Factors affecting assignment to psychotherapy. Major clinical guidelines universally recommend psychotherapy as a first-line treatment for PTSD and there is no question that it can be efficacious; nevertheless, how to decide when to refer patients with PTSD for psychotherapy and which psychotherapy to recommend to any given patient are not questions that are easily answered with the available evidence. There are limited data regarding how clinicians and patients navigate these issues.
Key factors that some experienced Veterans Administration (VA) therapists take into consideration when deciding whether or not to start exposure-based CBT include patient readiness (eg, willingness or commitment to engage, presence of coping skills, degree of suicidality, safety and stability in terms of home environment or lifestyle) and comorbid conditions (eg, substance use disorders, paranoia, personality disorders). They use different decision-making processes falling into two classes—chiefly provider-driven decision-making processes (eg, clinical judgment, decision-making tree) or patient-involved decision-making in which the clinician shares the best available evidence with the patient and assists him or her to achieve informed preferences.16
It has been suggested that, if psychotherapy is indicated, it may be important to start as soon as possible as to not collude with the avoidance features of PTSD. Furthermore, patient time and clinic resources may be best used by encouraging providers to use more shared decision-making with their patients, as opposed to making unilateral decisions about readiness and to providing preparatory information or instruction to them that may be redundant with exposure-based CBTs for PTSD.16
Shared decision-making regarding psychotherapy for PTSD is influenced by the attitudes of both the clinician and the patient toward the different options. Becker et al.17 randomly mailed surveys to 852 doctoral level psychologists (217 were returned and 207 were judged to be usable) to learn about their attitudes toward and use of exposure therapy for treating PTSD. Despite the extensive evidence supporting the efficacy of exposure therapy for PTSD, 69% of respondents had received no training in imaginal exposure (IE) for PTSD and only 17% and 11% reported current use of IE and in vivo exposure, respectively, for treating PTSD. Of those who reported using IE, less than half used it in the majority of their PTSD cases. Formal training in IE and greater experience in treating PTSD were associated with higher utilization of IE, although the majority of even the most experienced clinicians used it in a minority of PTSD cases. Contraindications to using IE cited by respondents included severe suicidality (85%), a comorbid psychotic disorder (85%), homicidality (81%), severe anger (54%), physical aggression by other to patient (52%), dissociation (51%), any comorbidity (37%), and a comorbid anxiety disorder (32%). The study authors17 noted that these contraindications had little to no supportive empirical evidence. The potential complications of using IE for treating PTSD endorsed as being likely to occur included increases in arousal (87%), re-experiencing symptoms (83%), dissociation (76%), substance abuse (75%), suicidality (75%), self-injury (68%), and desire to drop out of therapy (59%). Factors reported as being most likely to lead therapists to not use IE for treating PTSD included limited training (60%), preference for individualized over manualized therapy (25%), and concern that the patient would decompensate (22%). The results suggest that exposure therapy might be recommended by psychotherapists for PTSD more often if formal training in it becomes more widely disseminated and if further research clarifies what the contraindications to, and likely complications of, exposure therapy for PTSD are, as well as how to manage them.
The results of a recent analysis18 of the treatment preferences of patients (N = 110) in an RCT comparing prolonged exposure (PE), interpersonal psychotherapy (IPT), and relaxation therapy for chronic PTSD suggested that patients may also favor alternatives to exposure therapy in many cases. After the researchers gave individual people a description of each therapy approach, including the degree of empirical support (which was greatest for PE), 50% of the patients preferred IPT, 26% PE, and 26% relaxation therapy. On the other hand, 3%, 26%, and 16% were disinclined toward IPT, PE, and relaxation therapy, respectively. Patients preferring IPT had a higher prevalence rate of chronic trauma. Factors associated with preferring PE included earlier age and longer duration of the chief trauma, poorer baseline social functioning, and higher prevalence rates of sexual and interpersonal abuse and recurrent depression. Women were significantly more likely than men to express an opinion (ie, preference or disinclination) about prolonged exposure (odds ratio [OR] 4.59). Patients with lower-than-median education levels preferred relaxation therapy more than those with higher ones (OR 3.29), and those who expressed an opinion about it were more likely to have had psychotherapy in the past than those who did not. Interestingly, patients with no treatment preferences had higher mean clinician-rated PTSD symptoms at baseline and trended toward improving more than those with preferences. Those who received undesired therapy showed significantly lower treatment response rates relative to patients who did not. Among people with comorbid MDD, receiving undesired treatment was significantly associated with having worse clinician-rated PTSD symptoms after treatment relative to those who had no opinion or received desired treatment. The study may have been underpowered to identify other interactive effects of preference, disinclination, or lack of opinion regarding psychotherapy and treatment assignment on outcomes. Further research is needed to determine what accounts for the attitudes of patients with PTSD toward psychotherapy and how they affect their treatment choices and results.
Overview of the evidence for psychotherapy. The American Psychiatric Association, Institute of Medicine (IOM), International Society for Traumatic Stress Studies, NICE, National Health and Medical Research Council, and VA/Department of Defense (DoD) have developed treatment guidelines for the treatment of PTSD in adults using various methods. All of them agree that trauma-focused psychotherapies, including exposure therapy and cognitive therapy, are evidence-based, efficacious first-line treatments, and all rated eye movement desensitization and reprocessing (EMDR) a primary or secondary level of recommendation except for the IOM guidelines, which only endorsed exposure therapy as a first-line therapy and found insufficient evidence to recommend EMDR. Recognizing that these guidelines were becoming somewhat dated and had various methodologic shortcomings, Cusack et al.19 recently conducted a systematic review and meta-analysis of psychological treatments for PTSD including 64 RCTs (N = 5,595) that generally enrolled patients with severe or extreme, mostly chronic PTSD. The authors concluded that exposure therapy should be a first-line treatment based on the high strength of evidence and large magnitude of benefit. They also recommended cognitive therapy (including cognitive processing therapy [CPT]), mixed CBT, Narrative Exposure Therapy, and EMDR in view of their large effect sizes. EMDR was supported by low-strength evidence, whereas the others had moderate-strength evidence (Table 2). The recommended therapies frequently also reduced symptoms of depression or anxiety.
Efficacies of Selected Psychotherapies for PTSD
A recent review20 of RCTs of psychotherapy for military-related (eg, combat, military sexual trauma) PTSD found that effect sizes were large (0.78–1.10) in comparison to waitlists and treatment as usual for the most frequently studied approaches, CPT (five RCTs with N = 481) and PE (four RCTs with N = 402), but that they were only marginally better than nontrauma-focused psychotherapy (ie, present-centered therapy). Although 49% to 70% of patients attained clinically meaningful symptom improvement, the mean posttreatment PTSD scores remained at or above clinical cutoffs for the diagnosis of PTSD. Indeed, roughly two-thirds of patients retained their PTSD diagnoses and only 15% to 27% achieved remission after treatment with one of these therapies. Dropout rates were 16% to 35% and 13% to 39% for CPT and PE groups, respectively. Mean symptom scores remained at or above clinical cutoffs for the diagnosis of PTSD after treatment with the much less well-studied EMDR, although there were significant symptom reductions. EMDR was comparable to variants excluding eye tracking and to active nontrauma-focused psychotherapy. In another study, EMDR outperformed biofeedback relaxation and a waitlist condition in terms of loss of categorical diagnosis and the results were maintained at the 9-month follow-up visit. The authors concluded that the efficacy of EMDR remains largely based on civilian studies and that 12 sessions of manualized trauma-focused CBT or nontrauma-focused present-centered therapy is generally insufficient for treating military-related PTSD.
Psychotherapy for child abuse-related PTSD in adults. A recent meta-analysis21 of 16 RCTs (N ≈ 1,250) of psychotherapies for PTSD in adult survivors of child abuse found that the effect sizes were moderate compared to treatment as usual and moderate-to-large compared to waitlist controls. The effect sizes remained stable at the follow-up visit. In general, trauma-focused treatments were more efficacious than nontrauma-focused ones and treatments including individual sessions had larger effect sizes than group-only ones. As such, the best effects could be achieved with individual trauma-focused CBT.
Recognizing that CBT designed for PTSD did not specifically address interpersonal or social adjustment or emotion regulation dysfunction (eg, easy provocation, high reactivity to emotionally evocative stimuli, difficulty calming down) and that clinicians were avoiding using exposure therapy due to unfounded fears of potential adverse effects, Cloitre et al.22 developed a phase-based treatment for PTSD in which Skills Training in Affect and Interpersonal Regulation (STAIR) precedes standardized exposure therapy with the hypothesis that improvements of day-to-day life competencies might enhance the efficacy of, and protect against potential adverse consequences of, subsequent trauma processing. An RCT22 showed that this approach was superior to a waitlist condition for improving PTSD symptoms, emotional regulation, and interpersonal problems in women with PTSD related to child physical or sexual abuse (N = 58). A later RCT23 involving women with child abuse-related PTSD (N = 104) showed that narrative story telling (NST; ie, exposure therapy) preceded by STAIR (ie, STAIR/NST) was more beneficial and had fewer adverse effects than NST preceded by supportive counseling (SC/NST) and had similar adverse effects to STAIR followed by SC (ie, STAIR/SC). More specifically, STAIR/NST was superior to SC/NST in achieving lasting remission and loss of categorical diagnosis of PTSD and in reducing problems in emotion regulation, anger expression, and anxiety. Furthermore, it was better than the other treatment conditions for improving interpersonal problems. PTSD symptom improvements were more persistent in the STAIR/NST group than in the other two treatment conditions. In terms of negative effects, the STAIR/NST dropout and second phase symptom exacerbation rates were lower than those for SC/NST and similar to those of STAIR/SC. The results supported the notion that exposure therapy is safe and effective for women with PTSD related to child abuse, but could potentially be even more effective if preceded by dialectical–behavior therapy-like skills training such as STAIR. Cloitre et al.24 reanalyzed the data from this study and found that the STAIR/NST group showed greater PTSD symptom improvements at all baseline levels of dissociative symptoms (ie, derealization, depersonalization, amnesia). The proportions of people with high levels of residual dissociation at completion of treatment were about equal among the three treatment groups. Unlike those from the other groups, those with high residual dissociation from the STAIR/NST group continued to exhibit improvement in PTSD symptoms. Interestingly, people with high dissociation were less likely to drop out of treatment than those with low levels.
Safety and tolerability of exposure therapy. Cloitre et al.22–24 provided evidence that exposure therapy could be efficacious, tolerable, and safe, even for those with high levels of trauma-related dissociation. Other recent studies have addressed the tolerability and safety of exposure therapy, yielding results that are discrepant with commonly expressed concerns that exposure therapy might result in symptom exacerbation during treatment or in other ways worsen the outcome of PTSD. Thus, a study25 involving women veterans with PTSD (N = 192) who were victims of interpersonal violence (rape or sexual or physical assault at least 3 months prior) compared three CBT treatment conditions with differing doses of exposure (ie, PE, CPT, and CPT without trauma narrative writing) to understand the association between symptom exacerbations and treatment outcome and completion rates. They found that symptom exacerbation occurred in a minority of patients; was usually short-lived; was not significantly associated with the type of treatment, amount of exposure involved, or onset of exposure exercises; and did not stop patients from achieving substantial decreases in PTSD symptoms or predict worse outcomes. Of note, a history of childhood sexual or physical abuse was not associated with a greater likelihood of symptom exacerbation.
Bryan et al.26 studied active duty military personnel with PTSD (N = 108). With a focus on suicidal ideation and suicide, they compared trauma-focused group CBT with an active nontrauma-focused psychotherapy. They found that suicide-related outcomes were rare (even among those who entered treatment with suicidal ideation), similar in both treatments, and primarily associated with comorbid depression. Rates of suicidal ideation decreased over the course of treatment and new-onset suicidal ideation was rare in both groups. There were no suicide attempts during treatment or 12 months of follow-up in either group.
Although improving emotion regulation skills in patients with PTSD prior to exposure therapy may be helpful in some cases,23 it may not always be indicated. An RCT27 involving civilians with chronic PTSD (N = 200) found that sustained improvements in emotional regulation (ie, expressive suppression, cognitive reappraisal, and negative mood regulation) were achieved with 10 weeks of PE or sertraline (dose range, 0–300 mg per day; mean dose, 115 mg per day); those who had the poorest pre-treatment emotional regulation skills improved the most, emotion regulation skills were only moderately associated with PTSD symptoms, and numerous patients had emotional regulation skills within the normal range. The results suggest that those who enter PTSD treatment with good emotional regulation have little to gain from additional emotion regulation-focused interventions and that emotion regulation improvements can occur without them.
Virtual reality exposure-based therapy. A recent review28 including 12 trials (N = 197; range, 9–125) of virtual reality exposure-based therapy (VR-EBT) versus another treatment (eg, PE, IE, CPT, EMDR) or waitlist control for PTSD suggested it may have several advantages over traditional exposure-based therapies: (1) it may be more engaging than non-VR exposure for people with limited imaginal capacities; (2) it offers more confidentiality than in vivo exposures tasks; (3) therapists have a higher degree of control over the exposure experience and can therefore prevent unpredictable or unwanted stimuli or may repeat certain stimuli many times; (4) patients may be exposed to stimuli that otherwise would be difficult to access; and (5) it may be more acceptable to certain patients, improving their retention in therapy. VR-EBT has been shown to improve PTSD symptoms related to various traumas (eg, abuse, assault, robbery, mobbing, domestic violence, motor vehicle accident) and to be as efficacious as traditional treatments (eg, IE, in vivo exposure, CBT, present-centered therapy). Most studies completed to date have involved veterans, limiting generalizability. Currently, most therapists have limited access to VR-EBT resources and training.
Internet-based cognitive-behavioral therapy. An unblinded RCT29 comparing a web-based CBT intervention to a waitlist condition for improving PTSD symptoms in war-traumatized patients (N = 159) in Iraq found that the intervention had a large effect size (d = 0.92) for sustainably reducing PTSD symptoms and that intervention completers were significantly more likely to achieve symptom remission. These results provide low level evidence that a web-based CBT intervention can effectively treat PTSD even in an unstable and unsafe setting.
Eye movement desensitization and reprocessing. A recently completed meta-analysis30 of 11 studies (N = 424) comparing EMDR and CBT for treating PTSD in adults concluded that EMDR had a borderline significant, small-to-moderate effect size advantage over CBT for reducing intrusive and arousal symptoms, but that there was no evidence of a difference between the two in reducing avoidance symptoms. More rigorous and less heterogeneous studies are required to verify the finding.
Psychodynamic psychotherapy. In a recent review of the effectiveness of psychodynamic psychotherapies, Fonagy31 noted that there was one study of psychodynamic psychotherapy for PTSD that showed a significant reduction of intrusion and avoidance symptoms compared to waitlist controls, an effect similar to hypnotherapy and trauma desensitization. Three prior systematic reviews found insufficient evidence for the efficacy of psychodynamic psychotherapies in the treatment of PTSD. Subsequently, staff at the Unit for Treatment of Combat-Related PTSD of the Israel Defense Forces completed a naturalistic comparative effectiveness study32 of 24 weekly CBT (CPT + PE) sessions versus 50 weekly psychodynamic psychotherapy sessions for treating men with chronic combat-related PTSD (N = 243). Treatment group assignment was based on senior therapist consensus regarding which approach was most complementary to the patient's overall problem set. The patients in the two groups had similar demographics and exhibited significant and comparable improvement in terms of self-reported and clinician-rated PTSD symptoms, depression, and functional impairment, and their improvements were maintained after 8 to 12 months of follow-up. About one-third of patients in each group experienced remission of symptoms at follow-up. Nevertheless, mean group clinician-rated PTSD symptoms remained above threshold for diagnosis in both groups. Average improvement in functioning was from poor to fair. It could not be determined if the group outcomes were similar due to differences in the number of weekly sessions, nonspecific therapy factors (eg, therapeutic alliance, positive attitude, empathy), or the treatment assignment protocol used.
Interpersonal psychotherapy. Markowitz et al.33 conducted a 14-week RCT comparing IPT, which is a nonexposure-based, non-CBT known to be efficacious for MDD, to PE and relaxation therapy for the treatment of chronic PTSD of moderate or greater severity in unmedicated adult outpatients (N = 110; n = 2 combat veterans), half of whom had comorbid MDD. PE, IPT, and relaxation therapy each significantly improved clinician-rated PTSD symptoms and had large within-group effect sizes (d = 1.88, d = 1.69, and d = 1.32, respectively). PE was not statistically significantly superior to IPT (34.6 vs 29.1), but was relative to relaxation therapy (34.6 vs 22.4), in terms of reductions in mean clinician-rated PTSD symptom scores. IPT had a significantly greater response rate (63%) and a nonsignificant trend toward one relative to relaxation therapy (38%) and PE (47%), respectively. There were no significant differences in remission rates between the PE (26%), IPT (23%), and relaxation therapy groups (22%). Of note, PTSD symptoms improved faster in the PE group than in the IPT group. IPT and PE were significantly superior to relaxation therapy and similar in magnitude in improving self-reported PTSD symptoms, quality of life, social functioning, and interpersonal problems. Overall, dropout rates in the IPT (15%), prolonged exposure (29%), and relaxation therapy (34%) groups were not significantly different; however, dropout rates among people with co-morbid MDD in the PE group (50%) were significantly higher and trended toward being so relative to the relaxation therapy (26.7%) and IPT groups (20%), respectively. In contrast, dropout rates among people without MDD trended toward being higher in the relaxation therapy group (35.3%) compared to the PE (5.6%) and IPT (10%) groups. There were no significant differences between the groups in terms of adverse events. These results suggest that IPT may be about as efficacious as, and may have better retention rates than, PE among patients with PTSD, particularly those with comorbid MDD. It also adds to the growing evidence that intentionally including exposure may not be necessary for effective treatment of PTSD.
Mindfulness-based therapy. A recent review34 of mindfulness-based interventions included three small RCTs (N = 32–62) that evaluated the efficacy of mindfulness-based stress reduction (MBSR) techniques for PTSD. Results were positive overall, but there was no evidence that mindfulness-based techniques were more effective than standard stress-inoculation treatments and no clear conclusions could be drawn given the heterogeneity and lack of methodologic rigor of the studies. A subsequently reported RCT35 evaluating the efficacy of MBSR versus present-centered group therapy for PTSD among veterans (N = 116) found that the MBSR group faired better in terms of improvements in PTSD symptoms, quality of life, and mindfulness at the 2-month follow-up visit, but about one-half of the patients in each group continued to fulfill the diagnosis of PTSD in the same time frame. Overall, MBSR appeared to have a modest advantage over present-centered therapy. Significant methodologic problems with the study beg caution in interpreting these results.
Hypnotherapy. The results of two recent meta-analyses, one36 including six studies (N = 391) and another37 including four studies (N = 160), provided some evidence that hypnotherapy techniques could produce large effect sizes in the treatment of adults with PTSD.
Body-based therapies. Corrigan and Hull2 recently argued that body-based therapies (eg, somatic experiencing, sensorimotor psychotherapy) may have certain advantages for the treatment of PTSD, particularly complex PTSD, over the so-called talking therapies, such as PE, that focus on retraining the higher brain to regulate subcortical disturbances (eg, amygdala hyperactivity). They speculated that by not narrowly focusing on effortful regulation of distress, by acknowledging traumatic sensorimotor memories and their associated emotions that may not easily be expressed in words, and helping people to find feelings of safety in their bodies that body-based therapies may be more acceptable to certain patients and may allow for processing of material that otherwise would not be accessed or tolerated.
Along these lines, a review38 of emerging treatments for PTSD found moderate-strength evidence supporting acupuncture, emotional freedom therapy, mantra-based meditation, and yoga. These approaches are largely theoretically grounded in mind-body philosophies, do not directly address cognitions or emotions, and do not involve talk therapy. Larger, more rigorous RCTs with longer follow-up periods and dismantling studies are needed to verify the efficacies of these approaches and identify their active components, respectively.
Physical activity. A recent meta-analysis39 of four RCTs (N = 200) found that physical activity (discussed here for simplicity, although it may not be best categorized as a psychotherapy) was significantly more effective than placebo for reducing PTSD and depressive symptoms among those with PTSD. These findings suggest that physical activity should be an important aspect of the multidisciplinary management of PTSD as it may improve PTSD symptoms and protect against psychiatric and medical comorbidities. Physiotherapists and exercise therapists may be important providers of clinical exercise programs to patients with PTSD.
Psychotherapies for insomnia and nightmares in PTSD. Two of the most common and distressing symptoms of PTSD are insomnia and recurrent nightmares. They are reported by about 70% and 52% to 96%, respectively, of people with PTSD.40 Recurrent nightmares can persist for several decades.41 Evidence suggests that sleep disturbances exacerbate the waking symptoms of PTSD and may increase the risk for other serious problems (eg, substance abuse, poor health status, depression, suicide).40,41 However, few studies of first-line treatments for PTSD have adequately (eg, used validated sleep measures) examined their efficacy for improving PTSD-related sleep disturbances. An important point to bear in mind is that in any given case of PTSD, a sleep disturbance may represent a symptom of PTSD, a comorbidity (eg, obstructive sleep apnea, periodic limb movement disorder, rapid eye moment sleep behavior disorder, depression, anxiety, substance abuse), or a medication side-effect (eg, antidepressant-induced periodic limb movements).
One 8-week RCT (N = 45) comparing CBT for insomnia (CBT-I) to wait-list and two smaller uncontrolled trials (N = 5 and N = 8) supported the efficacy of CBT-I for improving insomnia. There is evidence from various, largely uncontrolled studies that imagery rehearsal (IR), including exposure, relaxation, and rescripting therapy, is efficacious in both civilians and veterans for long-term reductions in nightmares. Most studies of psychotherapy for PTSD-related sleep disturbances used some combination of CBT-I and IR, obscuring their individual effects. CBT-I and IR should be considered first-line treatments for PTSD-related insomnia and nightmares, respectively, and their combination may be helpful for many people (Table 3).40
Recommended Treatments for PTSD-Related Sleep Disturbances
Psychotherapy for PTSD with specific comorbidities. A number of recent studies have explored the efficacy of psychotherapeutic interventions for PTSD with specific comorbidities. One small uncontrolled pilot study42 of the efficacy of MBSR in veterans with PTSD and a history of mild traumatic brain injury (N = 9) showed a moderate effect size for improvements in attention and a large effect size for improvements in PTSD symptoms that were sustained at the 3-month follow-up visit, lending low level evidence that MBSR might be efficacious for simultaneously improving both neuro-cognitive and PTSD symptoms.
In addition, a single-blind RCT43 comparing PE, EMDR, and a waitlist control for treating patients with current PTSD and a history of a psychotic disorder (N = 155) found that PE and EMDR were superior to waitlist in terms of reduction of PTSD symptoms (effect sizes 0.78 and 0.65, respectively) and loss of categorical diagnosis (OR 3.41 and 3.92, respectively) during treatment. PE was also significantly superior in terms of likelihood of gaining full remission. Treatment effects were maintained in both groups at the 6-month follow-up visit. There were no significant differences between EMDR and PE in any outcome or in terms of dropout rates (20% vs 24.5%). The results suggested that both approaches could be effective for patients with psychotic disorders and that excluding people with psychotic disorders from PTSD trials, as is often the case, may not be justifiable.
Finally, a systematic review44 of 14 RCTs (N = 1,506) of psychological interventions for PTSD and comorbid substance use disorders, 13 of which were included in a meta-analysis, found low-quality evidence that individual trauma-focused CBT delivered together with a substance use disorder intervention was more effective than treatment as usual or minimal intervention for PTSD severity at posttreatment and follow-up or for level of substance use at posttreatment; however, there was an advantage for the latter at the 5- to 7-month follow-up visit. Evidence to support nontrauma-focused individual or group-based interventions was scarce.
General pharmacotherapy for PTSD. Several major treatment guidelines state that psychotherapy should generally be preferred over medication as the initial intervention for PTSD. Reasons to consider prescribing pharmacotherapy for PTSD include comorbidities (eg, depression, psychosis), lack of access to psychotherapy, failure to tolerate or respond to psychotherapy, and the personal preference of the patient (Table 4).45 Pharmacotherapy, like most psychotherapies for PTSD, typically provides some symptomatic relief, as opposed to remission. The best evidence for pharmacotherapy for PTSD is for the selective serotonin reuptake inhibitors (SSRIs) (ie, paroxetine, sertraline, fluoxetine) and serotonin-norepinephrine reuptake inhibitors (SNRIs) (ie, venlafaxine), with paroxetine and sertraline being the only medications approved by the US Food and Drug Administration (FDA) for the treatment of PTSD in adults. Generally speaking, initial trials of these medications should be conducted for at least 12 weeks. Switching between agents or increasing their doses may be helpful, but there is currently no strong evidence base to support or guide decisions regarding these interventions. It is important for the prevention of relapse that pharmacotherapy be continued after symptoms improve, although the optimal length of treatment is unknown for any pharmacotherapy for PTSD. Medications with the best side-effect profiles should be selected and those that are ineffective should be tapered and discontinued to avoid unnecessary polypharmacy.7
Reasons to Consider Prescribing Pharmacotherapy for PTSD in Adults
A recent meta-analysis45,46 of 39 double-blind, placebo-controlled RCTs (N = 6,171) that assessed the efficacy of pharmacological monotherapy interventions relative to placebo control groups for reducing PTSD symptoms concluded that SSRIs (19 RCTs, N = 3,350) were the sole drug class for which there was sufficient evidence to calculate an overall effect size (d = −0.27, P < .05) (Figure 1). The medications that were found to be statistically significantly better than placebo were fluoxetine, paroxetine, sertraline, and venlafaxine. Paroxetine was superior to placebo both by clinician- and self-rated PTSD symptom measures and had a small-to-moderate effect size (d ≈ −0.4 for both), whereas fluoxetine, venlafaxine, and sertraline were superior to placebo only for the former and had small effect sizes (ie, d = −0.24, d = −0.20, and d = −0.16, respectively) (Table 5). Amitriptyline, GR205171 (a neurokinin-1 antagonist), mirtazapine, and phenelzine were superior to placebo in single RCTs. Selected drugs for which evidence of efficacy is lacking included alprazolam, citalopram, escitalopram, imipramine, risperidone, olanzapine, topiramate, and valproate. Olanzapine and topiramate had moderate effect sizes that were statistically insignificant, possibly due to small sample sizes. There were no major differences between the agents in terms of number of people who dropped out of the studies, suggesting that the treatments were generally well-tolerated. The effect sizes for the pharmacological agents were small, both absolutely and relative to those of trauma-focused psychotherapies versus waitlist or treatment as usual.
Meta-analysis of selective serotonin reuptake inhibitors versus placebo for the treatment of posttraumatic stress disorder. Reprinted from Hoskins et al.46 with permission from The Royal College of Psychiatrists.
Effect Sizes of SSRIs and Venlafaxine Versus Placebo for Treating PTSD
PTSD is associated with increased noradrenergic tone, raising interest in the use of alpha-2 receptor agonists (ie, clonidine, guanfacine). There is only low-quality evidence (ie, case reports, one small pilot study) that clonidine may be helpful for certain symptoms (eg, nightmares, aggression) or as an augmenting agent, and two RCTs of guanfacine for the treatment of chronic PTSD in veterans (N = 99) did not support its efficacy as a monotherapy or augmenting agent.47 Clonidine is less specific and more sedating than guanfacine, which may, in part, explain why it is more frequently prescribed for PTSD. For targeting symptoms like arousal, aggression, or sleep disturbances, these agents have side-effect profile advantages over benzodiazepines and atypical antipsychotics.
With regard to atypical antipsychotic monotherapy, a 12-week randomized, placebo-controlled trial48 of quetiapine for chronic PTSD in veterans (N = 80) showed that quetiapine (dose range, 50–800 mg per day; mean dose, 258 mg per day) was superior to placebo in improving total CAPS (Clinician-Administered PTSD Scale) score, CAPS re-experiencing and hyperarousal subscale scores, Davidson Trauma Scale total score, Hamilton Depression Rating Scale and Hamilton Anxiety Rating Scale scores, Positive and Negative Syndrome Scale global psychopathology and positive symptom scores, and Clinical Global Impressions Scale score. The average CAPS score at the end of the trial was greater than 53; thus, this study provides early evidence that 12 weeks of monotherapy with quetiapine may be efficacious for but insufficient to bring about remission in a substantial proportion of cases of chronic PTSD in this population.
Empirical evidence for the efficacy of augmentation with atypical antipsychotics in PTSD treatment is weak; nevertheless, it is not unusual for risperidone to be prescribed in this manner. Along these lines, one recent double-blind, placebo-controlled, multicenter RCT49 involving veterans with antidepressant-resistant, military-related PTSD (N = 296) found that risperidone augmentation (up to 4 mg daily) of antidepressants was associated with significant but not clinically meaningful (ie, greater than 0.5 standard deviations in magnitude) decreases in clinician-rated re-experiencing and hyperarousal symptoms and with significantly more side effects. Unfortunately, paranoia and psychosis were not sufficiently common in the sample to determine if risperidone augmentation might reduce these symptoms.
Pharmacotherapy for insomnia and nightmares in PTSD. Prazosin, an alpha-1 adrenergic receptor antagonist, is the only medication with a grade of “recommended” for treating posttraumatic nightmares in the American Academy of Sleep Medicine's best practice guidelines (Table 3). Current VA/DoD Clinical Practice Guidelines for Management of PTSD give it a strength-of-evidence rating of B for treating nightmares.41 It may act by decreasing fear responses and disrupting consolidation of fear memories. Since Raskind et al.50 conducted a double-blind, placebo-controlled crossover study showing prazosin is efficacious for alleviating sleep disturbances in Vietnam veterans with chronic PTSD (N = 10), multiple RCTs have shown it to reduce nightmares and improve sleep quality in active duty military personnel, veterans, and civilians with PTSD.40 A recent meta-analysis41 of six RCTs (N = 240) of prazosin for treating sleep disturbances in adults with PTSD found that the effect sizes were large and medium-large for improving sleep quality, nightmares, and clinical global impression and for improving PTSD symptoms, respectively. After removing one study that accounted for the substantial between-study heterogeneity, the effect sizes decreased to medium. There is also low level evidence from a retrospective chart review study of veterans with PTSD that prazosin may decrease the number of non–nightmare-distressed awakenings.40 Prazosin is generally well-tolerated. The most common side effects, which occur in about 10% of cases, are dizziness and lightheadedness related to orthostatic hypotension, a problem that may be minimized by starting the medication at 1 mg or less at bedtime and titrating slowly and by separating its administration from that of phosphodiesterase inhibitors by at least 5 hours. Drowsiness is also a fairly common side effect. In most cases, side effects resolve or stabilize at tolerable levels with continued use.40,41 The above meta-analysis found that there were no statistically significant changes in blood pressure reported. The mean final dose in reported clinical trials ranged from 3 to 13 mg at bedtime.40 The optimal length of treatment and whether or not there are lasting benefits of prazosin treatment after discontinuation are not yet known. See Table 6 for a list of practical considerations for treating PTSD-related sleep disturbances with prazosin.
Tips for Treating PTSD-Related Sleep Disturbances with Prazosin
There is little evidence that SSRIs (eg, sertraline, paroxetine, fluoxetine), SNRIs (ie, venlafaxine), tricyclic antidepressants, or monoamine oxidase inhibitors improve PTSD-related insomnia or nightmares. Additional medications for which there is low-level evidence of efficacy for the treatment of PTSD-related sleep disturbances include low-dose cortisol, eszopiclone, trazodone, nefazodone, gabapentin, topiramate, adjunctive olanzapine,40 and adjunctive risperidone.51
Pharmacotherapy for PTSD with specific comorbidities. A recently completed Cochrane review52 of pharmacotherapy for anxiety and comorbid alcohol use disorders included one RCT (N = 88) that found desimpramine was superior to paroxetine for reducing comorbid alcohol abuse and had a lower associated drop-out rate, but they were equivalent for decreasing clinician-rated PTSD symptoms. Another RCT involving mostly civilians with moderate PTSD symptoms (N = 94) found that sertraline had no advantage over placebo in terms of reducing clinician-rated PTSD symptoms or alcohol abuse. Overall, there was insufficient evidence to draw any conclusions about the efficacy of pharmacotherapy for treating comorbid PTSD and alcohol use disorders.
A subsequent pilot RCT53 found that prazosin had greater efficacy than placebo for reducing alcohol abuse in adult outpatients with PTSD and comorbid alcohol dependence (N = 30; 30% veterans), but there was no advantage for reducing PTSD symptoms. A larger follow-up RCT54 involving mostly male veterans with PTSD and comorbid alcohol dependence (N = 96) failed to find an advantage for prazosin over placebo for treating either condition, consistent with numerous studies showing less robust treatment responses in veterans relative to civilians with PTSD. These results raise the possibility that alcohol dependence may block the benefits of prazosin for PTSD.
Lack of evidence for using benzodiazepines. The literature does not support the use of benzodiazepines in PTSD.10 There is some evidence that temazepam and alprazolam may actually worsen PTSD outcomes.14,28 Furthermore, benzodiazepines carry a number of risks including paradoxical disinhibition and abuse liability that alternative treatments do not.
Combined psychotherapy-pharmacotherapy for PTSD. A Cochrane review55 regarding the efficacy of combined pharmacotherapy and psychotherapy for the treatment of PTSD included three adult trials and concluded that the quality of the evidence was insufficient to determine if there were advantages or disadvantages of combined therapy over monotherapy with either modality. It noted, however, that the individual studies did not appear to support combination therapy as being more effective than monotherapy.
A more recent double-blind RCT56 (N = 37) comparing PE combined with paroxetine or placebo in terms of improving terrorism-related PTSD found that the paroxetine augmentation group had significantly greater improvements in PTSD symptoms (incident rate ratio 0.50), remission rates (OR 12.6), response rates, and quality of life after 10 weeks compared to placebo; however, among those who elected to continue for another 12 weeks of double-blind treatment with paroxetine or placebo alone, the remission rates for both groups were the same at 22 weeks. The paroxetine group had higher dropout rates (36.8% vs 27.8%), although the adverse events rate did not differ between the two groups. The authors speculated that the primary benefit of paroxetine might be to increase the efficacy of psychotherapy during active treatment.
There was no evidence for augmentation of CBT with D-cycloserine, a partial N-methyl-D-aspartate glutamate receptor agonist, being more efficacious compared with placebo augmentation as assessed by PTSD remission rates and symptoms, co-morbid depression, or quality of life by a recent Cochrane meta-analysis.57 It was superior, however, for improving anxiety at posttreatment but not at follow-up. One study included in the meta-analysis found that 100 mg of D-cycloserine combined with VR-EBT was better than VR-EBT alone for producing remission of terrorism-related PTSD at posttreatment and at the 6-month follow-up visit; however, a larger study found no advantage of combining VR-EBT with 50 mg of D-cycloserine over placebo.28 Further research is needed to resolve these contradictory results.