Prevention of Post-Traumatic Stress Disorder After Trauma: Current Evidence and Future Directions

Psychological debriefing was a widely used method in the 1980s–1990s which aimed at preventing long-term post-traumatic symptoms by promoting quick emotional processing of traumatic events shortly after trauma exposure [27]. Debriefing was offered to survivors of a potentially traumatic event without prior diagnosis or evaluation, exposure being considered as good-enough risk indicator. The method typically involved a single session within hours or a few days after trauma exposure, either in a group or individual, and included general education about trauma exposure and its effect, sharing and validation of individuals’ experiences, and preparation for future encounters [28]. The method has face validity and is still well-known and, therefore, may be expected by lay people when confronted with traumatic events. However, well-conducted studies showed no evidence of beneficial effects and even suggested that debriefing may have a negative effect on recovery [29-31]. After a negative Cochrane review was first published in 1997, most treatment guidelines have been updated to recommend against providing single session psychological debriefing on a routine basis for adults after trauma [32, 33].

Trauma-focused CBT can involve different strategies with distinct aims. Exposure-based CBT, exemplified by the prolonged exposure (PE) protocol [34], aims to achieve and maintain fear extinction through repeated exposure to trauma-related stimuli in a safe context, thereby providing a sense of control over reactions and reducing avoidance. Cognitive-based CBT challenges the patient’s beliefs about the meaning and current implication of the trauma. It does so in order to change the way patients react to trauma-related reminders, to remove behavioral restrictions and rules derived from the traumatic experiences, and to reduce negative appraisal of self and others. CBT is offered individually or in a small group, to people who report symptoms (otherwise there are no ‘intervention targets’), and typically involves several weekly sessions, homework, and in vivo training exercises. The treatment may continue for over 3 months and requires significant skills from the therapists.

Trials of exposure-based CBT have generally demonstrated moderately positive results in reducing PTSD or other symptoms in the long term (Table 1). Rothbaum et al. (2012) conducted a study using modified Prolonged Exposure (PE) in rape, assault, and motor vehicle accidents survivors around 12 h after trauma, and found lower PTSD symptoms in the intervention group at 4 and 12 weeks after trauma, mainly for sexual assault victims. The same cohort also showed that PE might mitigate symptoms of PTSD in genetically predisposed individuals [35, 36•]. Bryant et al. (2008) found 5 weeks of exposure-based CBT to be effective in reducing PTSD in participants who met acute stress disorder diagnostic criteria [37]. Bisson et al. (2004) found a reduction of PTSD symptoms at 13 months—but not 3 months after the traumatic events [38], while a small study with 3 weeks of PE did not find significant symptom improvement in the PE group compared to supportive counseling [39].

CBT without in-session exposure has shown effectiveness in some but not all studies. Sijbrandij et al. (2007) compared CBT to waitlist control subjects with acute PTSD and found that CBT accelerated recovery, but makes no long-term difference [40]. Shalev et al. (2012), found that cognitive therapy fared as well as prolonged exposure 9 months [41•] after trauma exposure. However, neither intervention separated from non-intervention at 3 years [42]. Individuals classified as non-remitting [14] in that cohort were also treatment refractory.

Modifications of clinician-administered CBT, meant to make treatment more affordable or accessible, showed varied results. Irvine et al. (2011) conducted a telephone-based CBT in patients with an implantable cardioverter defibrillator installed and reported significant improvements in PTSD symptoms in the CBT group [43]. Mouthaan et al. (2013) developed a self-guided internet-based intervention (Trauma TIPS) based on CBT to prevent the onset of PTSD symptoms. The result did not support the efficacy of Trauma TIPS [44]. Two negative studies of five sessions of early telephone-based CBT have recently been submitted for publication (O'Donnell ML, Shalev AY personal communication).

CBT is currently the mainstay of early prevention of PTSD. Several considerations, however, make its systematic implementation a major challenge. First, CBT might not be needed in a large proportion of symptomatic survivors. A meta-analysis of early interventions has indicated that CBT is only efficient in participants with diagnosable PTSD at treatment onset [45, 46], and results from the Jerusalem Trauma Outreach and Prevention Study (J-TOPS) have similarly shown that survivors with sub-threshold PTSD symptoms equally recover with or without CBT [41•]. A study by Rothbaum et al. [36•] suggests that efficacy of early CBT is strongly dependent on the type of traumatic event. CBT was most effective for sexual assault victims, had a marginal effect among accident victims, and was not effective for victims of physical assault. Additionally, CBT was found to equally reduce chronic PTSD symptoms when delivered 1 or 5 months after the traumatic event [41•] and its initial effect was conserved for 3 years [42].

CBT is consequently best positioned as a clinical intervention for identified and ascertained acute PTSD cases. It is optimally provided at some distance from the traumatic event, during which time survivors with transient symptoms recover. Survivors who recover with early CBT seldom relapse spontaneously but could remain sensitive to subsequent exposure. Importantly, early CBT leaves numerous survivors unimproved (e.g., 20 % of those treated in the J-TOPS) and thus should be supplemented by ‘second-step’ interventions, which unfortunately are very poorly mapped, if at all. CBT is therefore a ‘must try’ in symptomatic trauma survivors, for many of whom it may shorten symptom duration by months and years.

Hydrocortisone has been shown to be effective especially in patients who have never been treated for psychiatric disorders [48•]. A study recruited 64 trauma survivors in a level I trauma center and randomly assigned them to hydrocortisone and placebo group. At 3 months post trauma, no (0 %) hydrocortisone recipient and 3 (14 %) placebo recipients met full PTSD diagnostic criteria. PTSD symptom severity decreased over time in both groups, with hydrocortisone recipients reporting lower Clinician-Administered PTSD Scale (CAPS) scores than the placebo group (19.4 ± 4 vs.31.3 ± 3) [48•]. The underlying mechanism has not yet been established. One hypothesis is that hydrocortisone can facilitate extinction learning through both non-genomic and genomic effects [49]. Some also believe that high-dose exogenous hydrocortisone administered shortly after trauma may promote recovery through enhancing synaptic plasticity and connectivity. An animal model showed significantly increased dendritic growth and spine density, with increased levels of brain-derived neurotropic factor and decreased postsynaptic density protein 95 expression in steroid-treated stressed rats [57].

Propranolol is a beta-adrenergic antagonist that crosses the blood-brain barrier and, therefore, capable of reducing the central nervous system adrenergic drive associated with defensive threat responses. Experimental studies of propranolol in healthy subjects have shown that its administration prior to exposure to potentially traumatic narratives reduced the recollection of stressful elements of the narrative without affecting the general recall [58]. It was thereby positioned as a prime candidate to affect traumatic recall in PTSD. Early treatment with propranolol aims at preventing the over-consolidation of traumatic memories by blocking the memory-enhancing influence of stress hormones [59]. It therefore has to be started while memories of the trauma are still being formed and consolidated, preferably within hours of the traumatic event. An initial small-scale (n = 31) pilot study showed an efficacy of propranolol in reducing physiological responses to mental imagery of traumatic events 3 months after the event—but not on PTSD symptoms [60]. Two subsequent controlled studies [61, 62•], including one from the group that published the original study have also failed to show a preventive effect of propranolol. Because of its significant promise and strong theoretical basis, the documented gap between propranolol’s effect on physiological responses to trauma reminders (via mental imagery) and its lack of effect on PTSD symptoms might be interpreted as suggesting that the acquisition of traumatic memories in PTSD is not limited to amygdala-mediated threat conditioning and involves other modes of learning and memory [63].

Benzodiazepines are gamma-amino butyric acid agonists and thereby enhance inhibitory transmission in many areas of the brain. They are used as tranquilizers and sleep inducers—also interfering with long-term potentiation and therefore with learning — and, in acute administration, with memory acquisition. They were positioned as capable of reducing excessive trauma-related learning, but then mostly tried for possible preventive effects on PTSD during the aftermath of traumatic events despite having no known effect on retrograde recall. In a prospective case-control study of 13 trauma survivors treated with the benzodiazepines clonazepam or alprazolam and 13 control cases matched by gender and symptom severity, benzodiazepine-treated patients were three times more likely to have PTSD at 6 months [64], a result which was replicated in another sample [65]. A ‘predator stress’ study in rodents similarly found that administering diazepam shortly after predator - odor exposure enhanced the acquisition of long-term fear responses (avoidance of open maze and exaggerated startle) [69]. While the exact mechanism of benzodiazepines’ PTSD-enhancing effect is unknown, it is possible that these compounds interfere with extinction learning, a critical phase in threat-response extinction, particularly when administered hours and days after trauma exposure. Remarkably, current evidence on benzodiazepines’ effect relies on small case series, whereas these compounds are widely used to mitigate acute response to stress. Further evidence is clearly needed, including the use of benzodiazepines to affect traumatic recall within minutes or hours from trauma exposure, that is, within the putative memory consolidation phase.

Animal studies suggest that morphine can produce retrograde amnesia for contextual conditioned fear, possibly through decreasing cyclic adenosine monophosphate or activating N-methyl-D-aspartate receptors in the hippocampus [70]. Observational studies of hospital patients also suggested a possible beneficial effect of morphine administration within 48 h after trauma exposure to survivors who experience pain, reducing the likelihood of a PTSD outcome [71, 72]. A similar result was reported retrospectively in 696 military personnel with severe combat injury, in which not having PTSD was associated with higher likelihood of having received morphine—as per participants’ medical charts [73]. Given the retrospective nature of most studies, more research is needed to separate a specific effect of morphine from a generic ‘analgesic’ effect. Pain after trauma exposure is a potent predictor of PTSD. It is unclear, therefore, whether morphine has any preventive value in trauma survivors without physical pain.