The personal, social, and economic costs of posttraumatic stress disorder (PTSD) have stimulated enormous efforts over the past several decades into developing better strategies to reduce the adverse psychological effects of trauma. Much of this energy has been devoted to early intervention strategies. By early intervention, we mean interventions that are implemented in the initial hours, days, or weeks after trauma exposure. The goals of these approaches are variably to reduce the acute stress or to achieve secondary prevention to avert subsequent PTSD. In this chapter I will review the current strategies for early intervention, the evidence for these strategies, and the challenges that are facing the field.

Over the past three decades, there has been a very common trend to provide immediate psychological response to all survivors of trauma. This approach has been driven by the perception that all people are vulnerable to adverse effects of trauma, and if we do not provide interventions shortly after the trauma exposure, dire results will occur. However, this presumption of vulnerability has been shown to be wrong. Most people are highly resilient and do not require formal mental health interventions. People are able to adjust to the experience once the immediate threat passes and use their own resources or social networks to adapt successfully.

Despite this pattern, perhaps early interventions may facilitate adaptation or perhaps assist people who would otherwise develop psychiatric problems to avert these outcomes. Although there have been variants of early interventions for 100 years (Shephard 2000), including interventions provided to soldiers in the immediate aftermath of combat, they have been most popular since the 1980s. These interventions are most commonly referred to as ‘psychological debriefing’, and there are numerous iterations of them. The simplest means of evaluating the role of psychological debriefing is to review the history and efficacy of its most popular variant: Critical Incident Stress Debriefing (CISD). This programme was initiated in the 1980s by a firefighter, Jeffrey Mitchell, who argued that it could ‘generally alleviate the acute stress responses which appear at the scene and immediately afterwards and will eliminate, or at least inhibit, delayed stress reactions’ (Mitchell 1983, p. 36). CISD typically consists of a single debriefing session and is usually administered within several days of the trauma exposure. The session typically lasts between 3 and 4 h (Everly and Mitchell 1999). It formally comprises seven phases. The introduction phase introduces the format and clarifies that it is not intended to be psychotherapy. The fact phase asks participants to recount their accounts of the traumatic incident. The cognitive phase invites participants to describe their cognitive responses to the experience, with encouragement to be aware of initial thoughts about what occurred. Next, in the reaction phase, participants are encouraged to express emotional responses they have about the experience. In the symptom phase, the participants are asked to notice symptoms or reactions with the view of identifying stress reactions. The teaching phase involves normalising stress reactions. Finally, in the re-entry phase, a summary is provided of the debriefing session and any necessary referrals are offered. Although CISD was initially targeted toward emergency responders, it has subsequently expanded to encompass a much broader audience. Summarising the growth of CISD by the end of the 1990s, the authors wrote, ‘Mitchell’s CISD model of psychological debriefing is generally recognized as the most widely used in the world and is used across the greatest diversity of settings and operational applications’ (Everly and Mitchell 1999, p. 84).

Does CISD help people? There are many reports in the literature of CISD being well received and participants finding it helpful (Adler et al. 2008; Carlier et al. 2000). This does not amount to evidence, however. Numerous trials have been conducted that have assessed people who have and have not received CISD (or variants of it) and compared their functioning (typically in terms of PTSD levels) at subsequent follow-up assessments (for a review, see McNally et al. 2003). Overall, these studies indicate that CISD and related debriefing interventions do not result in reduced levels of posttraumatic stress relative to not receiving the intervention (Bisson et al. 2007).

Can debriefing do harm? This is a critical issue because it may be important to respect people’s natural adaption processes, and it can be unwise to interfere with these mechanisms. This issue has become particularly relevant in the light of some evidence that CISD may be harmful; this conclusion emerges from studies that found that those (particularly those who display marked PTSD symptoms initially following trauma) who received debriefing had worse PTSD than those who did not receive it (Bisson et al. 1997; Mayou et al. 2000). Another controlled study found that emotional debriefing led to a delayed recovery relative to educational input (Sijbrandij et al. 2006). Although these studies are limited by methodological flaws, it nonetheless raises concerns that global interventions may not be warranted. It has been suggested that activating trauma memories for a brief period after which the person is not seen repeatedly may not be helpful and may even further consolidate trauma memories (Bisson and Andrew 2007). Debriefing may also be harmful because it typically occurs without prior assessment, and so it involves a standard intervention for all regardless of individual differences in premorbid vulnerability, distress severity, or social context. Accordingly, treatment guidelines around the world generally recommend against this intervention (Foa et al. 2009).

If psychological debriefing, such as CISD, is out of fashion, then what is the preferred alternative? In light of evidence that a single-session intervention is not preventative of subsequent adaptation, more recent approaches simply aim at helping people cope with the acute response. The most common new approach is psychological first aid (PFA; Brymer et al. 2006). It is not really accurate to describe PFA as new because it actually contains many of the components of CISD and other forms of debriefing. It has tried to retain fundamental strategies without encouraging steps that may be unhelpful, such as encouraging emotional catharsis in the acute setting. PFA involves suggested strategies to provide safety, information, emotional support, and access to services, heighten expectancy of recovery, encourage utilisation of social support, and promote self-care. Like other forms of universal intervention, PFA does not commence with a formal assessment. This approach is now promoted in practice guidelines (Inter-Agency Standing Committee 2007), which is questionable since there is no evidence that it is beneficial. When one scrutinises PFA, it is difficult to determine how to evaluate its efficacy because it does not have clear and explicit goals that are readily tested. Whereas CISD explicitly aimed to reduce PTSD severity – which could be tested through trials – the PFA approach is somewhat unfalsifiable because assisting people to cope in the acute phase is difficult to operationalise.

In contrast to approaches that provide universal intervention to all trauma survivors, other approaches have adopted a targeted strategy that intends to focus on trauma survivors who are at high risk for subsequent PTSD. This framework presumes that we can identify people in the acute phase who will subsequently develop PTSD. Over the past several decades, much work has focused on acute predictors of longer-term PTSD – and in that time the enthusiasm for how confidently we can predict chronic PTSD has tempered a lot.

The main challenge for our field in this regard has been the recognition that acute stress reactions are not linearly related to PTSD reactions at subsequent points in time. Earlier work indicated that the initial spike in traumatic stress in the weeks after trauma exposure remitted markedly in the following months; this was shown in cohorts of survivors of rape (Rothbaum et al. 1992), nonsexual assault (Riggs et al. 1995), motor vehicle accidents (Blanchard et al. 1996), disasters (van Griensven et al. 2006), and terrorist attacks (Galea et al. 2003). This raised a challenge for early identification of trauma survivors at risk of developing PTSD because how do we disentangle transient stress reactions from the early reactions that are precursors of subsequent PTSD?

The problem of early identification is highlighted when we consider more recent research on the longitudinal course of posttraumatic stress reactions, which has underscored the complicated trajectories of posttraumatic response. Of course the most salient example of this is delayed-onset PTSD, which has traditionally been recognised as PTSD that develops at least 6 months after trauma exposure. Systematic reviews of the available evidence attest to the frequency of delayed-onset PTSD, with approximately 25 % of PTSD cases being delayed onset, with particular frequency in military populations following deployment (Smid et al. 2009). Increasing evidence tells us that the changing course of posttraumatic stress reactions can be influenced by ongoing stressors, appraisals people make, social factors, or health issues (Bryant et al. 2013). For example, in a study of survivors of Hurricane Katrina, rates of PTSD generally increased over the initial 2 years following the hurricane, which has been attributed to the ongoing stressors in the region arising from lack of infrastructure, poor housing, and lack of other necessary community resources (Kessler et al. 2008). This issue has been further highlighted by a body of evidence that has used latent growth mixture modelling to map the different trajectories that trauma survivors experience. This statistical strategy classifies homogenous groups in a population to identify class of variation over time, and rather than assuming that all people belong to a homogenous population, it tracks heterogenous patterns of response. Across a number of studies, four major trajectories have been noted: (a) a resilient class with consistently few PTSD symptoms, (b) a recovery class with initial distress then gradual remission, (c) a delayed-reaction class with initial low symptom levels but increased symptoms over time, and (d) a chronic distress class with consistently high PTSD levels. These trajectories have been noted in a range of trauma survivors, including traumatic injury (deRoon-Cassini et al. 2010), disaster (Pietrzak et al. 2013), and military personnel deployed to the Middle East (Bonanno et al. 2012). This highlights that it is not a straightforward task to identify who will eventually develop PTSD after trauma, and attempts at early intervention need to recognise that not all high-risk trauma survivors can be identified in the acute phase.

As distinct from universal interventions, more recent studies have focused on treating people with severe stress reactions. These have focused primarily on people with ASD, although some studies have treated acute PTSD. Early interventions can be divided into psychological and biological strategies. Much more work has been done in psychological interventions, arguably because of the substantial success of trauma-focused psychotherapies in treating chronic PTSD. Without doubt the front-line treatment of PTSD is exposure-based therapies, which is reflected in international treatment guidelines (Foa et al. 2009). Meta-analyses support these conclusions (Bradley et al. 2005; Roberts et al. 2009).

Early intervention following trauma has generally abridged standard CBT approaches by shortening them to five or six sessions while retaining the core content. The key commonality of these approaches is that they usually have a trauma-focused exposure as the centrepiece of treatment. Therapy usually commences with psychoeducation about the trauma responses and then focuses on anxiety management, exposure, and cognitive restructuring. Anxiety management techniques aim to reduce anxiety through a variety of techniques, including breathing retraining or relaxation skills or self-talk. Therapy usually gives most attention to prolonged exposure, which involves both imaginal and in vivo exposure. During imaginal exposure the patient is asked to vividly imagine their traumatic experience for prolonged periods, usually for at least 30 min. The therapist asks the patient to provide a narrative of their traumatic experience in a way that emphasises all sensory, cognitive, and affective details. In vivo exposure involves graded exposure to feared and avoided situations in which the patient is asked to stay in close proximity to fearful reminders of the trauma; this begins with minimally fearful situations and then increasing to more frightening situations. This approach is thought to be effective via a number of mechanisms, including extinction of initially conditioned fear responses, integration of corrective information, and self-mastery through management of exposure itself (Rothbaum and Mellman 2001; Rothbaum and Schwartz 2002). Cognitive restructuring is usually conducted following exposure and identifies the evidence for maladaptive automatic thoughts about the trauma, the person, and their future. This approach is based on much work showing that excessively negative appraisals in the acute period after trauma are strongly predictive of subsequent PTSD (Ehring et al. 2008).

Although there were some earlier attempts at early intervention (Frank et al. 1988; Kilpatrick and Veronen 1984), possibly the first study that attempted abridged forms of trauma-focused exposure approaches in the framework of early intervention was a study in which Edna Foa’s team provided brief CBT to sexual and nonsexual assault victims shortly after the assault (Foa et al. 1995). Participants received four sessions of CBT, and their responses were compared with matched participants who had received repeated assessments; although at posttreatment the CBT participants had less PTSD than those in the matched condition, this difference dissipated at 5-month follow-up. It should be noted that those receiving CBT had less depression and re-experiencing symptoms than the control participants.

One potential limitation of this study is that it focused on all trauma survivors who had symptoms severe enough to meet PTSD criteria (without the 1-month duration requirement), and we have seen that many people who are initially distressed may subsequently adapt regardless of intervention. In an attempt to address this issue, other studies have focused on people who meet criteria for ASD because of some evidence that most people who do display ASD are more at risk for subsequent PTSD (Bryant 2011). In an initial study adopting this approach, Bryant and colleagues randomised motor vehicle accident or nonsexual assault survivors with ASD to either CBT or nondirective supportive counselling (SC) (Bryant et al. 1998). Both interventions consisted of five 1½-h-weekly individual sessions. Emphasis in therapy was placed on imaginal and in vivo exposure and cognitive restructuring. Six months after treatment, there were fewer participants in the CBT group (20 %) who met criteria for PTSD relative to supportive counselling participants (67 %). A later study by the same group dismantled CBT by allocating ASD participants to five sessions of either (a) CBT (prolonged exposure, cognitive therapy, anxiety management), (b) prolonged exposure combined with cognitive therapy, or (c) supportive counselling (Bryant et al. 1999). This study found that at 6-month follow-up, PTSD was observed in approximately 20 % of both active treatment groups compared to 67 % of those receiving supportive counselling. A follow-up of participants who completed these two treatment studies indicated that the treatment gains of those who received CBT were maintained 4 years after treatment (Bryant et al. 2003a).

Since these early studies, a range of studies have followed that have essentially replicated these findings. One study randomised civilian trauma survivors (N = 89) with ASD to either CBT, CBT associated with hypnosis, or SC (Bryant et al. 2005). Hypnosis was employed because some theories posit that hypnosis may facilitate emotional processing in a condition that is characterised by dissociative symptoms (i.e. ASD) (Spiegel et al. 1996). Individuals with ASD have been shown to be particularly skilled at using hypnosis (Bryant et al. 2001), and so this study used hypnosis immediately prior to imaginal exposure with the suggestion to facilitate processing of trauma memories. All participants received the same number of sessions and identical length of exposure, with the exception that some participants received the hypnotic induction prior to the exposure exercise. Regarding those who completed treatment, more participants who received SC (57 %) had PTSD criteria at 6-month follow-up than those who received CBT (21 %) or CBT + hypnosis (22 %). Participants in the CBT + hypnosis condition did have fewer re-experiencing symptoms at posttreatment than those in the CBT condition, suggesting that the hypnosis may have facilitated the exposure component. Another study of ASD participants (N = 24) who sustained mild traumatic brain injury compared the relative efficacies of CBT and SC in people who lost consciousness as result of their traumatic injury (Bryant et al. 2003a). Fewer participants receiving CBT (8 %) met criteria for PTSD at 6-month follow-up than those receiving supportive counselling (58 %). In the largest study of ASD to date, Bryant and colleagues randomly allocated 90 civilian trauma survivors to either (a) imaginal and in vivo exposure, (b) cognitive restructuring, or (c) assessment only (Bryant et al. 2008b). Exposure therapy led to lower levels of PTSD, depression, and anxiety at 6-month follow-up compared to other conditions.

It is important to be aware that other teams have also demonstrated the utility of early intervention using trauma-focused interventions. Jon Bisson and his team randomised 152 injury survivors to receive four sessions of CBT or no intervention in the first 3 weeks after the trauma (Bisson et al. 2004). Those who received the active intervention had lower PTSD symptoms at 13-month follow-up. This approach focused on severe acute PTSD symptoms, rather than ASD, which is more consistent with the DSM-5 definition of ASD by not requiring the specific clusters of ASD symptoms. Similarly, early provision of CBT has been shown to be beneficial in other studies that have targeted trauma survivors with elevated PTSD levels (Lindauer et al. 2005).

Rather than using a prolonged exposure approach, several studies have focused on more cognitively oriented approaches. Echeburua and colleagues provided 20 participants with acute posttraumatic stress with either cognitive restructuring and coping skills training or progressive relaxation training (Echeburua et al. 1996). There were no differences between conditions at posttreatment; however, cognitive restructuring led to less severe PTSD symptoms 12 months later. An Israeli study provided two sessions of SC or CBT that aimed to promote memory reconstruction to facilitate recovery in 17 survivors of road accidents (Gidron et al. 2001). This study used an entry criterion of a resting heart rate higher than 94 beats per minute at admission to the emergency room (on the basis that elevated heart rate in the acute phase is predictive of PTSD; Bryant et al. 2008a; Shalev et al. 1998). Treatment was delivered by the telephone 1–3 days after the accident. Patients who received the CBT intervention had less PTSD 3–4 months after the trauma than did those who received SC.

In an important study that compared early and later intervention, Shalev randomised 242 patients who were admitted to an emergency department and met criteria for either full or subsyndromal ASD. Participants were allocated to either prolonged exposure, cognitive restructuring, wait list (who were subsequently randomised to exposure or cognitive restructuring after 12 weeks), escitalopram (SSRI), or placebo (Shalev et al. 2012). At 9-month follow-up, PTSD outcomes did not differ between prolonged exposure (21 %) and restructuring (22 %) conditions; in contrast, there were higher rates in the SSRI (42 %) and placebo (47 %) conditions. There were no longer-term differences between participants who received the early or later therapy. One important implication of this study is that early intervention is not essential for optimal outcome and one can provide therapy some time later and achieve the same outcomes in the long run. This is not to minimise the benefit of early intervention, however, because there are clear advantages in reducing the stress (and associated problems) that can occur in the intermediate phase after trauma.

Another study commenced exposure therapy very soon after trauma by administering it in the emergency room for trauma patients and then repeating it weekly over the following 2 weeks (Rothbaum et al. 2012). Compared to an assessment-only comparison condition, patients who received exposure therapy had reduced PTSD at the 3-month follow-up. The interesting implication that emerges from this study is that it indicates that exposure therapy can be safely commenced very soon after trauma exposure.

Other controlled trials that have applied CBT to acute PTSD have recruited patients within 3 months of the trauma and therefore not provided therapy in the very acute phase after trauma. These studies also point to the utility of early intervention using exposure-based approaches. Each of these studies shows moderate to large effect sizes of PTSD symptom reduction (Ehlers et al. 2003; Sijbrandij et al. 2007).

Notably, not all studies have reported beneficial effects of trauma-focused early intervention. One large study randomised 90 female assault survivors who had acute PTSD symptoms to either prolonged exposure, SC, or a repeated assessment within 4 weeks of trauma (Foa et al. 2006). Nine months after treatment, all participants had made similar gains in terms of reducing PTSD; this finding did not change when only those who met ASD criteria were included. Another null finding was found in a study that used a writing paradigm, in which 67 traumatic injury patients were randomised to either a trauma writing intervention group or an information control group (Bugg et al. 2009). There were no differences between the groups at either posttreatment or follow-up assessments.

In summary, early provision of trauma-focused therapies, and particularly those that encourage emotional processing, appears to be efficacious in preventing subsequent PTSD. This conclusion is supported by systematic review of the available evidence (Ponniah and Hollon 2009). It is important to qualify this claim, however, by noting that a substantive proportion of people do not respond to early intervention, and so it should not be regarded as a panacea for posttraumatic problems.

Apart from psychological interventions, a handful of studies have also explored the potential for early provision of pharmacological interventions. Most of these approaches are rooted in fear conditioning models in the weeks after trauma as key mechanisms that lead to PTSD (Rauch et al. 2006). These models posit that when a traumatic event (unconditioned stimulus) occurs, people respond with fear (unconditioned response); this elevated fear leads to strong associative conditioning between the fear response and the stimuli associated with the trauma. Reminders of the trauma (conditioned stimuli) can then trigger fear reactions (conditioned response), including re-experiencing symptoms. This strong response to the traumatic event involves the release of stress neurochemicals (including norepinephrine and epinephrine) into the cortex, leading to an overconsolidation of trauma memories. Sensitisation can occur from repetitive distress, which elevates sensitivity of limbic networks, and this leads to increasing reactivity to trauma-related stimuli (Post et al. 1995).