treating posttraumatic stress disorder with metacognitive therapy: a preliminary controlled trial

Treating Posttraumatic Stress Disorder With Metacognitive Therapy:A Preliminary Controlled Trial

Adrian Wells1 and Judith S. Colbear2

1University of Manchester and NTNU, Trondheim, Norway2Manchester Mental Health NHS Trust

Objectives: Exposure, trauma-focused cognitive therapy and eye-movement desensitisation andre-processing (EMDR) are effective treatments for posttraumatic stress disorder (PTSD) producingequivalent outcomes. How might the field advance? One way is to base new treatments on PTSDmaintenance mechanisms. A treatment that does this, metacognitive therapy (MCT), underwent pre-liminary controlled evaluation in this study. Method: Twenty participants aged 18 to 65 yearswith chronic PTSD were randomly allocated to either a total of 8 sessions of MCT or a delayed treat-ment control. Measures of PTSD, emotional symptoms, and underlying metacognitive variables wereobtained at pretreatment and posttreatment. Patients were followed-up at 3 and 6 months postinter-vention. Results: Statistically significant reductions in PTSD symptoms, depression, and anxiety atposttreatment were observed in the MCT group but not in the control group. Changes were maintainedover follow-up. The average number of sessions delivered was 6.4. Eighty percent of patients (intentionto treat) met clinical significance criteria for recovery based on the IES. Treatment was well toleratedwith only one (10%) dropout. Changes in thought control strategy hypothesized to be involved in themaintenance of PTSD were found. Conclusions: MCT appeared to be a brief treatment producinghigh recovery rates. The data add to existing uncontrolled evaluations and provide strong justificationfor future evaluation of this treatment against existing evidence-based interventions. C© 2012 WileyPeriodicals, Inc. J. Clin. Psychol. 68:373–381, 2012.

Keywords: ptsd; metacognition; metacognitive therapy; cognitive behavioral therapy; traumatic stress

Several approaches are effective in the treatment of posttraumatic stress disorder (PTSD). Thestrongest support is found for exposure, trauma-focused cognitive therapy and eye-movementdesensitisation and re-processing (EMDR; Bisson et al., 2007; Bradley, Greene, Russ, Dutra, &Wetsern, 2005). In meta-analyses, these approaches appear equally effective (Seidler & Wagner,2006). Bradley et al. (2005) reported improvement rates on intent-to-treat (ITT) samples: 41.5%for exposure, 37.6% for cognitive-behavioral therapy (CBT), and 51.8% for EMDR. Interestingly,improvement rates for the waitlist conditions were 10.3% and for supportive control 23.4%.In a meta-analysis of direct comparisons between manualized treatments, Benish, Imel, andWampold (2008) found no differences. Furthermore, dismantling or augmenting treatmentshas not revealed differential efficacy. When comparing exposure therapy alone and cognitiverestructuring without exposure, these two approaches were equally effective (Marks, Lovell,Noshirvani, Livanou, & Thrasher, 1998; Tarrier et al., 1999), and cognitive restructuring, whenadded to prolonged exposure, did not enhance treatment outcome (Foa et al., 2005). In addition,it has been shown that the outcome of EMDR is not enhanced by the use of eye-movements(Cahill, Carrigan, & Frue, 1999).

A lack of differences in outcomes suggests that no specific technique used in these approachesmay be crucial. For instance, even though most of them involve systematic exposure to memoriesand thoughts about trauma, cognitive restructuring without this element produces similar out-comes, suggesting exposure is not crucial. Despite new developments that require less exposure

Please address correspondence to: Adrian Wells, University of Manchester and NTNU, Trondheim,Norway. Division of Clinical Psychology, Rawnsley Building, MRI, Manchester, M13 9WL UK.E-mail: [email protected]

JOURNAL OF CLINICAL PSYCHOLOGY, Vol. 68(4), 373–381 (2012) C© 2012 Wiley Periodicals, Inc.Published online in Wiley Online Library (wileyonlinelibrary.com/journal/jclp). DOI: 10.1002/jclp.20871

374 Journal of Clinical Psychology, April 2012

or deliver this in ways that may be less distressing for patients and therapists alike, treatmentoutcomes have failed to improve.

One possible way to enhance treatment outcomes is to base treatment on psychological factorsinvolved in the maintenance of PTSD. This has been the impetus behind metacognitive therapy(MCT: Wells, 2000; Wells & Sembi, 2004a). This treatment does not require exposure, reliving,or challenging of thoughts about the trauma. MCT is based on a model proposing that PTSDsymptoms are normal in the immediate aftermath of a traumatic event. Symptoms such asintrusive thoughts, startle responses, and increased arousal are part of an intrinsic psychologicaladaptation process called the Reflexive Adaptation Process (RAP; Wells & Sembi, 2004a). Thegoal of this process is to orient cognition and action toward generating generic plans for copingwith threat. Symptoms subside when the RAP is complete and the person exits from threat-focused modes of processing. However, symptoms persist and therefore constitute PTSD whenthe individual cannot exit threat-related modes. This is caused by a pattern of cognition andcoping called the Cognitive-Attentional Syndrome (CAS). This comprises mental processes ofworry/rumination, threat-monitoring, and coping behaviours such as thought suppression andavoidance. The problem with these processes is that they extend threat-related thinking and so theindividual fails to down-regulate the anxiety program that is triggered by intrusive thoughts andreminders of the trauma. Instead, threat modes of responding are prolonged and strengthenedand cognition does not re-tune to the threat-free environment.

The pattern of responding with the CAS is supported by underlying metacognitions com-prising positive and negative beliefs about thinking. For example, positive beliefs include thefollowing: “If I worry about being attacked, then I’ll be able to avoid it”; “Only when I have acomplete memory of what happened then will I recover”; and “Focusing on all possible threatsin the environment will keep me safe.” Examples of negative metacognitions include: “I’ve lostcontrol of my mind”; “Unwanted thoughts mean I’ll never be normal again”; and “Thinkingabout the event could make me go crazy.”

MCT aims to help patients relate to their intrusive thoughts in new ways that reduce worryand rumination and remove threat monitoring and maladaptive coping. In service of this, itaims to modify metacognitive beliefs about rumination, worry, attention, and symptoms. Thistreatment has been described in detail in published treatment manuals (Wells, 2009; Wells &Sembi, 2004a). It differs significantly from exposure and cognitive therapy treatments. Theseapproaches use repeated exposure to trauma memories with increasing refinements in the levelof detail experienced or guided re-scripting of imagery. Cognitive techniques of challengingnegative thoughts about the self and the environment are also used. MCT does not have thisfocus; it does not require detailed information about the nature of the trauma. Instead, it focusesexclusively on the style of the patient’s reaction to intrusive thoughts, nightmares, and arousal.Beliefs are challenged but these are only in the metacognitive domain and not other domains.

MCT has undergone preliminary evaluation in a systematic A-B replication series (Wells &Sembi, 2004b) and an uncontrolled trial of chronic cases (Wells et al., 2008). In the Wells andSembi (2004b) study, 6 patients entered treatment after stable baseline periods. All participantsshowed a marked improvement in PTSD symptoms and general symptoms of anxiety anddepression and gains were maintained over long-term follow-up. In the trial of chronic cases(Wells et al., 2008), the mean duration of PTSD was 19.5 months and treatment was associatedwith large and significant improvement on measures of PTSD, anxiety, and depression. Theapplication of standardised recovery criteria to these data showed that 90% of patients wererecovered after treatment and approximately 89% were recovered or reliably improved at 6-month follow-up.

The present study was the next step in the treatment evaluation process and aimed to comparethe effectiveness of an eight-session MCT intervention to a waitlist control condition in chronicPTSD. Comparison to a no-treatment group is important because it controls for the effects oftime, regression to the mean, participation in a study, and the effects of repeated testing. Thisstudy was approved by the University of Manchester Committee on the ethics of research onhuman beings and by relevant local regional National Health Service (NHS) ethics, research,and development committees.

Metacognitive Therapy for PTSD 375

Table 1Patient Characteristics

Variable MCT (n = 10) WL (n = 10)

Gender: Female 6 5Male 4 5

Age (yrs) M 33.4 41.3SD 13.4 13.7

Index Trauma Assault 3 4Robbery 1 1Traffic accident 3 1Sexual assault 2 1Witness 1 1Work accident 0 2

Number of traumas Median 1.5 1(IQR) 1.0-2.25 1.0-3.0

Chronicity of PTSD (median months) (IQR) 13.0 15.56.0-27.0 9.5-45.0

Additional diagnosis: Minor DD 2 1Major DD 4 5GAD 2 1

Note. MCT = metacognitive therapy; WL = waitlist control; IQR = Interquartile range; PTSD = posttrau-matic stress disorder; DD = depressive disorder; GAD = generalized anxiety disorder; M = mean; SD =standard deviation.

Method

Participants

Inclusion criteria were males and females aged 18 years or older, meeting Diagnostic and Sta-tistical Manual of Mental Disorders, 4th Edition (DSM-IV; American Psychiatric Association,1994) criteria for PTSD. A minimum of 3-months duration of symptoms was required. Patientswere excluded if they reported current suicidality, psychosis, current alcohol, or substance de-pendence requiring prioritization, and/or required assessments and treatments that could notbe conducted without the aid of an interpreter. Our aim was to be as inclusive as possible withour eligibility criteria so that results would reflect what is possible in typical clinic settings.

One-hundred and nine participants were invited to take part (46 females, 63 males) afterscreening the waiting lists and referral details of patients at three NHS clinical psychology de-partments. Twenty-seven opted in (13 females, 14 males). Seven were excluded for the followingreasons: severe depression with suicidality (n = 3), PTSD not the primary problem (n = 3), andvoluntary withdrawal from the trial (n = 1). Twenty participants (11 females and 9 males) meteligibility criteria for the study. All participants had been referred to an outpatient psycholog-ical clinic by a general practitioner or psychiatrist and gave written informed consent to theirinvolvement in the study. Participant characteristics are summarized in Table 1.

Measures

The Structured Clinical Interview for DSM-IV, Axis 1 Disorders (SCID-I; First, Spitzer, Gibbon,& Williams, 1997) was used to standardise the assessment procedure and to determine a diagnosisof PTSD. Four self-report measures of symptom severity and one assessor scale served as primaryoutcome variables:

Posttraumatic Stress Diagnostic Scale (PDS; Foa, 1995). This is a 49-item self-report measure designed to aid diagnosis and assess symptom severity. It is an effective methodof screening for PTSD and has advantages over other self-report measures as it corresponds to


all six criteria for PTSD in the DSM-IV. The measure has good internal consistency (Cronbachalpha = .92) and an 82% sensitivity to correctly identify cases of PTSD and 76.7% specificity inidentifying noncases.

Impact of Events Scale (IES; Horowitz, Wilner, & Alvarez, 1979). The IES generatesscores for intrusion experiences (seven items), avoidance experiences (eight items) and a totalscore. Horowitz et al. (1979) reported that the measure is sensitive to change, and the internalconsistency of subscales was as follows: intrusions = .78 and avoidance = .82. Individuals ofvarious educational, economic, and cultural backgrounds have been able to use the scale.

Beck Depression Inventory - Second Edition (BDI-II; Beck, Steer, & Brown, 1996).This instrument assesses the presence and severity of depressive symptoms. It has been foundto have high reliability regardless of the clinical population. The internal consistency is .92 foroutpatients and .93 for college students.

Beck Anxiety Inventory (BAI; Beck & Steer, 1987). The presence and severity ofcommon anxiety symptoms are measured with this scale. The measure is devised to minimizeassociations with symptoms of depression (Beck, Epstein, Brown, & Steer, 1988), such as thosemeasured on the BDI-II, and it has high internal consistency (alpha = .92).

Assessor Rating. A rating scale was constructed to serve as objective clinical opinion,which comprised four items rated on 0-8 scales of frequency of re-experiencing, avoidance,increased arousal, and an overall rating of severity of PTSD. An independent clinician admin-istered this scale at pretreatment and posttreatment. The same scale was also completed by thetherapist. The aim in using this ad hoc measure was to provide a rapid observer rating that waseconomical to use and did not require training.

A secondary measure served as a marker to asses the effect of treatment on underlying causalprocesses variables (CAS):

Thought Control Questionnaire (TCQ: Wells & Davies, 1994). This is a self-reportmeasure designed to assess cognitive strategies that are used to control unwanted thoughts. Wewere specifically interested in the possible effect of treatment on the use of worry as a self-regulatory strategy in line with the metacognitive model. This scale has been widely used andhas good validity, and the Cronbach alpha of the worry subscale has been reported as .71 (Wells& Davies, 1994).

Participants were matched for gender and chronicity of PTSD. Chronicity was deemed“matched” if the number of months two participants had experienced symptoms fell withinthe same preset time intervals (i.e., 3-6 months, 6-12 months, in 6-month ranges to 5+ years).They were then randomly allocated to treatment or a waitlist condition by an individual whowas unrelated to the study and allocated by tossing a coin (heads = treatment; tails = control).Six females and four males entered treatment, and five females and five males entered the controlcondition. Participants taking psychotropic medication (n = 3; MCT; n = 4 waitlist control)were asked to maintain a stable dosage during treatment.

The main assessments were completed at pretreatment and posttreatment time intervals. Therewas a briefer mid-treatment assessment that we included should some patients not require orattend the full number of sessions. Participants in the treatment group were invited to return forfollow-up assessments at 3months and 6 months to examine the stability of treatment effects.Participants in the control condition waited 8 weeks prior to being reassessed and were thenoffered treatment.

Treatment

The treatment followed the core treatment manual (Wells, 2009; Wells & Sembi, 2004a). Thetherapist (JSC) received supervision from AW on each case to ensure adherence to the treatmentprotocol. Tape recordings of sessions were reviewed in supervision to monitor treatment fidelity


but no formal rating of such was made. There was no imaginal reliving, exposure, or challengingof thoughts and beliefs about trauma. Treatment comprised case formulation, socialization, andtraining detached mindfulness (note that this is not meditation). Here, patients were introducedto the idea that PTSD symptoms are part of psychological adaptation and “healing” andintrusive thoughts and memories must be left alone. Patients were helped to discriminate betweenan intrusive thought and memory and subsequent extended processing in the form of worry andrumination or thought suppression. The worry/rumination postponement experiment was thenintroduced, followed by challenging metacognitive beliefs concerning the uncontrollability anddanger of thoughts and the positive beliefs about the need to control thoughts and engage inextended thinking and threat monitoring. Later in treatment, attention modification was usedto curtail residual threat-monitoring tendencies and relapse prevention was addressed.

Results

Demographic Variables

Comparisons of the age of participants in each group, t(18) = 1.31, p = .21, number of traumasexperienced, z = −.39, p = .70, and chronicity of PTSD symptoms, z = −.38, p = .71, revealedno significant differences. Descriptive statistics are presented in Table 1.

Reliability of Assessor Ratings

The relationship between the therapist and assessor ratings was investigated with Spearmanrank order correlation. Relationships between each item at pretreatment and posttreatmentassessment were computed. At pretreatment (n = 20), the coefficients ranged from .36 (avoid-ance) to .73 (re-experiencing), while at posttreatment, the range was .70 (avoidance) to .91(re-experiencing). To reduce the number of subsequent analyses, a total symptom frequencyscore was obtained by summating the re-experiencing, avoidance, and arousal scores. At eachassessment the inter-rater correlation between the total scores were as follows: pretreatment =.63 and posttreatment = .84.

Pretreatment Symptom Scores

Independent samples t tests were conducted to compare the PDS, IES, BDI-II, BAI and assessorscores for the MCT intervention group and the waitlist control group at pretreatment. Thesescores are displayed in Table 2. There were no significant differences on the PDS, t(18) = 1.58,p = .13, the IES, t(18) = .60, p = .56, BDI-II, t(18) = .97, p = .35, the BAI, t(18) = −.61, p = .55,or overall assessor rating, t(18) = 1.24, p = .23. Similarly, there were no significant pretreatmentdifferences in TCQ-worry, t(18) = -.40, p = .69.

Attrition

One participant dropped out of treatment after session six; the person had completed mid-treatment assessment. Thus, there was a dropout rate of 10% from the MCT group. A secondparticipant entered further treatment after the end of treatment and was therefore lost to follow-up. Nine participants completed therapy. Not all participants required the full eight sessions.The mean number of sessions delivered was 6.4.

Primary Treatment Outcome

Analysis of the outcome measures at posttreatment was conducted using ITT method (lastobservation carried forward). Analyses were conducted using SPSS 15 for Windows and two-tailed significance applied throughout.

A series of mixed-model analyses of variance were run to determine if the MCT groupshowed significantly greater improvements in PTSD and related symptoms than the control


Table 2Descriptive Statistics: Means and Standard Deviations for Primary and Secondary Measures atPretreatment, Posttreatment/Wait, and Follow-Up (Intent-to-Treat Data)

Pretreatment Posttreatment Follow-up

MCT WL MCT WL 3m 6mMeasure (n = 10) (n = 10) (n = 10) (n = 10) (n = 10) (n = 10)

PDS M 32.7 37.3 16.8 34.1 11.9 9.1SD 7.4 5.6 16.4 4.7 16.7 12.1

IES M 53.2 56.2 20.5 54.8 18.1 17.5SD 12.1 10.1 18.1 12.3 26.2 23.4

BDI-II M 25.4 30.4 8.9 29.6 11.4 10.0SD 11.2 11.9 8.7 8.5 17.3 12.8

BAI M 29.6 26.7 12.9 28.3 11.8 9.3SD 10.8 10.5 12.7 14.8 18.0 11.7

Assessor M 6.2 6.8 3.9 6.7 – –SD 1.1 1.0 2.3 1.1 – –

TCQw M 12.1 11.4 9.7 13.4 – –SD 4.3 3.5 3.6 4.1 – –

Note. PDS = post traumatic stress diagnostic scale (severity); IES = Impact of events scale; BDI-II = Beckdepression inventory; BAI = Beck anxiety inventory; Assessor = assessor rating; TCQw = Thought controlquestionnaire- worry subscale; M = mean; SD = standard deviation.

group immediately after treatment. Standardized recovery rates were calculated and reported atposttreatment and follow-up for patients in the MCT condition.

On the PDS, the interaction effect between group and time was statistically significant, F(1,18) =8.07, p = .01, with a very large effect size (partial η2 = .31). However, the between-group varianceswere not similar for this variable. An independent samples t test was therefore conducted. Therewas a significant difference between the change scores for the MCT group (mean [M] = −15.9,standard deviation [SD] = 13.25) and the waitlist control group, M = −3.2, SD = 4.94; t(11.46) =2.84, p = .02.

On the IES (total score), the group x time interaction was statistically significant, F(1,18) = 33.9,p<.0005, with a very large effect size (partial η2 = .65). Analysis of BDI-II showed a significantgroup × time interaction, F(1,18) = 25.0, p < .0005, with a very large effect size (partial η2 = .58).The BAI interaction effect was also statistically significant, F(1,18) = 12.3, p = .003; partial η2

= .41. On the assessor rating of overall symptom severity, there was a significant interaction ofgroup x time, F(1,18) = 7.64, p = .01, partial η2 = .30.

The between group effect sizes at post treatment (Cohen’s d) were computed for the ITTsample using the pooled standard deviation. These results were as follows: PDS = 1.4, IES =2.2, BDI = 2.4, BAI = 1.2, Assessor = 1.6.

In summary, for each symptom outcome measure, the interaction terms showed that partici-pants in the MCT condition improved significantly more than the patients in the waitlist controlgroup. Inspection of the mean scores for the waitlist participants showed that there was littleimprovement in this group. Between group effect sizes showed that the effects for MCT werevery large. Descriptive statistics for the outcome variables are presented in Table 2.

Stability of Treatment Effects

The PDS, IES BDI-II, and BAI scores were analysed for treatment completers using repeatedmeasures t tests with Bonferroni adjustment for eight comparisons (p < .006) from posttreatmentto 3-month and 6-month follow-up. Using this corrected alpha at 3 months, there was no changefrom posttreatment in the PDS (M = 11.9, SD = 16.7, t(8) = 2.0, p = .08), IES (M = 18.1, SD =26.2, t(8) = 2.9, p = .02), BDI-II (M = 11.4, SD = 17.3, t(8) = .10, p = .92), and BAI (M = 11.8,


SD = 18.0, t(8) = 1.8, p = .10). At 6 months, there continued to be no change from posttreatmentin the PDS (M = 9.1, SD = 12.1, t(7) = .74, p = .48), IES (M = 17.5, SD = 23.4, t(7) = .36, p =.73), BDI-II (M = 10.0, SD = 12.8, t(7) = −1.6, p = .16), and BAI (M = 9.3, SD = 11.7, t(7) =.27, p = .79). The results show that treatment effects were maintained. For reference purposesdescriptive statistics for the ITT sample at 3-month and 6-month follow-up are presented inTable 2.

Clinically Significant Improvement and Recovery

Jacobson, Follette, and Ravenstorf’s (1984) criteria for assessing clinically significant changewere applied to the PDS and IES. Appropriate normative data for a nonclinical sample wasnot available for the PDS and therefore cutoff criterion “a” was used, which defines whetherthe participants score after treatment falls outside the range of the clinical population (reliablechange index = 8; cutoff < 22). For the IES criterion “c” was used based on the standarddeviation taken from Briere and Elliott (1998), which yielded a reliable change index = 11 andcutoff < 33.

At posttreatment, 70% of patients (ITT, n = 10) were recovered on the PDS, and 80% on theIES. For completers, these figures were 77.8% and 88.9%, respectively. At 3-month follow-up,among those who could be assessed (n = 8), one patient showed no change and seven wererecovered on both the PDS and IES. At 6-month follow-up (n = 8), one patient showed nochange on both the PDS and IES, one patient showed reliable improvement on these measures,and six patients were recovered on both. On an ITT basis, 60% of patients were recovered at 6months on the PDS and this was 80% at 6 months on the IES.

Secondary Outcome

The metacognitive model proposes that maladaptive coping strategies that block the RAP needto change for improvement to occur. A subscale tapping this dimension was examined: TCQ-worry. The interaction effect, F (1, 18) = 9.66, p = .006, was statistically significant with a verylarge effect size (partial η2 = .35). The between group posttreatment effect size (Cohen’s d) was.96. Descriptive statistics are presented in Table 2.

Discussion

This preliminary study aimed to test the effectiveness of MCT in the treatment of chronicPTSD and examine the effects on hypothesised underlying mechanisms. The results showedthat treatment was associated with significant effects as assessed by PTSD, anxiety, and moodmeasures. A large proportion (70%-80%) of participants met objective criteria for recoveryimmediately following treatment. Furthermore, recovery rates of 60%-80% (based on intentionto treat) for the PDS and IES were obtained at 6-month follow-up. These effects cannot beattributed to the effects of time or repeated measurement, as there was little or no improvementin the control group. However, we cannot partial out the effects of nonspecific treatment factorsin this type of design.

Treatment could be effectively delivered in a small number of sessions and it appears to havebeen well tolerated with only 1 dropout. These results demonstrate the efficacy of MCT and addto the recent literature evaluating MCT for PTSD (Wells & Sembi, 2004b; Wells et al., 2008) inwhich similar recovery rates were obtained.

Treatment appeared to lead to significant changes in the use of worry as a metacognitive copingstrategy. However, the direction of the causal relationship between metacognition change andsymptom change (or the presence of a causal relationship in the first case) remains unclear.Although the treatment aims to modify metacognition and levels of worry, it may be thatsymptom reduction leads to less need to worry, and the nature of these relationships remains to beexplored. However, data from prospective studies suggest that thinking style and metacognitionprecede the development of PTSD and depressive symptoms after trauma (Holeva, Tarrier, &Wells, 2001; Nolen-Hoeksema & Morrow, 1991).


MCT does not use exposure to memories of trauma, manipulation of trauma imagery, or chal-lenging of thoughts about trauma. In contrast, it helps patients respond to intrusive thoughtsin new ways that reduce extended trauma-related thinking. The results support the view thatexposure to trauma memories and challenging general thoughts and beliefs about the selfand world are not crucial determinants of treatment effects. The aim of MCT is to reducepreoccupation with danger and traumatic events by bringing specific thinking processes un-der adaptive control. The present results suggest this can be a highly efficient and effectiveapproach.

The limitations of this study include an absence of objective measurement of treatmentcompliance such as rating of audiotaped sessions. The use of such would allow for independentassessors to examine the level of adherence to the treatment manual. A further shortcoming isthe small sample size. Further research with larger sample sizes representing a wider range ofindex traumas would be advantageous to improve the generalizability of findings. Larger trialsprovide more precise estimates of treatment effects. However, small trials are of benefit as theygenerate justification for further investigations. There was reliance on a single therapist, whichis a limitation, and on a single assessor with limited data on assessor reliability. Follow-up wasrelatively short at 6 months and longer term follow-up would be more informative of stabilityof treatment effects.

The next stage in assessing MCT is to compare it with another active treatment for PTSDsuch as exposure, EMDR, or cognitive therapy. Supplementary analyses in future trials would beof benefit to allow further exploration of changes in metacognitive beliefs and thought controlstrategies.

References

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders. (4th ed.).Washington DC: Author.

Beck, A. T., Epstein, N., Brown, G., & Steer, R. A. (1988). An inventory for measuring clinical anxiety:Psychometric properties. Journal of Consulting and Clinical Psychology, 56, 893–897.

Beck, A. T. & Steer R. A. (1987). Manual for the Beck Anxiety Inventory. San Antonio, TX: The Psycho-logical Corporation.

Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Manual for the Beck Depression Inventory-II. San Antonio,TX: The Psychological Corporation.

Benish, S. G., Imel, Z. E., & Wampold, B. E. (2008). The relative efficacy of bona fide psychotherapiesfor treating posttraumatic stress disorder: A meta-analysis of direct comparisons. Clinical PsychologyReview, 28, 746–758.

Bisson, J. I., Ehlers, A., Mattews, R., Pilling, S., Richards, D., & Turner, S. (2007). Psychological treat-ments for chronic posttraumatic stress disorder: Systematic review and meta-analysis. British Journal ofPsychiatry, 190, 97–104.

Bradley, R., Greene, J., Russ, E., Dutra, L., & Western, D. (2005). A multidimensional meta-analysis ofpsychotherapy for PTSD. American Journal of Psychiatry, 162, 214–227.

Briere, J., & Elliott, D. M. (1989). Clinical utility of the Impact of Event Scale: Psychometrics in the generalpopulation. Assessment, 5, 171–180.

Cahill, S. P., Carrigan, M. H., & Frueh, B. C. (1999). Does EMDR work? And it so why? A critical reviewof controlled outcome and dismantling research. Journal of Anxiety Disorders, 13, 5–33.

First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1997). Structured clinical interview forDSM-IV axis I disorders - Patient edition (SCID-I/P, version 2.0, 4/97 revision). New York: BiometricsResearch Department, New York State Psychiatric Institute.

Foa, E. B. (1995). Posttraumatic stress diagnostic scale manual. Minneapolis: MN: National ComputerSystems, Inc.

Foa, E. B., Hembree, E. A., Cahill, S. P., Rauch, S. A. M., Riggs, D. S., & Feeny, N. C. (2005). Randomizedtrial of prolonged exposure for posttraumatic stress disorder with and without cognitive restructuring:Outcome at academic and community clinics. Journal of Consulting and Clinical Psychology, 73, 953–964.


Holeva, V., Tarrier, N., & Wells, A. (2001). Prevalence and predictors of acute stress and PTSD followingroad traffic accidents: Thought control strategies and social support. Behavior Therapy, 32, 65–83.

Horowitz, M., Wilner, N., & Alvarez, W. (1979). Impact of events scale: A measure of subjective stress.Psychosomatic Medicine, 41, 209–218.

Jacobson, N. S., Follette, W. C., & Ravenstorf, D. (1984). Psychotherapy outcome research: Methods forreporting variability and evaluating clinical significance. Behavior Therapy, 15, 336–352.

Marks, I., Lovell, K., Noshirvani, H., Livanou, M., & Thrasher, S. (1998). Treatment of posttraumaticstress disorder by exposure and/or cognitive restructuring - A controlled study. Archives of GeneralPsychiatry, 55, 317–325.

Nolen-Hoeksema, S., & Morrow, J. (1991). A prospective study of depression and posttraumatic stresssymptoms after a natural disaster: The 1989 Loma Prieta earthquake. Journal of Personality and SocialPsychology, 61, 115–121.

Seidler, G. H., & Wagner, F. E. (2006). Comparing the efficacy of EMDR and trauma-focused cognitive-behavioral therapy in the treatment of PTSD: A meta-analytic study. Psychological Medicine, 36, 1515–1522.

Tarrier, N., Pilgrim, H., Sommerfield, C., Faragher, E. B., Reynolds, M., Graham, E., & Barrowclough, C.(1999). A randomised controlled trial of cognitive therapy and imaginal exposure therapy for posttrau-matic stress disorder. Journal of Consulting and Clinical Psychology, 67, 13–18.

Wells, A. (2000). Emotional disorders and metacognition: Innovative cognitive therapy. Chichester, UK:Wiley.

Wells, A. (2009). Metacognitive therapy for anxiety and depression. New York, NY: Guilford Press

Wells, A., & Cartwright-Hatton, S. (2004). A short form of the metacognitions questionnaire: Properties ofthe MCQ-30. Behaviour Research and Therapy, 42, 385–396.

Wells, A., & Davies, M. I. (1994). The Thought Control Questionnaire: A measure of individual differencesin the control of unwanted thoughts. Behaviour Research and Therapy, 32, 871–878.

Wells, A., & Sembi, S. (2004a). Metacognitive therapy for PTSD: A core treatment manual. Cognitive andBehavioural Practice, 11, 365–377.

Wells, A., & Sembi, S. (2004b). Metacognitive therapy for PTSD: A preliminary investigation of a new brieftreatment. Journal of Behavior Therapy and Experimental Psychiatry, 35, 307–318.

Wells, A., Welford, M., Fraser, J., King, P., Mendel, E., Wisely, J., . . . Rees, D. (2008). Chronic PTSDtreated with metacognitive therapy: An open trial. Cognitive and Behavioral Practice, 15, 85–92.

treating posttraumatic stress disorder with metacognitive therapy: a preliminary controlled trial

Documents