Abstract
Davidson and Foa [1] identified traumatic simple phobia (TSP) as one of nine DSM-III-R disorders that may follow a trauma. In contrast to most simple phobias described by David Barlow as ‘objects of curiosity’ [2], the dread of retraumatisation is readily understandable. In fact, following a terrifying experience it is the absence of significant phobic symptoms rather than their presence that may require an explanation. The prevalence of TSP following a traumatic stressor, rather than the more wide-ranging psychopathology of post traumatic stress disorder (PTSD; DSM-IV) is unknown. The high frequency with which TSP is encountered in clinical practice suggests that it may not be uncommon. Munjac [3] reported that 30% of his series of driving phobics became phobic after a driving accident.
Assessment of TSP and fear of retraumatisation
A TSP may be a sole symptom following a trauma (see case 1), may persist after natural remission of the other symptoms of PTSD, or may prevail after treatment (cases 2 and 3). The terror and dread associated with fear of retraumatisation is inevitably associated with a state of hypervigilance for trauma-related cues. Hypervigilance is regarded by Beck [4] as an intensification of normal function, activating a ‘cognitive mobilization set’, sensory perceptual in nature, promoting a state of readiness for disengagement and escape. Thus, searching behaviour for signals of danger is activated when contextual signs of threat are perceived; once established, triggers tend to be stereotyped and predictable.
Clinically, traumatic phobias may be extremely resistant to extinction. The phasic hyperaroused state with a readiness to take evasive action at the slightest sign of danger, presents a marked contrast to the re-experiencing symptoms of PTSD, which are often associated with an over-aroused helpless state that inhibits adaptive functioning. This adaptive advantage may partially explain how as an isolated symptom it may endure considerably longer than other PTSD symptoms despite apparently adequate exposure.
Fear of retraumatisation as part of posttraumatic stress disorder
In PTSD, the fear of retraumatisation can be enmeshed with the unresolved affects associated with the traumatic memory. Commonly, flashbacks, helplessness and other reliving symptoms occur; then panic attacks and other concerns about control or distressing feelings such as anger, guilt, shame or worthlessness. In these cases, the phobic object may not be a signal that a trauma is likely to occur but rather a signal of imminent overwhelming affect.
Treatment of unresolved affects take precedence over treatment of fear of retraumatisation. Distressing affects are managed psychotherapeuti-cally and panic can be controlled using anxiety management skills training, interoceptive exposure and antidepressants. The terror associated with vivid reliving can often be reduced using therapeutic exposure. At first the patient confronts their traumatic experiences in an informal and unstructured way in the safety of the therapist's office. This may be followed by a more structured imaginal exposure program [5–6]. Caution should be taken with the use of these techniques [7–8].
As the treatment continues, in vivo exposure is added. Previously avoided situations are selected so that situations where the trauma could not possibly recur are dealt with first. Re-experiencing symptoms (Category B, DSM-IV) can be managed, without being complicated by the fear of retraumatisation. Initial in vivo targets might include going to the airport for a patient who had been in an air crash. If the exposure regimen is to be tackled in a systematic way, then dealing with the fear of retraumatisation component of the disorder occurs late in the treatment. It is at this stage that the absence of danger is impossible to guarantee.
Therapeutic exposure for fear of retraumatisation
Exposure to phobic situations where the trauma could recur presents special treatment difficulties. Commonly, these situations involve driving, or getting back to work. The patient is often too acutely aware that if a trauma has occurred once in a certain situation, it can occur again. Two questions will inevitably need to be dealt with: what is the probability of recurrence, and what is the emotional valency associated with a recurrence? Often the tendency to grossly overestimate the probability of being retraumatised can be countered cognitively. With treatment the patient accepts that the trauma is unlikely but the belief may remain that should the trauma recur, it would be the most dreadful experience imaginable. Motor accidents, robberies, random violence, industrial accidents, natural disasters occur unpredictably. Patients report that in these contexts voluntary inhibition of hypervigilance for danger cues in preparation for immediate evasive action is impossible.
The present study
The following case histories illustrate our clinical experience that, contrary to the traditional behavioural model of extinction of the fear response [9], the fear of retraumatisation and the associated hyper-vigilance responds poorly to in vivo exposure even with prolonged and repeated exposure sessions. We hypothesised that despite its apparent adaptive function, hypervigilance represents an anticipatory defensive response which limits engagement with the phobic stimulus, seriously disrupting exposure. It follows that inhibition of this component of the anxiety response could facilitate effective extinction of the fear response.
Case histories
Case 1
Two days after a front-end motor vehicle collision, a 39-year-old nursing supervisor developed phobic symptoms which increased over the next 2 years. Her symptoms were confined to hypervigilance and alarm whenever a car approached. Her hypervigilance was not associated with phobic avoidance. She drove to and from work for over 1 h every day and on four occasions she had gone for 6-h drives without improvement. As is typical with these patients, longer exposure sessions brought on a sense of exhaustion and demoralisation.
A 1-h in vivo assessment showed a fluctuating subjective units of discomfort level, reaching 10/10 as she approached intersections. She was unable to inhibit her hypervigilance and turn her attention towards neutral stimuli. When moved to the back seat with her view of the traffic hampered, her anxiety and sense of vulnerability increased despite the realisation that should an accident occur she was no more likely to be seriously injured. At the end of the 1-h assessment session, her anxiety and hyper-vigilance showed no sign of habituation.
Visual scanning was inhibited by wearing taped sunglasses. She initially reported an increase in anxiety (10/10) which lowered to 6/10 after 45 min. Total exposure time was 5 h. Improvement was maintained after the blindfold was removed. One month after treatment on a driving holiday with her husband, she reported falling asleep a number of times. At 12 months follow-up interview, improvement was maintained.
Case 2
Five months after a car accident, a 30-year-old man was treated for nightmares, flashbacks, irritability, insomnia, generalised anxiety, chronic pain disorder and depression. A multimodal treatment of sertraline 100 mg nocte, coupled with a variety of cognitive-behaviour therapy techniques resulted in a marked improvement in symptoms. His depression improved from 27 on the Beck Depression Inventory [10] to 7. Intrusiveness on the Impact of Event Scale [11] reduced from 30 to 11, and avoidance from 23 to 15. At this stage, disability was confined to discomfort while driving where he constantly scanned his left field of vision.
In vivo assessment showed anxiety scores from 9/10 to 0/10 depending on activity of the left field of vision. Attempting to inhibit his hypervigilance increased his sense of helplessness. Treatment involved taping his sunglasses while his father acted as a cotherapist. They drove for 1 h every day for 13 days. The tendency to monitor traffic by listening intently was inhibited by turning on the radio. Anxiety levels were recorded frequently.
Again the improvement generalised to travel without the blindfold. He reports becoming more comfortable discussing neutral topics in his car. As the hypervigilance improved, so did the startle reflex.
Case 3
Twelve months after a bank hold-up, a 23-year-old female bank teller was free of PTSD symptoms but was unable to return to work because of her fear of being held up again. A return to administration was partially successful in getting her back to work but hypervigilance with the general public prevented her returning to her preferred job as a teller. In the public area, any slight movement around the doors captured her attention and prompted immediate scrutiny of customers for signs of danger. Despite many hours of in vivo exposure coupled with relaxation therapy, there was no improvement and treatment was leading to demoralisation.
Hypervigilance during exposure was inhibited with earphones and strategic seat placement requiring her to face the wall. Initially, this was so distressing that she was allowed to turn and inspect the customers every 5 min then every 10 min. As in the first two cases, there was a stabilised increase in anxiety followed by a decrease. After a number of months, she returned to full-time work. The improvement was maintained 2 years later.
Discussion
This report illustrates how inhibiting hypervigilance during therapeutic exposure for the fear of retraumatisation increased treatment effectiveness in three cases. Strategic inhibition of hypervigilant behaviour, with a blindfold in cases 1 and 2, and with seat placement and audio distraction in case 3, resulted in an initial generalised increase in anxiety followed by a gradual reduction in anxiety. Extinction tended to occur across sessions rather than specifically within sessions. Generalisation of improvement to everyday situations occurred spontaneously and was maintained.
What could account for failure of these hypervigilant patients to habituate in everyday life and during standard therapeutic exposure? First, the increased anxiety is not sustained. Second, we believe that the hypervigilant component of the anxiety response represents an anticipatory defensive manoeuvre that results in nonfunctional exposure due to lack of engagement with the phobic stimulus. The initial increase in anxiety following hypervigilance inhibition could be seen as supporting this. Inhibiting hypervigilance deals with both problems by achieving a more persistent level of functional exposure or engagement and by preventing the defensive manoeuvre.
Patients find it excruciatingly difficult to voluntarily inhibit the impulse to continually keep watch on the danger zone. Inhibiting hypervigilance by blindfolding and audio-masking was an obviously helpful strategy for our patients. Perhaps teaching the patient to gain more self-control over their hypervigilance might make the need for such mechanical aids unnecessary in some cases, or at least reduce the distress that patients initially report with this exercise. One such adjunctive modality to teach self-control could be systematic desensitisation [12], where the exposure is imaginal, brief, minimally arousing and repeated. Greater acceptability by both patients and therapists may well be counterbalanced by its principal drawback of a greater commitment in therapist time.
It could be argued that there are parallels between the role hypervigilance is playing in reducing functional exposure in fears of retraumatisation and the role of safety behaviours in social phobia. Salkovskis [13] argued that despite everyday exposure to phobic cues, symptoms of social phobia were enduring because patients used defensive manoeuvres. For instance, with fears of shaking, defensive manoeuvres could take the form of gripping objects tightly, keeping rigid, breathing in a certain way, or substance use to lessen visible signs of anxiety; all of which reduce the persistence of exposure and the level of engagement. Experimental evidence supporting this model was provided by Wells et al. [14], who showed that the inhibition of safety behaviours in social phobia resulted in an improved outcome for therapeutic exposure.
There are various shortcomings of this report. There were no controls, few standardised assessments, and the observations depended on the clinician's own experience of the patient. Generalising from case reports that argue increased effectiveness for an accepted treatment by adding novel strategies could be unwarranted. Before treatment, all three patients were feeling helpless, so adding a new component to treatment must bring non-specific factors into play. A sense of demoralisation may well have been reduced by virtually any novel approach with a plausible rationale and the enthusiastic endorsement of an experienced therapist, promoting renewed efforts to confront and overcome fears.
Ultimately, the degree to which the increased efficacy is determined by these non-specific factors can only be satisfactorily determined by a controlled trial of hypervigilance inhibition, paying special attention to patient expectations, therapist behaviours and relationship factors. In the meantime, we believe that hypervigilance towards danger cues may be an important variable for clinicians to take into account, when assessing and treating fears of retraumatisation.
Footnotes
Acknowledgements
We would like to acknowledge helpful suggestions from Neil McConaghy and Jennifer Timewell.