-Language Pathology (Online), 31(1), 287-302. https://doi.org/10.1044/2021_ajslp-21-00110
Saigh, P.A. (2004). A structured interview for diagnosing Posttraumatic Stress Disorder: Children’s PTSD Inventory. PsychCorp.