Prolonged Exposure Therapy

Neil P Roberts et al. (2019) · European journal of psychotraumatology

This systematic review and meta-analysis explored which early psychological treatments, started within three months of a traumatic event, reduce post-traumatic stress symptoms. Post-traumatic stress disorder (PTSD) involves intrusive memories, avoidance and heightened alertness that persist after trauma, and acute stress disorder (ASD) is a similar early reaction in the first month. The authors updated an earlier 2009 review, pooled 61 randomized controlled trials, and prepared it as part of the International Society for Traumatic Stress Studies treatment guidelines. Trials were split into three groups: help offered to everyone exposed to a trauma, help for people already showing symptoms, and help for people diagnosed with ASD or PTSD. Included trials tested a wide range of approaches, including trauma-focused cognitive behavioural therapy (CBT-T, which uses written, imagined or real-life exposure to the trauma), cognitive therapy without exposure, brief EMDR (eye movement desensitization and reprocessing, a therapy in which a person recalls a trauma while following guided eye movements), cognitive processing therapy, internet-based self-help and stepped collaborative care. Most participants were adults recruited after accidents, assaults, injury or intensive care. Symptom severity was compared using standardized mean differences (SMD, where negative favors treatment), and the authors rated evidence quality with GRADE. For people not screened for symptoms, no intervention was clinically better than usual care. For people with symptoms, CBT-T outperformed waitlist (SMD -0.63) and supportive counselling (SMD -0.61) after treatment, and the gap was larger in diagnosed ASD or PTSD (SMD -0.89 and -0.75). Cognitive therapy versus waitlist reached SMD -0.68, and brief EMDR reached SMD -2.50, though that result came from four small trials with only 84 participants and very low quality evidence. CBT-T and cognitive therapy performed similarly (SMD -0.19). Benefits were less consistent at 7 to 12 months, and cognitive processing therapy showed no effect in two trials. The authors judged the CBT-T, cognitive therapy and EMDR findings clinically important, whereas significant effects for collaborative care, brief dyadic therapy and a self-guided internet program in unscreened individuals were not judged clinically important. Most ratings were low or very low quality, and few trials reported adverse events. The authors suggest that CBT-T, cognitive therapy and brief EMDR are supported for symptomatic people and that routine intervention for everyone exposed is not supported. They note limits: diverse studies, many small trials, high attrition, inconsistent follow-up and thin safety data. They also caution that the large EMDR effect rests on very few participants, that pooled estimates are uncertain because most GRADE ratings were low or very low, and that non-publication of negative results cannot be excluded.

Barbara Olasov Rothbaum et al. (2012) · Biological psychiatry

This randomized pilot study with 137 civilian trauma patients explored whether starting psychotherapy within hours of a traumatic event, before the memory of the event fully consolidates (the process by which an experience becomes a stable, long-term memory), could prevent posttraumatic stress disorder (PTSD) from developing. The intervention was a modified version of prolonged exposure (PE), a therapy that reduces conditioned fear responses by having a person repeatedly revisit a feared memory in a safe setting until the fear response diminishes, a process known as fear extinction. Because most PTSD treatments begin only after symptoms are already established, this trial tested the therapy as prevention rather than as treatment of chronic illness. Participants were adults aged 18 to 65 who arrived at a Level One trauma center emergency department within 72 hours of a qualifying trauma; 137 were randomly assigned to three one-hour sessions of modified PE, delivered one week apart and beginning in the emergency department, or to assessment only. The modified PE sessions involved guided recounting of the trauma memory, discussion of the traumatic experience, and take-home exposure exercises. Blinded evaluators measured posttraumatic stress severity with a structured clinical interview (the PSS-I) and depression with a standard self-report questionnaire (the BDI-II) at 4 and 12 weeks after enrollment. Participants who received the intervention reported significantly lower posttraumatic stress severity than the assessment-only group at 4 weeks (19.09 vs. 24.54 points, p = 0.02 for the overall treatment effect) and 12 weeks (15.47 vs. 20.33 points), representing a medium-sized benefit (Cohen's d of 0.34 to 0.38, a standard measure of effect strength where 0.2 is considered small and 0.8 is considered large). By 12 weeks, significantly more intervention participants no longer met PTSD criteria than controls (74% vs. 53%, p = 0.04). Depression scores at 4 weeks were also significantly lower in the intervention group (p < 0.01, d = 0.35). The benefit was largest among rape victims, who showed a large effect (d = 0.70) at 4 weeks, notable because sexual assault carries among the highest known risk of developing PTSD. The intervention did not significantly reduce symptoms tied to participants' earlier, unrelated traumas, suggesting its effect was specific to the event that brought them to the emergency department. The findings suggest that exposure-based psychotherapy can be safely delivered within hours of trauma, in an emergency department, before a traumatic memory becomes consolidated. The authors note important limitations: no baseline PTSD measure was collected, since assessing symptoms so soon after trauma was judged invalid, so only between-group comparisons were possible; the sample was pilot-sized and not powered for smaller effects; and roughly a quarter to a third of participants did not complete follow-up assessments. The authors describe the results as establishing feasibility and call for larger replication trials.

Veerle Oosterbaan et al. (2019) · European journal of psychotraumatology

This systematic review and meta-analysis explored whether psychological help offered within 3 months of a sexual assault can reduce or prevent posttraumatic stress disorder (PTSD), a condition in which distressing memories, avoidance and heightened alertness persist after trauma. The authors searched six databases for studies from 1980 to April 2018, and excluded psychological debriefing because earlier work found it harmful. They report that this narrow question had not been reviewed before. Seven studies with 350 participants met the criteria and tested five different early interventions, a small evidence base that the authors describe as diverse in design and timing. The included studies were mostly of women in high-income countries, with average ages of 22 to 34, recruited through hospitals, forensic examinations or counselling centres. Interventions ranged from a single 9- or 17-minute educational video to a 113-minute EMDR session (eye movement desensitization and reprocessing, in which a person recalls a trauma while following guided eye movements), a modified prolonged exposure session (structured, repeated recalling of the trauma), cognitive restructuring with relaxation, and six weeks of cognitive processing therapy. Two authors independently screened studies and rated risk of bias; four studies compared with standard care were pooled. Pooling four studies, early intervention did not clearly reduce PTSD severity at 1 to 6 weeks (g = -0.28, 95% CI -0.57 to 0.02; g is a standardized effect size, where negative favors intervention), though all four leaned that way. At 2 to 12 months, a small effect just reached significance (g = -0.23, 95% CI -0.46 to 0.00), with no measured heterogeneity. Two studies that compared PTSD diagnoses found no difference between groups. Safety data were thin: only two studies reported adverse events, and none used an active strategy to detect harm, though author contact turned up no serious events. All seven studies were judged at high risk of bias, and true randomization occurred in only two of the five studies that compared treatment groups, which the authors link to a high risk of selection bias. The authors suggest that early interventions may lead to lasting reductions in PTSD severity after sexual assault, but state that definite conclusions about safety, efficacy and PTSD prevention are impossible with so few, heterogeneous, high-risk-of-bias studies. They also note that the studies used older PTSD definitions from DSM-III-R and DSM-IV. They call for trials that actively track harm, report PTSD diagnoses as well as symptom scores, compare early with later treatment, and consider cost and access for survivors.