• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Comparison of the Trauma Symptom Checklist for Children, UCLA PTSD Index, and Child Behavior Checklist in children with a trauma history

ProQuest Dissertations and Theses, 2009
Dissertation
Author: Melissa V Broome
Abstract:
The purpose of this dissertation was to study a host of PTSD assessment-related problems in children with a trauma history (N = 110) who were seeking treatment at a community mental health clinic. Exploratory factor analyses using the trauma-related and non-trauma-related subscales on the Child Behavior Checklist (CBCL; Achenbach 1991), UCLA PTSD Index (Pynoos 1998) , and the Trauma Symptom Checklist for Children (TSCC; Briere 1996) were conducted. Results indicated that in children aged 7 to 11, but not in older children aged 12 to 17, the UCLA PTSD Index and the TSCC trauma-related scales formed a trauma factor. The CBCL "trauma" scale did not load onto this trauma factor. Although there were no racial differences on the TSCC "PTS" scale, African-American children were more likely than Caucasian children to have clinical elevations on the UCLA PTSD Index "PTSD overall severity score"; Caucasian children were more likely than African-American children to have clinical elevations on the CBCL "trauma" scale. These differences were partially accounted for by an estimate of household income, however, there continued to be a trend indicating that there were racial differences on clinically significant elevations on the UCLA PTSD Index "PTSD overall severity score" question and the CBCL "trauma" scale. Also, the TSCC "PTS" scale performed significantly above chance and had moderate specificity and high sensitivity when compared with the UCLA PTSD Index "PTSD full diagnosis likely" question. The CBCL "trauma" scale performed significantly above chance and demonstrated moderate specificity and moderate sensitivity when contrasted with the UCLA PTSD Index "PTSD full diagnosis likely" question. However, the TSCC "PTS" scale performed better when compared to the UCLA PTSD Index "PTSD full diagnosis likely" question than the CBCL "trauma" scale did when compared to the UCLA PTSD Index "PTSD full diagnosis likely" question. Lastly, secondary analyses indicated that children in this sample were unlikely to meet DSM-IV criteria for avoidance cluster symptoms. However, African-American children were more likely than Caucasian to have a clinically significant number of avoidance symptoms. These findings indicate that many of the trauma focused instruments appear to adequately, but not ideally, assess for children's PTSD symptoms. Future directions and limitations of this study are discussed.

Broome, Melissa, 2009, UMSL, p. 3 Table of Contents I. Abstract 2

II. Background A. Overview of Childhood Sexual Abuse 5 B. Posttraumatic Stress Disorder (PTSD) in children 6 C. PTSD measures for children 16

III. Study Objectives & Hypotheses 22 A. Definition of ROC Curve Analysis Terms (Table 1) 26

IV. General Methods 27 A. Participants 27 i. Summary of Participant Demographic Characteristics (Table 2) 29 ii. Number/Percent of Children who Identified the Below Traumas as their Primary and Percent of Children who Endorsed Experiencing a Single Trauma vs. Multiple Traumas (Table 3) 31 iii. Means and Standard Deviations for TSCC scales (Table 4) 33 iv. Means and Standard Deviations for CBCL scales (Table 5) 34 v. Means and Standard Deviations for UCLA PTSD Index (Table 6) 34 B. Instruments 35 i. Use of Extrapolated Income and Non-Extrapolated Income in African- American and Caucasian Children (Table 7) 36 C. Procedure 39

V. Results 40 A. Aim 1 41 i. Summary of Exploratory Factor Analysis Results for CBCL, TSCC, and UCLA PTSD Index scales (N=110) (Table 8) 43 ii. Summary of Exploratory Factor Analysis Results for CBCL, TSCC, and UCLA PTSD Index scales, ages 7-11 (N=54) (Table 9) 45 iii. Summary of Exploratory Factor Analysis Results for CBCL, TSCC, and UCLA PTSD Index scales, ages 12-17 (N=56) (Table 10) 47 B. Aim 2 49 i. Clinically Significant and Non-Clinically Significant Elevations on the UCLA PTSD Index “PTSD Full Diagnosis Likely” Question Among African-American and Caucasian Children (Table 11) 50 ii. Clinically Significant and Non-Clinically Significant Elevations on the CBCL “trauma” Scale Among African-American and Caucasian Children (Table 12) 50 iii. Clinically Significant and Non-Clinically Significant Elevations on the TSCC “PTS” Scale Among African-American and Caucasian Children (Table 13) 51 C. Aim 3 52 i. ROC Curve for the TSCC when Compared to the UCLA PTSD Index, with the Diagonal Line Representing Chance Performance of the TSCC

Broome, Melissa, 2009, UMSL, p. 4 (Figure 1) 55 ii. Performance Measures of the TSCC “PTS” to Detect PTSD Status Relative to the UCLA PTSD Index Diagnosis of PTSD (Table 14.1) 56 iii. Performance Measures of the TSCC “PTS” to Detect PTSD Status Relative to the UCLA PTSD Index Diagnosis of PTSD (Table 14.2) 56 iv. Performance Measures of the TSCC “PTS” to Detect PTSD Status Relative to the UCLA PTSD Index Diagnosis of PTSD (Table 14.3) 57 v. ROC Curve for the CBCL “trauma” Scale when Compared to the UCLA PTSD Index, with the Diagonal Line Representing Chance Performance of the CBCL “trauma” Scale (Figure 2) 59 vi. Performance Measure of the CBCL “trauma” Scale to Detect PTSD Status Relative to the UCLA PTSD Index Diagnosis of PTSD (Table 15) 60 vii. Means, Standard Deviations, and z-scores for the TSCC “PTS” Scale, CBCL “trauma” Scale, and UCLA PTSD Index Grouped by PTSD Diagnosis as Determined by the UCLA PTSD Index “Full PTSD Diagnosis Likely” Question (Table 16) 61 D. Secondary Analyses 62 i. Clinically Significant and Non-Clinically Significant Elevations on the UCLA PTSD Index Avoidance Scale Among African-American and Caucasian Children (Table 17) 64

VI. Discussion 64

VII. References 87

Broome, Melissa, 2009, UMSL, p. 5

Comparison of the Trauma Symptom Checklist for Children, UCLA PTSD Index, and Child Behavior Checklist in Sexually Abused Children

Child Sexual Abuse CSA is a serious and pervasive problem (Ackerman, Newton, McPherson, Jones & Dykman, 1998). However, the reported extent and prevalence of CSA varies according to the type of sample studied (e.g., volunteers, college students, sampling across a population of children) and how the data are collected (e.g., in-person interview, questionnaire, telephone interview) (Goldman & Padayachi, 2000). Estimates of the prevalence rate of CSA range from 0% to 19% for males and 11% to 45% for females (Badgely, Allard, McCormick, Proudfoot, Fortin, Ogilvie, Rae-Grant, Gelinas, Pepin & Sutherland, 1984; Finkelhor, Hotaling, Lewis & Smith, 1990; Goldman & Padayachi, 2000). Also, children’s socioeconomic status (SES) is a factor that has been found to influence risk for exposure to traumatic events. Studies have found that children from lower SES categories are more likely to experience various types of trauma, such as CSA, than children from higher SES categories (Duncan, 1996; Pfefferbaum, 1997; Sadowski & Friedrich, 2000; Shaunesey, Cohen, Plummer & Berman, 1993). CSA victims display more psychological problems than non-CSA victims, including; somatic complaints, low self-esteem, aggressive behavior, hyperactivity, delinquent behavior, self-injuring behavior, substance abuse, sexualized behavior, fear, suicidal thoughts and actions, nightmares, and learning problems (Kendall-Tackett, Williams & Finkelhor, 1993). Individuals who experience CSA are also at an increased risk for many types of psychological disorders such as PTSD (Bennet, Hughes, & Luke, 2000; Kendall-Tackett et al., 1993).

Broome, Melissa, 2009, UMSL, p. 6 Posttraumatic Stress Disorder in Children PTSD is a serious and relatively common disorder exhibited by children who have experienced CSA. Studies have found that between 21% to 64% of CSA victims have PTSD (Deblinger, McLeer, Atkins, Ralphe & Foa, 1989; Kiser, Heston, Milsap & Pruitt, 1991; McLeer, Deblinger, Atkins, Foa & Ralphe, 1988; Timmons-Mitchell, Chandler-Holtz, & Semple, 1997). PTSD consists of three clusters of symptoms: reexperiencing, hyperarousal, and avoidance. These symptoms arise after experiencing a traumatic event that included death, possible death, bodily injury, or danger to an individual’s “physical integrity.” Adults’ reaction to the traumatic event must include “intense fear, helplessness, or horror”; children must have either the same type of reaction or “disorganized or agitated behavior” (DSM, Fourth Edition, Text Revision, 2000). The reexperiencing cluster includes: recurrent and intrusive upsetting thoughts and/or remembrances about the event (including repetitive play in children), recurring nightmares about the event (including general nightmares in children), reliving the event, severe distress when exposed to cues that represent the event, and heightened physical reactions to cues that represent the event. The hyperarousal cluster includes: difficulty maintaining sleep, irritability, problems concentrating, hypervigilance, and heightened startle response. The avoidance cluster includes: attempts to avoid emotions and cognitions related to the event, attempts to avoid places and people that are related to the event, failure to remember main details of the event, lessened interest in activities that were previously enjoyed, disinterest in other people, restricted affect, and feeling they have a foreshortened future (DSM, Fourth Edition, Text Revision, 2000). These avoidance symptoms are commonly categorized into two groups: symptoms that directly involve the trauma (i.e., failure to remember main details of the event), and symptoms that involve

Broome, Melissa, 2009, UMSL, p. 7 constriction of affect and the child’s general environment (i.e., disinterest in other people). Research has shown that differences exist in the type of avoidance symptoms and how avoidance symptoms are displayed between children and adults (Scheeringa, Zeanah, Drell & Larrieu, 1995; Scheeringa, Zeanah, Myers & Putnam, 2003; Trickett, Reiffman, Horowitz & Putnam, 1997). One example is that children may tend to avoid reminders of the trauma that are not included in the DSM-IV criteria of emotions, cognitions, places, and people that are related to the event (Coates & Schechter, 2004). Children may also behaviorally exhibit a “constriction of play” instead of voicing a decreased interest in activities that were previously enjoyed (Coates & Schechter, 2004). Also, children may withdraw socially, which could be conceived of as a behavioral example of articulating feelings of disinterest in other people. Children also may display regression in developmental level. This can be conceptualized as an avoidance symptom and may be a behavioral representation of numbing and a diminished participation in important events. (Scheeringa, Peebles, Cook & Zeanah, 2001; Scheeringa, Zeanah, Drell et al., 1995; Scheeringa, Zeanah, Myers et al., 2003) A key point is that it is extraordinarily difficult to asses children’s avoidance symptoms through verbalization, and thus these symptoms must often be assessed through their behaviors (Scheeringa & Zeanah, 1995). Thus, there may be conceptual and practical issues to understanding how children can present the DSM-IV criteria. In order to qualify for a diagnosis of PTSD, children or adults must exhibit at least one reexperiencing cluster symptom, at least three avoidance cluster symptoms, and at least two hyperarousal cluster symptoms. (DSM, Fourth Edition, Text Revision, 2000). Children or adults can have various combinations of symptoms in each of these clusters, and can exhibit uneven distribution of symptoms among clusters (i.e., a child may display every hyperarousal symptom but only the minimum number of avoidance and reexperiencing symptoms). Unfortunately, a

Broome, Melissa, 2009, UMSL, p. 8 significant problem exists with children qualifying for a diagnosis of PTSD based on DSM-IV criteria. Research has shown that it can be difficult for children to meet enough DSM-IV criteria to qualify for this diagnosis, even after undergoing quite severe trauma (Scheeringa, Zeanah, Myers et al., 2003). It has been postulated that this is because DSM-IV criteria for PTSD are not developmentally sensitive, as PTSD was originally conceptualized to classify the difficulties some adult male war veterans had after combat exposure (Scheeringa, Zeanah, Drell et al., 1995; Scheeringa, Zeanah, Myers et al., 2003; Trickett, Reiffman, Horowitz & Putnam, 1997; Yehuda & McFarlane, 1995). Although children do display intense reactions to traumatic stress, applying PTSD criteria that were developed on adult males to children is problematic. PTSD and Children’s current developmental stage and age at onset of trauma The types of trauma symptoms displayed by children, and how these symptoms are manifested, tend to differ markedly from adults who have experienced a trauma. These differences are tied to the verbal capabilities, cognitive abilities, and independence level that are unique to children in various developmental stages. Therefore, examining PTSD symptomatology from a developmental perspective is essential. However, in many studies children’s current developmental stage is not clearly delineated from their age at onset of the trauma. Symptoms that are intended to be associated with age of onset of abuse will be presented, but the symptoms will be the current symptoms the children display. This is problematic because many studies do not indicate if the children are in the same developmental stage they were in when the abuse first occurred. If the children have progressed from one developmental stage to another, it cannot be concluded that their current symptoms are associated with their age at onset of the trauma, but rather that they are a reflection of the symptoms associated with the children’s current developmental stage. Due to the confounding

Broome, Melissa, 2009, UMSL, p. 9 of age of trauma onset and current developmental stage, a clear distinction between these two factors is difficult. Children tend to have several trauma reactions that are not seen as frequently in adults. Children who have relatives with mood disorders are more likely to exhibit mood symptoms than younger children or adults (Perry, 1994). They also are more likely than adults to exhibit “posttraumatic reenactment play” (Coates & Schechter, 2004; Pynoos, 1990). Some literature indicates that after experiencing a traumatic event, children are more likely to exhibit behavioral extremes, such as hostile or withdrawn behavior, than older children (Manly, Jungmeen, Rogosch & Cicchetti, 2001; Pynoos, 1990). Also, it has been suggested that psychological symptoms such as emotional lability and impulsivity, which are hypothesized to be linked to heightened sympathetic nervous system activity, are features of childhood PTSD (Perry, 1994). Additionally, it has been hypothesized that some traumatized children are misdiagnosed with Attention Deficit/Hyperactivity Disorder (ADHD) when these difficulties (attention, aggressive behavior) are better conceptualized as being related to a manifestation of their PTSD hyperarousal (Pynoos, 1990). Another finding is that children’s flashbacks tend to be less complex than adults (Pynoos, 1990). In children, the avoidance symptoms are most likely to be expressed differently from the traditional DSM-IV symptom criteria. Children appear to be more prone to develop restricted affect and to somaticize their symptoms (e.g., have frequent headaches or stomach aches; Pynoos, 1990). Somaticizing might indicate that a child is avoiding thoughts or feelings about the trauma, and instead focusing his or her distress on more concrete, physical ailments. Dissociative symptoms also appear to be more common in children than in adults (Trickett, Reiffman, Horowitz & Putnam, 1997). Thus, it appears that avoidance symptoms are important

Broome, Melissa, 2009, UMSL, p. 10 and fairly common symptoms that occur in children. It also appears that these avoidance symptoms tend to manifest quite differently in children than in adults. PTSD and Duration of CSA Duration of CSA is a factor that potentially impacts the rate of PTSD in children who have been abused; a substantial body of literature remains divided as to whether the duration of CSA produces an incremental impact on children. A study involving 90 children who had suffered CSA, approximately half of whom met criteria for PTSD, found that the duration of abuse was positively correlated with a greater likelihood of meeting criteria for PTSD (Wolfe, Sas & Wekerle, 1994). This study divided duration of abuse into three categories: “isolated”, under one year, and more than one year. Duration of more than one year was the category statistically associated with increased likelihood of developing PTSD. A particular strength of this study was that it used a sample of CSA victims that closely represented typical CSA victims in regards to duration of CSA and whether the perpetrator was in a primary care-giving role. Additionally, in this study the duration of abuse was correlated to frequency of abuse. However, as the authors noted, this study did contain a higher than typical proportion of CSA that involved the perpetrator using, or threatening to use force. Therefore, it is uncertain if the association between duration of abuse and likelihood of developing PTSD would apply to CSA victims that did not experience the use of, or threatened use of force. Furthermore, all children in this study were currently involved in the criminal justice system relating to the CSA, and therefore this additional stressor may have increased the likelihood of these children developing PTSD. A study by Mennen & Meadow (1995) did not find that the duration of CSA was significantly correlated with higher levels of psychopathology such as: depression, anxiety, and self-esteem. The CSA in this study was generally perpetrated by males whom the girls knew

Broome, Melissa, 2009, UMSL, p. 11 (including approximately half in a primary care giving role), and lasted on average for three years (with a standard deviation of three years). These abuse characteristics make a straightforward interpretation of these results somewhat challenging. This type of CSA is not typical, as most CSA is characterized by a much shorter duration and does not involve males in a primary care-giving role (Mennen & Meadow, 1995; Wolfe et al., 1994). Thus, it is difficult to apply these findings about the effects of duration of abuse to children whose CSA markedly differed from the girls in this study. The mean duration of abuse of three years might indicate that the data was skewed towards long abuse duration, possibly obscuring the effects of duration. CSA might be so devastating that the first few instances have a profoundly damaging effect on children, and subsequent instances do not significantly increase the survivors’ distress. Therefore, a study investigating duration not defined in terms of years but rather defined in terms of incidents might be better able to tease out the effects of duration. Therefore, although this study is informative, it is not definitive. In summary, although it appears as if longer duration of CSA is associated with poorer outcomes, this association may only exist for children whose CSA fits certain criteria. Nevertheless, this is an important factor to consider. PTSD and Severity of CSA The severity of the abuse is another factor that contributes to the development and expression of symptoms following trauma. Whether the CSA included intercourse or penetration, whether force was used during the CSA, and the relationship of the child to the perpetrator are three abuse characteristics that are frequently used to assess CSA severity. The literature generally indicates that children who experience CSA that involved penetration suffer more severe psychological distress than children who experience CSA without penetration

Broome, Melissa, 2009, UMSL, p. 12 (Briggs & Joyce, 1997; Mennen, 1995; Mennen & Meadow, 1995). A study involving females found that those whose CSA involved penetration had significantly lower levels of self-esteem as well as higher levels of anxiety and depression than the girls whose CSA did not involve penetration (Mennen & Meadow, 1995). A study of 73 adult female CSA survivors found that when the abuse included sexual intercourse, the women had a significantly higher PTSD score than women who did not experience abuse with sexual intercourse (Briggs & Joyce, 1997). Specifically, these women had higher scores on portions of the measure assessing “dissociation”, “intrusive thoughts”, and “hyperarousal”. However, generalizing these results to children is problematic; this study used only adult females and was measuring current symptomatology, not the symptomatology displayed by these women when they were children. The mean age of participants in this study was 31.5, and the authors defined CSA as abuse taking place under age 16. Therefore, for most participants a substantial amount of time had passed since the CSA. Nonetheless, this study is useful because the findings correspond to several studies using children as participants (e.g., Mennen, 1995; Mennen & Meadow, 1995). Thus, the preponderance of research indicated that there is an association between CSA involving penetration and more serious psychological symptoms. Research has also indicated that severity of psychological distress is associated with the use of force during the abuse (Mennen & Meadow, 1995; Wolfe et al., 1994). A study involving 90 children who had suffered CSA, approximately half of whom met criteria for PTSD, found that the use of force, or being told force would be used was positively correlated with the proportion of children who met criteria for PTSD (Wolfe et al., 1994). Previously Mennen and Meadow (1995) found that this characteristic was associated with higher scores on measures of

Broome, Melissa, 2009, UMSL, p. 13 depression and self-esteem and a non-significant trend for higher scores on a measure of anxiety. Therefore, CSA involving intercourse, penetration, and whether force was employed appear to be associated with the development and expression of symptoms following trauma. Another measure of severity of CSA that has been extensively investigated is the relationship of the child to the perpetrator. Unfortunately, as with many factors related to CSA, a straightforward connection between the relationship of the child to the perpetrator and severity of psychological symptoms does not appear to exist. Although studies have found that CSA perpetrated by a biological father or an adult who took on a father role are associated with more negative outcomes (McLeer et al., 1988), other studies have not found this direct relationship (Mennen, 1995; Mennen & Meadow, 1995). Studies have delineated how abuse related factors (e.g., whether the abuse involved penetration, use of force) appear to be differently associated with more negative outcomes depending in part on whether the CSA perpetrator was the father or not (Mennen, 1995; Mennen & Meadow, 1995). Abuse involving penetration was associated with greater severity of symptoms when the abuse was perpetrated by a biological father or an adult who took on a father type role. However, when the perpetrator did not have this relationship with the child, this association did not exist (Mennen & Meadow, 1995). Additionally, the relationship between use of force and severity of symptoms only existed when the CSA perpetrator was not a biological father or an adult who took on a father type role (Mennen & Meadow, 1995). Therefore, it appears that this factor encompasses too many other variables (e.g., amount of trust in the perpetrator, is the perpetrator responsible for ensuring the child’s safety) to allow an unambiguous interpretation of role of this factor. Although it appears likely that this is a significant factor in determining outcomes after CSA, not being able to cleanly operationally define this factor makes it exceedingly difficult to explain its association

Broome, Melissa, 2009, UMSL, p. 14 with children’s psychological functioning following CSA. Clearly more research is needed to disentangle the critical elements inherent in perpetrator status. Racial differences in PTSD and Psychological Distress Another complicating factor in the development of PTSD and trauma-related symptoms in children is that the rates of psychological distress and disorders appear unequally distributed across races. Minorities who experienced a childhood or adulthood trauma, such as CSA, child abuse, sexual assault, motor vehicle accident, war trauma, natural disasters, serious medical illness, victim of crime, or war trauma, were more likely to develop PTSD in childhood or adulthood than Caucasians, according to a meta-analysis of 77 articles (Brewin, Andrews & Valentine, 2000). This finding held up even after the authors accounted for the fact that the minority group had a lower SES status and therefore increased risk for PTSD. However, a contrasting meta-analysis found that minority children appeared to be at a lower risk for developing psychological disorders and distress than Caucasian children after accounting for SES (Samaan, 1998). Several notable differences exist between these two meta-analyses. Only individuals who had experienced a trauma were included in the Brewin et al. study, whereas the articles used in the Samaan study included both clinical and community samples. Factors associated with the development and expression of PTSD were the focus of the former study, and race and SES status as risk factors for general pathology were the focus of the latter study. Lastly, significantly more articles were used in the Brewin et al. study than the Samaan study, thus increasing the power of the Brewin et al. study and possibly helping ensure this sample was more representative of individuals who experience a trauma than the Samaan study. Therefore, given the focus and scope of the two meta-analyses, it is likely that the former study provides more valuable and specific information about trauma-related symptoms than the latter study.

Broome, Melissa, 2009, UMSL, p. 15 In addition to the above meta-analyses, several individual studies provide illumination into the possible role that race has in trauma-related psychological distress in children. Physically abused minority children also appear to be at higher risk for developing depressive, anxious, and somatic symptomotology than Caucasian children (Lau, Huang, Garland, McCabe, Yeh, & Hough, 2006). African-American, Asian, and Latino children have also shown a trend for higher depression scores on the Children’s Depression Inventory (CDI) than Caucasian children (Siegel, Aneshensel, Taub, Cantwell & Driscoll, 1998). The possible increased risk for psychological distress among minority children is complicated by the possibility that some of the instruments used to measure psychological distress, such as the CBCL, might be biased to over-pathologize minority children (McCarty & McMahon, 2003; Sandberg et al., 1991). The above studies should be viewed with caution, as several studies have found that nonclinical minority children tend to have significantly higher elevations on the CBCL, including the depression scales, than nonclinical Caucasian children (McCarty & McMahon, 2003; Sandberg, Meyer-Balhburg & Yager, 1991). Thus, it appears that the CBCL might be constructed in such a way that minority children would have higher elevations than Caucasian children, even if both groups were experiencing an equal level of symptoms. Therefore, it cannot be conclusively determined that the traumatized minority children truly were experiencing more distress than the traumatized Caucasian children, as the measures used to assess their symptoms might also tend to over-pathologize minority children, as the CBCL is suggested to do. Therefore, although the preponderance of research indicates minority children are more likely to have elevated scores on measures assessing for psychological distress, there are conflicting findings. Given the paucity of trauma specific literature, more research should be conducted to determine possible differences in trauma

Broome, Melissa, 2009, UMSL, p. 16 reactions among minority and Caucasian children. This research should be conducted with attention to factors such as: SES, duration of trauma, severity of trauma and possible instrument biases. Measures used to assess PTSD in children As the above literature review illustrates, PTSD expression is quite varied; a number of complex and interacting factors influence the development and manifestation of these symptoms. Given this varied expression of PTSD in children, and the significant number of children who experience PTSD symptoms following a trauma, it is imperative that the measurement tools used to assess PTSD accurately assess the expression of PTSD symptoms in children. Three commonly used measures to assess PTSD symptoms in children are the Trauma Symptom Checklist for Children (TSCC), the UCLA PTSD Index, and the Child Behavior Checklist (CBCL). A widely noted difficulty that exists across these instruments is that measuring internalizing symptoms (i.e., avoidance cluster symptoms) in children is a difficult task (Kurt & Merrell, 1998). This is largely due to children’s limited verbal ability which hinders their ability to communicate their thoughts and feelings. Self-report measures, such as the TSCC and the UCLA PTSD Index, are generally used to elucidate the internalizing difficulties children are experiencing. (Kurt & Merrell, 1998) However, the use of self-report measures can be problematic because they rely on the children to possess the ability to comprehend their thoughts and feelings as well as express them accurately (Kurt & Merrell, 1998; Sadowski & Friedrich, 2000). UCLA PTSD Reaction Index The UCLA PTSD Index for the DSM-IV is a self-report measure for children age 8-18 based on DSM-IV criteria. This measure has questions that first assess for the presence of a

Broome, Melissa, 2009, UMSL, p. 17 trauma and the child’s reaction to the trauma, then specifically assess for DSM-IV PTSD symptoms (Steinberg, Brymer, Decker & Pynoos, 2004). It is frequently used to assess for trauma-related symptoms in children in the United States and abroad who have experienced a variety of traumatic events, and can be used to determine if a diagnosis of PTSD is probable as it maps directly onto DSM-IV PTSD criteria (Steinberg et al., 2004). However, a critique of this measure is that children express PTSD symptoms that can differ quite markedly both from adults and from current DSM criteria, which is based on adult presentation of the disorder (Coates & Schechter, 2004; Manly et al., 2001; Perry, 1994; Pynoos, 1990; Scheeringa, Peebles et al., 2001; Scheeringa, Zeanah, Drell et al., 1995; Scheeringa, Zeanah, Myers et al., 2003; Trickett et al., 1997; Yehuda & McFarlane, 1995). Therefore, measures that are strictly based on current DSM criteria may not be especially helpful or valid measures of PTSD in children. These measures may not cover the variety of presentations of psychological distress manifested by children after experiencing a traumatic event. Additionally, convergent validity of this instrument was measured using other child PTSD measures that also rely heavily on current DSM criteria. Although convergent validity was high, this is not necessarily an indication that this measure is truly assessing the full spectrum of PTSD presentation in children. Another possible problem with this measure is that it might be relatively insensitive to avoidance symptoms, as children may not possess the necessary cognitive and verbal skills to accurately endorse DSM-IV based avoidance questions (Scheeringa & Zeanah, 1995). Some examples of this possible endorsement problem are the statements, “I have trouble feeling happiness or love” and “I have trouble feeling sadness or anger”, assessing for the avoidance symptom of “restricted range of affect” (DSM, Fourth Edition, Text Revision, 2000). Given children’s cognitive and verbal limitations, it might be difficult for them to realize

Full document contains 95 pages
Abstract: The purpose of this dissertation was to study a host of PTSD assessment-related problems in children with a trauma history (N = 110) who were seeking treatment at a community mental health clinic. Exploratory factor analyses using the trauma-related and non-trauma-related subscales on the Child Behavior Checklist (CBCL; Achenbach 1991), UCLA PTSD Index (Pynoos 1998) , and the Trauma Symptom Checklist for Children (TSCC; Briere 1996) were conducted. Results indicated that in children aged 7 to 11, but not in older children aged 12 to 17, the UCLA PTSD Index and the TSCC trauma-related scales formed a trauma factor. The CBCL "trauma" scale did not load onto this trauma factor. Although there were no racial differences on the TSCC "PTS" scale, African-American children were more likely than Caucasian children to have clinical elevations on the UCLA PTSD Index "PTSD overall severity score"; Caucasian children were more likely than African-American children to have clinical elevations on the CBCL "trauma" scale. These differences were partially accounted for by an estimate of household income, however, there continued to be a trend indicating that there were racial differences on clinically significant elevations on the UCLA PTSD Index "PTSD overall severity score" question and the CBCL "trauma" scale. Also, the TSCC "PTS" scale performed significantly above chance and had moderate specificity and high sensitivity when compared with the UCLA PTSD Index "PTSD full diagnosis likely" question. The CBCL "trauma" scale performed significantly above chance and demonstrated moderate specificity and moderate sensitivity when contrasted with the UCLA PTSD Index "PTSD full diagnosis likely" question. However, the TSCC "PTS" scale performed better when compared to the UCLA PTSD Index "PTSD full diagnosis likely" question than the CBCL "trauma" scale did when compared to the UCLA PTSD Index "PTSD full diagnosis likely" question. Lastly, secondary analyses indicated that children in this sample were unlikely to meet DSM-IV criteria for avoidance cluster symptoms. However, African-American children were more likely than Caucasian to have a clinically significant number of avoidance symptoms. These findings indicate that many of the trauma focused instruments appear to adequately, but not ideally, assess for children's PTSD symptoms. Future directions and limitations of this study are discussed.