Chi20418 811.819

A Pilot Randomized Controlled Trial of Combined Trauma-Focused CBT and Sertraline for Childhood JUDITH A. COHEN, M.D., ANTHONY P. MANNARINO, PH.D., JAMES M. PEREL, PH.D., Objective: To examine the potential benefits of adding a selective serotonin reuptake inhibitor, sertraline, versus placebo, to trauma-focused cognitive-behavioral therapy (TF-CBT) for improving posttraumatic stress disorder and related psychological symptoms in children who have experienced sexual abuse. Method: Twenty-four 10- to 17-year-old female children and adolescents and their primary caretakers were randomly assigned to receive TF-CBT + sertraline or TF-CBT + placebo for 12 weeks. Results: Both groups experienced significant improvement in posttraumatic stress disorder and other clinical outcomes from pre- to posttreatment with no significant group  time differences between groups except in Child Global Assessment Scale ratings, which favored the TF-CBT + sertraline group. Conclusions: Only minimal evidence suggests a benefit to adding sertraline to TF-CBT. A drawback of adding sertraline was determining whether TF-CBT or sertraline caused clinical improvement for children with comorbid depression. Current evidence therefore supports an initial trial of TF-CBT or other evidence-supported psychotherapy for most children with PTSD symptoms before adding medication. J. Am. Acad. Child Adolesc. Psychiatry, 2007;46(7):811Y819. Key Words: cognitive- behavioral therapy, posttraumatic stress disorder, randomized trial, sertraline.
Posttraumatic stress disorder (PTSD) is a potentially the full disorder (Carrion et al., 2002). Left untreated, serious mental health condition that may develop PTSD can lead to difficulties that last into adulthood, following traumatic events such as childhood sexual such as increased suicide attempts, hospitalizations, abuse. Children who have significant PTSD symptoms substance abuse, depression, and dissociation (Warshaw without meeting full diagnostic criteria have been shown et al., 1993). For survivors of child abuse, these to have comparable functional impairment to those with problems are associated with increased use of healthcare resources (Walker et al., 1999). To minimizesuffering and prevent these long-term deleterious out- comes, it is important to identify and test a range of Drs. Cohen and Mannarino and Ms. Staron are affiliated with the Drexel optimal treatments for childhood PTSD symptoms.
University College of Medicine, Allegheny General Hospital, Department of Trauma-focused cognitive-behavioral therapy (TF- Psychiatry, Pittsburgh; and Dr. Perel is with the University of Pittsburgh Schoolof Medicine, Western Psychiatric Institute and Clinic, Pittsburgh.
CBT) is recognized as an effective treatment for This project was funded by an Independent Scientist grant from the National childhood PTSD symptoms (Cohen et al., 2000; Institute of Mental Health (K02MH01938) to Dr. Cohen. Sertraline tablets Saunders et al., 2004). In six randomized, controlled and identical pill placebo were donated to this project by Pfizer. The authorsthank the project therapists, research coordinator, independent evaluator, and trials (RCTs) for sexually abused and multiply consultants for their assistance, as well as Jeffrey Bridge, Ph.D., Rachel San traumatized children (reviewed in Cohen et al., Pedro, M.S.W., and Ann Marie Kotlik for their many contributions to this 2004), TF-CBT was superior to comparison or control project. Most of all, they thank the children and parents who participated in thisproject.
conditions in decreasing PTSD symptoms as well as Correspondence to Dr. Judith A. Cohen, Four Allegheny Center, Eighth Floor, improving depression, anxiety, and externalizing behav- Pittsburgh, PA 15212; e-mail: jcohen1@wpahs.org.
ioral problems. In the largest study of TF-CBT to date 0890-8567/07/4607-0811Ó2007 by the American Academy of Child (Cohen et al., 2004), 229 sexually abused children ages 8Y14 and their nonoffending parents were randomly J. A M. A CA D. CHILD ADOLE SC . P SYC HIA TRY, 46: 7, JULY 2007 Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
assigned to receive 12 sessions of TF-CBT or child- The study was thus reconfigured to assess the potential centered therapy (CCT). Outcomes with TF-CBT were benefits and risks of adding an SSRI, sertraline, versus superior for almost all measures, with effect sizes placebo to TF-CBT in a double-blind RCT design. Our ranging from medium to large between the two groups primary questions were whether adding sertraline to TF- on the three PTSD clusters and total PTSD symptoms.
CBT could significantly improve children_s PTSD Despite this success, 21% of the TF-CBT group symptoms or global functioning. Our second question continued to carry a diagnosis of PTSD. Although this was whether the children who received sertraline would compared favorably to the CCT group, these findings experience more frequent and/or more severe side effects.
suggest that for a minority of children, supplementing Exploratory questions were whether the addition of TF-CBT with additional interventions may be neces- sertraline produced a more favorable (i.e., faster) response in PTSD symptoms during the course of treatment, and Several RCTs have demonstrated that pharmacolo- whether the addition of sertraline produced differential gical interventions alone, particularly selective seroto- beneficial responses in terms of other symptoms typically nin reuptake inhibitors (SSRIs), can successfully lead to assessed in sexually abused cohorts.
remission in adult PTSD diagnosis (Brady et al., 2000;Davidson et al., 1996; van der Kolk et al., 1994).
Potential mechanisms for this response include thatsertraline indirectly stimulates the 5-HT1A receptor and thus would be expected to be anxiolytic (Schreiber et al., Funding for this project was received in April 2001 with subject 1998) and that sertraline has been shown to stimulate recruitment beginning in January 2002. The project proposed to the CNS neurosteroid/GABAa system (Griffin and recruit a final cohort of 80 subjects with 40 in each treatment Mellon, 1999). No RCTs have yet been published for condition (TF-CBT + sertraline versus TF-CBT + placebo).
pharmacological treatments of childhood PTSD.
Information became available in 2003 that children may beexperiencing increased suicidal ideation while taking SSRIs The typical design for a pharmacological RCT is to (Mitka, 2003). Public warnings about antidepressant medication randomly assign participants to medication or placebo were issued by the U.S. Food and Drug Administration in 2003, in a double-blind fashion, such that some children and the eventual Bblack box[ warnings were included on SSRImedications in 2004 (U.S. Food and Drug Administration, 2004).
receive only placebo intervention. Concerns could be Families we were attempting to recruit for the present study were raised about comparing a medication condition to a provided with updated information as it was issued on Web sites by placebo condition when addressing a potentially serious the U.S. Food and Drug Administration and the AmericanAcademy of Child and Adolescent Psychiatry with regard to disorder such as PTSD, particularly when an effective potential side effects of sertraline. Due to families_ concerns, the psychosocial treatment such as TF-CBT is available, total number of subjects recruited for this project was far below what despite compelling cases having been made for the benefits of conducting placebo-controlled trials (Leon, Subjects were consecutively referred children ages 10Y17 years who had experienced contact sexual abuse that was confirmed by 2000; Quitkin and Klein, 2000). These concerns Child Protective Services (CPS), law enforcement, or a profes- diminish if children in both conditions also receive a sional independent forensic evaluator, who met all study criteria psychosocial treatment with proven efficacy such as TF- and agreed (along with a custodial parent or other caregiver) toparticipate in the study. Referral sources included CPS, police, CBT, although this design would decrease the like- victim advocacy centers, pediatric care providers, mental health lihood of detecting differences between the medication care providers, and self-referrals. No recruitment ads were placed.
and placebo conditions (because the majority of both Inclusion criteria were having sexual abuseYrelated PTSDsymptoms (defined as at least five PTSD symptoms on the groups would be expected to respond to TF-CBT alone, Schedule for Affective Disorders and Schizophrenia for School- thus in effect reducing the sample size). This project as Age Children-Present and Lifetime version [K-SADS-PL] with at originally designed proposed to evaluate the relative least one symptom in each of the three PTSD clusters and efficacy of sertraline versus TF-CBT for treating PTSD clinically significant impairment), having a parent or caregiverwho was available to give consent and participate, and being symptoms in sexually abused children by comparing 10Y17 years of age. Exclusionary criteria were non-English TF-CBT versus sertraline versus placebo. Because the speaking; schizophrenia or other active psychotic disorder; mental above concerns were raised, the design was revised to retardation or pervasive developmental disorder (all due to therequirement to receive TF-CBT, a cognitive-oriented psychother- provide TF-CBT to all participants and to randomize apy); or taking current psychotropic medications. Youths with children to receive either sertraline or placebo.
substance use were not excluded from the study (in an attempt to J. A M. ACA D. CHILD ADOL ESC. P SYC HIATRY, 46:7, JULY 2007 Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
include a representative sample of sexually abused youths with study before the third treatment session due to residential PTSD), but those with current substance dependence were not Demographic information is presented in Table 1. Demographic A CONSORT flowchart for this study is presented in Figure 1.
characteristics of the final cohort included in the study did not Of the 24 participants who met the required PTSD criteria and were significantly differ from the 31 children who did not meet criteria randomly assigned to treatment, two did not complete the study, leaving a final treatment cohort of 22 subjects. The two participants According to consensus interview on the K-SADS-PL, 15 who did not complete treatment (one in each condition) left the (68.2%) of the subjects met criteria for diagnosis other than PTSD, Fig. 1 Flowchart. MR = mental retardation; SUD = substance use disorder; RCT = randomized controlled trial; PTSD = posttraumatic stress disorder; TF-CBT =trauma-focused cognitive-behavioral therapy.
J. A M. A CA D. CHILD ADOLE SC . P SYC HIA TRY, 46: 7, JULY 2007 Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
with seven subjects (63.5%) in the TF-CBT + sertraline condition random number sequence to the TF-CBT + placebo or TF-CBT + and eight (72.7%) in the TF-CBT + placebo condition meeting sertraline condition. This was a double-blind procedure whereby no criteria for at least one comorbid diagnosis. All but one of these one directly involved in the study (i.e., parents, children, children met criteria for major depressive disorder (MDD). Other independent evaluator, therapists, or child and adolescent psychia- diagnoses included general anxiety disorder, substance abuse not trist) knew which condition the children were assigned to otherwise specified, oppositional defiant disorder, panic disorder, As determined by the K-SADS-PL, this cohort had experienced a mean of 3.0 different types of previous traumas in addition to sexual abuse, including the following: 10 (45.4%) serious accidents, 2 (9%) The following instruments were administered by the independent disasters, 6 (27.2%) violent crime, 17 (77.3%) traumatic death or evaluator: K-SADS-PL (Kaufman et al., 1996) is a semistructured life-threatening illness, 4 (18.2%) domestic violence, 2 (9.1%) interview administered independently to child and parent to assess physical abuse, and 4 (18.2%) other PTSD-level traumas.
the presence of DSM-IV psychiatric disorders. For the purposes ofthis study only, the PTSD section was readministered posttreat- ment; the other sections were administered pretreatment to assesschildren for exclusionary criteria only. For the present sample, Upon referral, a telephone screen was conducted and the interrater reliability Cohen_s . for K-SADS-PL-PTSD diagnostic possibility of participating in the study was discussed with parents.
status (yes/no) was 0.92. The CGAS (Shaffer et al., 1983) is a 0- to Those who agreed were scheduled for an initial assessment, which 100-point scale on which an independent rater evaluates the child_s was conducted by an independent evaluator. This evaluator was global impairment. The CGAS is highly correlated with children_s trained in administration of the K-SADS-PL and the Children_s general functional competence and is not highly correlated with Global Assessment Scale (CGAS). The evaluator continued to be other measures of symptom severity (Green et al., 1994). For the randomly checked throughout the course of the study to ensure present sample, Pearson_s correlation interrater reliability of the ongoing interrater reliability with regard to both instruments. At the CGAS was 0.90. The CGAS was completed pretreatment and at initial evaluation, children and parents completed the initial assessment instruments described below and those who qualified Children completed the following instruments pre- and for admission to the study read and signed informed assent/consent posttreatment. The Children_s PTSD Symptoms Scale (CPSS; forms. Children were randomized according to a computerized Foa et al., 2001) is a 24-item self-report measure of PTSD. Test-retest reliability for the CPSS is 0.84. The Mood and FeelingsQuestionnaire (MFQ; Costello and Angold, 1988) is a 33-item self-report measure of child and adolescent depression. Test-retest reliability for the MFQ is 0.85. Clinical cutoff score for thisversion of the MFQ is 28. (Note: Children additionally completed the CPSS and MFQ at weeks 3, 5, and 8.) The Screen for Children_s Anxiety Related Emotional Disorders (SCARED; Birmaher et al., 1997) is a 38-item self-report measure for children_s anxiety symptoms with good (0.86) test-retest reliability.
Clinical cutoff score for the SCARED is 25. The Children_sAttributions and Perceptions Scale (Mannarino et al., 1994) is a self-report instrument in which children report their child abuseYrelated attributions and perceptions related to self-blame, feeling different from peers, feelings of not being believed, and loss of interpersonal trust. Test-retest reliability for the Children_s Attributions and Perceptions Scale is 0.75.
Parents completed the following instruments pre- and posttreat- ment. Child Behavior Checklist (Achenbach, 1991) is a 118-item parent report form for children_s behaviors. Reliability is 90.90 for the broadband scales used in this study. Beck Depression Inventory II (Beck et al. 1996) is a parent self-report of depression. Test-retestreliability for the BDI-II is high (0.93). Cutoff score for a nonclinical population on the BDI is 12. The Parent_s Emotional Reaction Questionnaire (Mannarino and Cohen, 1996) is a self-report measure of parental emotional distress related to the child_s sexual abuse; test-retest reliability is 0.90; the Parental Support Ques- tionnaire (Mannarino and Cohen, 1996) is a self-report ques- tionnaire of parental support of the child and attributions about responsibility for the abuse. Test-retest reliability for the Parental The child and adolescent psychiatrist assisted parents and children in completing the following form at each medication evaluationsession. The Side Effects Form for Children and Adolescents Mean no. of months since most recent abuse (SEF-CA; Klein, 1998) is a child- and caretaker-informed form to J. A M. ACA D. CHILD ADOL ESC. P SYC HIATRY, 46:7, JULY 2007 Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
be completed regarding psychotropic medication side effects. Each item is completed regarding frequency and severity.
Data analyses were conducted using SPSS version 14. A series of repeated-measures analyses of covariance (ANCOVAs) were conducted with the covariate in each analysis being the pretreatmentscore for the outcome score posttreatment under analysis. The Identical pill sertraline (25 and 50 mg) and pill placebo were ANCOVAs of the K-SADS-PL-PTSD were conducted separately provided free of charge to this study by Pfizer. Pfizer had no because those were the only data collected through a semistructured other involvement in the study. At the initial evaluation session, a interview. The CGAS is not highly correlated with measures of psychiatric evaluation was performed. The same child and symptomatology (Green et al., 1994), and for this reason, a separate adolescent psychiatrist performed this evaluation for all of the children in the study and conducted subsequent medication Due to the large number of other overlapping child and parent management appointments. The initial psychiatric evaluation self-report measures included in the study, a multivariate analysis of included obtaining relevant psychiatric history, obtaining permis- covariance was conducted to determine whether overall differences sion to receive the child_s primary care provider records including between the two treatment groups were present from pre- to the most recent physical examination; inquiring about possible posttreatment, before performing ANCOVAs on individual instru- pregnancy if appropriate; reviewing current medications; review- ments. These multivariate analysis of covariances and ANCOVAs ing current and past medical conditions; and reviewing sertraline only included pre- and posttreatment data. To examine potential information (mechanisms of action, risks, benefits, side effects, differences between the two groups with regard to side effects, data dose, and administration). The child and parent were given from the SEF-CA were compared using a two-tailed Fisher exact written information about sertraline, and a urine sample was test. To examine suicidality, responses to BI thought about killing obtained for urine drug screen, urinalysis, and urine pregnancy myself [ on the MFQ were examined separately at sessions 1, 3, 5, 8, test (if postmenarchal). The child and parent then were and 12. Chi-square analyses were performed to compare differences interviewed individually by the child and adolescent psychiatrist in responses between the two groups.
The child and parent were seen for medication checks at sessions 1, 3, 5, 8, and 12, with sertraline/placebo dose titrated as follows.
Pills were started at 25 mg/day for 3 days for all subjects, withinstructions to increase the dose to 50 mg/day to be continued at a There were no significant differences between the dose of 50 mg/day for the first 2 weeks. At the second appointment(week 3), subjects completed the CPSS and MFQ at the medication two groups on any demographic variables at pretreat- management session. Subjects whose CPSS indicted clinically ment. As shown in Table 2, on the K-SADS-PL-PTSD, significant improvement of symptoms from their initial score, although there was a highly significant effect for time, and/or whose side effects were judged sufficiently severe to warrant there were no significant group  time differences with holding the dose at 50 mg/day were instructed to maintain this dose.
Subjects whose symptoms were not judged to be significantly regard to improvement in PTSD symptoms from pre- improved and/or whose side effects were not judged to be severe were to posttreatment. However, effect sizes were medium in instructed to increase the dose to 100 mg/day for the next 2 weeks.
Dose titration continued in this manner at each medicationcheck, to a maximum of 200 mg/day. Mean maximum dose reached There was a significant group  time interaction for the TF-CBT + sertraline group was 150 mg/day (range 50Y200 with regard to the CGAS (Table 3). The largest mg/day). Mean maximum dose reached for the TF-CBT + placebo single incremental improvement in CGAS scores in group was 172.73 mg/day (range 50Y200 mg/day). Dose for theTF-CBT + placebo group was higher at each point in treatment the TF-CBT + sertraline group occurred between weeks from week 5 onward. This difference in dose between the two 3 and 5, a time during which the pharmacological action groups approached significance ( p = .078).
of sertraline could be expected to become noticeable.
The multivariate analysis of covariance showed a significant effect for time (F = 13.59, p < .001) but not TF-CBT is described in detail elsewhere (Cohen et al., 2006; for group (F = 0.986, p = .51) or group  time (F = 1.23, Deblinger and Heflin, 1996; TF-CBT Web, 2005). The primary p = .37). However, because of the significant time effect, components are summarized by the acronym PRACTICE: parenting we examined which instruments changed significantly skills, psychoeducation, relaxation, affect modulation, cognitiveprocessing, trauma narrative, in vivo mastery of trauma reminders, over time by conducting univariate ANOVAs for the conjoint child-parent session, and enhancing safety, healthy sexu- entire cohort. The MFQ, SCARED, CPSS, CBCL ality, and future development. In the present study TF-CBT was Total, Parent_s Emotional Reaction Questionnaire, provided over the course of 12 weeks, in 12 parallel sessions forchildren and their parents/caretakers. TF-CBT was provided by one BDI, and Parental Support Questionnaire showed of two randomly assigned clinicians. Both therapists were licensed significant (<.05) change from pre- to posttreatment.
master_s degreeYlevel social workers. Treatment sessions were The two groups showed no significant differences audiotaped for adherence to the TF-CBT model, and all sessionswere required to meet 990% adherence to the model through the use with regard to side effects at any point in the study as measured by the SEF-CA. They also showed no J. A M. A CA D. CHILD ADOLE SC . P SYC HIA TRY, 46: 7, JULY 2007 Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
Repeated-Measures ANCOVA for PTSD Symptoms Note: ANCOVA = analysis of covariance; PTSD = posttraumatic stress disorder.
* p < .001.
significant difference with regard to responses to the improvement on any instrument. The numbers of question about suicidal ideation on the MFQ at any participants attaining clinically meaningful improve- point during the study, with equivalent proportions of children experiencing decreased suicidal ideation in PTSD Diagnosis. Two of the participants were in the both groups during the course of the study. No no-PTSD diagnosis category at pretreatment (both in participant in either group developed new onset of the TF-CBT + sertraline group). At posttreatment, 14 suicidal ideation during the course of the study.
additional participants had moved into the no-PTSDdiagnosis or nonclinical range in this regard (eight TF-CBT + sertraline; six TF-CBT + placebo).
Adverse events were defined as suicide attempts, Global Impairment Status. At pretreatment, all 22 reportable child abuse episodes, drug overdoses, or participants were in the clearly impaired (CGAS <60) psychiatric hospitalizations. One participant in the range on the CGAS. At posttreatment, 15 had moved TF-CBT + sertraline group was hospitalized overnight into the not clearly impaired range on the CGAS (nine during week 12 for oppositional defiant disorder TF-CBT + sertraline, six TF-CBT + placebo).
symptoms after running away from home. This child Other Clinical Instruments. On the SCARED, four had a history of oppositional defiant disorder and participants (two in each group) scored in the running away, which preceded this treatment study.
nonclinical range (score <25) at pretreatment; at No other adverse events were reported during the posttreatment 13 additional participants had improved into the nonclinical range (eight TF-CBT + sertraline;five TF-CBT + placebo).
On the MFQ, nine participants scored in the There was no significant difference between the two nonclinical range (score <27) at pretreatment (three treatment groups with regard to clinically meaningful TF-CBT + sertraline; six TF-CBT + placebo); at CGAS-Sertraline 45.09 (5.24) 50.09 (5.36) 59.45 (6.82) 61.18 (8.62) 66.64 (10.12) 46.64 (5.03) 51.64 (4.23) 50.64 (9.12) 54.73 (8.84) 59.55 (9.70) Note: ANCOVA = analysis of covariance; CGAS = Children_s Global Assessment Scale.
* p < .01; ** p < .001.
J. A M. ACA D. CHILD ADOL ESC. P SYC HIATRY, 46:7, JULY 2007 Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
posttreatment, 13 additional participants had moved (ANCOVA), calculated that at least 32 treatment into the nonclinical range (eight TF-CBT + sertraline; completers per group would be needed to achieve five TF-CBT + placebo). With regard to suicidal ideation adequate power when comparing the TF-CBT + placebo at pretreatment, five participants responded Btrue[ to the versus and TF-CBT + sertraline groups. Clearly the final question BI thought about killing myself[ (four TF-CBT sample size of this study is too small to allow us to draw + sertraline, one TF-CBT + placebo). At posttreatment, any definitive conclusions about possible benefits or lack no participants responded Btrue[ to this question.
thereof of adding sertraline to TF-CBT for this On the CBLC Total score, at pretreatment four participants had scores in the nonclinical range (T score The current findings are promising in that only <60; three TF-CBT + sertraline, one TF-CBT + placebo).
one adverse event (likely unrelated to medication) At posttreatment eight additional participants had occurred among children receiving sertraline, despite moved into the nonclinical range (four in each group).
the fact that they received substantial doses of this On the BDI eight parents had scores in the normal medication (mean maximum dose of 150 mg/day).
range at pretreatment (four in each group); at posttreat- Among the 11 children and adolescents participating ment, six more had moved into the normal range (four in this study who received sertraline, none developed TF-CBT + sertraline, two TF-CBT + placebo).
significant suicidal symptoms. Although this is a smallnumber of children on which to base any conclusionsof safety, the fact that the side effects reported were also relatively low is encouraging. It is possible that This study attempted to conduct an RCT to evaluate the early components of TF-CBT, which provide the benefit of adding sertraline to TF-CBT for sexually individualized relaxation and affective modulation abused children with PTSD symptoms. This cohort interventions, helped to reduce the prominence of was not representative of sexually abused children requesting clinical treatment due to the refusal of many This study points out the substantial difficulty of families to consider pharmacological treatments for recruiting a clinically representative sample with which their children, and conclusions drawn from this study to conduct a child psychopharmacological RCT among must be evaluated in this context. There were no traumatized children. One third of the potential cohort significant group  time differences between the two pool (59/176) refused to consent to participate in this groups with regard to remission of PTSD symptoms or study, largely based on the possibility of receiving active other clinical symptomatology. The only statistically medication. Although it is impossible to prove that this significant difference in outcome between the groups difficulty was related to parental concerns regarding the was on the CGAS, and this clinically observed potential association between SSRIs and suicidality in improvement corresponded with the time at which children, it is interesting to note that 20 of 24 of the clinical improvement may be expected to occur based subjects who were randomized in this treatment trial on the pharmacological properties of sertraline. The were recruited within the first 18 months of recruit- fact that both groups improved significantly and ment (January 2002YJuly 2003), with only four equivalently with regard to not only PTSD but also additional subjects being successfully recruited from depression, anxiety, behavior problems, and a variety of July 2003 to January 2006. If there was a systematic parental scores supports the idea that the treatment sampling bias in the present cohort, then one could effect observed was due to TF-CBT interventions rather reasonably speculate that it was more positively than sertraline. However, this finding must be taken in predisposed toward psychotropic medication use in the context of the fact that the study was inadequately powered to detect significant differences between thetwo treatment groups.
An initial power analysis using the Sample Power 1.0 Limitations of this study include the small sample program of Borenstein et al. (1997), which allowed for size, the self-selected nature of the cohort, and the lack the need to enter covariates, a medium effect size, and of a representative sample with regard to gender or the need to conduct one-way fixed effects of covariance ethnic diversity other than African American youths.
J. A M. A CA D. CHILD ADOLE SC . P SYC HIA TRY, 46: 7, JULY 2007 Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
combination with psychotherapy. It is also important to Although children in both groups experienced note that in many cases, combined TF-CBT + sertraline equivalent improvement in all domains except CGAS can be provided as a short-term or time-limited treat- scores, at the end of the study, decisions had to be made ment modality. Most participants (7/11) who received as to whether TF-CBT + sertraline participants should this modality were able to tolerate tapering off a continue taking sertraline or taper off this medication.
relatively high dose of sertraline within 1Y3 months, Because seven of these participants met criteria for including three participants who had comorbid MDD.
MDD at the start of the study, and all of these In conclusion, this study found that for a small and participants responded positively to treatment with self-selected group of multiply traumatized sexually regard to improvement in depression scores, it was not abused girls, TF-CBT + sertraline was not superior to possible to determine whether this response was due to TF-CBT + placebo except with regard to CGAS TF-CBT, sertraline, or a combination of both treat- outcomes. However, because the study was statistically ments. The practice parameters in place at the time underpowered, no definitive conclusions can be drawn suggested that continuation treatment be maintained about the potential benefit or lack thereof of adding for at least 6 months following remission (American sertraline to TF-CBT for this population. A potential Academy of Child and Adolescent Psychiatry, 1998), drawback of starting combined treatment was that once and thus in cases in which MDD symptoms have been a child responded positively, it was not possible to know prominent, it was difficult to justify discontinuing the whether the response was to psychotherapy, medication, medication once it was started. This was particularly or both, and this may have led to unnecessary true when depressive symptoms intermittently in- prolongation of pharmacological treatment. Due to creased after the end of the study (e.g., when legal the failure of this pilot trial to demonstrate a clear proceedings occurred related to sexual abuse charges).
benefit of adding sertraline, treatment in most cases In other cases, the child and adolescent psychiatrist should begin with TF psychotherapy with proven suggested tapering medication, but the child and/or efficacy such as TF-CBT. Medication should be added parent were reluctant to do so due to concerns about only when and if an individual child_s situation symptoms returning if medication was discontinued.
Although TF-CBT treatment ended after the 12thtreatment session, 4 of the 11 TF-CBT + sertraline Disclosure: The authors have no financial relationships to disclose.
participants continued sertraline for more than 4months. The remaining seven tapered and discontinuedsertraline within 1Y4 months after the end of the study.
Had these children received TF-CBT + placebo, some Achenbach TM (1991), Integrative Guide for the 1991 Child Behavior may have recovered without needing medication. To Checklist (CBCL) 4Y18, YSR and TRF Profiles. Burlington: University ofVermont, Department of Psychiatry illustrate this, although there were seven participants in American Academy of Child and Adolescent Psychiatry (AACAP) (1998), the TF-CBT + placebo group with comorbid MDD, to Practice parameters for the assessment and treatment of children andadolescents with depressive disorders. J Am Acad Child Adolesc Psychiatry our knowledge, only one participant in the TF-CBT + placebo group later required ongoing SSRI treatment.
Beck AT, Steer RA, Brown GK (1996), Manual for the Beck Depression Inventory II. San Antonio, TX: Psychological Corporation Birmaher B, Khetarpal S, Brent D et al.(1997), The Screen for Child Anxiety Related Emotional Disorders (SCARED): scale construction and Because this study did not demonstrate a clear benefit psychometric characteristics. J Am Acad Child Adolesc Psychiatry36:545Y554 of adding medication to TF-CBT, based on current Borenstein M, Rothstein H, Cohen J (1997), Sample Power 1.0 (software).
evidence it appears that for most children with PTSD, Brady KT, Pearlstein T, Asnis GM et al. (2000), Double-blind placebo- including those with comorbid MDD, a trial of initial controlled study of the efficacy and safety of sertraline treatment of TF-CBT or other evidence-based trauma-focused posttraumatic stress disorder. JAMA 283:1837Y1844 psychotherapy alone is warranted before adding medi- Carrion VG, Weems CF, Ray R, Reiss AL (2002), Toward an empirical definition of pediatric PTSD: the phenomenology of PTSD symptoms cation. Those with significant MDD or serious psy- in youth. J Am Acad Child Adolesc Psychiatry 41:166Y173 chosocial stressors (e.g., subsequent legal involvement) Cohen JA, Berliner L, March JS (2000), Treatment of children and adolescents. In: Effective Treatments for PTSD, Foa EB, Keane TM, may require medication at a later time, alone or in Friedman MY, eds. New York: Guilford, pp 106Y138 J. A M. ACA D. CHILD ADOL ESC. P SYC HIATRY, 46:7, JULY 2007 Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.
Cohen JA, Deblinger E, Mannarino AP, Steer R (2004), A multisite ran- Mannarino AP, Cohen JA (1996), Family-related variable and psychological domized trial for sexually abused children with symptoms of posttrau- symptom formation in sexually abused girls. J Child Sex Abuse matic stress disorder. J Am Acad Child Adolesc Psychiatry 43:393Y402 Cohen JA, Mannarino AP, Deblinger E (2006), Treating Trauma and Mannarino AP, Cohen JA, Berman SR (1994), Children_s Attributions and Traumatic Grief in Children and Adolescents. New York: Guilford Perceptions Scale: a new measure of child sexual abuse-related factors.
Costello EJ, Angold A (1988), Scales to assess child and adolescent depression: checklists, screens, and nets. J Am Acad Child Adolesc Mitka M (2003), FDA alert on antidepressants for youth. JAMA 290:2534 Quitkin FM, Klein DF (2000), What conditions are necessary to assess Davidson JRT, Malik ML, Sutherland SM (1996), Response characteristics antidepressant efficacy? Arch Gen Psychiatry 57:323Y324 to antidepressants and placebo in posttraumatic stress disorder. Int Clin Saunders BE, Berliner L, Hanson RF (2004), Department of Justice Office of Victims of Crime Treatment Guidelines for Sexually and Physically Deblinger E, Heflin AH (1996), Treating Sexually Abused Children and their Abused Children. Available at: www.musc.edu/cvc. Accessed February 7, Non-Offending Parents: A Cognitive Behavioral Approach. Thousand Schreiber R, Melon C, DeVry J (1998), The role of 5-HT receptor subtypes Foa EB, Treadwell K, Johnson K, Feeny NC (2001), The Child PTSD in the anxiolytic effects of selective serotonin reuptake inhibitors in the Symptom Scale: a preliminary examination of its psychometric proper- rat ultrasonic vocalization test. Psychopharmacology 135:383Y391 Shaffer D, Gould MS, Brasic J et al. (1983), A Children_s Global Assessment Green B, Shirk S, Hanze D, Wansrath J (1994), The Children_s Global Scale (CGAS). Arch Gen Psychiatry 40:1228Y1231 Assessment Scale in clinical practice: an empirical evaluation. J Am Acad TF-CBT Web (2005) A Web based learning course for Trauma-Focused Cognitive Behavioral Therapy. Available at: www.musc.edu/tfcbt.
Griffin LD, Mellon SH (1999), Selective serotonin reuptake inhibitors directly alter activity of neuroteroidogenic enzymes. Proc Natl Acad Sci U.S. Food and Drug Administration (2004), FDA launches a multi-pronged strategy to strengthen safeguards for children treated with antidepressant Kaufman J, Birmaher B, Brent DA et al.(1996), Schedule for Affective Disorders and Schizophrenia for School-Aged Children-Present and van der Kolk BA, Dreyfuss D, Michaels M et al.(1994), Fluoxetine in Lifetime Version (K-SADS-PL): initial reliability and validity data. J Am posttraumatic stress disorder. J Clin Psychiatry 55:517Y522 Walker EA, Unutzer J, Rutter C et al. (1999), Costs of health care use by Klein RG (1998), Side Effects Form for Children and Adolescents. New women HMO members with a history of childhood abuse and neglect.
York: New York State Psychiatric Institute. Available from rklein@ Warshaw MG, Fierman E, Pratt D et al. (1993), Quality of life and Leon AC (2000), Placebo protects subjects from nonresponse: a paradox of dissociation in anxiety disordered patients with histories of trauma or Stress Predicts Brain Changes in Children: A Pilot Longitudinal Study on Youth Stress, Posttraumatic Stress Disorder, andthe Hippocampus Victor G. Carrion, MD, Carl F. Weems, PhD, Allan L. Reiss, MD Objective: Does stress damage the brain? Studies of adults with posttraumatic stress disorder have demonstrated smallerhippocampal volumes when compared with the volumes of adults with no posttraumatic stress disorder. Studies of children withposttraumatic stress disorder have not replicated the smaller hippocampal findings in adults, which suggests that smallerhippocampal volume may be caused by neurodevelopmental experiences with stress. Animal research has demonstrated that theglucocorticoids secreted during stress can be neurotoxic to the hippocampus, but this has not been empirically demonstrated inhuman samples. We hypothesized that cortisol volumes would predict hippocampal volume reduction in patients withposttraumatic symptoms. Patients and Methods: We report data from a pilot longitudinal study of children (n = 15) with history ofmaltreatment who underwent clinical evaluation for posttraumatic stress disorder, cortisol, and neuroimaging. Results:Posttraumatic stress disorder symptoms and cortisol at baseline predicted hippocampal reduction over an ensuing 12- to 18-month interval. Conclusions: Results from this pilot study suggest that stress is associated with hippocampal reduction in childrenwith posttraumatic stress disorder symptoms and provide preliminary human evidence that stress may indeed damage thehippocampus. Additional studies seem to be warranted. Pediatrics 2007;119:e509Ye516.
J. A M. A CA D. CHILD ADOLE SC . P SYC HIA TRY, 46: 7, JULY 2007 Copyright @ 2007 American Academy of Child and Adolescent Psychiatry. Unauthorized reproduction of this article is prohibited.

Source: http://www.youthandfamilyservices.org/wp-content/uploads/2013/10/Cohen-et-al-2007-A-pilot-randomized-controlled-trial-of-comb.pdf