A 28-Year-Old Woman With Panic Disorder

DR BURNS: Ms M is a 28-year-old woman with symptoms of palpitations and chest discomfort. She is a full-time student and works 2 jobs. She has Massachusetts free care insurance coverage and receives her primary care from Dr G, a general internist at the Beth Israel Deaconess Medical Center.

When Ms M first came to see Dr G, she reported that during the prior 3 weeks she experienced 4 episodes of palpitations and chest discomfort that occurred both at rest and with exertion. Following 2 of these episodes, she went to the emergency department for treatment. There, she was told that her symptoms were due to anxiety and panic.

Ms M has experienced anxiety since early childhood and first underwent psychotherapy when she was 9. The symptoms included anxiety and worry, which are not present now. Her anxiety was treated with a variety of antidepressants, without improvement. However, her symptoms significantly improved with clonazepam. Ms M also began receiving acupuncture and found it helpful. At the time of the recent episodes, she was neither taking clonazepam nor receiving acupuncture treatments.

Dr G prescribed lorazepam for Ms M and referred her to a social worker. A medical evaluation also was undertaken to determine the cause of her symptoms. This included a complete blood cell count and thyroid hormone tests, results of which were within normal limits. An electrocardiogram performed when Ms M was having symptoms revealed sinus tachycardia. Results of a patient-requested HIV test were negative. Ms M continued to have difficulty accepting that her physical symptoms were due to panic and felt that something else must be wrong. Due to formulary changes, her medication was switched to clonazepam.

Ms M has no other past medical or surgical history. Although a frequent social drinker, she has decreased her alcohol use because it worsened her panic symptoms. She does not use any illicit drugs. Ms M smokes up to 1 pack of cigarettes daily. Her biological family history of alcohol abuse and mood and anxiety disorders is unknown, because she was adopted.

Gradually, with the assistance of her social worker, Ms M has been able to accept that her symptoms are caused by panic. Currently, she experiences an episode every couple of weeks. She admits to becoming anxious about her physical symptoms, but feels that the anxiety follows the onset of symptoms. She is usually able to control her symptoms with relaxation techniques and only occasionally needs to take medication (clonazepam, 0.5 mg).

I always had trouble dealing with people in social situations. I was really shy. I don't think anyone diagnosed me with anxiety until I was in my 20s. Then, the first psychiatrist put me on antidepressants and said my anxiety was being caused by depression. And that didn't ring true to me at all. I said, "No, the anxiety is causing depression." And he said, "No, that's not the case. That's never the case." Eventually, I found a therapist who said, "Absolutely, you have anxiety that's causing depression," and started treating me with clonazepam, which really helped.

I went off the antidepressants and just stayed on the clonazepam, and I was fine for a long time. At one point, I started going to acupuncture, which really helped, and then I went off all medications completely. I got it into my head that I was fine, and I was for a while. Then out of nowhere, I had all this stress in my life and this whole new episode happened.

My heart was racing and I got pains in my arms and my legs. All my symptoms were very physical. There was not really any emotional component until I started panicking over the symptoms, which I think probably made it worse. So, I went to the emergency room. I couldn't calm down and couldn't bring my heart rate back down. I've had anxiety my whole life, so I feel like I'm pretty good at controlling it. And I couldn't control it at that point.

I don't understand why this came out at age 28, if it had never before manifested this way. The symptoms that I had before were never physical, and this time they were only physical. I didn't have a thought in my head that made me panic. Out of nowhere, I had physical symptoms. I couldn't relate that to anxiety and had never experienced anything like that. I was terrified. I thought I wasn't going to make it to the hospital in time. I thought I was dying.

The most helpful thing has been that Dr G never discredited anything I said and always listened to me. She was willing to test for any physical problems, and she never said it's all in your head, which I guess it is. Instead of telling me I was wrong, she managed to prove it to me. She started doing all kinds of tests—testing me for everything possible—and said, "You have nothing," and that helped.

The first time I met Ms M, she had experienced 3 or 4 episodes of significant palpitations, anxiety, and feeling like she was going to die. While being appropriately concerned that this was a physical problem, she was open to the suggestion that it could be panic disorder. Other patients are more interested in having many medical tests for the physical symptoms that panic brings on. For her, I think I was able to order very few tests. We evaluated her for some physiologic causes of palpitations, and over time, we convinced her that her symptoms are from panic disorder.

My questions for Dr Gorman are how should I monitor her long term for this disorder now that she's stable? Should she take medication? How often can someone have panic disorder, be treated, recover a bit, and then re-present with a different clinical picture? How far should primary care physicians go in the medical evaluation of these patients?

What is the definition, epidemiology, pathophysiology, and natural history of panic attacks and panic disorder? How is the diagnosis established? How can the primary care physician distinguish it from other underlying medical and psychiatric problems? What diagnostic testing is recommended? When should a primary care physician consider pharmacologic treatment and with what agent(s)? What nonpharmacologic treatments are available? When should a patient be referred to a mental health professional? What do you recommend for Ms M?

DR GORMAN: In many respects, Ms M represents a classic case of panic disorder. Like most patients, she had a lifelong history of anxiety before her first panic attack. She also notes a lifelong problem with social situations and shyness, raising the question of comorbid social anxiety disorder. At one point, an incorrect assessment was made ascribing her anxiety to depression, when, as she notes, anxiety preceded any mood disturbance and it is not clear that she ever suffered from clinical depression.

Panic disorder is probably the most extensively studied of the anxiety disorders, but it is not the most common. With a lifetime prevalence of 3.5%,¹ it is seen less frequently than social anxiety disorder or generalized anxiety disorder. However, having a panic attack is one of the greatest motivating factors that leads a patient to seek medical attention. Emergency department and primary care physicians are well acquainted with the patient who believes he or she is in the midst of a catastrophic medical event, but no pathology can be found. Hence, the dramatic nature and frequent presentation of panic attacks in the medical setting also prompted great interest in panic disorder.

Ms M had just this presentation. Her first attack led to a trip to an emergency department where only sinus tachycardia was detected. She was prescribed diazepam and, like most panic disorder patients having their first attack, hoped that the experience might be "a one-time thing." Unfortunately, the next attack, 1 week later, was even worse and once again led her to the emergency department where she wondered, "What is wrong with me physically?"

The current definition of panic disorder is derived from the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR).² For the diagnosis, the patient should experience recurrent panic attacks, at least some of which should be unlinked in the patient's view to any external event or situation. These are "spontaneous" panic attacks and are distinguished from "provoked" or "situationally bound" attacks, which also occur in panic disorder and are linked to specific environmental cues.² The panic attack itself is characterized by a sudden crescendo of autonomic arousal and fear, lasting approximately 10 to 30 minutes. At least 4 of a possible 13 symptoms should be present for a "full-blown" attack, although "limited symptom" attacks also occur and are clinically important ( Article ).

A. The patient must have both:
 1. Recurrent unexpected panic attacks
 2. At least 1 of the attacks followed by at least 1 month of ≥1 of the following:
  a. Persistent concern about having additional attacks
  b. Worry about the implications of the attack or its consequences
  c. A significant change in behavior related to the attack
B. A panic attack is defined as a discrete period of intense fear or discomfort with ≥4 of the following symptoms that develop abruptly and peak in intensity within 10 minutes:
 1. Palpitations, pounding heart, or accelerated heart rate
 2. Sweating
 3. Trembling or shaking
 4. Sensations of shortness of breath or smothering
 5. Feelings of choking
 6. Chest pain or discomfort
 7. Nausea or abdominal distress
 8. Feeling dizzy, unsteady, lightheaded, or faint
 9. Derealization or depersonalization
 10. Fear of losing control or going crazy
 11. Fear of dying
 12. Paresthesias
 13. Chills or hot flushes

Reprinted with permission from the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision. Copyright 2000, American Psychiatric Association.

A substantial number of individuals in the general population will experience at least 1 panic attack³ ; hence, panic attacks are a necessary but not sufficient part of the diagnosis of panic disorder. At least 1 attack must be followed by at least 1 month of either worry and anticipation about further attacks, phobic avoidance, or a change in behavior because of the attack. Hence, panic disorder is defined both by the experience of having panic attacks and by the emotional and behavioral consequences of having an attack.

For unclear reasons, panic disorder is twice as common in women as in men.⁴ Its prevalence is fairly constant around the world,⁵ making it unlikely that it is a consequence of any particular cultural or ethnic factor. The usual age of onset is late adolescence or early adulthood, but there are many descriptions of onset of panic disorder in children.⁶

Although the etiology of panic disorder is unknown, a number of important observations have led us closer to understanding the basic pathophysiology. There is no question that panic disorder is familial, with the risk of having panic disorder elevated several-fold among first-degree relatives of individuals with panic disorder.⁷ Several studies examining panic attacks among twins have found a significantly higher concordance rate between monozygotic than dizygotic twins.^{8 - 9} Therefore, the propensity to develop panic disorder seems to be partially under genetic control. It is unlikely, however, that a single gene is involved, and finding chromosomal markers or verifying candidate genes for panic disorder so far has been frustrating.

One interesting childhood disorder has been linked to adult panic disorder and helps to validate genetic models. Kagan et al¹⁰ identified a characteristic among children identified as early as 2 years of age called "behavioral inhibition to the unfamiliar." These children are generally developmentally and intellectually normal, but when confronted with a novel stimulus, they become exorbitantly fearful and avoidant. They also manifest increased heart rate, decreased heart rate variability, and increased levels of salivary cortisol.¹¹ Such children are likely to be the offspring of parents with panic disorder¹² and have a greater chance than noninhibited children of developing panic disorder.¹³ It has been speculated, therefore, that behavioral inhibition may be a genetically mediated behavioral trait that is the precursor to panic disorder.

Despite strong evidence for a genetic aspect to panic disorder, the condition cannot be ascribed entirely to inheritance. Even among monozygotic twins, the concordance rate for panic is less than 50%, making it almost certain that environmental factors are involved. Studies have varied in their success at identifying such factors, but early life separations from parents¹⁴ and disruption of normal child-parent relationships¹⁵ have been shown to increase the risk. Interestingly, another childhood disorder, separation anxiety, has been linked to panic disorder.¹⁶ Such children fear separation from their parents and can become afraid and even refuse to attend school. Adults with panic disorder fear being alone, and the threshold for having a panic attack is raised considerably if a trusted companion is present. Hence, separation phenomena may play a role in causing panic disorder. Most now believe that panic disorder occurs in a genetically predisposed individual who is exposed to significant life stress.

Whether inherited or not, Ms M tells us that her anxiety problems date from early childhood. She recalls always having been shy and averse to social situations, and by 9 years of age she was in psychotherapy. When she was 13 years old, she had trouble going to school and once again was sent for psychotherapy. She was not given an anxiety disorder diagnosis until her 20s.

Neurobiological research of panic disorder initially rested on the finding that several substances have the capacity to provoke attacks only in patients with panic disorder. In addition to sodium lactate,^{17 - 18} the list includes yohimbine,¹⁹ carbon dioxide,²⁰ cholecystokinin,²¹ norepinephrine,²² and doxapram.²³ Specificity is not as great as once believed, with studies showing that patients with posttraumatic stress disorder are sensitive to the effects of lactate and yohimbine,^{24 - 25} and that patients with premenstrual dysphoric disorder are sensitive to lactate and carbon dioxide.^{26 - 27} Furthermore, the list of agents associated with panic attacks has grown so large that it is difficult to find any neurobiological common denominators among them that can give insight into basic pathophysiology.

One current view is that patients react with heightened anxiety and physiological responses to agents like yohimbine and carbon dioxide because of hypersensitivity of the brain's intrinsic fear network. In experimental animals, the circuits that underlie that behavioral and physiological expression of fear are well worked out and include the amygdala and its brain stem projections, the hippocampus, and the medial prefrontal cortex.^{28 - 29} Neuroimaging studies are beginning to show that these circuits are also critically involved in human fear and may be abnormally sensitive in anxiety disorder.³⁰

The diagnosis of panic disorder is made by patient history. Although laboratory tests can rule out medical conditions masquerading as panic disorder, there is no definitive test to confirm the diagnosis. The typical presentation is of a young person who complains of the sudden onset of troubling physical symptoms. The patient will often say that "out of the blue" she began experiencing rapid heartbeat, chest pain, difficulty breathing, tremulousness, sweating, and a feeling that things are not real. She will also express a belief that she is dying or at least "going crazy" and a concern that she will do something out of her control, such as accidentally drive a car over a bridge. Although such patients usually have been anxious their entire lives and may have a history of previous depression or substance abuse, the panic attack is the overwhelming element of the presentation. At this stage, the patient will generally not identify typical symptoms of depression, such as loss of appetite or interest in things, sleep disturbance, or suicidal ideas.³¹

An important feature of panic disorder that complicates the diagnosis is the high rate of comorbidity with other psychiatric illnesses. Patients with panic disorder frequently have 1 or more other anxiety disorders,³² including social anxiety disorder and generalized anxiety disorder. Substance abuse is higher among panic patients than the general population.³³ Although some of this may be an attempt to self-medicate, according to some studies it is more common for alcohol abuse to precede the onset of panic.³⁴ Ms M also has a history of excessive alcohol use. Finally, as many as 70% of patients with panic disorder also will have depression at some time in their lives.³⁵ Indeed, 1 clinician apparently told Ms M that her anxiety was secondary to depression. So common is the coexistence of panic disorder and depression that scientists have begun to explore the possibility that common genes and/or neurobiological substrates may be involved in both disorders.

The primary care clinician's first concern is to rule out potentially life-threatening medical illness. This is important, but panic disorder should never be a diagnosis of exclusion. Rather, it should be placed in the initial differential diagnostic scheme. Otherwise, patients tend to be told at the end of a lengthy medical evaluation that "there is nothing wrong that we can find," delaying the diagnosis and treatment of panic disorder.

Several medical conditions can present with paniclike symptoms. These include acute myocardial ischemia, cardiac arrhythmia, hyperthyroidism, asthma, pheochromocytoma, and opiate or alcohol withdrawal.³⁶ Usually, careful history taking can rule out most of these, but physical examination, routine blood chemistries, electrocardiography, and thyroid function tests are a reasonable approach to being sure. Unless these indicate possible disease, brain imaging studies, electroencephalography, continuous cardiac monitoring, and echocardiography are not generally needed.³⁶ Although mitral valve prolapse may occur more frequently in patients with panic disorder than in the general population,³⁷ this is probably a coincidence and does not affect the clinical course or treatment decisions.³⁸

A number of medical problems are associated with panic disorder including chronic respiratory disease, irritable bowel syndrome, and migraine headache.^{39 - 41} Although panic attacks are not immediately life-threatening, several studies have shown that anxiety disorder increases the risk for cardiovascular morbidity and mortality.^{42 - 43} Patients with panic disorder may have an altered ratio of parasympathetic to sympathetic innervation to the heart, manifested by reduced heart rate variability,⁴⁴ which may in part predispose to cardiac disease later in life. Hence, the physician should not hesitate to institute treatment for the patient with panic disorder.

Obtaining a positive family history of mood and anxiety disorder will help to confirm the diagnosis. Clinicians also should inquire about recent significant life stress. Although patients often insist the attacks are "out of the blue" at first, many will identify stressors that may be important to consider later in treatment.

It is often difficult to convince a patient who believes she is having a heart attack that her problem is really panic disorder. On the other hand, studies have shown that patients with chest pain who have normal cardiac angiography often turn out to have panic disorder and respond to treatment.⁴⁵ Because of a close relationship with the patient, the primary care physician is often the best person to suggest the diagnosis of panic disorder. Generally, the physician should explain that panic disorder is a medical condition characterized by excessive activity of parts of the brain that control fear and autonomic nervous system responses. Like hypertension, the cause is not known, but treatment is almost always successful.

Delaying treatment of panic disorder has several consequences. As panic attacks continue, many patients become fearful of situations in which they have had previous panic attacks or in which help would not be easily accessible in the event of an attack. Patients begin avoiding these situations. Classically, patients avoid buses, trains, planes, cars, or any other conveyance from which exit could be delayed or impossible. They attempt to sit near the aisle at church or in the theater. In the worst-case scenario, patients refuse to go anywhere unless accompanied by someone who can bring them to the nearest emergency department in case an attack begins. This is called "agoraphobia." As mentioned above, the onset of depression and substance abuse may also occur if panic disorder is not treated. Several studies have shown that the suicide attempt rate increases dramatically in panic disorder patients who are also depressed.⁴⁶

Ms M's anxiety disorder was extremely debilitating, as is usually the case.⁴⁷ She missed work often and eventually stopped driving and had difficulty leaving the house. She is clearly improved now, thanks in large part to a physician who, as Ms M puts it, "never discredited anything I said." By acknowledging the reality of her physical sensations and the resultant terror they produced, Dr G was able to engage Ms M in a rational process of medical work-up ending in the diagnosis of panic disorder.

Dr G was extremely sensitive to the need to balance Ms M's fears of physical illness with the need to make an anxiety disorder diagnosis and initiate appropriate treatment. Ms M notes that instead of discounting her symptoms by telling her it was all in her head, Dr G was able to convince her that her problems were psychiatric and needed treatment. Interestingly, Ms M believes this was accomplished by "doing all kinds of tests," but Dr G states that she actually "was able to order very few tests." Most likely, Dr G's patient approach to Ms M's physical complaints helped Ms M feel that adequate medical investigation had been undertaken without the need to embark on a laboratory and radiologic adventure.

Dr G treated Ms M by referring her to a therapist and prescribed benzodiazepines, both of which resulted in positive therapeutic effect. Indeed, the most important part of this so far successful treatment has been the excellent relationship Dr G established with Ms M. As is true with all medical illness, including psychiatric conditions, a positive relationship between a caring physician and the patient is critical for the rest of the treatment to proceed.

Two modalities have been shown by rigorous clinical trials to be effective for panic disorder, cognitive behavioral therapy (CBT) and medication.⁴⁸ They are equally effective and the decision about which to recommend for any given patient is not straightforward.

Cognitive behavioral therapy rests on the principle that patients with panic disorder develop catastrophic misinterpretations about ordinary somatic sensations.⁴⁹ For example, a minor ache in the chest is interpreted as a heart attack. The fear triggers further autonomic nervous system activation and a generalized feeling of terror. Cognitive behavioral therapy for panic disorder generally involves approximately 12 sessions over the course of 3 months.⁵⁰ During the sessions, patients are given information about the nature of panic attacks, emphasizing that they are not life-threatening. Most therapists also offer breathing retraining in which patients are taught to avoid hyperventilation during anxious moments by the technique of paced abdominal breathing. Cognitive therapy involves teaching patients like Ms M to challenge their catastrophic misinterpretations and develop a more realistic understanding of the significance of somatic sensations. Some therapists also use interoceptive deconditioning, in which somatic sensations like dizziness or rapid heartbeat are deliberately provoked over and over again until the patient realizes that they are not dangerous but merely expected physiological responses.⁵¹ Finally, for phobic patients, in vivo desensitization is used to reduce avoidant behavior by exposing the patient to increasing doses of the phobic stimuli.⁵²

Many randomized controlled trials have shown that approximately 70% of panic disorder patients will respond to CBT with a decrease in panic attacks and phobic behavior and overall improvement in function.⁵³ It has been claimed that successful CBT leads to long-term remission,⁵⁴ although more recent studies suggest that relapses do occur in the months following the discontinuation of CBT.⁵⁵ It is critical for the primary care physician to understand that CBT for panic disorder is a technique that requires specific training. Not all therapists have such training and may not be qualified to treat patients with panic disorder.

Based on a number of randomized placebo-controlled, double-blind studies, it is clear that medications of many classes are effective for panic disorder. These can and should be prescribed by the primary care physician if medication is chosen as the intervention (Table 1). Antidepressants of the tricyclic and monoamine oxidase inhibitor classes were first shown to be specifically effective for panic disorder.^{56 - 57} The patient need not be depressed to benefit from tricyclic antidepressants or monoamine oxidase inhibitors. Studies indicate that about 70% of patients respond to them after about 4 to 6 weeks of treatment.⁵⁸ Because of their many adverse effects, however, they are rarely used as first-line therapy.

As established by studies published in the 1980s,⁵⁹ benzodiazepines are effective in treating panic disorder. Alprazolam was the first medication ever approved by the Food and Drug Administration (FDA) for the specific treatment of panic disorder, and clonazepam was later similarly approved. Studies show that in treating panic disorder, both medicines are effective with few adverse effects.^{59 - 60} Other benzodiazepines are probably also effective.⁶¹ It is often very difficult to discontinue benzodiazepines because of withdrawal symptoms, however, and they are rarely used as monotherapy. Rather, low doses of benzodiazepine may be prescribed simultaneously with an antidepressant to achieve rapid relief of panic disorder. Usually, once the antidepressant begins working, the benzodiazepine is tapered and discontinued. However, some patients respond only to benzodiazepines and in this case their use is appropriate.

Most now agree that selective serotonin reuptake inhibitors (SSRIs) are the first-line medication treatments for panic disorder.⁶² Data suggest that all 5 SSRIs now available in the United States are effective,^{63 - 65} but only paroxetine⁶⁶ and sertraline⁶⁷ are approved for this indication by the FDA. Patients with panic disorder are usually started on one-half the starting dose used to treat depression, and then the dose is increased to comparable doses. Four to 6 weeks is usually required for a response. Following response, patients should continue taking the medication for at least 1 year.⁶² Data indicate that patients who discontinue medication sooner than this are at risk for relapse, but no data currently exist to indicate if there is ever a time when medication can be stopped and response maintained.^{68 - 70}

There has been much debate about whether medication or CBT is the best treatment for panic disorder. In a recent study by our group,⁷¹ both treatments were equally effective after 12 weeks of treatment compared with placebo, although among responders there was an advantage for medication therapy. During a subsequent 6-month maintenance phase, most placebo responders failed to maintain response, but patients receiving active treatment remained well. Also, in this stage, an advantage for combined medication and CBT treatment over monotherapy emerged. Finally, in a subsequent follow-up phase when all treatments were discontinued, patients who received CBT remained well at higher rates than those who received medication. The study suggests that medication may give the best initial response, that combined therapy may be better than either alone, and that the response to CBT is more durable than the response to medication.

The issue now is whether Ms M and her physician should be satisfied with her progress. She is not missing work as much and seems to be functioning. However, she does acknowledge still having panic attacks, with chest pain and shortness of breath, even though she now feels she can "control them." She still feels anxious, particularly when she is not busy, and "the thought processes in my head just go on and on." At these times, she takes clonazepam.

Although she is better, Ms M is still clearly symptomatic and therefore at risk to have future difficulties. Hence, there are a number of things to consider for further treatment. First, although her therapy has been helpful, it is not clear what kind of therapy was administered. As mentioned above, it is important that she receives a course of well-proven CBT that might help her, among other things, to control her anxious thoughts. We should hear more about the nature of the therapy and plans for the future.

Second, I would advocate a trial of an SSRI for Ms M. Apparently, in the past she was briefly given an antidepressant and did not benefit, but it is not uncommon for patients to have received doses that are too low for too short a period of time, or even antidepressants with no proven efficacy for panic disorder.

Although Ms M does not recognize any stressful life events that may have precipitated her problems, there are ample indications that this is not necessarily the case. We are not told anything about her personal and social life, but she did have intense fears of HIV infection. It would be very important to find out the basis for such fears and whether they relate to any difficulties with sexual or romantic adjustments. We are told relatively little about her career aspirations, but do know that she is attempting to work 2 jobs to get through school. This seems like a very demanding schedule and one that could be associated with considerable stress. Finally, Ms M has a history of alcohol abuse, although we do not know the biological family history. We assume that she began drinking to treat her anxieties, but this is not always the case and it is important to explore this further. We know that she had a very positive response to a physician who listened and understood her, and this raises the question about whether any issues from her childhood are important.

With approximately 3.5% of the population having panic disorder, there are not nearly enough psychiatrists to treat all the patients who need intervention. Hence, only by having the primary care physician initiate therapy will most patients who need treatment receive it. In this case, a combination of medication prescribed by the primary care physician and a referral to a therapist worked quite well, although it is likely that more can still be done to help Ms M.

DR G: Ms M is smoking up to a pack a day of cigarettes and feels that the cigarettes are treating her symptoms to a certain extent. Both she and I would like to work on this issue. I feel like my hands are tied right now because her panic disorder is more stable, and I don't want to upset the balance. How can I help her quit smoking in the setting of her panic disorder?

DR GORMAN: Whether or not this would upset her improvement in terms of anxiety disorder, this is obviously a critical thing to attempt for her health in general. I would favor prescribing an SSRI until it is clear she is substantially better, then using a combination of nicotine patch and CBT for this specific problem. Bupropion is not an effective agent for panic disorder and can sometimes make anxious patients more anxious.⁷² However, in my experience, I have found the combination of bupropion with an SSRI in patients who have anxiety disorder can often help with smoking cessation without exacerbating their anxiety.

A PHYSICIAN: How long should patients stay on benzodiazepines?

DR GORMAN: I try to avoid prescribing benzodiazepines if I can. The first thing I say to patients is there are 2 ways to treat this illness, psychotherapy and/or medication, and if you choose medication, I would prefer that you take an antidepressant. However, I think it is inhumane to deny a benzodiazepine to a person who is wracked with anxiety, because it clearly is the quickest, safest way to make them feel better. I try to educate the patient that I don't regard the benzodiazepine as the definitive treatment and that I will want to try to stop it as soon as possible. Studies show that clonazepam is the best choice because it's probably the easiest to stop after a few weeks and is effective in low doses.⁷³ Also, it has a relatively long half-life and usually can be given once or twice daily. Some evidence from the literature on CBT suggests that benzodiazepines impair the ability of patients to benefit from CBT, whereas antidepressants do not.⁷⁴ That is another reason to think twice about using them. If, however, you diligently try all the other solutions and only benzodiazepines work, treating patients with them is much better than having these patients sit at home and be unable to function.

A PHYSICIAN: Would you continue SSRIs indefinitely?

DR GORMAN: After about 1 year, I try to taper the dosage. SSRIs present the opposite situation than used to occur with tricyclics. People who responded to tricyclics after a few months would say, "I can't stand this stuff anymore. The side effects are horrible, when can I go off?" After 1 year of an SSRI, patients often say, "Why do I have to stop? I feel fine." I usually tell patients that there is a good possibility, particularly if they've had behavior therapy, that they may be able to stop the medication and still feel fine. If they don't feel fine, then I may keep them on SSRIs indefinitely after that.