Chapter 1
Cognitive therapy for traumatic stress reactions
An introduction
Nick Grey
Post-traumatic stress disorder (PTSD) is a common and disabling reaction to traumatic experiences. While clinical and research efforts have been focused on PTSD, other post-traumatic psychological difficulties include depression, panic disorder, and phobias. Cognitive therapy is a successful treatment for many disorders, including PTSD. Recent clinical guidelines have recommended the use of trauma-focused cognitive behaviour therapy for the treatment of PTSD (National Collaborating Centre for Mental Health (NCCMH), 2005). However, such guidelines, and the research trials on which their conclusions are based, do not cover all possible presentations of PTSD and traumatic stress reactions. Therefore clinicians are always working with some clients âbeyond the guidelinesâ. In such cases clinicians need to apply skills flexibly in empirically guided clinical interventions (Salkovskis, 2002). This involves careful assessment of phenomenology, individualized formulation, and the use and further development of strategies derived from efficacious and effective treatments. This casebook is intended as a resource for clinicians in devising such interventions.
This chapter introduces some basic information on PTSD, other psychological reactions to traumatic events, and associated comorbidity. It presents Ehlers and Clarkâs (2000) cognitive model of PTSD, the main treatment approaches that are derived from it, and reflections on how case and treatment descriptions can be used by clinicians to improve their practice.
TRAUMATIC STRESS REACTIONS
What is trauma?
The study of post-traumatic stress symptoms has often been a controversial area, subject to scientific, political, and legal influences (see Brewin, 2003), with some concern that the term âtraumaâ is used too loosely, and colloquially even, and as such the term becomes meaningless. A formal diagnostic definition of a traumatic event requires that the individual âexperienced, witnessed, or was confronted with an event or events that involved actual or threatened death or serious injury, or a threat to the physical integrity of self or othersâ and that the personâs âresponse involved intense fear, helplessness, or horrorâ (American Psychiatric Association (APA), 1994).
Post-traumatic stress disorder
In order to meet formal DSM-IV diagnostic criteria (APA, 1994), following a traumatic event an individual needs to have one re-experiencing symptom, three avoidance or numbing symptoms, and at least two hyperarousal symptoms. However, factor analyses of traumatic stress symptoms have indicated that a four-factor structure (re-experiencing, avoidance, numbing, and hyperarousal) is a better fit to the available data than a three-factor structure (with avoidance and numbing combined together as in DSM-IV) (Foa, Riggs, & Gershuny, 1995).
A core feature of PTSD is the presence of intrusive memories of the event(s). Typically these intrusions are in the form of visual mental images but can also occur in other sensory modalities (Hackmann, Ehlers, Speckens, & Clark, 2004). One feature that distinguishes these traumatic memories from other autobiographical memories is that they are experienced as happening ânowâ rather than as a memory of the past (Ehlers & Clark, 2000; Hackmann et al., 2004). Degree of ânownessâ of intrusive memories is a good predictor of chronic post-traumatic stress disorder after assault (Michael, Ehlers, Halligan, & Clark, 2005). Clinically there is a need to differentiate intrusions of (aspects) of the traumatic memory itself from rumination on the event(s) or sequelae of the event(s).
Epidemiology
The largest sample, from the US National Comorbidity Survey, found rates of exposure to traumatic events of 61% in men and 51% in women (Kessler, Sonnega, Bromet, Hughes, & Nelson, 1995). Rates of exposure in some non-western societies are higher due to greater exposure to natural disasters and warfare. In this sample the risk of developing PTSD in response to a traumatic event was 8% for men and 20% for women. Events such as rape and torture are associated with higher rates of PTSD than events such as accidents and natural disasters. Lifetime prevalence rates of PTSD in western community samples are usually around 5â10%. Kessler et al. (1995) found lifetime prevalence in women of 10.4% and in men of 5.0%. In a valuable epidemiological study in survivors of war or mass violence who were randomly selected from community populations, de Jong et al. (2001) found prevalence rates of PTSD of 37% in Algeria, 28% in Cambodia, 16% in Ethiopia, and 18% in Gaza. Higher rates of PTSD are found in refugees and asylum-seekers who have fled from their country of origin. Turner, Bowie, Dunn, Shapo, and Yule (2003) examined a large group of Kosovan Albanian refugees in the UK and found 49% met criteria for PTSD.
Risk factors
While experiencing symptoms such as nightmares and flashbacks in the aftermath of traumatic events is common, most people recover from the early appearance of traumatic stress symptoms without any formal intervention and it is a subgroup that go on to develop chronic PTSD. Two thorough meta-analyses have provided strong evidence for particular risk factors for the development of PTSD such as post-trauma support and life stress, and peritraumatic processes such as dissociation (Brewin, Andrews, & Valentine, 2000; Ozer, Best, Lipsey, & Weiss, 2003). Studies also indicate the role of post-traumatic cognitions as important predictors of the development of PTSD following road traffic accidents and assaults (e.g., Ehlers, Mayou, & Bryant, 1998). Recent longitudinal prospective studies have investigated the role of disorder-specific cognitive predictors derived from cognitive models of PTSD, depression and phobias in predicting the severity of symptoms of these three disorders following motor vehicle accidents (Ehring, Ehlers, & Glucksman, 2008) and assaults (Kleim, Ehlers, & Glucksman, 2008). These two studies showed that depression, phobia, and PTSD are correlated but distinct, and that symptoms are best predicted by the cognitive factors from the respective disorder-specific model, rather than other established predictors.
Other traumatic stress reactions
PTSD can only be formally diagnosed 1 month after the traumatic event. Within the first month individuals may meet diagnostic criteria for acute stress disorder (ASD) if they have the requisite number of symptoms, similar to those in PTSD, but also specifically requiring the presence of three dissociative symptoms.
If an individual has symptoms characteristic of PTSD without meeting the criterion A for the traumatic stressor, DSM would currently classify this as an adjustment disorder. A common example is the reaction to relationship break-ups or workplace bullying, in which no criterion A event has occurred but intrusive memories and nightmares relating to these events may occur.
In the PTSD literature a differentiation is often made between Type I trauma and Type II trauma. Type I trauma is essentially a one-off traumatic event such as a road traffic accident, assault, or natural disaster. Type II trauma refers to prolonged, repeated traumatic events such as repeated abuse or torture. Such circumstances may lead to more complicated traumatic stress presentations. Herman (1992) refers to this as âcomplex traumaâ characterized by poor affect and impulse regulation, dissociation, somatization, and pathological patterns of relationships.
It has also been suggested that borderline personality disorder (BPD) is better conceptualized as a âcomplex traumaâ reaction. Certainly there are similarities in the criteria for BPD and âcomplex traumaâ. Furthermore, those people who could be diagnosed with BPD also often experience traumatic stress symptoms. However, epidemiological studies demonstrate that many individuals meet criteria for BPD without meeting criteria for PTSD, and that they are more likely to also meet criteria for a mood disorder, particularly depression, rather than PTSD (Zanarini et al., 1998).
The utility of the term âcomplex traumaâ is currently unclear. It is used in differing ways, all of which try to describe some sense of difficulty or profound impact on the client not fully captured by PTSD. It is preferable to describe the actual problems or symptoms an individual may have and to use an idiosyncratic psychological formulation. Models of depression, PTSD, and other anxiety disorders may be helpful in planning treatment approaches.
Epidemiology of other disorders following trauma
There is less research in this area as the focus has been on PTSD. In a very large sample of US veterans, PTSD was the most common disorder following trauma (13%), followed by any other anxiety disorder (6%), adjustment disorder (6%), and depression (5%) (Seal, Bertenthal, Miner, Sen, & Marmar, 2007). Twenty-five per cent of the sample had at least one diagnosis and of those 56% had more than one. In an Australian sample following physical injury and using conservative methodology, the most common disorders at 12 months post-injury were PTSD (10.4%), depression (10.1%), any substance use disorder (6.5%), phobia (3.6%), and panic disorder (2.3%) (OâDonnell, Creamer, Pattison, & Atkin, 2004). Half had comorbidity, most commonly depression and PTSD.
Comorbidity following traumatic experiences
The high levels of comorbidity in people with PTSD is acknowledged in DSM-IV (APA, 1994). The most common comorbid conditions are affective disorders (37â49%), substance-use disorders (27â45%), and other anxiety disorders (e.g. panic disorder, 13%) (Breslau, Davis, Andreski, & Peterson, 1991; Creamer, Burgess, & McFarlane, 2001; Kessler et al., 1995). It is unsurprising that there is high comorbidity because many symptoms overlap with other diagnoses. In most cases of comorbid depression or substance-use disorders, the PTSD was primary (Chilcoat & Breslau, 1998). In a large community sample in Chile, 71% of men and 90% of women who met criteria for PTSD also had another lifetime diagnosis (Zlotnick et al., 2006). In a review of comorbidity profiles, Deering, Glover, Ready, Eddleman, and Alarcon (1996) found that they differ according to the type of trauma experienced and the population studied. For example, the rates of substance-use disorders among combat veterans with PTSD is higher than those with PTSD from other traumatic events, and trauma involving physical suffering may be more likely to lead to somatization in PTSD.
ASSESSMENT
A comprehensive reference text addressing issues of assessment of traumatic stress reactions is Keane and Wilson (2004). Probably the âgold standardâ for assessing PTSD is the structured interview Clinician Administered PTSD Scale (CAPS; Blake et al., 1990), and for a range of diagnoses the Structured Clinical Interview for DSM-IV (First, Spitzer, Gibbon, & Williams, 1996). However, many people either do not meet specific diagnostic criteria, or meet criteria for many disorders. It is important to pay close attention to phenomenology and use an individualized formulation (e.g., Tarrier, 2006). Clinicians should not assume that intrusions indicate PTSD because intrusive memories occur in other disorders such as depression (Reynolds & Brewin, 1999) and, more broadly, intrusive images occur across many, if not all, disorders (see Holmes & Hackmann, 2004).
Self-report questionnaires also provide very helpful information and should be used to monitor progress session-by-session. Commonly used questionnaires for traumatic stress symptoms are the Revised Impact of Events Scale (IES-R; Weiss & Marmar, 1997) and the Posttraumatic Diagnostic Scale (PDS; Foa, Cashman, Jaycox, & Perry, 1997). Although beyond the scope of this chapter and book, there is particular interest in measures that may be used to screen large numbers of people, such as following disasters or terrorist attacks, in order to better direct available therapeutic resources. One such is the 10-item Trauma Screening Questionnaire (Brewin et al., 2002), which focuses on the re-experiencing and hyperarousal symptoms. Recent research suggests that other symptom combinations may work better and highlights the need for cross-validation research (Ehring, Kleim, Clark, Foa, & Ehlers, 2007). All self-...