A 46-Year-Old Man With Anxiety and Nightmares After a Motor Vehicle Collision

DR SHIP: Mr M is a 46-year-old man who has experienced physical and emotional symptoms following a recent motor vehicle collision. He lives with his wife near Boston and works as an artist. He has managed care insurance.

Mr M's history includes 3 unrelated motor vehicle collisions. The first occurred overseas in December 1999. At that time, he was a backseat passenger in a taxicab struck by another vehicle. He had a thoracic compression fracture and required a prolonged course of rehabilitation. He recovered from this event except for some back and neck spasms. In June 2001, Mr M was a belted driver when his car was rear-ended. His neck and back pain and spasms returned, and he returned to physical therapy. In late January 2002, he was a belted front-seat passenger in a car struck head-on by another vehicle. The air bag deployed. He sustained 3 fractured ribs and multiple contusions; his back and neck pain returned.

Since that time, Mr M has had recurrent thoughts about the collision during the day and difficulty sleeping at night. He has trouble concentrating, feels pervasive anxiety and fear, and often finds it difficult to leave the house. He fears being in automobiles, worries about whether another collision might occur, and has been hesitant to resume his previously frequent bicycling. He reports a tremendous decrease in his energy and finds himself easily exhausted. He feels "fragile" and as if he has lost his previous confidence and sense of control.

Mr M's past medical history is significant only for knee surgery. He has used oxycodone/acetaminophen, diazepam, and naproxen for pain and muscle spasms, but he currently takes only occasional aspirin. He has no known drug allergies. Mr M does not smoke and uses no recreational drugs. He drinks alcohol occasionally, about 3 beers per week.

On physical examination, Mr M appears healthy and has normal vital signs. He is alert, oriented, and articulate. His cranial nerve examination results are normal. He has mild cervical muscular tenderness and spasm.

Mr M reports improvement in his physical symptoms with physical therapy, swimming, and time. Mr M has encountered men with the diagnosis of posttraumatic stress disorder (PTSD), and he is skeptical about the term and its application. He was surprised to learn that his primary care physician, Dr Z, was concerned about him having PTSD.

I have had 3 car accidents in the past about 2½ years, none of which I've had any control over. I told myself it was just a freak thing, but this has bothered me a great deal. Besides nightmares initially—about being in a car—then it became something more general. I think about things being out of control, things happening to me and not having any choice.

I'm fearful of a great many things that I have not been fearful of since I was a child. Sleeping has been a major problem since the time of the accident. Not just the nightmares, but not being able to find any comfortable position. Part of it is having a chunk of my life taken away—there was nothing I could really do for a period of time. I just have this feeling of being very fragile, and it has affected my confidence.

It is so important that my primary care doctor is able to make me feel that he's there to listen. The coping issues are difficult for me to talk about since I feel like I am and should be so self-reliant.

After this last accident, I unfortunately have a feeling that it's not over. I know the odds of these things happening are so small, and yet they can't be so small if they've happened so frequently. The biggest concern for me is getting control over my life again, and [to figure out] how to do it as quickly as possible.

The third event occurred recently. Mr M had a couple of broken ribs, and not surprisingly it reexacerbated a lot of his musculoskeletal injuries, stiff neck, and muscle spasms. But perhaps more significant to him is what it has triggered in terms of his response. He has nightmares about the events. He has complained of difficulty leaving the house, fear that this may recur, as well as a diffuse fear and anxiety.

It was later during unrelated activity that I bumped into a description of posttraumatic stress disorder and thought to myself, "Whoa, this guy has a lot of these symptoms." I didn't initially entertain that diagnosis, because I have, perhaps as he has, for better or for worse, a bit of a prejudicial reaction to that diagnosis, because so many patients carry the diagnosis very chronically. It's seldom that you see a patient in the acute phases of this particular illness.

How much of an onus is there on early diagnosis—should we be talking to patients about this whenever they have sustained an episode of trauma? Many patients are reluctant to seek psychiatric therapy, such that primary care doctors often find themselves thrust into treating psychiatric conditions that we perhaps don't feel entirely adept treating. To what extent should we try to treat these ourselves? To what extent is pharmacological therapy helpful for these patients, and if so, which medications?

What are the definition, symptoms, and epidemiology of PTSD? What are typical inciting events? What are the risk factors for development of PTSD? When should primary care physicians consider PTSD in the differential diagnosis? What can primary care physicians do to care for these patients, and when and to whom should they refer? What does the future hold? What do you recommend for Mr M?

DR STEIN: Mr M suffers from anxiety and nightmares consequent to a motor vehicle collision (MVC). In fact, this was the third MVC in which he was involved—all in a little over 2 years. In the first collision he was seriously injured, incurring neck and back fractures; the subsequent collisions have aggravated the injuries and reactivated his pain and his fears about being in a car. He thinks a lot about the collisions, has insomnia with nightmares about driving, and has trouble concentrating. When he cannot avoid getting in a car, he finds that he feels very nervous. The collisions have also challenged his basic view of himself as a strong, competent individual. He sees himself as "fragile" and the world as a somewhat more unpredictable place than he had previously envisioned. More than a month after the most recent MVC, Mr M is on the mend, but he is still bothered and to some extent limited by these symptoms. In light of this presentation, Dr Z was right to consider the diagnosis of PTSD.

Posttraumatic stress disorder is characterized by a constellation of distressing and/or impairing symptoms that occur after experiencing, witnessing, or being confronted with a traumatic stressor event that involves actual or threatened death or serious injury or threat to the physical integrity of self or others ().¹ Events that fall short of these criteria, even if viewed as "traumatic" by the individual (eg, loss of a job; divorce), do not qualify. The diagnosis of PTSD is not made until 1 month has passed since exposure to the traumatic stressor; if PTSD-like symptoms of sufficient intensity are seen prior to this, the diagnosis of acute stress disorder may be applied. Posttraumatic stress disorder is considered "acute" when it has been present for 1 to 3 months and "chronic" when it persists longer.¹

Criterion A
The person experiences a traumatic event in which both of the following were present:

Criterion B

The traumatic event is persistently reexperienced in any of the following ways:

Criterion C

Persistent avoidance of stimuli associated with the trauma and numbing of general responsiveness (not present before the trauma) as indicated by at least 3 of the following:

Criterion D

Persistent symptoms of increased arousal (not present before the trauma) as indicated by at least 2 of the following:

Criterion E

Symptoms from criteria B, C, and D last for more than 1 month.

Criterion F

The disturbance causes clinically significant distress or impairment in social, occupational, or other important areas of functioning.

*Reproduced with permission from the American Psychiatric Association.¹

Empirical research suggests that PTSD should be thought of as having 4 categories of symptoms: reexperiencing, avoidance, numbing, and hyperarousal symptoms.^{2 - 3} Avoidance and numbing are grouped in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV), resulting in 3 symptom clusters. Almost pathognomonic of PTSD is reexperiencing symptoms, or the tendency to relive the event in thoughts and dreams. Recollections of the trauma intrude on the person's consciousness, and exposure to reminders of the event result in a reexperiencing of the terror. Avoidance of situations that resemble or remind the individual of the trauma can lead to functional disability. Amnesia for important parts of the trauma is also considered a form of avoidance. Emotional numbing often occurs, wherein the individual feels so changed (or, indeed, damaged) by the trauma that he or she is unable to experience joyful or tender or loving feelings; the toll on social and family relationships can be devastating. Hyperarousal symptoms, including irritability, reduced concentration, hypervigilance, and an exaggerated startle response, are frequent concomitants. But perhaps the most pervasive and troubling of the hyperarousal symptoms is insomnia, which, when compounded with nightmares, can rob the individual of any respite from the torment.

Mr M had symptoms characteristic of PTSD by virtue of his exposure to a life-threatening event to which he responded with intense fear and horror. A month after the injury, he still had reexperiencing (nightmares about driving, intrusive frequent thoughts of the crash throughout the day), avoidance (avoiding driving and bicycling by walking), and hyperarousal (difficulty concentrating and insomnia) symptoms. The only category of symptom that Mr M did not have is emotional numbing. It is unclear from the history if Mr M met full DSM-IV diagnostic criteria for PTSD or had a subthreshold presentation (sometimes referred to as "partial PTSD").

In US and Canadian general population samples, the 1-year prevalence of PTSD is in the range of 2% to 4%, with lifetime rates approaching 10% in some studies.^{4 - 6} Women make up approximately two thirds of cases of PTSD in the community^{4 ,6 - 7} ; this is likely due to their increased exposure to traumatic events with a high propensity for inducing PTSD (eg, interpersonal violence including rape and intimate partner abuse)⁸ combined with an apparent 2- to 4-fold increased susceptibility to PTSD once exposure has occurred.^{9 - 10} Hispanic individuals also seem approximately twice as likely to develop PTSD as persons of other racial or ethnic backgrounds.^{7 ,11}

Though previously known by names such as "shell-shock" or "war neurosis," PTSD has long been recognized as a consequence of combat experience. Indeed, military trauma remains a serious source of suffering and chronic disability due to PTSD in soldiers and peacekeepers alike.¹² However, PTSD does not occur exclusively in veterans. Studies subsequent to recent terrorist attacks have underscored the pervasive psychological sequelae of traumatic stress in the community. Approximately one third of adult survivors of the Oklahoma City bombing had PTSD when assessed 6 months after the disaster.¹³ A national survey of 560 US adults 3 to 5 days after the terrorist attacks of September 11 found that 44% reported 1 or more substantial symptoms of stress (eg, being upset by reminders or having intrusive recollections of the attacks; trouble concentrating or sleeping).¹⁴ A representative survey of 1008 residents of Manhattan 5 to 8 weeks after the attacks found that 7.5% met DSM-IV criteria for PTSD, with the rate reaching 20.0% in persons who were very near the World Trade Center.⁷ A survey of a nationally representative cross-sectional sample of 2273 adults 1 to 2 months after the attacks found that the point prevalence of probable PTSD (according to DSM-IV criteria) was much higher in the New York City metropolitan area (11.2%) compared with that in the rest of the country (4.0%).¹⁵ And, most recently, a nationwide longitudinal Web-based survey of US adults was conducted during the first month following the September 11 attacks, with follow-up assessments obtained at approximately 2 and 6 months after the attacks. Seventeen percent of the population outside New York City reported symptoms of PTSD 2 months after the attacks, with the rate decreasing to 5.8% at 6 months.¹⁶ Although the focus of the consequences of terror has recently shifted to the United States, the horrific mental and physical health consequences of armed conflict and terrorist attacks in residents of and refugees from regions where war, persecution, or civil insurgence are endemic cannot be overestimated.^{17 - 18}

From a public health perspective, however, the major burden of illness associated with PTSD in the United States in recent decades comes not from military actions or from terrorist attacks—to which fewer than 1 in 20 US adults have been directly exposed⁴ —but from other forms of serious trauma that are much more likely to be experienced by the populace. These experiences include exposure to childhood sexual trauma, violent crime (including intimate partner violence),^{19 - 20} and injury.^{4 ,6} Foremost among the latter category is involvement in MVCs.

Mr M is among the multitude of Americans who are injured in automobile crashes. In 2000, more than 3 million people were injured in MVCs, an injury rate of 1161 per 100 000.²¹ In the same period, 41 821 people died in MVCs in the United States—an average of 115 persons each day.²¹ The annual economic cost associated with these crashes, including the costs of injury and property damage, exceeds $190 billion.²² Approximately one third of individuals involved in potentially life-threatening MVCs will develop PTSD.^{23 - 24} What is not quantified in the literature is the cost of emotional consequences that Mr M and patients like him experience subsequent to these collisions.

Following severe traumatic stress, 30% to 50% of people will develop acute (ie, lasting <1 month) stress symptoms,¹⁴ but less than half of these will fail to recover and have their symptoms crystallize into PTSD.²⁵ In prospective studies of people involved in collisions presenting for emergent care, the presence of acute stress symptoms (such as sleep problems, anxiety or panic attacks, intrusive thoughts of the trauma, and feeling dazed) in the first few weeks after the trauma predicted persistent posttraumatic distress (odds ratios of 1.5-2.0).^{24 ,26 - 28} Those with pronounced physiologic hyperarousal, as manifested by tachycardia, after trauma have been suggested to be at increased risk for developing PTSD,^{29 - 30} but this finding has not been widely replicated.³¹ Most people will recover spontaneously or with minimal intervention from most forms of trauma. However, a minority (10%-25% in most prospective studies) of individuals, distinguished by their propensity to have severe acute stress symptoms in the immediate postdisaster period, will not recover spontaneously and will need more intensive intervention.²⁵ Clinical experience suggests that the earlier in the posttrauma period this intervention can be provided, the better.

Differences in characteristics of the trauma and among individuals are thought to explain variations in outcome following exposure to traumatic stress. Severity of physical injury has been an inconsistent predictor of PTSD development after serious physical trauma such as that sustained in an MVC, whereas female sex and perceived threat are associated with a 2- to 3-fold increase in the risk for PTSD.^{28 ,32} Mr M, who expressed a strong belief that he "could have died," had this latter risk factor for PTSD. He also may have been at increased risk for PTSD by virtue of anxiety problems (fears and phobias) earlier in life, because several epidemiological studies have found that preexisting depressive or anxiety disorders modestly increase risk for PTSD.^{16 ,33 - 34} Although it is unclear from Mr M's history whether he had PTSD following either of the first 2 collisions, his current presentation is consistent with the finding that prior trauma exposure increases risk for PTSD symptoms on subsequent traumatization.^{23 ,35}

Mr M's presentation to his physician was fairly typical, in my experience. He did not present seeking help for his posttraumatic stress symptoms. Rather, his symptoms were uncovered in the course of a postinjury medical visit when he mentioned to Dr Z that he was having trouble sleeping. Further inquiry uncovered nightmares about the event and led Dr Z to consider PTSD highly in the differential diagnosis.

Posttraumatic stress disorder in primary care is associated with increased use of health care resources,³⁶ driven at least partly by care-seeking for physical complaints. In epidemiologic and clinical samples, trauma victims are at several-fold greater risk than those not traumatized for reporting a wide range of somatic symptoms, including (but not limited to) chronic headaches, irritable bowel, and pelvic pain.^{37 - 40} The presence of unexplained medical symptoms or chronic pain should therefore raise the suspicion for PTSD. In women, questions directed to eliciting the possibility of intimate partner violence are particularly valuable.⁴¹ Sleep problems, though certainly not diagnostic of PTSD, are so often seen in this disorder^{42 - 43} that it behooves the primary care physician to ask patients with insomnia about "terribly stressful events," nightmares, and other reexperiencing symptoms. Positive responses signal the need for more detailed inquiry and assessment.

There are no diagnostic tests for PTSD. Although alterations in hypothalamic-pituitary-adrenal axis and autonomic nervous system function have been reported in many patients with PTSD and other trauma-related disorders,^{44 - 47} measurement of these parameters has yet to find clinical diagnostic utility. Brain imaging studies, while revealing possible abnormalities in brain structure in PTSD (including reduced volume of the hippocampus, though this is controversial)⁴⁸ and the involvement of memory and emotional circuits in the expression of PTSD,^{49 - 51} are not yet of clinical diagnostic or prognostic value.

Like Mr M, many individuals with PTSD symptoms may be 1 or 2 symptoms short of the diagnostic criteria set forth in DSM-IV,¹ but this should not prevent a diagnosis of PTSD if it best describes the clinical picture in conjunction with clinically meaningful distress and/or impairment. Although PTSD is defined by its relationship to traumatic events, it is certainly not the only psychological disturbance seen in the wake of trauma. Depression is another (perhaps even more) common outcome of exposure to serious trauma,^{7 ,52 - 54} as are other anxiety problems (eg, driving phobias, panic disorder).⁵⁵ Distinguishing PTSD from depressive or other anxiety disorders can be clinically challenging; the presence of reexperiencing symptoms (eg, recurrent thoughts of the traumatic event, nightmares, flashbacks) is a strong diagnostic clue. Furthermore, PTSD is comorbid with depressive or other anxiety disorders in about half of cases seen in primary care,⁵⁴ often making an "either-or" decision moot. Clinical experience, buttressed by the findings of a recent survey of 2608 persons seeking information about anxiety problems, suggests that PTSD symptoms are associated with an approximately 2-fold increased rate of suicidal ideation.⁵⁶ Mr M reported some transient depressive symptoms that were not severe enough to warrant a comorbid diagnosis of major depression. Substance abuse, particularly alcoholism, is more prevalent among those with PTSD; it may precede the onset of PTSD (and in some instances may have contributed to the traumatic injury) or may be a later development.^{57 - 58} In either case, the physician should initiate referral for appropriate substance abuse treatment.

Although as many as 10% of patients in general medical settings have PTSD by DSM-IV criteria, it often goes undetected.^{36 ,59 - 60} Barriers to recognition may include failure to query patients about traumatic experiences and low awareness of PTSD among physicians, whose prior associations with PTSD may be confined to their experiences treating Vietnam veterans with this condition. A study that conducted in-depth, open-ended interviewing with survivors of serious traumatic injury found that, contrary to popular belief, nearly all were relieved and gratified to have the opportunity to disclose their experiences of trauma and to discuss their perceived needs following the injury.⁶¹ Questions about prior trauma, particularly sexual abuse, must be asked with respect and compassion, knowing that some patients may decline to answer. But these questions must be asked.

Both Mr M and Dr Z commented that their prior negative connotations to PTSD had led them initially to reject this diagnosis. Some in the medical community perceive that PTSD was invented by lawyers and their litigious clients.⁶² Although misuse of the diagnosis for monetary gain may occur on occasion, to say that some persons have not been seriously damaged as a result of traumatic stress is to turn a blind eye to their suffering. This stigma may present yet another obstacle to recognition on the part of physicians and an impediment to disclosure or willingness to accept treatment on the part of patients. Disability is a common consequence of PTSD,^{32 ,56 ,63 - 64} and PTSD-related injury claims often arise.⁶⁵ An example is the recent suit on the part of combat veterans in the United Kingdom against the Ministry of Defence, in which the claimants allege inadequate attempts to prevent or treat their posttraumatic stress symptoms.⁶⁶ Although it can be a burden, physicians who care for traumatically injured patients should be willing to complete necessary paperwork and/or testify on the status of their patients when required, or risk further imperiling patients' recovery.

No evidence-based guidelines exist for the diagnosis and treatment of PTSD in primary care. The following recommendations are therefore primarily based on clinical experience and extrapolation from lessons learned in the treatment of other mental disorders (eg, depression) in primary care. The primary care physician has the potential to play a central role in early identification and treatment of posttraumatic stress symptoms.⁴¹ By eliciting a history of reexperiencing, avoidance and numbing, and hyperarousal symptoms in the aftermath of serious trauma, the diagnosis of PTSD can be uncovered. From a holistic therapeutic standpoint, the physician can listen and provide emotional support, help the patient to normalize or reintegrate his or her experience, encourage the patient to enlist help from friends and family, promote the use of support groups and self-help materials, and direct the patient to sources of high-quality educational and informational material (eg, National Institute of Mental Health, National Center for PTSD).^{67 - 68}

Given the evolution of PTSD symptoms over time, the primary care physician is in an excellent position to assess patients longitudinally following trauma and to determine if acute stress symptoms are waning, remaining static, or worsening. In the latter circumstances, the physician has the option of starting treatment or referring the patient to a mental health professional. Which route to take will largely depend on patient and provider preferences, as well as the often-frustrating realities and disparities of insurance coverage.

Primary care physicians with psychopharmacological and counseling expertise can manage many cases of PTSD, particularly those of recent onset. In some cases, the physician will provide pharmacological management of anxiety and depressive symptoms and pain, working closely with a mental health professional who provides the specific psychotherapeutic management. When PTSD is more chronic, derived from childhood sexual abuse or other forms of serious childhood maltreatment, associated with serious personality disturbance and/or substance abuse, or complicated by suicidal ideation or psychotic symptoms, referral to a mental health specialist is advisable.

The evidence for effective treatment of PTSD has expanded substantially in the last decade. However, the main modes of treatment, psychotherapy and pharmacotherapy, have not been compared in randomized controlled trials (RCTs), nor have large-scale effectiveness studies (eg, in primary care or other "real-world" settings) been conducted in PTSD. Results of RCTs to date are expressed in terms of effect sizes to permit rough comparisons. Effect sizes represent the degree of nonoverlap between the scores in the treatment and control groups. An effect size of 0.0 indicates complete overlap of the 2 groups; effect sizes of 0.5, 1.0, and 1.5 indicate nonoverlap of 33%, 55.4%, and 70.7%, respectively.⁶⁹

Psychotherapy generally centers on helping the individual address and overcome feelings of fear, shame, and vulnerability.⁷⁰ Cognitive behavioral therapies (CBTs), which feature prolonged and repeated exposure (ie, to thoughts, mental images, and reminders of the trauma), appear to be particularly effective. Three RCTs of cognitive behavioral therapy are especially informative in this regard. An RCT involving 87 men and women with chronic PTSD compared 10 weekly sessions of prolonged exposure (imaginal and live) alone, cognitive restructuring alone, or the 2 interventions combined vs relaxation therapy. The first 3 interventions had effect sizes that were similarly large (≈1.5) compared with relaxation.⁷¹ Another RCT involving 96 women with chronic PTSD who had been assaulted compared 9 twice-weekly sessions of either prolonged exposure, stress inoculation training, or both vs a wait-list control. The 3 interventions were similarly effective and all 3 were superior (effect sizes ≈1.5) to the control.⁷² Finally, an RCT of 168 women with chronic nightmares and PTSD associated with sexual assault compared 3 sessions of imagery rehearsal therapy with a wait-list control. Again, the intervention group did substantially better (effect size ≈1.0) than the control group.⁷³ A recent meta-analysis concluded that eye movement desensitization and reprocessing (EMDR), a controversial therapy, is no more—or less—effective than other exposure techniques.⁷⁴

Some (but clearly not all) psychosocial treatments also hold great promise to secondarily prevent PTSD in persons exposed to trauma. Despite its widespread use, psychological debriefing for trauma victims has been deemed ineffective, according to a recent Cochrane review of the published literature.⁷⁵ A study of 45 individuals involved in MVC or non–sexual assault with prominent acute stress symptoms demonstrated that 5 sessions of prolonged exposure therapy (a critical element of CBT) reduced rates of PTSD to approximately 1 in 5, compared with 2 in 3 who received only supportive counseling.⁷⁶ These very promising results with CBT to prevent PTSD require broader testing in multisite RCTs with various samples of trauma survivors before their adoption can be recommended as standard practice.

Pharmacotherapy of PTSD is also increasingly supported by good evidence of efficacy. Sertraline and paroxetine are now approved by the Food and Drug Administration (FDA) for use in PTSD, and 4 RCTs support this indication. Sertraline administered for 12 weeks was moderately more effective (effect sizes ≈0.5) than placebo in 187 individuals⁷⁷ and in 208 individuals⁷⁸ with PTSD. Paroxetine administered for 12 weeks was also moderately more effective (effect sizes ≈0.5) than placebo in 2 trials, one with 551 individuals with chronic PTSD⁷⁹ and the other in 307 individuals with PTSD.⁸⁰ While not FDA-approved for PTSD, fluoxetine also has shown benefit compared with placebo in two 12-week RCTs. In an RCT of 53 civilians with PTSD, fluoxetine administered for 12 weeks showed a marked (effect size ≈1.0) improvement over placebo,⁸¹ while in a study of 301 persons with PTSD, of whom 81% were men and 48% were combat veterans, fluoxetine showed a moderate (effect size ≈0.4) improvement over placebo.⁸²

Clinical experience with antidepressants from other classes, such as venlafaxine extended-release⁸³ or nefazodone,⁸⁴ suggests that these may also be useful. Still, pharmacotherapy for PTSD has considerable room for improvement, as experience suggests that the current best treatments provide clinical benefit to 50% to 60% of patients, but result in full remission of symptoms in less than a third.^{77 - 82} Sleep disturbance in PTSD can be vexingly difficult to treat, and, despite the absence of evidence supporting their utility, benzodiazepines are often used for this purpose, as they were prescribed for Mr M. Promising in this regard is the α-1 antagonist prazosin, which has been shown in uncontrolled studies to enhance sleep and reduce nightmares in patients with PTSD.⁸⁵ Some patients with PTSD may similarly benefit from treatment with atypical antipsychotic agents, either alone or possibly in conjunction with selective serotonin reuptake inhibitors (SSRIs), as shown in an RCT of 19 patients with combat-related PTSD.⁸⁶

As noted in a recent review of the topic,⁸⁷ the evidence base for PTSD treatment is a work in progress. From the few RCTs for which effect sizes can be estimated, it would appear that CBTs have an edge over pharmacotherapy. Furthermore, whereas pharmacotherapy works as long as the medication is taken (in one RCT, relapse rates were only 5% over a 28-week period in 46 patients who were randomly assigned to continue sertraline treatment, compared with 26% in 50 patients who were randomly assigned to placebo),⁸⁸ the effects of CBT persist beyond the actual treatment itself (seen up to 6 and 12 months, respectively, in 2 RCTs).^{71 - 72} However, the putative advantages of one treatment modality over the other are at present speculative, as no published study has compared acute or long-term outcomes of pharmacotherapy with any psychosocial treatment, or with both treatments combined. These are areas in which additional research is sorely needed.

Pain is an area where PTSD intersects with medical illness, and primary care physicians are ideally situated to play a key role in pain management. Indeed, inadequately treated pain following physical injury can redirect the patient's thinking toward any injury and its consequences, thereby mutually maintaining or exacerbating the pain and the posttraumatic stress symptoms.⁸⁹ Mr M experienced a worsening of his neck and back pain sustained in the first motor vehicle injury 2 years prior to the most recent injury. Dr Z's close attention to pain management, including the use of analgesics and the encouragement of prompt resumption of activity and appropriate exercise, undoubtedly has contributed to Mr M's rapid recovery. Indeed, a naturalistic study of 24 young burn patients with PTSD suggests that ample pain control (perhaps through opioid-mediated pathways) can improve outcomes⁹⁰ ; this effect remains to be replicated in RCTs.

I base my comments on Mr M's prognosis primarily on clinical experience, as individual predictors of response to specific treatments (or to no treatment) are largely unknown. Mr M's psychiatric symptoms (anxiety, fearful avoidance, concentration and sleep problems) were intense in the first 4 to 6 weeks following the collision, and had they persisted, I would have strongly recommended that he seek either CBT or pharmacotherapy (most likely with an SSRI). But, as Mr M reports more recently, his symptoms have abated without specific psychological or pharmacological intervention and he has resumed his precollision level of functioning. At present, he reports some residual fears about driving or cycling, but no ongoing avoidance. With his symptoms resolving well with time and the support and empathic ear provided by Dr Z, Mr M's prognosis is good. In my opinion, he is not in need of additional specialist-provided psychotherapy or pharmacotherapy at this time. However, he may nonetheless benefit from having the opportunity to work with a therapist to help resolve his feelings of vulnerability subsequent to the collision and to help extinguish his residual phobias. Although such intervention does not seem essential to his immediate recovery, it may benefit Mr M's long-term health and welfare and may help "protect" him from a recurrence of posttraumatic stress symptoms if or when he is next exposed to a potentially traumatic event.

The prognosis is also excellent for the state of science in this field. Advances in understanding the basic mechanisms for posttraumatic fear acquisition and maintenance, and the neurobiology of stress-responsive brain systems,^{51 ,91 - 93} are changing the way researchers think about the pathophysiology of PTSD. These advances have made it possible to consider and begin to test pharmacological approaches to secondary prevention of PTSD.⁹⁴ Needed desperately, however, are larger RCTs that enroll representative samples of persons with PTSD at various stages of illness and effectiveness studies that attempt to determine to what extent treatments initially honed and tested in rarified research settings prove useful and acceptable in more generalizable clinical settings. Promising in this regard is a recent study of CBT administered by non–research therapists to 91 patients with PTSD resulting from a car bomb in Northern Ireland. The results suggest that positive findings obtained in research settings may well generalize to the "frontlines."⁹⁵

A PHYSICIAN: For PTSD, is psychotherapy more effective than medication? What is the difference between SSRIs, therapy, and placebo for these patients?

DR STEIN: As yet, there are no head-to-head trials of specific forms of psychotherapy vs medication. For patients with PTSD, I would not recommend restricting treatment to pharmacotherapy. Even if the medication can diminish the symptoms, or if in some cases the symptoms disappear, the medicine does not help people to understand what happened to them. It does not help them to put it into perspective; it does not teach them how to go on and put their lives back together. That is where psychotherapy plays a role in PTSD that medication alone cannot.

DR Z: My impression is that medications may not be helpful in many situations when we are dealing with PTSD in the primary care setting. I appreciate the compliments about an empathic ear, but I don't frequently find myself in a position to have that much time these days to spend with patients. However, it was very difficult to try to have someone see Mr M promptly for any counseling. What are your thoughts about availability of people in the psychiatric field to help patients with PTSD?

DR STEIN: I understand what you are saying. But I wouldn't want to write off medications, first of all, because they can help. The primary care physician can assist by prescribing antidepressant medications appropriately. This can make a tremendous difference, for both PTSD and possible depression. Primary care physicians often do not have time to provide counseling, and in fact, most psychiatrists don't either. Most psychiatrists also manage medications and work with social workers, psychologists, and other mental health professionals who provide the counseling. What is available, unfortunately, really depends on what kind of insurance the person has and what kind of access he or she can gain to mental health care. It varies tremendously between communities. Often people who have very little in the way of insurance can get very little in the way of counseling, and that is a big problem for which I don't have an answer.

A PHYSICIAN: How do you deal with insurance if someone has "partial" PTSD? That may be a problem, and I hope those of you who are working on the next DSM will fit that in.

DR STEIN: It is a problem that many people fall 1 or 2 symptoms short of having PTSD. The DSM was never meant to guide insurance or any sort of legal outcomes. Despite everyone saying that, it has unfortunately become something that people have reified and actually deny benefits for. Clinical and epidemiologic studies show that levels of disability are generally indistinguishable between persons who meet full and "partial" PTSD criteria.^{5 ,56} Scientifically, there is thus every reason to say that the criteria should change. There will be plenty of political opposition to that, I can tell you, but perhaps they will.

A PHYSICIAN: I believe DSM assigns to the clinician discretion about whether the severity of symptoms is sufficient to overcome a lack of number for almost any disorder. So you can give a diagnosis of major depression for someone with 3 or 4 symptoms if he or she is disabled enough. I think the same would probably be true for PTSD, although on a research basis it may be more complicated.

DR STEIN: I really encourage you to diagnose PTSD when it looks like PTSD and not be counting symptoms. The DSM guidelines were never meant to be used that way—particularly not in primary care. The primary care version of DSM-IV does not emphasize the enumeration of symptoms.⁹⁶ You may need to go to bat with insurers regarding individual diagnoses, however.

A PHYSICIAN: Can you comment on PTSD in refugees, specifically when current living situations may prolong their anguish?

DR STEIN: Research shows very high rates of PTSD in refugees,^{17 - 18} many of whom have been exposed to horrific levels of trauma. Current living situations may perpetuate PTSD symptoms. We talk about PTSD as if it occurs when something bad happens to someone and then the person gets PTSD. The truth is that for many people, something bad happens and it keeps happening. Childhood sexual abuse, domestic violence, and the trauma faced by many refugees are examples. The syndrome that we describe here often does not do justice to the kind of chaos present in people's lives. We may need to change some of the criteria for characterizing this condition.

A PHYSICIAN: Could you comment on the overlap between PTSD and borderline personality disorder? Frequently for people who might otherwise in the past have been diagnosed with borderline personality disorder, PTSD is a more "attractive" diagnosis for a variety of reasons.

DR STEIN: The relationship between PTSD and borderline personality disorder is controversial. My view is that many people who get borderline personality disorder have something that probably wouldn't have manifested were it not for traumatic events happening to them when they were children. The definition of borderline personality disorder fits for some people. However, many of those people also have PTSD. I think that clinicians should try to focus on what can be treated. For some people, pharmacotherapy can help with mood lability symptoms. And if the PTSD diagnosis lets you provide pharmacotherapy when the borderline personality diagnosis wouldn't, then that's beneficial. Nevertheless, for most people who have those 2 disorders together, whatever pharmacotherapy is provided, it is still crucial to work with a therapist who can try to help the individual cope with some of the personality dysfunction that is attached to that early trauma.