The treatment literature of the past twenty years reflects an enormous interest in discovering the most effective psychological therapy for clients with a diagnosis of posttraumatic stress disorder, PTSD. The overall aim of this paper is to critically evaluate current cognitive models of PTSD and literature on the effectiveness of cognitive behavioural therapies to treat this disorder based on these models.
Definitions of PTSD
In the Fourth edition of the Diagnostic and Statistical Manual of Mental Disorders, DSM-IV (American Psychiatric Association, 1994) ‘trauma’ is defined as: (a) The person experienced, witnessed or was confronted with an event that involved actual or perceived threat to life or physical integrity; and (b) the person’s emotional response to this event included horror, helplessness or intense fear. Foa and Meadows (1997, p. 450).
In DSM-IV psychological symptoms of PTSD are categorised into three cluster symptoms: re-experiencing, avoidance/numbing and increased arousal, which arise after the person is exposed to a traumatic stressor. The recurrent re-experiencing symptoms e.g. flashbacks, nightmares, intrusive thoughts, have been considered the hallmark of PTSD (e.g. Foa & Rothbaurn, 1992). The second cluster includes avoidance of trauma-related stimuli and numbing of general responsiveness e.g. deliberately avoiding trauma-related stimuli and symptoms of emotional numbing (Foa, Hearst-Ikeda, & Perry, 1995; Litz, 1993). The latter are considered distinguishing features of PTSD (Foa & Meadows, 1997). The third symptom cluster includes increased arousal e.g. hypervigilance, exaggerated startle response, difficulty sleeping and irritability (APA, 1994).
Current Government Guidelines on the treatment of PTSD
Determining effective and efficient treatments for PTSD has become a priority in light of the condition’s prevalence and the many techniques and interventions available. The National Institute for Clinical Excellence, NICE, reviewed the most robust outcome research and produced guidelines, to inform and guide clinical practice for the psychological treatment of PTSID in adults (NICE, 2005). The guidelines were based on an independent, systematic, rigorous and multistage process of identifying, reviewing and appraising evidence for the effective treatment of PTSD. These guidelines conclude that individuals with PTSD should receive either trauma focused Cognitive Behavioural Therapy, TFCBT or Eye Movement Desensitisation and Reprocessing, EMDR. However, a distinction is made between single incident trauma and more complex presentations, and the guidelines suggest increasing the total number of sessions accordingly. Although the guidelines appear helpful for the treatment of single incident PTSD, they are arguably not as informative for treatment approaches for a large group of individuals with ‘complex’ PTSD. This presents difficulties for the clinician and client in deciding the most effective therapeutic options.
Cognitive Behavioural Therapy (CBT) is the most extensively researched therapy for individuals with PTSD (Foa & Meadows, 1997) and many studies support its efficacy in reducing symptom severity (e.g. Foa et al., 1995; Foa & Jaycox, 1996; Foa, Rothbaurn, Riggs, & Murdock, 1991; Resick & Schnicke, 1992; Richards, Lovell, & Marks, 1994; Thompson, Charlton, Kerry, Lee, & Turner, 1995). However, CBT for PTSD encompasses diverse techniques. These include exposure procedures, cognitive restructuring procedures, and combinations of both these techniques.
Exposure therapy is based on the premise that imaginal exposure (IE) to the trauma or feared situation, leads to symptom reduction. The theory argues prolonged activation of traumatic memories leads to emotional processing of the affective information, habituation of anxiety and integration of corrective information (Foa et al., 1995). Numerous studies have demonstrated that treatment based on exposure therapy is efficacious in reducing PTSD (e.g. Foa et al., 1999; Frueh, Turner, Beidel, Mirabella, & Jones, 1996; Keane, Fairbank, Cadell, & Zimmering, 1989).
Foa, Rothbaum, Riggs, and Murdoch (1991) investigated exposure therapy, stress inoculation (a type of Anxity Management Treatment, AMT), supportive counselling, and a non-treatment group in the treatment of PTSD as a result of rape. Clinical ratings of symptoms and standardized psychometric tests were examined before and after treatment as well as at a 3-month follow-up. The stress inoculation intervention showed greater results than the counselling and non-treatment conditions at post-test. However, at the follow-up, the individuals participating in exposure therapy showed more improvements of PTSD symptoms than individuals in the other groups.
Research has investigated the efficiency of exposure therapy compared to different methods of treatment. For instance, Tarrier et al. (1999) investigated exposure therapy and cognitive therapy in the treatment of individuals with PTSD arising from several different traumatic incidents. The two groups demonstrated noteworthy decrease in PTSD symptoms that was still present at the 6-month follow up. Although results were positive for both groups, there was no non-treatment control against which these two active treatments could be evaluated.
Similarly, Foa et al. (1999) compared exposure therapy to AMT and then combined the two treatments. These three groups were compared to a non-treatment control group. All three of these treatments effectively reduced symptoms of rape-related PTSD and resulted in functional improvement. There were no differences among the three treatment groups on outcome measures, but all three groups improved more than the non-treatment comparison group did.
In a study that once again compared exposure therapy to cognitive therapy, Marks, Lovell, Noshirvani, Livanou, and Thrasher (1998) examined these two treatments alone and in combination in outpatients with PTSD secondary to a wide range of traumatic events. A relaxation therapy condition was employed as the primary comparison group. All three active treatment conditions showed significant improvement, and greater improvement than that observed in the relaxation group. The three active treatments did not differ from one another on the key outcome variables.
Several investigations have advanced the field of PTSD treatment, even though the methodology utilized in the outcome study limited the conclusions that could be drawn. Frank and Stewart (1983) reported the effects of systematic desensitization on women who had been raped and who developed significant psychological symptomatology. Compared to an untreated comparison group, those women treated with graduated exposure improved most on a range of anxiety and depression symptom measures.
Richards, Lovell, and Marks (1994) compared imaginal and in vivo exposure in a randomized study of survivors of diverse traumatic events. At the 12-month follow-up, patients reported consistent reductions in PTSD symptoms and improved social adjustment. These data further substantiate the effectiveness of exposure therapy for some individuals, and also suggest that improvements in symptoms are also reflected in critical domains of life functioning. In summary, the existing data support the use of exposure therapy in the treatment of PTSD. In a previous review of this literature, Solomon, Gerrity, and Muff, (1992), (Sited in Shapiro, 1995) derived the same conclusion from data available at that time. Similar conclusions were drawn by Otto, Penava, Pollack, and Smoller (1996) in a more recent review of the literature.
In what may ultimately prove to be an important lesson for the treatment of individuals exposed to traumatic events, Foa, Hearst-Ikeda, and Perry (1995) examined the efficacy of a brief intervention to prevent the development of chronic PTSD. For women who had been recently raped, the authors developed a program based upon that which worked so well in earlier trials with chronic PTSD. Exposure therapy figured prominently in the package of treatments assembled. This package also included elements of education, breathing retraining, and cognitive restructuring. When individuals receiving the package were compared to a matched control group, this study found that at 2 months after intervention only 10% of the treated group met criteria for PTSD, while 70% of the untreated comparison group did.
As information continues to grow on exposure therapy, there is a distinct need for studies to examine combinations of treatments, to employ measures that assess social and occupational functioning, and to address the impact of treatments on comorbid psychological conditions. Clearly, the available efficacy studies demonstrate the value of extending the use of exposure therapies to patients with PTSD. However future studies assessing the generalization of exposure therapy from laboratory trials to clinical settings would be particularly useful.
When exposure therapy has been compared to other forms of cognitive therapy, such as cognitive restructuring (see below), it has proved to be more successful in reducing PTSD. Tarrier et al., (1999) compared Cognitive Therapy (CT) with imaginal exposure therapy (IE) for 72 people with chronic PTSD, and concluded that there was no significant difference between the two groups initially or at 12 month follow up. Participants recruited were obtained from a sample of referrals to primary and secondary mental health services and voluntary services, indicating that they were representative of a genuine clinical sample. However, 50% of the sample remained above clinical significance for PTSD symptoms after treatment was completed, although this dropped to 25% at six-month follow-up. This lack of improvement may have been influenced by participant’s failure to attend sessions regularly. Furthermore, those who did not show improvement rated the therapy as ‘less credible’ and were rated as ‘less motivated’ by the therapist. Therefore, it is argued that motivation for therapy and regular attendance plays an important role in outcome of therapy regardless of treatment model. A further limitation of this study was that no control group was used and non-specific treatment factors and spontaneous remission could also account for the improvements in reported symptoms.
Cognitive restructuring is based on the theory that identifying and modifying catastrophic and unrealistic interpretations of the traumatic experience leads to symptom reduction. Recent models have emphasised the importance of correcting cognitive distortions in the adaptive recovery of people following trauma (Ehlers & Clarke, 2000).
Ehlers, Clark, Hackmann, McManus, and Fennell (2005) utilized cognitive therapy based on the cognitive model of PTSD (see Ehlers & Clarke, 2000). From this model, the aim of therapy is to modify excessively negative appraisals, correct the autobiographical memory disturbance and to remove the problematic behavioural and cognitive strategies. In a randomised controlled trial, twenty-eight participants who were referred to a community mental health team were diagnosed with PTSD. Fourteen participants were randomly allocated to immediate cognitive therapy or a 13-week waiting list condition. Those receiving cognitive therapy had 12 weekly treatment sessions, based on the Ehlers and Clarke (2000) model of trauma focused CBT. Participants completed self-report measures of PTSD symptoms, depression, anxiety and also completed the Sheehan Disability Scale (APA, 2000). Measures were completed pre and post treatment and at 6 month follow up. Results found that CT for PTSD was superior to a 3-month waiting list condition on measures of PTSD symptoms, disability and associated symptoms of anxiety and depression.
This study had no dropouts, which is a significant improvement on other studies, which Yielded high dropout rates. (e.g. Tarrier et al., 1999). Participants displayed a positive change in cognitive appraisals. The Ehlers and Clarke (2000) model suggest that two other pathways of change, change in autobiographical memory of the trauma, and dropping of maintaining behaviours and cognitive strategies as integral in reducing symptoms of PTSD. Although the treatment addressed these other two factors, these have not been systematically measured, so it is difficult to conclude whether clients experienced a change in these two areas.
Further analysis indicated that demographic, trauma and diagnostic variables did not predict treatment outcome, suggesting that the treatment is applicable to a wide range of trauma survivors. However, the degree in variation of trauma and small sample numbers suggests that this finding would not be present in a larger sample. Co-morbid depression and previous trauma history, which was present in over half the sample, did not negatively affect outcome.
Combinations of therapy
Resick and Schnicke (1992) have proffered a multidimensional behavioural treatment package for women who have rape-related PTSD. This package, entitled “cognitive processing therapy” (CPT), combines elements of exposure therapy, Anxiety Management Training (AMT), and cognitive restructuring. The cognitive therapy component of CPT involves addressing key cognitive distortions found among women who have been assaulted. In particular, these authors have designed interventions for addressing difficulties in safety, trust, power, self-esteem, and intimacy in the lives of survivors. In a preliminary evaluation of CPT, the authors compared outcomes at pre-treatment, post-treatment, 3 months follow-up, and 6 months follow-up for a treatment group and a non-treatment comparison group (no random assignment was used). On clinician ratings and psychometric inventories of PTSD, the individuals receiving CPT improved markedly. At the post-treatment assessment, impressively, none of the treated patients met criteria for PTSD.
In a recently completed study, Resick, Nishith, and Astin (2000) reported on a comparison of CPT and exposure therapy in the treatment of rape-related PTSD. In general, the two treatments were equally effective and more effective than a non-treatment control condition. CPT did also seem to reduce comorbid symptoms of depression, as well as those of PTSD.
Combination treatments that include an array of cognitive-behavioural strategies have the advantage of addressing multiple problems that people with PTSD may exhibit, as well as incorporating techniques that have considerable empirical support in the clinical literature. Keane, Fisher, Krinsley, and Niles (1994) described a treatment package including exposure therapy, AMT, and cognitive restructuring as central features of their approach to treating PTSD. This package employs a phase oriented approach to treating severe and chronic PTSD that includes the following six phases: (1) behavioural stabilization; (2) trauma education; (3) AMT; (4) trauma focus work; (5) relapse prevention skills; and (6) aftercare procedures.
Although this approach has clinical appeal, it wasnaa‚¬a„?t until psychologists Fecteau and Nicki (1999) examined such a package in a randomized clinical trial for PTSD secondary to motor vehicle accidents that the impact of a combination package such as that proposed by Keane et al. (1994) was assessed. Their intervention consisted of trauma education, relaxation training, exposure therapy, cognitive restructuring, and guided behavioural practice. Patients were randomly assigned to the intervention or to a non-treatment comparison group and received some 8aa‚¬”10 sessions of individualized treatment. The results of the intervention were successful as measured by clinical ratings, self-report questionnaires, and a laboratory-based psycho-physiological assessment procedure. Described by the authors as clinically and statistically significant, these treatment effects were maintained at the 6-month follow-up assessment.
Bryant, Moulds, Guthrie, Dang, and Nixon (2003) studied the effects of IE alone or IE with CR in the treatment of PTSD. They hypothesised that CR combined with IE would result in greater PTSD symptom reduction than exposure alone, which in turn would have greater benefits than a supportive counselling condition. Fifty-eight civilian trauma survivors, diagnosed with PTSD as measured by Clinician Administered. PTSD Scale, version 2, CAPS-2, (Blake et al., 1995) were randomly allocated to one of the three conditions. Each participant received eight weekly 90-minute sessions of either IE, CR and IE or supportive counselling. Participants completed assessments at pre and post treatment and 6 month follow up. These measured PTSD symptoms and psychopathology. Forty-five participants completed treatment and analysis indicated that dropouts had higher scores for depression, avoidance and higher catastrophic cognitions than those who completed. Results indicated that participants receiving both IE and IE/CR had greater reductions in PTSD symptoms and anxiety than supportive counselling (SC).
The major finding of this study was that therapy involving IE and CR leads to greater reductions in CAPS-2 intensity scores than therapy involving IE alone. Furthermore, those receiving IE/CR, but not IE alone, reported fewer avoidance, depression and catastrophic cognitions than those receiving SC. The results from this study indicated that the combination of IE and CR are effective in reducing symptoms of PTSD. It can be argued that the reasons why IE/CR may have been more effective than augmented treatments in the past (e.g. Foa et al., 1999) was that the study carefully controlled for the amount of time actively spent on each treatment component. Furthermore, participants were instructed on CR before commencing IE so they understood the rationale behind the techniques prior to addressing the strong emotional components of IE. This may have increased their understanding and belief that it was a credible treatment approach.
The finding that CR enhanced the treatment gains of IE may have been mediated by several possible mechanisms. IE and CR may involve common elements, including processing of emotional memories, integration of corrective information and development of self-mastery (Marks, 2000). Combining both interventions may provide the individual with greater opportunity to benefit. CR may have lead to greater symptom reduction as it specifically addressed identification and modification of maladaptive cognitions that may contribute to maintenance of PTSD and associated problems (Ehlers & Clarke, 2000). Paunovic and Ost (2001), compared treatment outcome data for CBT and exposure therapy for sixteen refugees with PTSD. The authors excluded those who became too distressed in the initial interview, expressed ‘a lack of confidence’ in the therapist or were misusing alcohol or drugs. Results indicated there was no significant difference between participants completing CBT or exposure therapy, being similar to Tarrier et al’s (1999) findings.
Criticisms of Paunovic and Ost (2001)’s study are that participants did not use a self-report trauma measure, so although results are positive, there is no clear analysis of whether participants felt their trauma symptoms decreased as a result of the treatment. Further, it is not possible to generalise these findings to traumatised refugees in general, as this work is unique. Working with the use of an interpreter raises several ethical and sensitive issues, as the participant must be able to develop a therapeutic alliance with the therapist and trust the interpreter (Tribe, 2007). It could be argued that participants may have been experiencing a greater degree of trauma, not least because they had not yet learned the native language.
The most effective CBT programs appear to be those that rely on repeated exposure to the trauma memory (Foa et al., 1999; Foa et al., 1991; Foa & Rothbaum, 1992) on cognitive restructuring of the meaning of the trauma, (Ehlers & Clarke, 2000) or a combination of these methods, (Resick & Schnicke, 1992). Importantly, studies have concluded that trauma focused CBT is more effective than supportive counselling (Blanchard et al., 2003; Bryant et al., 2003).
Whilst the studies reviewed have helpfully added to our understanding of PTSD there are numerous limitations of the applications of the findings. One in particular is an over-reliance on non-clinical samples of participants such that many claims of clinically effective therapy have been made from research with participants who were not within mental health systems, and despite having PTSD symptoms had not actively sought treatment.
In addition, dropout rates in studies are high, particularly for those studies that did not use a clinical sample. This might have skewed the evidence particularly with approaches that used exposure-based therapy. Furthermore, most of the studies reviewed screened out those individuals experiencing the greatest amount of distress, avoidance and co-morbidity.
Therefore results are biased towards those clients who were able to tolerate treatment and whose symptoms were not as chronic. Indeed, inclusion and exclusion criteria appear to have a great impact on outcome of treatment.
For example, studies with a strict inclusion criteria (e.g. no co-morbidity, substance misuse, self harm) appear to have significant improvements, whilst other studies i.e. Kubany et al., (2003), allowed participants to continue with other therapy while embarking on their therapy. This makes it methodologically difficult to ascertain exactly what has been effective in reducing PTSD symptoms. As inclusion and exclusion criteria are idiosyncratic across studies, it makes it difficult to draw general conclusions regarding treatment effectiveness with a clinical population across studies.
Studies often chose to focus therapy on identified groups, e.g. police officers. However, clients who experience PTSD do not form a homogeneous group and further, the symptoms experienced may be diverse even within a sample of individuals who have experienced the same trauma.
Treatment studies often do not control for other factors that may be important contributing factors in outcome such as the role of education, quality of the therapeutic relationship, therapeutic alliance and other nonspecific factors.
The literature was generally from American, British or European sources although clearly trauma is intercultural. This raises issues about how different cultures interpret ‘PTSD’, an essentially Western concept, and also whether the treatments advocated would be effective cross-culturally.
Previous research has strongly indicated that PTSD is not an appropriate term to use in non-western situations (Summerfield, 1997), hence therapeutic approaches need to account for this. It is not clear in the majority of the research when the participant experienced the trauma, and at what point therapy started. Frequently these characteristics are omitted from studies, therefore making it difficult to compare effectiveness of studies. It is important to consider the types of clients who have been represented in the research and to look at whether it is representative of those who seek treatment. Finally, very little has been reported on the impact of other difficulties an individual is experiencing as PTSD can have a wide ranging impact on an individuals quality of life and functioning and most often clients have more complex presentations. Only very few studies reviewed controlled for this variable (see Ehlers et al., 2005). This is an inherent difficulty when completing research with a trauma population as within research it is important to obtain a sample that have a similar degree of difficulties in order to assess treatment efficacy.
Several papers have evaluated different types of therapy according to particular groups. However, it appears that ‘one size’ does not fit all in relation to PTSID. In particular the issues of culture and gender are of importance (see Liebling & Ojiambo-Ochieng, 2000; Sheppard, 2000). Individual formulations of presenting problems and contexts, which informs therapy that is adapted to suit individual client’s needs, may in fact be more helpful. It remains important to consider individual differences and client choice when offering trauma therapy.
Trauma therapy outcome studies are limited by the fact that sufferers usually have other mental health problems alongside PTSD such as depression or social anxiety. Evaluation of effective treatment of trauma survivors therefore might need to go beyond medical diagnostic categories as most of the research excludes clients with co-morbid problems. A multifaceted intervention, based on clients’ own views, which addressed these other difficulties, may help reduce relapse and improve long-term efficacy of any PTSD treatment. As outlined in the methodological limitations section, much of the research reviewed has not used a genuine clinical sample, there are high dropout rates, widely variable inclusion and exclusion criteria, and the heterogeneity of PTSD has perhaps not yet been accounted for. It is therefore difficult to ascertain what is specifically helpful or effective within the treatment components. This seems to be the next area for consideration in research.
Further research into the optimal length of treatment and timing of therapy, the effect of co-morbidity and the differing effects of individual and group therapy approaches for traumatised clients are required. Further controlled research is needed to ascertain if the types of therapies reviewed can provide long term lasting effects in reducing PTSD symptomatology.
Currently the empirical data is generally limited to the assessment of short term, focused interventions, and it would be helpful to have controlled studies on longer-term treatment for more complex trauma cases. Further research would benefit from considering the clients views and experiences of therapy, this perspective was lacking in the literature reviewed. Service user and carer perspectives are beyond the scope of this review, however they have been highlighted as an important consideration within the NICE guidelines and therefore require further consideration in future research.
There appear to be at least three treatments with excellent empirical support for treating PTSD; exposure therapy, cognitive therapy or a combination of these methods. These three approaches have excellent empirical support in well-controlled clinical trials, manifest strong treatment effect sizes, and appear to work well across diverse populations of trauma survivors. However future studies to examine the effectiveness of these approaches in clinic settings are warranted.
There is much to be learned about the treatment of PTSD. It is certain there will be no simple answers for treating people who have experienced the most horrific events life offers. Undoubtedly, combinations of treatments as proposed by Keane et al. (1994) and Resick and Schnicke (1992) may prove to be the most powerful interventions.
PTSD research in this area is only in the earliest stages of its development.
Finally, an assumption about the uniformity of traumatic events has been made in the literature in general. Although it is reasonable to speculate that fundamental similarities exist among patients who have experienced diverse traumatic events and then develop PTSD, whether these patients will respond to clinical interventions in the same way is an empirical question that has yet to be addressed. Studies posing a question such as this would be a welcome addition to the clinical literature: Will people with PTSD resulting from combat, torture, genocide, and natural disasters all improve as well as those treated successfully following rape, motor vehicle accidents, and assaults? This is a crucial issue that requires additional scientific study in order to provide clinicians with the requisite evidence supporting the use of available techniques.
Research on the prevalence of exposure to traumatic events and the prevalence of PTSD has mainly been carried out in the United States. Yet there are fundamental errors in assuming that these prevalence rates apply even to other Western, developed countries. Studies that examine the prevalence of PTSD and other disorders internationally are clearly warranted. Implicit in this recommendation is the need to examine the extent to which current assessment instrumentation is culturally sensitive to the ways in which traumatic reactions are expressed internationally. Much work on this topic will be required before definitive conclusions regarding prevalence rates of PTSD internationally can be drawn.
Studies of the effectiveness of the psychological treatments across cultures and ethnic groups are also needed. What may be effective for Western populations may be inadequate or possibly even unacceptable treatment for people who reside in other areas of the world and who have different world views, beliefs, and perspectives. This issue will need to be more closely examined before we can draw definitive conclusions.
It is suggested that despite the type of treatment provided to individuals with trauma there is ultimately a need for a flexible, integrative approach to treatment in order to deal with the complex and varying needs of individual trauma survivors. A range of outcomes has been found with the types of approaches outlined in this review, it is unclear who will respond best to which treatment approach. However, what is important in determining the success of any psychological treatment of PTSD is that it is dependent upon establishing and maintaining a therapeutic alliance that is strong enough for the client to experience as safe and trusting for positive emotional change to occur.