Combining group-based exposure therapy with prolonged exposure to treat U.S. Vietnam veterans With PTSD: A case study

  • Published on

  • View

  • Download

Embed Size (px)


<ul><li><p>Journal of Traumatic StressAugust 2012, 00, 14</p><p>BRIEF REPORTCombining Group-Based Exposure Therapy With Prolonged</p><p>Exposure to Treat U.S. Vietnam Veterans With PTSD: A Case StudyDavid J. Ready,1,2 Edward M. Vega,1 Virginia Worley,1 and Bekh Bradley1,2</p><p>1Department of Veterans Affairs Medical Center-Atlanta, Decatur, Georgia, USA2Emory University Department of Psychiatry and Behavioral Sciences, Emory University School of Medicine, Atlanta, Georgia,</p><p>USA</p><p>Group-based exposure therapy (GBET) of 16-week duration was developed to treat combat-related posttraumatic stress disorder (PTSD)and decreased PTSD symptoms in 3 noncontrolled open trials with low attrition (0%5%). Group-based exposure therapy has not producedas much PTSD symptom reduction as Prolonged Exposure (PE) within a U.S. Veterans Affairs PTSD treatment program, although PE hadmore dropouts (20%). This pilot study was of a model that combined key elements of GBET with components of PE in an effort to increasethe effectiveness of a group-based treatment while reducing its length and maintaining low attrition. Twice per week, 8 Vietnam combatveterans with PTSD were treated for 12 weeks, with an intervention that included 2 within-group war trauma presentations per participant,6 PE style individual imaginal exposure (IE) sessions per participant, daily listening to recorded IE sessions, and daily in vivo exposureexercises. All completed treatment and showed Significant reductions on all measures of PTSD with large effect sizes; 7 participants nolonger met PTSD criteria on treating clinician administered interviews and a self-report measure at posttreatment. Significant reductionsin depression with large effect sizes and moderate reductions in PTSD-related cognitions were also found. Most gains were maintained6 months posttreatment.</p><p>The United States Department of Veterans Affairs (VA) treatsmore Vietnam veterans for PTSD than any other organization,and the development and dissemination of effective treatmentapproaches for these veterans is essential (Karlin et al, 2010).One treatment that was developed in the Atlanta VA MedicalCenter PTSD treatment program and used with Vietnam vet-erans is group-based exposure therapy (GBET; Ready et al.,2008); there have been three GBET open trials with no controlgroups. Two of these were conducted at the Atlanta VA PTSDprogram (Ready et al., 2008; Ready et al., 2012) and one study(Mott, Sutherland, Williams, Ready, &amp; Teng, 2012) was con-ducted at the Houston VA PTSD program. Although initial datafrom these studies suggested that GBET was an effective treat-</p><p>Correspondence concerning this article should be addressed to David J. Ready,Mental Health Service Line (116A), VA Medical Center- Atlanta, 1670 Clair-mont Road, Decatur, GA 30033. E-mail: David.Ready@va.govThis research was supported by the Atlanta VA Medical Centers Mental HealthService Line and Research Service Line. Those identified researchers hereinare employees of the Department of Veterans Affairs. The views expressed inthis article do not necessarily represent the views of that Department or of theUnited States.Published 2012. This article is a US Government work and is in the publicdomain in the USA. View this article online at wileyonlinelibrary.comDOI: 10.1002/jts.21734</p><p>ment, GBET had several important limitations. One was theamount of provider time GBET required (224 clinician hoursover 16 weeks to treat 10 veterans). Second, initial unpublishedprogram evaluation data gathered at the Atlanta VA (Bradleyet al., 2012) suggested that although both GBET and individualProlonged Exposure (PE; Foa, Hembree, &amp; Rothbaum, 2007)reduced PTSD symptoms, veterans receiving individual PE im-proved more than those receiving GBET. We also, however,found an individual PE dropout rate of approximately 20% ina sample of 101 primarily Vietnam veterans. This rate wasconsistent with attrition rates (20.5%) in overall PE research(Hembree et al., 2003). In contrast, GBET data from multi-ple studies showed lower attrition (0%5%). In addition, therewas empirical and theoretical support for the effectiveness ofgroup psychotherapy in the treatment of PTSD (Beck &amp; Coffey,2005; Morland, Hynes, Mackintosh, Resick, &amp; Chard, 2011).Therefore, the Atlanta VA PTSD treatment program developeda modified version of GBET that used less group treatment timeand that added the number of individual imaginal exposure (IE)sessions (6) that has been shown to ameliorate PTSD symp-toms in some populations (e.g., Coffey, Stasiewicz, Hughes, &amp;Brimo, 2006). The goal of this model was to improve treatmentoutcome with a more-efficient group treatment approach thathad low attrition rates.</p><p>1</p></li><li><p>2 Ready et al.</p><p>Method</p><p>Participants and Procedures</p><p>Participants were referred to treatment by the Atlanta VAMCPTSD treatment team. Inclusion criteria were combat-relatedPTSD and ability to recall a focal combat-related trau-matic experience. Exclusion criteria were active substanceabuse/dependence, severe cognitive impairment, or psychosis.Only 8 of the 20 screened veterans were not excluded and begantreatment. Exclusions were because of transportation problems,declining trauma-focused treatment, having health problems,carrying a diagnosis of psychosis, or lacking a focal trauma.This cohort was typical of Vietnam combat veterans treatedwithin the Atlanta VAMC PTSD clinic. The average age was63.75 years; 75% were African American. Five received VAdisability payments for PTSD. Each had been treated by a VApsychiatrist for at least 5 months prior to this treatment.</p><p>Measures</p><p>Measures were chosen based on demonstrated reliability andvalidity (citations below) as well as ease and efficiency ofadministration. Measures were completed during the firstweek of treatment, 1 week after completion, and 6-monthsposttreatment.</p><p>The PTSD Symptom Scale-Interview and Self-Report ver-sions (PSS-I and PSS-SR) are reliable and valid PTSD assess-ment instruments (e.g., Foa, Riggs, Dancu, &amp; Rothbaum, 1993)assessing symptoms over the past 2 weeks. The PSS-I was ad-ministered by treating clinicians. Participants also completedthe Posttraumatic Stress Disorder Checklist (PCL; Weathers,Litz, Herman, Huska, &amp; Keane, 1993), Beck Depression Inven-tory 2 (BDI-2; Arnau, Meagher, Norris, &amp; Bramson, 2001), andthe Posttraumatic Cognitions Inventory (PTCI-36; Foa, Ehlers,Clark, Tolin, &amp; Orsillo, 1999). One veteran did not completethe PSS-SR or the PTCI at pretreatment and another completedfollow-up assessment by phone (due to health problems) anddid not complete the self-report instruments at follow-up (hecompleted them posttreatment).</p><p>Treatment was provided by two psychotherapists and in-cluded three treatment phases. Phase 1 (2 weeks) consisted ofgroup therapy twice weekly for 4 hours a day (including par-ticipants eating lunch together and a break), and focused ondeveloping group cohesion, the rationale for exposure-basedtreatments, and in vivo exposure exercises. Participants devel-oped in vivo exposure hierarchies and set individual weeklyin vivo exposure goals. Phase 1 also included a specific groupexercise, focused on building group cohesion, in which eachparticipant telephoned all other participants and asked pre-determined general questions (e.g., favorite movie). Answerswere reported in the following group. Participants also made20-minute presentations to the group regarding their prewarlives.</p><p>Phase 2 (8 weeks) consisted of once weekly group therapysessions, and one individual IE session each week for par-</p><p>ticipants who had not made an IE presentation during grouptherapy that week. In the first hour of group therapy, in vivoexposure and IE homework compliance was reviewed. Next,two participants presented their combat-related index trauma(6090 minutes each). Across the full treatment each partici-pant completed two group presentations and six individual IEsessions. All IE sessions (group and individual) were recorded.Participants were instructed to listen to recordings of their mostrecent IE session daily. Similar to PE, after two IE sessionsfocused on describing the entire index trauma, the focus shiftedto hot spots (aspects of the index trauma associated with highestremaining distress).</p><p>The last 2 weeks reverted to the Phase 1 (group only) sched-ule. This phase focused on guilt, grief, and relapse prevention.As in GBET, these sessions included an imagined funeral for afallen comrade, a healing ceremony, a presentation about beinga survivor rather than a victim, and emphasis on not returningto avoidance and isolation.</p><p>Results</p><p>Data were analyzed using paired sample t tests and effect sizeswere calculated using Cohens d (see Table 1). Participantswere excluded from specific analyses in cases where they didnot complete a measure; single missing items on a measurewere replaced utilizing within-subject/measure mean substitu-tion. Due to data characteristics associated with small samplesizes, the data were also analyzed using the Wilcoxon signedranks test; results were similar to the parametric tests presentedbelow. Data indicated significant reductions were found on thePSS-I at both posttreatment, t(7) = 6.00, p = .001, and on pre-treatment versus 6-month posttreatment follow-up, t(7) = 6.33,p&lt; .001. Both self-report measures showed parallel results: forthe PCL, t(7) = 3.93, p = .006 at posttreatment and t(6) = 4.26,p = .005 at follow-up, and for the PSS-SR, t(6) = 3.52, p = .012at posttreatment and t(5) = 2.56, p = .050 at follow-up. Sevenparticipants no longer met DSM-IV PTSD criteria as measuredby the PSS-I and also fell below the proposed PCL diagnosticcutoff score of 50 (Weathers et al., 1993) at posttreatment. Fiveparticipants no longer met DSM IV criteria on the PSS-SR. Neg-ative posttraumatic cognitions were found to be reduced on thePTCI at posttreatment, t(6) = 3.14, p = .02, and follow-up, t(5)= 2.65, p = .045. BDI-II Depression symptoms were lower atboth posttreatment, t(7) = 4.87, p = .002, and follow-up, t(6) =3.45, p = .014. All participants completed treatment and therewas high attendance. One participant missed five sessions andthe rest missed fewer than three sessions of the 24.</p><p>Discussion</p><p>This was a pilot test of a treatment that added individual IEto a shortened group intervention that included key GBET el-ements. We intended to develop a more effective and abbrevi-ated group intervention with low dropouts. This small sample</p><p>Journal of Traumatic Stress DOI 10.1002/jts. Published on behalf of the International Society for Traumatic Stress Studies.</p></li><li><p>Combining Group and PE to Treat War-Related PTSD 3</p><p>Table 1Pretreatment, Posttreatment, and 6-Month Follow-up Measures With Effect Sizes</p><p>1. Pretreatment 2. Posttreament 3. Follow-up 1. to 2. 1. to 3.</p><p>Variable M SD M SD M SD d d</p><p>PSSI 32.75 7.61 7.38 8.23 10.50 5.29 3.20 3.39PCL 56.00 7.95 36.63 14.82 37.00 9.95 1.63 2.11PSS-SR 31.14 6.74 16.25 9.69 15.43 9.16 1.78 1.95BDI-2 28.46 9.48 12.02 10.98 6.42 12.27 1.60 1.10PTCI 123.29 33.76 105.50 32.39 92.43 36.85 0.54 0.87</p><p>Note. N = 8. PSSI = PTSD Symptom Scale Interview; PCL = PTSD Checklist; PSS-SR = PTSD Scale Self Report; BDI-2 = Beck Depression Inventory-2; PTCI =Posttraumatic Cognitive Inventory.p .05. p .01.</p><p>showed significant posttreatment reductions in PTSD symp-toms, self-reported depression, and cognitions related to PTSD.These improvements were largely maintained at follow-up. De-pending on the assessment instrument used, the percentage ofparticipants no longer meeting PTSD diagnostic criteria at post-treatment ranged from 63% (PSS-SR) to 88% (PSS-I and PCL).These levels of symptom change were similar to those foundin studies of individual PE (e.g., Powers, Halpern, Ferenschak,Gilihan, &amp; Foa, 2010; Tuerk et al., 2011) and other CBT PTSDtreatment approaches (Cahill, Rothbaum, Resick, &amp; Follette,2009). At 6-month follow-up, three participants had been dis-charged from our program to primary care, four continued onlyin medication management, and one required additional treat-ment for substance abuse and depression. Though the samplesize was too small to reliably assess dropout rates, our attritionrate (0) was similar to prior GBET studies (0%5%).</p><p>Although some have expressed concerns about harm from vi-carious exposure to others trauma narratives in psychotherapygroups, we found no evidence of this (Ready et al., 2012). Ourteam interviewed over 200 GBET completers who were at least6 months posttreatment and none reported any lasting negativeimpact from hearing about other veterans traumatic experi-ences, and almost all reported lasting benefits from sharingand hearing about others military-related trauma. These vet-erans frequently described developing lasting friendships withother group members due to the sense of community that oftencame with sharing similar painful experiences. Many elementsof GBET including the initial phone call exercise, participantshaving lunch together, length of treatment, and group presen-tation of military-related traumas were designed to promote asense of community among group members.</p><p>As in other studies, this case study had many limitations.As a case study, the number of participants was small andthe data included only one treatment group with no compari-son group. In addition to the small sample size, missing dataalso further reduced the sample size for some analyses. Sec-ond, treating clinicians conducted assessments; therapist biasand participants desires to please therapists could have influ-</p><p>enced outcomes. Third, the participants self-selected to be inthis treatment; the majority of individuals screened declinedto participate, limiting generalizability. This may, in part, ac-count for differences in the presented data and studies in whichparticipants were randomly assigned to groups or intent-to-treat samples, as well as potentially affecting dropout rates.Lastly, although this cohort was similar to Vietnam veteranstreated at the Atlanta VA PTSD programs, these veterans mightnot be representative of veterans treated in other VA PTSDprograms.</p><p>In terms of resources, compared to the initial GBET protocol,this treatment was modestly better. Both models utilized cother-apists; the current model required 184 clinician hours to treateight participants, and the original GBET model required 224clinician hours to treat 10 participants. The number of hours wascloser to those required to treat eight individuals in individualPE (156 based on our programs average of 13 sessions to com-plete individual PE). Although the current case study data didnot allow for a direct comparison to other treatment approaches(either prior GBET or individual PE), they did provide supportfor further research on this models impact on combat-relatedPTSD. This study also provides preliminary support for in-vestigating if adding a group component to PE might reducedropouts with older veterans while simultaneously increasingsocial support.</p><p>ReferencesArnau, R. C., Meagher, M. W., Norris, M. P., &amp; Bramson, R. (2001). Psy-</p><p>chometric evaluation of the...</p></li></ul>