Faculty of Social and Behavioural Sciences
Graduate School of Child Development and Education
Running head: EFFECTIVENESS OF TRAUMA TREATMENT
The effectiveness of trauma-treatment
for young children
Dieke Folkeringa (10301488)
Research Master Child Development and Education Thesis 2
Supervisors: Hülya Kosar-Altinyelken (UvA); Judith Ensink (de Bascule) 17 June 2016
1 Abstract
Younger children are most vulnerable to the impact of traumatic events. Trauma treatment may prevent long-term negative effects and may lead to improved well-being. Limited research has been conducted on the effectiveness of trauma treatment for younger children. The current study aims to provide insight on the rate of trauma treatment responsiveness for children under the age of eight, and on what factors are associated with trauma treatment effectiveness. Additionally, it is investigated whether trauma treatment effectiveness can be predicted when background variables are already known. In total, parents of 32 children filled out various questionnaires and were interviewed (DIPA). Several play observations took place to observe the parent-child interaction. Lastly, interviews were conducted with clinicians and parents to provide information on treatment expectations, on whether the researched effective components are used in practice, and whether treatments are combined or tailored to the client. Cross-tabulations,independent-samples Mann-Whitney U tests, and independent t-tests were conducted to compare responders (n = 16) to non-responders (n = 16). Logistic regression was used to explore whether background variables or questionnaire data were able to predict trauma treatment effectiveness. Results showed that significant relationships were found between trauma treatment effectiveness and comorbid externalizing or avoidant behavior. Non-responders were more likely to show externalizing behavior and avoidant behavior. Results of the logistic regression indicate that the main model accounted for 77% of the variance in trauma treatment responsiveness. The interviews revealed that various challenges arise in trauma treatment for young children, that therapists make
treatment adaptations for children in the researched age-group and that opinions vary on the feasibility to stick to the treatment protocols. The results of this explorative study need to be interpreted with caution due to the small sample size and the low power.
2 Acknowledgements
I would like to express my sincere appreciation to those that have contributed to this research. Thank you Hülya Kosar-Altinyelken and Judith Ensink for supervising me in this process and for your continuous guidance and monitoring. I am really grateful for all the valuable
feedback and learning opportunities you provided, for asking critical questions and for answering my questions. Thank you Judith for providing research suggestions when the original research plan turned out not to be feasible to carry out due to the limited time-span. Thank you Hülya for educating me about how to analyze interviews and for sharing your expertise along the way. I feel honored to have had the chance to work with both of you and feel grateful to have learned so much from you. I would like to thank de Bascule for offering me the opportunity to conduct the research at their organization. Lastly, I would like to thank the children, parents and therapists that participated in the current study. Specifically, I would like to thank the therapists and parents whom I interviewed. Thank you for your valuable contribution and for sharing your opinions with me.
3 The effectiveness of trauma-treatment for young children
Between 14% to 72 percent of children reported traumatic events in a range of studies (Alisic, van der Schoot, van Ginkel, & Kleber, 2008; Silverman, et al., 2008; Snyder, et al., 2012). About 0,5-2.2% of these children develops posttraumatic stress disorder (PTSD) and 13.4% of the children suffers from posttraumatic stress symptoms (PTSS) (Alisic, et al., 2008; Copeland, Keeler, Angold, & Costello, 2007; Perkonigg, Kessler, Storz, & Wittchen, 2000).
Children are most vulnerable to the impact of traumatic events due to a threat on stagnating social-emotional and cognitive development, changes in brain development after experiencing trauma, and dependence on caregivers (Beer, & de Roos, 2004). Young children are at even greater risk of adverse psychological outcomes than older children due to limited coping skills, limited language skills to encode and describe the event or reactions, strong caregiver dependence, and due to undergoing a rapid period of emotional and physiological development (Barnett, & Hamblen, 2015; De Young, Kenardy, & Cobman, 2011).
Treatment of PTSD is crucial to prevent long-term negative effects such as changes in brain development (e.g. inhibited growth and development of the hippocampus and
neocortex), impairment in social and educational functioning, and the development of other anxiety, substance-abuse or mood disorders. Additionally, treatment of PTSD may lead to improved well-being (De Young, et al., 2011; Glaser, 2001; Kar, 2011; Senneseth, Alsaker, & Natvig, 2012; Wang, Tsay, & Bond, 2005). Although not much research has been
conducted on trauma treatment for children under the age of eight, some researchers found it to be effective for younger children (Durlak, Fuhrman, & Lampman, 1991; Scheeringa, Weems, Cohen, Amaya-Jackson, & Guthrie, 2011).
More research needs to be conducted to investigate the effectiveness of trauma treatments for younger children. The current study will examine the rate of trauma treatment responsiveness of children younger than 8 years in a Dutch outpatient setting and compares
4 treatment responders to non-responders on background variables and questionnaire data with the aim to provide insight on why children do or do not respond to trauma treatment.
Additionally it is researched whether trauma treatment responsiveness can be predicted when background variables (e.g. type of trauma, behavioral problems) are already known.
This study is of importance to the practical field since it can be investigated whether young children respond to trauma treatment and whether this can be predicted by background variables. An advantage of the current study is that it provides information from various research methods (quantitative and qualitative) and various perspectives: caregivers (questionnaires, interviews), clinicians (interviews), and the researcher (coding the play observation). Additionally, the set-up of the study allows for both opinions as well as behaviors to be taken into account. Combining quantitative and qualitative information will provide more insight into trauma treatment effectiveness than by using solely one method. By interviewing clinicians, it can be investigated whether the researched effective components are used in practice, and whether treatments are tailored to the client. Interviewing parents provides more knowledge on treatment expectations. The results of the study will contribute to improving trauma-care for young children; if more is known about the effectiveness of offered trauma-treatments, clinicians and treatment centres can possibly adapt their treatment options. The current research may also contribute to theories on this subject, for example on how effective trauma treatment is for young children, what factors are significantly related to effective trauma treatment, and on challenges therapists experience. The thesis begins with reviewing theoretical approaches and previous studies on trauma treatment. Topics that will be discussed are comorbid problems, trauma treatment for young children, and
responsiveness to treatment for young children. The methods section that follows explains the types of data, the methodological choices and the sample. Lastly, the quantitative and
5 Theory
DSM-5
In the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) PTSD is included in trauma- and stressor-related disorders. The PTSD diagnostic clusters are: re-experiencing (e.g. distressing recollections), avoidance (e.g. avoid thoughts, feelings or conversations), and arousal (e.g. irritability). The duration of the disturbance is more than one month and the disturbance causes distress or impairment in one’s social interactions, capacity to work or other main areas of functioning (American Psychological Association, 2013). A preschool child-specific PTSD diagnosis was added because children perceive, interpret, express emotions and organize memories differently than adults, which impacts their experience of- and reaction to trauma (Scheeringa, Zeanah, Drell, & Larrieu, 1995). Moreover, a dissociative-PTSD subtype was added. Several researchers advocated for this since 13.7%-30% of individuals in their studies reported dissociation (Armour, Karstogt, & Richardson, 2014; Hagan, Hulette, & Lieberman, 2015; Wolf, et al., 2012). Detecting dissociation is of great importance since dissociation has been proposed to be an indicator of impaired information processing during trauma and impairs subsequent processing of trauma memories such that it effects symptoms of PTSD (Kenardy, et al., 2007).
Comorbidity
Traumatized children are frequently diagnosed with comorbid conditions. They may experience dissociation, in which the child disengages from stimuli in the external world in order to conserve energy during stressful experiences. Stressful affects are not experienced in consciousness and the individual’s sense of self and sense of relation to others may become dissociated (Schore, 2013). Chronic trauma exposure may lead to an overreliance on
dissociation as a coping mechanism that can aggravate difficulties with affect regulation, behavioral management, and self-concept. Dissociation places a child at risk for internalizing
6 and attachment problems, learning difficulties, victimization, and trauma exposure (Cook, et al., 2005; Hagan, et al. 2015).
Moreover, children diagnosed with PTSD are more likely to have comorbid
depression, anxiety, ADHD, and ODD (Cohen, A., Berliner, & Mannarino, 2010; De Young, Kenardy, Cobham, & Kimble, 2012). PTSD may also be related to relationship disturbance through dissociation. This is because dissociation may lead to asynchrony in- and withdrawal from interaction. Additionally, arousal could contribute to an agitated response to negative or ambiguous partner behavior (Vickerman, & Margolin, 2007; Wekerle, et al., 2001).
Trauma may also lead to interpersonal difficulties and disturbances in child-caregiver attachment, possibly due to reduced trust in the caregiver, withdrawal of affection, or child emotional dysregulations or behavioral changes (De Young, et al., 2011). However, attachment styles also interact with trauma reactions and coping styles of the child and caregiver (Barnett, & Hamblen, 2015). When traumatized, a child with poor attachment is more likely to show behavioral problems and is more difficult to soothe which may lead to more caregiver frustration (Barnett, & Hamblen, 2015). Secure attachment could be
complicated when the caregiver serves as a trigger to the child’s PTSS. Moreover, possible caregiver trauma symptoms might interfere with being able to attend sensitively to one’s child (Barnett, & Hamblen, 2015). Severely disrupted attachment is associated with lifelong risk of physical disease and psychosocial dysfunction due to increased susceptibility to stress, inability to regulate emotions without external assistance, and altered help-seeking (Brenning, & Braet, 2013; Cook, et al., 2005).
Child-caregiver attachment has an important role in regulating stress hormone production. Children who experience a secure relationship have a more controlled stress hormone reaction when upset or frightened. However, children with disorganized or insecure relationships demonstrate higher stress hormone levels. The increased incidence of elevated
7 cortisol levels may alter the development of brain circuits in ways that make some children less capable of coping with stress (Loman, & Gunnar, 2010; Milot, Ethier, St-Laurent, & Provost, 2010; National Scientific Council on the Developing Child, 2005; Schore, 2013).
Trauma treatments
Both trauma-focused cognitive behavioral therapy (TF-CBT) and eye movement desensitization reprocessing (EMDR) have been found to be effective in treating trauma-symptoms for older children and adults (Acarturk, et al., 2015; Chemtob, Nakashima, & Carlson, 2002; Chen, Zhang, & Liang, 2015; Cohen, Deblinger, Mannarino, & Steer, 2004; Jaberghaderi, Greenwald, Rubin, Zand, & Dolatabadi, 2004; Rodenburg, Benjamin, de Roos, Meijer, & Stams, 2009; Van den Hout, Muris, Salemink, & Kindt, 2001; Verlinden, & Lindauer, 2015; Weiner, Schneider, & Lyons, 2009).
CBT is based on the assumption that maladaptive or irrational beliefs and attitudes, cognitive thoughts and images and processing influence problem behavior. The therapist’s aim is to help children to identify maladaptive cognitions, reality-test them, and challenge irrational thoughts or teach new coping skills and hence promote control and self-efficacy (Grave, & Blissett, 2004; Cohen, & Mannarino, 2008). Stepped Care TF-CBT (SC-TF-CBT) is an adapted version of TF-CBT and aims to address barriers to treatment (Bringewatt, & Gershoff, 2010). SC-TF-CBT is designed to be accessible (limited in-office sessions), cost-effective (< therapist time), efficient (no full treatment for early responders) and parent-led (Salloum, et al., 2014; Salloum, Scheeringa, Cohen, & Storch, 2015).
At least two main theories exist on the working mechanisms of EMDR. Some
researchers state that EMDR is guided by the Adaptive Information Processing (AIP) model, and propose that PTSD symptoms result from unprocessed information of the traumatic memory to an adaptive state (Shapiro, & Maxfield, 2002; Van Etten, & Taylor, 1998). Following the AIP model, the goal of EMDR is to process distressing memories, reducing
8 their persistent effects and allowing clients to develop more adaptive coping mechanisms. This is done by having clients recall distressing images while receiving bilateral sensory input (e.g. side to side eye-movements). Dual-attention stimuli are believed to enhance information processing (Shapiro, et al., 2002). Other researchers state that EMDR is guided by the working memory theory which proposes that one’s working memory is able to conduct several tasks at the same time but has limited attention capacity. Hence, by executing one task, the performance on another task is strained (Maxfield, Melnyk, & Hayman, 2008). Side to side eye-movements may function as a distraction from focusing only on the traumatic memory. The created distance causes desensitization of the memory; the memory loses its emotional component and vividness as it is transferred to long-term memory (Engelhard, van den Hout, Janssen, & van der Beek, 2010; Gunter, & Bodner 2009; Maxfield, et al., 2008).
Another treatment that is used to treat trauma-symptoms in children is Words and Pictures, which consists of creating an illustrated narrative for children, that helps them understand events that are difficult for adults around them to talk about. The emphasis is on parents and therapist co-constructing a narrative that uses the family’s own words to describe ‘How did things get to be the way they are?’ The treatment engenders hope and focuses on potential family strengths (Hiles, Luger, Fox, & Essex, 2008; Turnell, & Essex, 2006).
Effectiveness of trauma treatments for younger children
Although not much research has been conducted on trauma treatment for younger children (< age 8) some researchers found trauma treatment to be effective for children in this age-group. Durlak and colleagues (1991) found that the pre-occupational group (age 5-7) and the concrete operational group (age 7-11) did not differ significantly in effect size in CBT. Puff and Renk (2015) demonstrated in a case study of a 5-year-old boy with PTSD that CBT was effective in treating PTSD symptoms. Scheeringa and colleagues (2011) found TF-CBT is feasible for 3-6 year old children with PTSD symptoms (n= 32 waitlist versus n= 32
TF-9 CBT). Salloum and colleagues (2015, aheadofprint) compared SC-TF-CBT to TF-CBT in a sample of 53 children and found no significant differences on PTSS severity, internalizing symptoms, changes in diagnostic status, treatment response, and remission.
The EMDR technique is also adapted in order to be better-suited for children (Ahmad, & Sundelin-Wahlsten, 2008; Lovett, 1999; Schlattmann, 2006). Several studies have been conducted on overlapping age groups (< age 8 as well as well as > age 8). Kemp, Drummond, & McDermott (2009) found 4 sessions of EMDR treatment (n = 13) to be effective in
comparison to a six-week wait-list control condition (n = 14) for children (age 6 - 11) who experienced motor vehicle accidents. Ahmad, Larsson, & Sundelin-Wahlsten (2007) found EMDR (n = 17) to be effective in reducing PTSD symptoms for children (age 6-16) when compared to waitlist condition (n = 16). De Roos, et al., (2011) conducted a RCT in which 52 children (age 4-18) were randomly assigned to either CBT (n = 26) or EMDR (n = 26) after an explosion of a fireworks factory. 23% were aged 4-6 years, 46% fell within the 7-12 years age-range, and 31% were aged 13-18. Children received up to four individual treatment sessions. Both treatments reduced PTSD scores; the results were maintained at three-month follow-up. With EMDR, fewer sessions were required to reach treatment gains.
Responsiveness to treatment
Several challenges may arise in the trauma treatment of young children. First of all, young children are extremely vulnerable due to dependence on caregivers, which may impact trauma treatment. Due to the young age of the child, parents are frequently involved in trauma treatment. Possible disturbances in child-caregiver attachment may hence be a challenge for treatment. Dissociation may also inhibit treatment from succeeding, due to possible cognitive difficulties and neuropsychological impairments, or because activation of the fear network is required in order to change maladaptive beliefs (Resick, Suvak, Johnides, Mitchell, & Iverson, 2012). Additionally, when someone is detached from the traumatic
10 experience(s) and its associated feelings, maladaptive beliefs cannot be changed. It is
important to establish a safe environment for therapy, since the safety of the therapeutic environment allows for increase in trust and reduced withdrawal and sensory distortion (Kreidler, Zupancic, Bell, & Longo, 2009). According to Grave and Blissett (2004), CBT may not be effective for young children since young children do not infer internal causation yet, which is fundamental in CBT since a child needs to understand the causal connection between cognitions and behaviors in order for CBT to be successful. Moreover, children can vary widely in their ability to sit still, focus, manage their impulses, cooperate, comprehend vocabulary or abstract ideas, and express themselves in words (Wesselmann, et al., 2012). Externalizing child behavior may also be a challenge for treatment, since externalizing child behavior is associated with more therapeutic relationship problems and higher drop-out from therapy when compared to children with internalizing problems. Problems with establishing a positive therapeutic relationship may especially arise in early stages of treatment, when psychopathology has not fully been addressed yet (Garcia, & Weisz, 2002; Zorzella, Muller, & Cribbie, 2015).
Although challenges may arise in treating young children, some researchers claim that trauma treatments are needed in order for changes in the brain to occur necessary to develop appropriately afterwards. Hence, well-functioning brain systems that respond to stress appropriately are essential to survival (National Scientific Council on the Developing Child, 2005; Perry, 2009). Zantvoord, Diehle, and Lindauer (2013) stated that TF-CBT influences brain regions involved in fear conditioning, extinction learning and possibly working memory and attention regulation. Additionally, effective EMDR treatment may lead to changes in brain activation patterns and facilitate information processing. Stickgold (2002) proposed that EMDR treatment reverses the breakdown of normal memory processing that initially leads to the development of PTSD. The reorienting of attention activates brain mechanisms (anterior
11 cingulate, the superior colliculus) that permit changes in memory processing required to integrate traumatic memories and moderate the experience of real versus perceived threat (Levin, Lazrove, & van der Kolk, 1999; van der Kolk, Burbridge, & Suzuki, 1997).
According to De Vries, Peak and Lawson (2008) an advantage of TF-CBT is that it is flexible and open for creativity, and children are allowed to learn techniques to deal with stress before dealing with their own personal trauma. Cohen and colleagues stated that eight sessions is the minimal number of CBT treatment sessions (Cohen, Mannarino, Kliethermes, & Murray, 2012). Several researchers have found that fewer EMDR sessions are needed compared to TF-CBT for achieving similar treatment results (Adler-Tapia, & Settle, 2009; De Roos, et al., 2011; Rodenburg et al., 2009). Van der Kolk and colleagues (2007) found 8 sessions of EMDR to be sufficient in traumatized people (Van der Kolk, et al., 2007). In many studies, 3 to 8 EMDR sessions have been found to be effective (Chemtomb, et al., 2002; Jaberghaderi, et al., 2004; Soberman, Greenwald, Rule, 2002). Diehle and colleagues (2015) have, in consultation with A. Mannarino, adapted the original 12 sessions version of the TF-CBT protocol and fitted the modules into eight sessions. They found 8 treatment sessions
(EMDR/TF-CBT) to be successful in treating trauma symptoms for children above the age of eight. Since this study is conducted in the same outpatient facility and was found to be
effective, the number of eight sessions will be used to classify treatment responders (≤ 8 sessions) and non-responders (≥ 8 sessions).
The current study
The main research questions that guide the current study are: (1) What is the rate of trauma treatment responsiveness for children under the age of eight in a Dutch outpatient setting? (2) Do treatment responders and non-responders differ significantly on background variables and child behavior and how can these differences be explained? (3) Could trauma treatment responsiveness be predicted when significant differences between responders and
12 non-responders on background variables and child behavior are already known? It is expected that dissociation occurs significantly more often amongst treatment non-responders in
comparison to treatment responders. Based on the fact that many researchers state the high incidence of dissociation in young children and that dissociation is included in PTSD measures, it may be that a dissociation PTSD profile will arise in this study. Moreover significantly more child externalizing behavior is expected amongst treatment non-responders. Lastly, a positive parent-child relationship is expected to be associated with a higher chance on successfully completing trauma treatment.
Method Procedure
Children were recruited at the trauma center of the department of child and adolescent psychiatry, de Bascule, of the Academic Medical Centre in Amsterdam. Recruitment took place from January 2014 to January 2016. Quantitative as well as qualitative data were gathered. As standard procedure at the Bascule, parents fill out the 1) Child Behavior
Checklist (CBCL), 2) Children’s Revised Impact of Event Scale (CRIES-OV), and 3) Trauma Symptom Checklist for Young Children (TSCYC) questionnaires, 4) are interviewed
(Diagnostic Infant and Preschool Assessment (DIPA)), and 5) are observed during a play-observation between caregiver and child. Parents or the legal guardian sign(s) informed consent in order for this information to be used for research purposes. In addition to the measures that are already used, the main researcher developed two brief qualitative interviews to interview parents and therapists before the start of treatment.
As standard procedure a meeting takes places at the end of the diagnostics phase, in which the therapist advices parent(s) and child on the next steps. In case trauma-treatment is advised, the therapist decides with his or her clinical impression what treatment will take place, depending on what treatment they expected to be better fitting to the specific child (e.g.
13 due to type of trauma or age). Hence, clinical practice did not allow an RCT at the time of the study. In case trauma-treatment was advised, a researcher was present at the advisory meeting to explain the goal of the study and to give an information folder about the study. If the parent(s) and child were willing to participate, an appointment was scheduled for the brief interview with parents and the parent(s) or legal guardian signed the consent forms before the interview took place. Informed consent of the children was not asked considering the age (3 to 8 years). Parents and children did not benefit directly from the research. Participants were allowed to terminate participation or ask researchers to delete the data at any time without giving any reason and without affecting the trauma-treatment as provided to the child.
The data as conducted during this study are strictly confidential. All questionnaire- and interview information were coded anonymously. Only the researchers have access to the data and the encryption key. The data are stored for 15 years at de Bascule. Parents and clinicians were informed of this in the information letter and consent form. The likelihood of incidental findings is very small due to the focused approach on trauma-symptoms. In case of an incidental finding, an independent commission that includes a lawyer, ethicist, laboratory specialist (molecular biologist / geneticist), clinician and clinical geneticist (medical
specialist), will discuss it. Afterwards, relevant incidental findings will be communicated with the general practitioner who will then discuss it with the parents and child.
Participants
The participants of the current study are traumatized children from three to eight years of age, their parent(s) or caregiver(s), and psychologists working at the trauma center of de Bascule. Due to time considerations and the amount of time needed to gather qualitative data, the sample size is 32 children. Children were included if they met the following criteria: age between 3 and 8 years, command of the Dutch language, exposure to at least one single traumatic event, and partial or full post-traumatic stress disorder as reported by the caregiver
14 (DIPA). Due to the clinical nature of the sample and the lack of guidelines regarding the scope of the effect sizes, it was decided not to compute the required sample size a priori. Hence the current study can be classified as an explorative one.
The mean age of the children that participated in the study was 6.21 years (SD = 1.33, min 3.40 – max 7.94) and 55.9% were boys. Although most children (n = 30, 88.2%) had Dutch citizenship, only about one-third (n = 10, 30.3%) were ethnic Dutch, meaning that both parents are Dutch. Most children had at least one parent of a non-Dutch nationality (e.g. Surinamese, (North-)African, or European). Legal custody was either allotted to both parents (41.2%), to the mother (32.4%) or to a legal guardian (20.6%). Two-thirds (64.8%) of the children lived in a foster family and one third lived with one or both biological parents (23.5% with mother, 8.8% both parents, 2.9% with father). The majority of the therapists (n-5, 83.3%) that were interviewed were female.
Measurements
Quantitative data
The main construct that will be measured in the current research is responsiveness to treatment. The included categories were: responders and non-responders. Responders can be defined as children that needed a maximum of 8 treatment sessions; Non-responders needed more than 8 treatment-sessions. Drop-outs were included in the non-responders category. For children in this category treatment had commenced but the children dropped out of treatment due to various reasons (e.g. no stable home situation).
The presence of PTSD symptoms and problem behavior were measured by means of the CRIES-OV, CBCL, TSCYC questionnaires, and the DIPA interview. The DPICS will be used to code parent and child behavior. The Children’s Revised Impact of Event Scale parent version (CRIES-OV, Children and War Foundation, 1998; Olff, 2005) is a 13-item
15 which they have observed the items in their children’s behavior during the past week using a four-point Likert-like scale (0 = not at all, 1 = rarely, 3 = sometimes, 5 = often). The total score indicates the severity of the posttraumatic stress response (Verlinden, et al., 2014).
The Child Behavior Checklist (CBCL, Achenbach, & Rescorla, 2001) is an
informant-report questionnaire assessing behavioral competency and behavioral problems in children within the previous six months. The CBCL/1.5-5 version consist of 99 items, the CBCL/6-18 version consists of 113 items. Eight syndrome scales are assessed: attention problems, anxious/depressed, depressed, somatic complaints, social problems, thought problems, rule-breaking behavior and aggressive behavior. Moreover, six DSM-oriented scales can be assessed: affective problems, anxiety problems, somatic problems, attention deficit/hyperactivity problems, oppositional defiant problems, and conduct problems. Lastly, the questionnaire contains 20 social competency items to obtain parents’ reports of the amount and quality of their child’s social competence. Behaviors are rated on a three-point scale (0-Not true, 1-Somewhat or sometimes true, 2-Very true or often true).
The Trauma Symptom Checklist for Young Children (TSCYC, Briere, et al., 2001) is a 90-item caretaker-report questionnaire that evaluates acute and chronic posttraumatic symptomatology and other psychological consequences of traumatic events in children aged 3-12 years. The TSCYC contains eight clinical scales: anxiety, depression, anger/aggression, PTSS-intrusion, PTSS-avoidance, PTSS-arousal, dissociation, and sexual concerns.
Symptoms are rated on a four-point scale (0-Not at all, 1-Sometimes, 2-Often, 3-Very often) according to how frequently the symptom occurred in the previous month.
The DIPA (Scheeringa, 2010) is a 1-2 hourcaregiver-interview for assessing DSM-IV syndromes in children aged 6 years or younger; hence the normative data are limited to children in this age-group. The interview covers symptoms in thirteen diagnostic categories in modules that each have functional impairment ratings. The categories consist of
16 posttraumatic stress disorder, attention-deficit/hyperactivity disorder, bipolar disorder, major depression disorder, conduct disorder, oppositional defiant disorder, reactive attachment disorder, social phobia, specific phobia, separation anxiety disorder, generalized anxiety disorder, sleep disorder, and obsessive compulsive disorder. Symptoms are linked to the DSM-IV criteria and test-retest for the PTSD module witha sample of 50 young children was found to be high (.87) (Scheeringa, 2010; Scheeringa, & Haslett, 2010).
The Dyadic Parent-Child Interaction Coding System (DPICS, Eyberg, & Robinson, 1983) is a behavioral observation system designed to assess the quality of parent-child social interactions. The DPICS will be used to code parent and child behavior during a 25-minute structured play observation.The coding manual was translated in Dutch (Abrahamse, Niec, Solomon, Junger, & Lindauer, submitted); the third edition of the DPICS was used (Eyberg, McDiarmid-Nelson, Duke, & Boggs, 2005).During the play observation, parent and child are observed in three 5-minute situation that require an increasing degree of parental control. In the first situation, Child-Led Play (CLP), parents are instructed to follow their child’s lead and to play along with the child’s chosen activity. This situation is followed by Parent-Led
Play (PLP), in which parents are instructed to tell their child it is the parent’s turn to choose
the activity and to play along with the parent. The last situation is Clean-Up (CU) in which parents are instructed to tell their child to clean up the toys without parental assistance. The DPICS coding on the parent’s and child’s verbal and nonverbal behaviors (as videotaped) were conducted after the research master student was trained to 80% agreement with an expert coder for all categories. The normative data presented in the DPICS manual are limited to parent-child dyads with children in the 3- to 6-year-old age range. For the present study eight categories were used. Six categories are based on the professional research manual (Eyberg, Nelson, Ginn, Bhuiyan, & Boggs, 2013) and were used by Abrahamse and colleagues (submitted) to study treatment effects of Parent-Child Interaction Therapy (PCIT).
17 The parent categories consisted of the percentage of Positive Following (including Behavior Descriptions, Reflections, Labeled Praises, Unlabeled Praises divided by the total number of parent verbalizations; coded in CLP only), the percentage of Negative Leading (including Commands, and Negative talk divided by the total number of parent verbalizations, coded in CLP only), Demandingness (the sum of Commands, coded in all three situations), Praise (the sum of Praises, coded in all three situations), and the added category percentage of Neutral talk (including Questions and neutral talk, coded in all three situations). It was decided for the Negative Leading category not to include ‘questions’ since in the current study questions were not necessarily perceived as negative (in contrast with the study of Abrahamse and colleagues, in which parents were specifically encouraged not to ask questions). The child categories were Inappropriate Behavior (including Negative talk, Negative Touch, Yell, and Whine, coded in all three situations), percentage of Compliance (coded in PLP and CU only) and child Neutral Talk (including Questions and Prosocial talk, coded in all three situations). It is important to examine the parent-child interaction because attachment styles interact with trauma reactions and coping styles of the child and caregiver (Barnett, & Hamblen, 2015).
Qualitative research
Data were extended with file analyses which were also conducted at de Bascule. The researcher inspected file information (in the User data system) of the participants to find out what type of treatment and how many treatment sessions the children received.
Moreover, two brief interviews were developed and conducted before treatment. The parent-interview consisted of four questions and aimed to give an insight into the treatment expectations. The clinician-interview consisted of seven questions and aimed to give an insight into treatment considerations and treatment expectations. The interviews lasted about 5-10 minutes each; only the researcher and the interviewee were present in a meeting room. Lastly, a short interview was held with each clinician about the treatment protocol which
18 lasted about 20-30 minutes each. During each interview, only the researcher and the therapist were present in the therapists’ room. Therapists’ experience varied (some had over 20 years of experience whereas younger therapists did not).
Analyses
Questionnaire and play observation data (before treatment) were described and compared between responders and non-responders. Analyses were performed in IMB SPSS Statistics (Version 20). The Fisher’s Exact Test was used to test associations between trauma treatment effectiveness and the following categorical variables: type of trauma, gender, cultural background, and current living situation. Subsequently, independent-samples Mann-Whitney U tests were used to compare responders to non-responders on various continuous variables: age, CRIES scores experiencing, avoidance, arousal), TSCYC scores
(re-experiencing, avoidance, arousal, dissociation), DIPA scores (PTSD, dissociation) and CBCL scores (internalizing and externalizing behavior). Additionally, independent t-tests were conducted to compare responders to non-responders on continuous variables (see previous sentence). This test was used because it has higher power than nonparametric tests such as the Mann-Whitney U test, and it is interesting to observe whether the t-test results overlap with those of the nonparametric tests. Effect sizes were computed for all continuous variables (.2 small, .5 medium, .8 large) and categorical variables (.1 small, .3 medium, .5 large).
Subsequently, logistic regression was used to investigate whether trauma treatment responsiveness could be predicted when background variables were already known.
Concerning the qualitative part of the study, data were recorded in written notes. The texts were read for subtracting emerging themes. A list of main codes and subcodes were subtracted from the interviews. The main codes subtracted from the general interviews with clinicians consisted of ‘feasibility to stick to the treatment protocol’, ‘treatment preference’, ‘challenges upon starting treatment’, and ‘combining treatments’. The main codes subtracted
19 from the parent- and clinician interviews on specific children consisted of ‘Need for
stabilization phase’, ‘treatment preference’ and ‘treatment expectations and consent’. Ethical considerations
Informed consent was obtained from parents and also signed by the researcher. The acquired information was coded anonymously in order to assure anonymity and
confidentiality. Participants were allowed to terminate participation or ask researchers to delete the data at any time without giving any reason and without affecting the trauma-treatment as provided to the child.
Results
Quantitative data
File analyses
Descriptive information on background variables as compared between treatment responders versus non-responders at baseline can be found in Table 1. 53.1% of the children were male, and 50% of the children were treatment responders. Of the non-responders two children dropped out of treatment. The majority of the children (84.4%) had experienced chronic trauma. In total, 13 children (40.6%) received TF-CBT, 11 children EMDR (34.4%), 3 children (9.4%) Words & Pictures, and 2 children (6.3%) received a combined treatment.
Table 1
Baseline descriptive statistics; responders versus non-responders on background variables
Responders n = 16 Non-responders n = 16 t DF p r/d Background variables
Sex of foster child n (%) 8 (50.0%a) 7 (43.8% a) . e . e .723 0.06
Type of trauma n (%) 12 (75.0%b) 15 (93.8%b) . e . e .311 0.27
Cultural background n (%) 7 (43.8%c) 3 (18.8% c) . e . e .244 0.30**
Nationality n (%) 14 (87.5%c) 14(87.5%c) . e . e .513 0.20
Living situation n (%) 9 (56.3%d) 12 (75%d) . e . e .264 0.20
Age of foster child M (SD) 6.38 (1.24) 6.26 (1.49) .24 30 .811 0.09
Note: a: female; b: chronic trauma; c: Dutch cultural background; d: with foster parent; e: no value due to the use of Fisher’s
20
Questionnaire data
For descriptive information on child behavior see Table 2. Based on the DIPA, dissociation symptoms were present in 37.5% of the cases. When looking at the relation between background information and questionnaire data and treatment responsiveness, independent-samples Mann-Whitney U tests revealed significant relationships between treatment responsiveness and TSCYC avoidance and CBCL externalizing behavior. Non-responders were more avoidant in comparison to Non-responders. Additionally, in contrast with children who did respond to trauma treatment, non-responders were more likely to show externalizing behavior. The same significant results were found by conducting independent t-tests (see table 2 for the results). After effect sizes were computed for all variables, achieved power was computed, ranging from .08 to 0.81 for the questionnaire data.
Table 2
Baseline descriptive statistics; responders versus non-responders on child behavior variables
Responders n = 16 Non-responders n = 16 n n t DF p r/d Child behavior CRIES Re-experiencing M (SD) 6 11.83 (7.44) 9 8.44 (5.03) 1.06 13 .309 0.53** CRIES Avoidance M (SD) 6 10.67 (6.25) 9 7.11 (6.23) 1.08 13 .299 0.57** CRIES Arousal M (SD) 6 14.83 (4.89) 9 13.00 (4.24) 0.77 13 .453 0.40
CRIES Total score M (SD) 6 37.33 (17.02) 9 28.56 (12.92) 1.14 13 .276 0.58**
TSCYC Re-experiencing M (SD) 10 13.50 (3.44) 5 15.00 (7.31) -0.55 13 .591 0.26 TSCYC Avoidance M (SD) 10 12.80 (1.48) 4 16.00 (2.45) -3.06 12 .010* 1.58*** TSCYC Arousal M (SD) 11 18.55 (3.98) 5 20.00 (3.39) -0.71 14 .492 0.39 TSCYC PTSS total M (SD) 10 45.10 (6.30) 4 47.75 (6.65) -0.70 12 .497 0.40 TSCYC Dissociation M (SD) 10 13.30 (2.98) 4 15.25 (5.38) -0.88 12 .394 0.44 CBCL Internalizing M (SD) 8 60.31 (11.86) 7 67.50 (7.78) -1.36 13 .196 0.72** CBCL Externalizing M (SD) 8 59.81 (8.04) 7 71.86 (8.55) -2.81 13 .015* 1.45*** DIPA PTSS M (SD) 14 0.93 (0.48) 13 0.77 (.44) 0.90 25 .375 0.35 DIPA Dissociation M (SD) 14 0.50 (0.52) 15 0.33 (0.49) 0.89 27 .381 0.39
21
Play observations
In total, 11 play observations were conducted. The play observation results, as compared between responders and non-responders, can be found in Table 3. The descriptive data mostly overlaps with the general descriptive information (See participants). The children that participated in the play observations were on average 6.94 years old (SD = 1.20, min 4.18 – max 7.94) and most of them had Dutch citizenship (90.9%), although of mixed origin (81.2% mixed origin). In the majority of the cases (63.6%), parents had legal custody; in two-third (63.6%) of the cases the child lived in a foster family. Most of the children (n = 9, 81.8%) were boys, whilst most of the caregivers (n = 10, 90.9%) were female. All children that experienced a single traumatic event happened to receive a play observation, leading to a higher average (36.4%) of children having experienced single trauma versus 63.6% of the children having experienced chronic trauma. No significant relationships were found at the 0.05 significance level. After effect sizes were computed for all variables, achieved statistical power was computed, ranging from .07 to .38 for the play observations.
Table 3
DPICS scores treatment responders versus non-responders
Responders n = 4 Non-responders n = 7 t DF p d U Child behavior Compliance (%) 50.36 (26.80) 53.26 (17.71) -0.22 9 .832 0.13 .527 Neutral talk (%) 93.07 (3.34) 91.52 (7.80) 0.37 9 .719 0.26 1.00 Inappropriate behavior M (SD) 5.25 (2.63) 4.71 (7.22) 0.14 9 .892 0.10 .315 Parenting stress Positive Following (%) 15.03 (7.96) 12.02 (5.70) 0.73 9 .481 0.43 .412 Negative Leading (%) 9.11 (4.78) 13.43 (9.96) -0.80 9 .444 0.55** 1.00 Neutral talk (%) 69.96 (8.26) 71.32 (6.77) -0.30 9 .772 0.18 1.00 Praise M (SD) 18.76 (14.41) 12.71 (6.97) 0.95 9 .364 0.53** .648 Demandingness M (SD) 30.75 (18.43) 18.42 (5.62) 1.86 9 .095 0.91*** .240
22
Overall results
In general, achieved power was very low for most variables and high for the two variables (externalizing behavior, avoidance) that were significantly related to trauma
treatment effectiveness during independent t-tests and independent-samples Mann-Whitney U tests. Overall, medium effect sizes were found for cultural background, CRIES
re-experiencing, CRIES avoidance, CBCL internalizing behavior, DPICS Negative leading, and DPICS praise. Large effect sizes were found for TSCYC avoidance, CBCL externalizing behavior, and DPICS Demandingness.
Subsequently, a standard logistic regression analyses was performed to examine whether treatment responsiveness can be predicted if certain background characteristics were already known (see Table 4). For the main model, it was decided to use one block of
predictors. The model comprised the TSCYC avoidance and CBCL externalizing behavior variables (Table 2). The model accounted for about 77% of the variance in trauma treatment responsiveness. Results also showed that the model was able to correctly predict 71.4% of the responders and 100% of the non-responders.
Table 4
Logistic regression for model including demographics (n = 11)
95% CI for odds ratio
Variables B S.E. Wald
Chi-square p Lower Odds ratio Upper Main model Constant -0.56 0.63 0.80 .372 .571 First block CBCL externalizing behavior 0.42 0.35 1.43 .232 .77 1.52 2.99 Second block TSCYC avoidance 3.09 2.83 1.20 .275 .085 22.00 5676
Note: Main model: First block: Nagelkerke R2: .51χ2(4) = 21.213, df=1, p < .05; Second block: Nagelkerke R2: .77; χ2(1) =41.961, df=1, p < .05
23 Qualitative data
General interview with clinicians
In total six interviews were held with the six trauma therapists at de Bascule. Four main themes were subtracted from the interviews: Feasibility to stick to the treatment protocol, treatment preference, challenges for trauma treatment, and combining treatments.
Feasibility to stick to the treatment protocol
Guidelines have been developed for the treatment of posttraumatic stress disorder. EMDR and TF-CBT treatments have found to be effective for the treatment of posttraumatic stress disorder in case the treatment follows the standardized treatment. For young children, a total of 16 TF-CBT sessions is stated in the TF-CBT protocol, no maximum number of sessions is stated for EMDR. Since the protocol for older children state 8 sessions as the total number of sessions, it was asked whether 8 sessions would be sufficient for young children as well. Deviations from the protocol have not been researched and thus cannot be defined as effective. However, in clinical practice there seems to be a debate as to whether the exact protocol has to be followed or whether deviations are allowed to be made in the best interest of the client. Several therapists stated that they sometimes wonder whether the use of a measurement instrument (e.g. questionnaires/protocol) is for research purposes or to help the client; these therapists are in favor of using their clinical experience to decide what would be in the best interest for the client. The discussion on the feasibility to stick to the protocol focused on six major aspects: type of trauma, age of the child, type of treatment, attachment problems, pre-treatment program, and goal of treatment.
Type of trauma
The type of trauma of the child arises as an important denominator for the feasibility to stick to treatment. A distinction should be made between a single trauma and complex trauma. Whereas some children experienced a single traumatic event (e.g. an accident, or
24 being a witness of a shooting) in their lives, others experienced multiple traumatic events (e.g. (witness of) long-term physical and/or psychological abuse in the home situation) over a longer period of time. Most therapists argued that it is much more difficult to stick to the treatment protocol when a child experienced complex trauma. For complex trauma the number of sessions as written down in the protocol may not be sufficient. Also, in case of physical abuse in the home, parents may benefit from more parental guidance (psycho-education about the trauma, how parents can help) than described in the protocol. Most therapists mentioned the presence of a difficult target group (chronic trauma, often less stable home situation) at the specialized care institution where the interviews were held. Therapists argued that in less specialized care, more cases of single trauma may arise. Children are usually referred to specialized care if less specialized care is not sufficient.
Age of child
The age of the child is another important denominator for the feasibility to stick to the protocol. Interestingly different opinions were expressed. Some therapists argued that
younger children have less memories of the event(s) and hence require fewer therapy sessions than older children. Other therapists were of the opinion that more therapy sessions are
required for younger children since they have a shorter attention span and have less
developed cognitions in comparison to older children. For that reason,treatment for younger children may require a more playful approach (e.g. more games) which takes up more time.
Type of treatment
Type of trauma treatment was another important dimension for the feasibility to stick to the number of sessions in the protocol. Most therapists mentioned that the EMDR
treatment requires fewer sessions. One therapist added that with TF-CBT more time is spent on stabilization during the first sessions and remarked that:
25
not just measure trauma treatment, you measure stabilization and some sessions of
trauma treatment versus trauma treatment.’
It must be noted, however, that for some therapists the choice of therapy was also based on the type of trauma (e.g. single trauma: preference EMDR). For this reason, it cannot be extrapolated whether solely the type of treatment is associated with the feasibility of sticking the protocol because for some therapists the type of treatment is based on the type of trauma.
Attachment problems
Attachment problems arose as a challenge for the feasibility to stick to treatment. According to many therapists, the EMDR and TF-CBT treatments do not focus enough on attachment problems of the children and on attachment figures. For that reason, therapists require more time to get the child with attachment difficulties to trust the therapist. Hence, according to some therapists, an extra trajectory is acquired before treatment can take place so that the child can get to know and to trust the therapist.
Pre-treatment program
Most therapists mentioned that a pre-treatment program is associated with the feasibility of the trauma treatment protocol. Several therapists stated that in case a child already received another treatment that is also offered at the same treatment centre, the child may be better prepared for trauma treatment. The multidimensional treatment foster care for preschoolers (MTFC-P) treatment program, for example, offers guidance to caregivers. Additionally, MTFC-P provides training to the child in regulating emotions and aims to motivate them for treatment. Hence, these therapists claim that this overlaps with the first few sessions of TF-CBT and may be the reason why these children need fewer treatment sessions because they are familiar to several concepts already.One therapist wondered:
“Will there be a difference in the number of treatment sessions between children that received the MTFC-P program before treatment versus children that did not?”
26 One therapist claimed that the specific type of treatment should not be advised in case the child is not ready yet for treatment. If it is clear that a longer stabilization phase is
required, then that should take place (and be clearly communicated) before the actual treatment starts. The therapist reasoned that the treatment relationship is based on the trust that the specific treatment is helping. Having a long trajectory ‘as part of treatment’ does not help in achieving trust.
Goal of treatment
Multiple therapists claimed that the goal of treatment is also associated with the feasibility to stick to the treatment. In case the goal of treatment is to treat the worst trauma, it may be possible to stick to the protocol and to fewer number of sessions. However, the same therapists felt that at times it was in the best interest of the child on the long term to use more sessions to also boost the child’s self-esteem, or totruly unravel the trauma. A few therapists stated that they require more treatment sessions than described in the protocols due to personal style or because they are afraid they might miss important information.
One therapist asked:
“How extensive do we make the trauma narratives in TF-CBT? If we incorporate the whole story, it takes more time. In case we keep it short, the trauma narrative may
become a compilation of bits and pieces, and is more like EMDR on the TF-CBT
level. […] Some children that receive short trauma therapy (as in the protocol), came
back to specialized care because it turned out they still had trauma symptoms. […]
Are children that received longer trauma treatment not referred back to specialized
care? Do they have fewer complaints later on in life?
Another therapist asked:
27
do we adapt the therapy to each child and allow for longer trauma narratives in order
to also improve the self-image of the child?”
The quotes above show there is unclarity in therapists’ adaptations to treatment (e.g. more extensive than in the protocol) and on its effects on child outcomes.
Treatment preference
Several therapists have a preference for EMDR treatment for young children, because young children have more fragmented images of the traumatic event(s) than older children. Young children usually do not have the complete story from head to tail so the fragmented images can be used during EMDR and less so in TF-CBT. One therapist wondered whether young children are able to get to the core of the TF-CBT treatment already because she doubted whether young children have sufficient capabilities to distinguish between cognitions. Another therapist stated that young children have limited verbal capabilities which are required for TF-CBT; hence if the therapist purely sticks to the treatment protocol, much information would be missing that is required for the trauma narrative. One therapist stated to have a preference for a combination of treatments (EMDR in combination with storytelling or Words and Pictures). The therapist prefers to complement EMDR by a story based on facts; this story can be made by the caregivers and can provide the child with a story of events that happened in his/her life. The fact that the caregiver can contribute to the story might improve the attachment between the child and caregiver.
Challenges for trauma treatment
Therapists discussed a number of challenges to trauma treatment for young children, such as motivation for treatment, safety of the child, and fitting children into the same straightjacket. Firstly, some therapists stated that motivating young children for treatment is more difficult than older children, since young children may have less insight in why
28 treatment may be important. A therapist mentioned that it is also more difficult in general to motivate the type of clients that come to the trauma centre, due to the type of population. Due to the high incidence of complex trauma, which is sometimes intergenerational and
sometimes stems from the family system, it is harder to motivate the family system for treatment. Adding to this, a therapist stated:
“The protocol shows what aspects to incorporate in the treatment but does not state how to prepare and/or motivate a child and his/her family for treatment. Preparing
them or making them able to show up for treatment takes time. This time is not
included in the protocol.”
Secondly, therapists stated that in order for the treatment to be allowed to take place, the child has to be safe. This is because the child needs to have a stable home situation whilst treatment takes place, since treatment can evoke unpleasant memories and the child needs to have support from caregivers in a safe environment in order to deal with the aspects therapy may evoke. In case there is no safe home situation at the time of treatment, clinicians will not start treatment. This is also a reason why some treatments do not start. Thirdly, therapists
wondered whether it was possible, at all, to fit children into the same straightjacket and whether that straightjacket works for everyone.
“When children do not fit into the protocol, you have to find a different way in an attempt to help them. […] Hence, circumstances force you to adapt your plan.”
To illustrate, some adjustments need to be made for children with additional psychiatric problems. In case a child is diagnosed with for example ADHD, this may cause difficulties in the start-up to treatment and during treatment (e.g. the child has a shorter attention span and is less able to sit still). Hence, some therapists make adjustments to the treatment for children that have accompanying psychiatric problems. It may take more time in general to complete
29 treatment for a child with comorbid problems. Several therapists stated being interested in knowing more about which children do not fit in the protocol and why this may be the case.
Combining treatments
All therapists agreed that combining treatments is a sensitive topic. Some therapists revealed that they at times combined treatments (e.g. TF-CBT with one session of EMDR on the most vivid ‘image’; TF-CBT; EMDR in combination with storytelling/Words & Pictures). In contrast, other therapists stated very clearly that they are against the combination of
treatment (e.g. two psychotherapeutic interventions). The latter advocated that when a therapist is of the opinion that several treatments could take place, it should be argued which of the treatments has priority and should take place first. Within the modules of one treatment there is room to make minor adjustments (e.g. spend an extra session on emotions or
cognitions). These therapists argued that one treatment should take place at a time because it is otherwise not possible to research what treatment (aspect) is effective.
Additionally, one therapist stated that adding Words & Pictures during TF-CBT may actually interfere with the trauma narrative of TF-CBT since there may be less room for the child to share how he or she experienced certain events if the Words & Pictures story is shared already. A therapist that is not against combining these treatments stated that Words & Pictures merely state the facts (e.g. you were picked up by a woman from the foster care system) and that the trauma narrative can then focus on the experiences of the child.
Clearly, different opinions have arisen about whether or not to mix treatments. One therapist claimed that therapists have different views toward combining treatments because of differences in 1) personal experiences, 2) level of knowledge about treatments, and 3) the type of clients. Some clients already acquired a lot of stabilization in the pre-treatment trajectory and caregivers required tips and tricks (e.g. MTFC-P). Thus, trauma treatment can start immediately and fewer sessions are required for a successful trauma treatment.
30
Interviews on specific treatments of children
In total eight interviews were held with clinicians and three interviews were held with parents. Several topics arose in these interviews.
Treatment preference for specific children
Whereas therapists may have their personal treatment preference, they may also have their treatment preference for a specific child, based on for example the child’s experienced trauma, trauma symptoms, or previous health care experiences.
For most of the children, the therapists had a preference for EMDR treatment for various reasons. Firstly, many of these children experienced single trauma and the therapists felt EMDR treatment was more suitable to treat single trauma and would take less time than TF-CBT (sometimes only 1 or 2 sessions would suffice to treat a few traumatic images). Secondly, EMDR would be more suitable for the short attention-span of young children. Thirdly, EMDR could be useful for treating preverbal trauma, since it focuses less on a complete trauma narrative but more on a few images. Fourthly, in case children had clear images of traumatic events, these images could then be applied during the bilateral
stimulation of EMDR. Lastly, a therapist mentioned to have a preference for EMDR so that a specific form of EMDR, named Attachment EMDR, could be used, in which the therapist also works on attachment problems. During attachment EMDR, caregivers are able to take part in the treatment which may be helpful for the relationship between child and caregiver.
Therapists named several reasons why TF-CBT would not be suited for the same children. For example, the specific child did not need as much time on the regulation of emotions or there were no worries about parenting skills (as are both addressed more in TF-CBT). In cases of preverbal trauma, the children would not be able to talk and draw/ write about the traumatic event(s). Some children are aware of what they have experienced but do not want to talk about it or make a narrative. Moreover, some therapists argued that some
31 children are in general not talkative or do not like to draw. One therapist doubted whether a specific child was able to distinguish between cognitions.
For some other children, therapists had the preference for TF-CBT, because the child was able to draw pictures for the story and that a story would help structure the past events that the child has experienced. A therapists stated that TF-CBT may give the opportunity for the family to be involved in the treatment, and that parents are able to receive separate guidance in case they want to (more so with TF-CBT than with EMDR).
For one child who had experienced preverbal trauma, the therapist preferred Words & Pictures treatment because she thought the treatment could answer the child’s questions about her past and include various people (biological/foster family, youth-care) in the process.
Caregivers did not have a preference for specific trauma treatment for their child. They stated that they have faith in the therapist for deciding what treatment would be most suitable for the child. All therapists started the trauma treatment that they preferred for each specific child.
Treatment expectations and consent
Expectations of caregivers were varied, yet mainly pertained to the improvement in their child’s well-being and healing. Specific expectations mentioned by them included: The child becomes aware of the connection between trauma and the child’s subsequent behavior, the trauma fades to the background, the child has fewer trauma symptoms, learns to trust again, gains improved self-image, and shows more positive behavior. One caregiver stated that he/she expected that the child would be taken seriously by the therapist.
All caregivers agreed with the chosen treatment that their child was about to receive for various reasons: EMDR may work due to the child’s focus on external stimuli; or TF-CBT may not work since the child does not like to talk about the traumatic event. Therapists attempt to get caregivers on board with their treatment choice, since caregivers’ approval is
32 important in order for a treatment to be successful. When deciding what treatment a child is about to receive, therapists aim to include caregiver’s opinions since caregivers can provide valuable information. For instance, when a caregiver informs that the child has vivid images, but does not like to write or talk much about the event, the therapist may opt for EMDR.
One caregiver worried that treatment may be too confronting or may evoke other trauma’s. The same caregiver did not deem treatment to be necessary since no problems were experienced at home. However, the caregiver consented with treatment for the child, since school reported behavior problems (aggression towards other children).
Need for stabilization phase
For about half of the children, the therapists did not deem a stabilization phase as required. Stabilization refers to extra help in case children are not stable enough for trauma treatment immediately. The aim of stabilization can be defined as the establishment of safety in physical, cognitive-behavioral, interpersonal, and social areas of functioning (ter Heide, Mooren, Kleyn, de Jongh, & Kleber, 2011; Herman, 1992). During stabilization, children for example receive interventions aimed at the body (e.g. relaxation exercises) or at cognitive, behavioral, and emotional control (e.g. attention exercises, or emotion regulation exercises) (ter Heide, et al., 2011). The therapists stated several reasons why some children did not require stabilization: the child 1) has a stable home situation and supporting caregivers, 2) is motivated for treatment, 3) is able to talk about the trauma, 4) can name basic emotions, 5) and can make connection between the traumatic event and trauma symptoms.
Several reasons were stated why about half of the children required stabilization, for example because of an instable home situation or child behavior problems (anger,
dissociation). Several children received a long start-up trajectory to treatment (8-10 sessions) since they were not able to talk about the trauma, were initially not motivated for therapy, and needed time to learn to trust the therapist.
33 One child did not require a stabilization phase before treatment but did receive some extra modules prior to Words & Pictures treatment. According to the therapist:
‘[The child] is not instable but very vulnerable. He has anxiety and is insecure. For that reason it was decided to give the child some extra tips and tricks before the
official trauma treatment started.’
Discussion
In the current study the rate of trauma treatment responsiveness was investigated and treatment responders were compared to treatment non-responders on various variables. The first research question aimed to investigate the rate of trauma treatment responsiveness of children under the age of eight in a Dutch outpatient setting. Descriptive statistics of the current study showed that 50% of the children were responders to treatment.
The second research question focused on comparing treatment responders to treatment non-responders on background variables as well as child behavior data (questionnaire data, play observations) and to provide information on why these differences exist (qualitative data). A significant relationship was found between avoidance (TSCYC) and treatment responsiveness whereby non-responders were more avoidant (according to caregivers) in comparison to responders. This result was expected, since avoidance could explain why a child may drop out of treatment or why it takes more time for the treatment to be effective (e.g. to avoid thinking of the trauma) (Resick, et al., 2012). Additionally, a significant relationship was found between externalizing behavior and treatment responsiveness. In contrast with children who did respond to trauma treatment, non-responders were more likely to show externalizing behavior. It may be the case that children with externalizing problem behavior may have more difficulties with the therapeutic relationship. Difficulties may especially arise in early stages of treatment, when psychopathology has not fully been addressed yet. As for the therapist, more time may be required to deal with the behavioral
34 problems before trauma treatment can commence (Garcia, & Weisz, 2002; Zorzella, Muller, & Cribbie, 2015).
The third research question focused on investigating whether trauma treatment responsiveness can be predicted when significant differences between responders and non-responders on background variables and child behavior are already known. Logistic regression showed that in the current study, trauma treatment responsiveness could be predicted by TSCYC avoidance and CBCL externalizing behavior.
Although it was hypothesized that non-responders would show significantly more dissociation in comparison to responders, this turned out not to be the case. However, when comparing the rate of dissociation in the current study (37.5%) to rates that have been reported in previous research (13.7%-30%) it can be stated that a dissociative-PTSD subtype may indeed be common in young children. This result is of importance since dissociation may affect trauma symptoms as well place children at risk for various difficulties (Armour, et al., 2014; Cook, et al., 2005; Hagan et al., 2015; Kenardy, et al., 2007; Wolf, et al. 2012).
Although a less positive caregiver-child interaction was expected for non-responders, no significant differences were found between responders and non-responders on the DPICS scales. This may be due to the small sample size of the play observation data. Perhaps different results would have arisen when type of caregiver (biological versus foster parent) would have been controlled for. The interaction between a child and its biological parent may differ from the interaction between a child and its foster parent. Potentially, children may show more extreme behavior (very resistant versus very obedient) in the interaction with foster parents when compared to the interaction with biological parents. This may be the case because the child may completely resist the placement (e.g. prefers to live with his/her parents) or want to be accepted so much that the child shows more conforming behavior and does not dare to resist for the fear of being rejected (Singer, Doornenbal, & Okma, 2004).
35 By interviewing clinicians and parents, it was investigated whether the researched effective components were used in practice, whether they were found to be effective and whether treatments are combined or tailored to the client. The caregiver-interviews revealed that caregivers did not have a preference for specific trauma treatment for their child and that their treatment expectations pertained to the improvement in their child’s well-being and healing. The interviews with clinicians led to the identification of several themes. Most therapists seemed to have a preference for EMDR treatment for the studied age-group, since they were of the opinion that young children have more fragmented images of the traumatic event(s) than older children, which can be used during the EMDR technique. It turned out that there seems to be a debate as to whether the exact protocol has to be followed and
whether all children should be fit in the same straightjacket or whether deviations are allowed to be made in the best interest of the client, for example based on the age of the child, type of trauma, presence of attachment problems, type of treatment. Interestingly enough, it was found that several therapists tailor the treatment and do not stick strictly to the protocol. They for instance combined components of multiple treatments (TF-CBT and EMDR, or Words & Pictures and EMDR). Several therapists said being interested in knowing more about which children do not fit in the protocol and why this may be the case. Additionally questions arose about the long-term effects of treatment duration (shorter versus longer) and whether children that received more treatment sessions would be referred back to specialized care less and have fewer complaints later on in life when compared to children that received fewer
treatment sessions. This shows it may be interesting for clinical practice when more research is conducted on what children do fit in the treatment protocols and which children do not and to study the long-term effects of a shorter versus longer treatment to provide more insight on the matter which may potentially lead to treatment adaptations for clients with different profiles. Although the therapists at the treatment centre decided which trauma treatment the
