Therapists often raise concerns that the emotional arousal experienced by trauma survivors undergoing exposure therapy may be extremely distressing and even damaging. Indeed, several clinical researchers have expressed reservations about the safety of exposure therapy in the treatment of at least some populations with PTSD (e.g., Cloitre et al., 2002; Kilpatrick & Best, 1984; Pitman et al., 1991). Two potential safety issues, in particular, have gained attention in the literature: (1) exposure therapy may exacerbate the very PTSD symptoms that it is designed to ameliorate; and (2) although
PTSD symptoms may be alleviated, other psychological symptoms (e.g., drinking, depression, guilt) may worsen.
For years, the primary evidence for the dangerousness of exposure therapy has been a paper by Pitman et al. (1991) that described six cases of combat veterans whose PTSD symptoms worsened after treatment by imaginal exposure. However, the study from which the case series was obtained did not include a control condition; therefore it is unknown how many veterans would have experienced an acute exacerbation of their symptoms during the study period had they not received treatment. Moreover, in the full sample, fear and physiological arousal as well as guilt, sadness, and anger were decreased after exposure therapy (Pitman et al., 1996).
More recently, Tarrier et al. (1999) conducted a randomized controlled trial comparing imaginal exposure with CT and reported that overall, the two treatments produced comparable outcomes on measures of PTSD prevalence and severity, anxiety, and depression. However, significantly more participants in the imaginal exposure group (31%) than in the CT condition (9%) exhibited "symptom worsening" at posttreatment. Taken at face value, these data would seem to support concerns about the safety of exposure therapy in the treatment of PTSD. However, several considerations lend doubt to this conclusion. First, the operational definition of "symptom worsening" was a posttreatment PTSD severity score that was greater than the corresponding pretreatment score by 1 or more points; the mean increase in PTSD severity scores was not reported. Given that an increase of just 1 point is within the measurement error of the instrument (CAPS), this definition may not reflect actual symptom worsening (for an extended discussion, see Devilly & Foa, 2001). Second, Tarrier et al. did not include a waiting-list condition; therefore it is not possible to determine whether the rates of "symptom worsening" observed in the imaginal exposure condition represented an increase, decrease, or no difference from what would have been observed if treatment had been withheld. Third, the group differences were not apparent on measures of depression and anxiety, areas of psychopathology that are correlated with PTSD. Finally, the group differences that were found on the PTSD measure at posttreatment were not found at the follow-up assessment.
Subsequent research has failed to support the safety concerns about exposure therapy raised by Pitman et al. (1991) and Tarrier et al. (1999). Taylor et al. (2003) investigated symptom worsening following treatment in a study comparing a group treated with imaginal plus in vivo exposure to a group treated with EMDR and a group treated with relaxation training. Rates of symptoms worsening were uniformly low across all three conditions (0%, 7%, and 7%, respectively). Similarly, Gillespie, Duffy, Hackman, and Clark (2002) administered a treatment that combined exposure and CR and found no symptom worsening. Cloitre et al. (2002) investigated the efficacy of a treatment involving sequentially combined skills training in affect and interpersonal regulation (STAIR), based on principles of dia lectical behavior therapy (DBT; Linehan, 1993), compared with imaginal exposure, to treat PTSD in female victims of childhood abuse. Applying the Tarrier et al. definition of "symptom worsening," Cloitre et al. reported that 4.5% of patients receiving STAIR/imaginal exposure had some increase in PTSD severity following treatment, compared to 25% in the waiting-list group. Although limitations on the design of this study preclude conclusions about whether or not the low rate of symptom worsening can be attributed to treatment with STAIR prior to administering exposure, the results do illustrate that treatment with exposure therapy does not result in symptom worsening.
Cahill, Riggs, Rauch, and Foa (2003b) analyzed data from the Foa et al. (1999) study of PE versus SIT versus PE/SIT versus waiting list and our recently completed study comparing PE alone to PE with CR (PE/CR) versus waiting list (Foa et al., 2002a). Of 162 participants who completed one of the active treatments, only one (0.6%) showed symptom worsening, defined as an increase in PTSD severity by 1 or more points on the PTSD Symptom Scale—Interview (PSS-I), the primary outcome measure. In the waiting-list condition, 3 out of 39 participants (7.7%) showed symptom worsening. Cahill et al. (2003b) also investigated symptom worsening on self-report measures of depression and anxiety. Only 6 out of 159 participants receiving active treatment (3.8%) showed an increase on depression, compared to 11 out of 36 waiting-list participants (30.6%). For general anxiety, the corresponding numbers were 12 out of 159 active treatment participants (7.5%) and 13 out of 34 waiting-list participants (38.2%). Combining across measures, there was a total of 16 out of 159 active treatment participants (10.1%) who showed worsening on one or more measures, compared to 20 out of 35 waiting-list participants (57.1%). Across the active treatments, rates of symptom worsening on at least one of the three measures were 6.8% for PE alone, 6.8% for PE/CR, 10.5% for SIT alone, and 27.3 for PE/SIT.
In summary, the results from the studies described above suggest that the rates of symptom worsening associated with exposure treatments are generally very low and that exposure therapy is not associated with a greater risk of symptom worsening than other forms of treatment. Indeed, results from the studies that included waiting-list controls would suggest that, if anything, withholding treatment rather than providing active treatments is associated with greater symptom worsening.
Another often-expressed concern is that the emotional distress associated with exposure therapy leads to high rates of dropout from treatment (Cloitre et al., 2002). Again, there is no empirical evidence for this concern. In a recent meta-analytic study, Hembree et al. (2003a) found no difference in the dropout rates from exposure therapy alone (20.5%), SIT or CT alone (22.1%), exposure therapy combined with SIT or CT (26.9%) or EMDR (18.9%), though active treatment did have higher dropout rates than did control conditions (11.4%). Overall, participants tolerate exposure therapy at least as well as other forms of CBT.
Was this article helpful?
It seems like you hear it all the time from nearly every one you know I'm SO stressed out!? Pressures abound in this world today. Those pressures cause stress and anxiety, and often we are ill-equipped to deal with those stressors that trigger anxiety and other feelings that can make us sick. Literally, sick.