More than you wanted to know about panic attacks
INTRODUCTION ? Panic attacks and panic disorder are common problems in the primary care setting. Patients with panic disorder may present with classic discrete episodes of intense fear that begin abruptly and last several minutes to an hour; however, they more commonly complain of one or more physical symptoms. By learning to recognize the DSM-IV criteria for panic disorder, the associated behavioral avoidance patterns, and psychiatric comorbid disorders such as major depression and agoraphobia, the primary care physician can make an accurate diagnosis and effectively treat patients with this disorder.
EPIDEMIOLOGY ? A large study of psychiatric illness in the general population found that a surprisingly large number of the community respondents (7.3 percent) experienced infrequent panic attacks without meeting full DSM-IV criteria for panic disorder [1]. The 12-month prevalence of panic disorder was 3.2 percent in women and 1.3 percent in men, with a lifetime prevalence of 5 and 2 percent, respectively. Other reports have found that a somewhat higher prevalence of panic disorder exists in the primary care setting, in the range of 4 to 6 percent [2-4]. Single attacks, which do not meet criteria for panic disorder, are much more common, occurring in up to one-third of individuals at some point in their lifetime [5,6].
Patients with panic disorder may have cognitive, affective, and somatic symptoms. The somatic symptoms often predominate and mimic other medical disorders. Thus, it is not surprising that most patients present to medical rather than mental health providers. In one report, for example, 35 percent of patients with panic disorder presented to their family physician, 32 percent to a hospital emergency department, and only 26 percent to a mental health setting [7].
Healthcare utilization ? Unexplained medical symptoms are common in the primary care setting. The magnitude of this problem was illustrated in a 3-year study of primary care physicians which examined the 10 most common symptoms prompting patient visits to primary care providers: an organic cause for the symptoms was found in only 10 percent of individuals at the end of a one year follow-up period [8]. These 10 symptoms, which included complaints common in patients with panic disorder such as chest pain, fatigue, dizziness, headache, and abdominal pain, prompted nearly one-half of all physician visits.
Many patients with unexplained medical symptoms have panic disorder; they are often unsatisfied with a negative physical work-up and repeatedly seek care for continuing frightening symptoms. In one report, for example, 40 of 57 patients (70 percent) saw an average of 10 physicians before finally receiving a diagnosis of panic disorder [9]. Others have shown that these patients have a significant increased utilization of medical services (eg, consultation rates, specialty physicians referrals, emergency department visits, hospital admissions, medication prescriptions, and laboratory tests) compared to age- and gender-matched controls; their extensive use of resources often precede the diagnosis of panic disorder by as long as 10 years [10].
These data are consistent with several analyses from the Epidemiological Catchment Area data which found that patients with panic disorder had the largest odds ratio for high utilization of medical services when compared with controls without psychiatric illness (8.2 for men and 5.2 for women) [11,12]. Utilization of primary care services among patients with panic disorder is roughly three times that of most patients in the United States health care system [13].
Panic disorder also reduces the quality of life and function in affected patients and their families. Decrements in familial, social, and vocational functioning occur, comparable to that seen with major depression [14,15]. In one report, the number of disability days taken by patients with anxiety disorders was significantly greater than in those with diabetes, cardiac disease, or renal disease [4].
PATHOPHYSIOLOGY ? Several neuroanatomic areas (including the amygdala, cingulated gyrus, midbrain raphe, locus ceruleus, and hippocampus) and a number of neurotransmitters (including norepinephrine, serotonin, and GABA [gamma-aminobutyric acid]) have been the focus of research into the pathophysiology of panic attacks and panic disorder. Other lines of research have focused on the brainstem as the neural trigger for panic attacks, suggesting that patients may inherit brainstem loci that are hyperexcitable (accounting for sensitivity to lactate, increased carbon dioxide levels, and yohimbine in provoking panic attacks in experimental situations) [16]. The prefrontal cortex, an area of the higher brain involved with learning and complex emotions, has been viewed as a neuroanatomical substrate for phobic avoidance in panic disorder [16].
Amygdala ? The amygdala, a structure in the temporal lobe, probably serves a central role in coordinating mammalian fear behaviors and responses. It receives afferent projections from the sensory and association cortices, the thalamus, and the hypothalamus, as well as from nuclei responsible for noradrenergic, dopaminergic, and serotonergic neurotransmission. Efferent output from the amygdala via the locus ceruleus projects to the following areas:
bullet The lateral hypothalamus, which is responsible for fear-induced sympathetic activation (tachycardia, increased blood pressure, sweating, piloerection and pupillary dilatation)
bullet The paraventricular nucleus of the hypothalamus, mediating the "hormone stress response" (eg, corticotropin-releasing hormone)
bullet The dorsal nucleus of the vagus, leading to fear-induced increases in urination, defecation, and bradycardia
Electrical stimulation of the locus ceruleus produces fear behaviors in animal models; decreased fear behaviors result from bilateral lesions of the locus ceruleus [17].
Anterior and posterior cingulate cortices ? The anterior and posterior cingulate cortices have been implicated in the modulation of anxiety in electrophysiologic and brain-mapping studies in human and nonhuman primates [18]. A study using quantitative PET image analysis found a marked reduction in the density of the serotonin type 1a receptor (a specific serotonin receptor) in the anterior and posterior cingulate areas and in the midbrain raphe of patients with panic disorder compared with controls [18]. Stimulation of the serotonin type 1a receptor in these areas regulates the synthesis and release of serotonin.
Hippocampus ? The septohippocampal area has also been proposed as an integrator of incoming novel and unpleasant stimuli [19]. This area has intimate interconnections with the temporal lobe; it is believed that the septohippocampal system predicts the next sensory event to which the organism is exposed, checks whether it actually occurs, and inhibits behavior if there is a mismatch or if the event is aversive.
This biologic system has many similarities to the cognitive model of panic disorder which suggests that panic symptoms result from errors in cognitive appraisal, leading to increased levels of anxiety, arousal, and somatic complaints that result in a vicious cycle of anxiety symptoms. Supportive research has demonstrated that electrical stimulation of this region most commonly produces sensations of fear in the awake human subject [20].
Genetic and environmental factors ? First-degree relatives of patients with panic disorder meet criteria for the disorder in 18 to 41 percent of cases [21-23]. Twin studies have shown a higher concordance for monozygotic than dizygotic twins (31 and 0 percent, respectively, in one series) [24].
The modest concordance among monozygotic twins suggests that environmental factors also play a role in the etiology of panic disorder. Patients with panic disorder recall more childhood fears, more anxiety in childhood, and have a higher rate of grossly disturbed childhood environments than controls [25]. Patients also report significantly more stressful life events than controls in the six months prior to developing anxiety attacks (see below).
CLINICAL MANIFESTATIONS ? Panic attacks are characterized by the sudden onset of intense apprehension, fear or terror, and by the abrupt development of specific somatic, cognitive, and affective symptoms (show table 1). In a study of 55 primary care patients with panic disorder, presenting complaints included the following [26]:
bullet Cardiologic ? 39 percent (chest pain in 22 percent, tachycardia in 25 percent)
bullet Neurologic ? 44 percent (headaches in 20 percent, dizziness in 18 percent, faintness and pseudoseizures in 9 percent)
bullet Gastrointestinal - 33 percent (epigastric pain in 15 percent)
Paresthesias, irritable bowel symptoms, and shortness of breath were also frequently seen.
It is important to appreciate the relationship between panic disorder and chest pain. In a review of the literature, panic disorder was present in 30 percent or more of patients with chest pain who have no or minimal coronary disease; it can also coexist with coronary disease [27]. (See "Diagnostic approach to the patient with chest pain", section on Psychogenic/psychosomatic causes of chest pain).
Patients with panic disorder also may have a higher prevalence of other disorders. These include asthma, labile hypertension [28], mitral valve prolapse, and migraine headaches compared to controls [29].
Panic attacks often occur at times of significant life stress. Various controlled studies have found that these individuals have a higher frequency than controls of stressful life events that connote danger and threat [30], events viewed as uncontrollable or undesirable and that cause severe lowering of self esteem [31], and events that involve the death or severe illness of a friend or relative [32].
Panic attacks may also occur after a physical illness, accident, trauma, rape, or assault, and after endocrinologic changes (eg, hyperthyroidism). Some patients selectively minimize or deny stressful life events, directing their focus toward somatic symptoms; alternatively, they perceive their distress as secondary to the frightening physical sensations.
Natural history ? Panic disorder is a recurrent or chronic disease in the majority of cases. In a review of sixteen studies using modern diagnostic criteria, most patients had improvement of symptoms, but few experienced complete resolution [33]. Comorbid agoraphobia, major depression, and personality disorders were found to predict a poorer outcome. In a second study, 55 patients with panic disorder initially evaluated in a specialty clinic were reinterviewed 15 to 60 months after naturalistic treatment in the community [34]. Most patients had improved, but only 5 were asymptomatic at follow-up.
DIFFERENTIAL DIAGNOSIS ? There is significant psychiatric comorbidity in individuals with panic disorder:
bullet One-third to one-half of patients meet DSM-IV criteria for major depression at initial presentation, while 60 to 90 percent have had one or more lifetime episodes of major depression [35-39].
bullet Approximately 40 percent meet DSM-IV criteria for agoraphobia (anxiety and avoidance of places from which escape might be difficult or help unavailable in the event of panic symptoms) [1]. (See "Overview of phobic disorders").
bullet Social phobia coexists with panic disorder in 10 to 50 percent of patients [40].
bullet Civilian patients with post-traumatic stress disorder (PTSD) commonly have combinations of symptoms of PTSD, panic disorder, and major depression. (See "Overview of post-traumatic stress disorder").
The majority of patients in the primary care setting who meet criteria for generalized anxiety disorder have a primary diagnosis of panic disorder or major depression.
Another problem in some patients with panic disorder (16 percent in one series) is a history of substance abuse [41]. Alcohol and sedative hypnotics are sometimes used in desperation to control symptoms of panic disorder. These agents have a short-lived anxiolytic action, but are subsequently associated with exacerbation of anxiety and panic attacks during declining blood levels. In addition, coexisting alcohol abuse with recurrent withdrawal can lead to a kindling effect on central controls of the sympathetic nervous system, resulting in more frequent and severe panic attacks [42].
Somatization disorder ? One of the most difficult psychiatric diagnoses to differentiate from panic disorder is somatization disorder. In one study, for example, 32 of 78 women (41 percent) with somatization disorder also met criteria for panic disorder [43]. (See "Somatization").
Three types of somatization are commonly seen in the primary care setting [44]:
bullet Acute transient somatic symptoms in response to stress
bullet Subacute somatization in patients with major depression and panic disorder
bullet Chronic somatization in patients with developmental histories of physical, sexual, and emotional abuse and emotional neglect
The last group of patients has a sense of powerlessness interpersonally. They may use somatic symptoms to manipulate a spouse, avoid intimacy, or to obtain disability or prescribed medication (eg, opiates and benzodiazepines) [45].
Treatment of panic disorder in patients with subacute somatization frequently cures the physical symptoms and the tendency to be hypochondriacal. In contrast, treatment of panic disorder or major depression in those with chronic somatization may only lead to a 30 to 40 percent reduction in somatic symptoms because of chronic social stressors and maladaptive patterns of coping with stress.
Medical disorders ? The possibility of organic etiologies should always be considered prior to making the diagnosis of panic attacks. A number of conditions can mimic panic symptoms, such as angina, arrhythmias, chronic obstructive pulmonary disease, temporal lobe epilepsy, pulmonary embolus, asthma, hyperthyroidism, and pheochromocytoma (show table 2). Treatment side effects must also be considered in patients with certain medical disorders. Examples include hypoglycemia in the patient with diabetes and toxic serum aminophylline concentrations in the patient with asthma.
DIAGNOSIS ? The DSM-IV criteria for panic disorder focus upon three areas of symptom characteristics (show table 1):
bullet Recurrent unexpected panic attacks
bullet At least one month of persistent concern or anxiety over the possibility of additional attacks
bullet A significant change in behavior in relation to the panic attacks
Agoraphobia may become a comorbid problem when the persistent concern and worry of having another attack and avoidance become significant. (See "Overview of phobic disorders").
A correct diagnosis of panic disorder is made in 95 percent of cases if anxiety or depression are the presenting complaints; however, the likelihood of correct diagnosis decreases to 48 percent with a primarily somatic presentation [46]. This is an important distinction because, in the primary care setting, approximately 90 percent of patients with panic disorder have mostly somatic symptoms [47].
History and physical examination ? A complete medical and psychiatric history will lead to an accurate diagnosis of panic disorder in most cases. The history taking process should begin in an open-ended manner and be unhurried.
Information should be elicited about current life stress, separations, recent deaths, patient concerns and fears, interpersonal problems, recent substance abuse, use of medications, and caffeine intake. The patient should also be asked about avoidance patterns which have developed since the onset of panic attacks. Involving family members can provide valuable information about precipitating events or may help clarify the current symptomatology.
Differentiating patients with concurrent somatization disorder requires questions regarding the existence of chronic somatization symptoms since adolescence, a family history of females with somatization disorder, a family history of males with antisocial personality disorder or substance abuse problems, chaotic family histories, childhood sexual or physical abuse or emotional neglect, and abuse of prescription drugs, street drugs, or alcohol.
A thorough physical and neurologic examination is essential to rule out an organic cause of symptoms.
Diagnostic testing ? Limited laboratory testing such as thyroid function tests, complete blood count, and a chemistry panel are sufficient in most cases. Other laboratory and radiological tests may be indicated in more complicated cases. An electrocardiogram is required in anyone with significant cardiovascular symptoms, although it is likely to have a low yield in female patients under 40 years of age.
Costly and extensive diagnostic evaluations have low yield. As an example, 20 to 30 percent of patients who undergo coronary arteriography for chest pain are found to have normal coronary arteries. (See "Syndrome X: Angina pectoris with normal coronary arteries"). In one report, over 40 percent of such patients met the DSM-III-R criteria for panic disorder compared with 6.5 percent of those with chest pain and positive arteriograms [48].
Similar findings were noted in another study of 300 patients evaluated for pheochromocytoma [49]. Only one had the disease, while 40 percent met the criteria for panic disorder compared with 5 percent of hypertensive controls.
TREATMENT ? Treatment of panic disorder varies depending upon the stage of development of the disorder (show table 3). Primary care providers will often see patients after the first panic attack, or in the earliest stages of the syndrome before phobic avoidance, anticipatory anxiety, or agoraphobia have developed. In the case of infrequent attacks in which there is no phobic avoidance, education about anxiety, supportive psychotherapy to deal with current life stressors, and instruction in relaxation techniques may be all that is necessary.
Providers can help patients who are experiencing a panic attack by reassuring them that they are not going to die, and that the panic attack will soon subside. Helping the patient relax in a quiet room either in the presence of a provider or a significant other, with emphasis on slowing down one's breathing can be helpful. Patients subjectively experience shortness of breath during a panic attack, when in fact they are usually hyperventilating. Pointing this out during or after the attack may help prevent a vicious cycle during the current or subsequent panic attacks in which the distressed patient hyperventilates instead of consciously slowing their breathing. In an emergency room or other clinical setting, sublingual, intravenous, or oral benzodiazepine (eg, lorazepam) may be used in addition, to help abort a panic attack.
Once panic attacks have become more frequent and phobic avoidance has developed, general therapeutic guidelines should be followed:
bullet Patients should be offered either pharmacologic treatment or cognitive behavioral therapy which have been shown in controlled trials to be equally effective in the treatment of panic disorder [50,51]. For more severe cases, or for cases that are more complex (e.g. panic disorder with comorbid PTSD or depression), pharmacotherapy as a first-line treatment is recommended.
bullet The basic behavioral and physiologic aspects of panic disorder should be discussed with the patient to help reframe a potentially frightening syndrome into a common treatable disorder in which the patient's "alarm system" trips indiscriminately. All side effects of treatment and the length of treatment should be discussed. Educational material or appropriate references about the course of the illness and treatment should be provided (eg, NIH publications: "Understanding Panic Disorder" ? NIH-93-3509; "Getting Treatment for Panic Disorder: Information for Patients, Families, and Friends" ? NIH-94-3642). These pamphlets and other information can be obtained through a toll-free number, 1-888-8-ANXIETY, or through a Web site
http://www.nimh.nih.gov/healthinformation/anxietymenu.cfm. The Anxiety Disorders Association of America (ADAA;
www.adaa.org or 301-231-9350) is a good resource through which patients and providers can obtain information about resources and qualified providers in their area. These measures often help to reassure the patient that he or she is not "going crazy."
bullet Once panic attacks are alleviated, the patient should be encouraged to reenter situations that have been avoided as a result of panic attacks, such as crowded places, parties, or driving on the freeway.
bullet A thorough psychosocial review should address personality traits, maladaptive coping styles, and current life stress, all of which can contribute to persistence of panic attacks. The patient may need referral to a marital or family therapist if the attacks develop in the context of marital or family strife. Similarly, referral for adjunctive psychodynamic or interpersonal therapy is indicated when chronic personality traits, low self-esteem, poor interpersonal relationships, and rejection sensitivity are prominent features.
bullet Psychiatric referral is indicated in patients with panic disorder who do not respond to adequate therapy within several months, who have serious comorbid psychiatric disorders, or who have serious suicidal ideation with suicidal intent or a plan.
Cognitive behavioral therapy ? A number of well performed, randomized clinical trials have established the clinical efficacy of cognitive-behavioral treatments in patients with panic disorder. (See "Psychological treatment of psychiatric disorders", section on Panic disorder.) Cognitive behavioral psychotherapy focuses on correcting maladaptive cognitions that play a role in amplifying normal bodily sensations so that they are experienced as frightening and uncontrollable [52]. As an example, increased heart rate after walking up a flight of stairs may provoke thoughts that one is having a cardiac arrhythmia or heart attack, which in turn provoke a greater increase in heart rate and respiratory rate.
The efficacy of cognitive behavioral therapy in patients with panic disorder was demonstrated in a double-blind, placebo-controlled trial of 312 patients who were randomly assigned to receive imipramine (up to 300 mg/day), cognitive behavioral therapy alone, placebo, cognitive behavioral therapy plus imipramine, or cognitive behavioral therapy plus placebo [51]. In the acute phase of therapy (first three months), both cognitive behavioral therapy and imipramine were superior to placebo and were associated with a similar number of responders; combining therapy did not confer any significant additional benefit acutely. Among patients who responded, imipramine was associated with a higher quality response than cognitive behavioral therapy. Following another six months of maintenance therapy, there appeared to be an advantage for those who received combination therapy, although this advantage did not persist once treatment was stopped. Overall, patients who were treated initially with cognitive behavioral therapy alone did best in the follow-up phase after all therapy had stopped, emphasizing the point that panic disorder is a chronic illness that often requires indefinite therapy (see below), and that some of the skills learned in psychotherapy can be used once the therapy has stopped.
Medical therapy ? There are four classes of medications that are equally useful and significantly more effective than placebo for panic disorder: serotonin reuptake inhibitors (SSRIs); tricyclic antidepressants (TCAs); benzodiazepines; and monoamine oxidase inhibitors (show table 4).
Serotonin reuptake inhibitors ? Selective SSRIs are the first-line treatment of choice for panic disorder in the primary care setting [53]. Controlled studies have supported the efficacy of fluvoxamine, paroxetine, sertraline, citalopram, and fluoxetine [54-58]. As an example, in a report of 75 patients with moderate to severe outpatient panic disorder, 13 of 23 (57 percent) receiving fluvoxamine were rated moderately improved or better at four weeks versus 8 of 20 (40 percent) given cognitive therapy and 5 of 23 (22 percent) receiving placebo [55]. Freedom from panic attacks was reported in 43, 25, and 4 percent of patients, respectively.
In another report, 278 patients with a mean of 9.5 to 11.6 full panic attacks during the screening period were randomly assigned to double-blind treatment with a 10-week course of placebo or paroxetine at a dose of 10, 20, or 40 mg/day [56]. During the last two weeks of the study, 67.4, 65.2, and 86 percent of the patients taking the various doses of paroxetine, respectively, were free of full panic attacks compared with 50 percent of the placebo-treated individuals.
Jitteriness, restlessness, agitation, headache, gastrointestinal symptoms (diarrhea and nausea), and insomnia are common side effects with SSRIs (show table 5). Thus, lower daily doses should initially be prescribed (eg, 5 to 10 mg of paroxetine; 12.5 to 25 mg of sertraline; 25 mg of fluvoxamine; 20 to 40 mg of fluoxetine, 20 to 40 mg of citalopram) for approximately one week, and then gradually titrated up to full doses. Therapeutic doses to completely alleviate panic attacks in most cases are 20 to 40 mg of paroxetine, 50 to 200 mg of sertraline, 100 to 300 mg of fluvoxamine, 20 to 40 mg of fluoxetine, and 20 to 40 mg of citalopram.
An initial therapeutic response typically occurs within four to six weeks of SSRI therapy. If there is no response by 8 to 12 weeks at a maximum therapeutic dose, the patient should be given a second trial of another antidepressant (eg, a different SSRI or TCA) or should be referred to a psychiatrist.
A significant percentage of men and women develop sexual side effects after several weeks or months of SSRI therapy; these adverse effects can lead to medication discontinuation. The management of SSRI-induced sexual dysfunction is discussed separately. (See "Pharmacology and use of antidepressants", section on Sexual dysfunction).
Tricyclic antidepressants ? Tricyclic antidepressants (TCAs) appear to be as effective as SSRIs for the treatment of panic disorder [59]. Although cheaper, TCAs are not generally considered first line therapy because of associated adverse effects [60]. Nevertheless, patients who have responded well to TCAs for depression or anxiety in the past and in whom side effects are not limiting can be treated with this class of medications. Clomipramine and imipramine have been shown to be effective in controlled studies [61,62]; clinical experience suggests that other TCAs are also efficacious. Nortriptyline has fewer side effects than most tricyclic antidepressants.
As with SSRIs, TCAs should be started at low doses (eg, 10 to 25 mg of imipramine) with gradual increases every four to five days until panic attacks cease. Often patients will not tolerate the side effects of the TCAs (show table 5); switching to an SSRI, which usually has fewer side effects, is indicated in these circumstances.
Benzodiazepines ? High potency benzodiazepines offer an alternative treatment for patients who are unable to tolerate or who do not benefit from an adequate trial of SSRI agents or TCAs. Alprazolam, lorazepam, and clonazepam have all been found to be more effective than placebo for the treatment of panic disorder [62,63]. Starting doses of clonazepam 0.25 to 0.5 mg TID, alprazolam 0.25 mg QID, and lorazepam 0.5 mg TID can be increased gradually every one to three days until panic attacks cease. Between dose "rebound" is minimized by using longer half-life benzodiazepines such as clonazepam.
Patients with polydrug or alcohol abuse, chronic pain disorders, and severe personality disorders should not be prescribed benzodiazepines in light of potential problems with abuse.
Monoamine oxidase inhibitors ? Monoamine oxidase inhibitors are at least as effective as tricyclic antidepressants for the treatment of panic disorder [64]. However, their use in the primary care setting is limited because of significant side effects such as orthostatic dizziness, and the need for a tyramine-restricted diet to prevent possible hypertensive crises. (See "Diagnosis and treatment of pheochromocytoma in adults", section on Other causes of sympathetic overactivity").
Combination therapy ? Benzodiazepines can be prescribed in the severely symptomatic patient while concomitantly initiating treatment with an SSRI or TCA. A slow taper of the benzodiazepine is started (10 to 20 percent per week) once the antidepressant begins to take effect.
Support for a combination therapy approach in patients with severe symptoms was derived from a study of 34 patients with panic disorder who were randomly assigned to receive sertraline (target dose 100 mg/day) plus either clonazepam (0.5 mg TID) or placebo for four weeks, followed by clonazepam taper for three weeks and then discontinuation for the remaining five weeks of the trial [65]. There were significantly more responders in the clonazepam than the placebo group after week one (41 versus 4 percent) and week three (63 versus 32 percent).
Collaborative care of panic disorder ? Randomized, controlled trials of pharmacologic and psychotherapeutic treatments of panic disorder have largely been tested in tertiary care settings with selected populations (ie, efficacy trials). Randomized trials of health services approaches to more effectively deliver these treatments to primary care patients have been completed (ie, effectiveness trials). Despite highly effective medications and psychotherapy, few primary care patients who are accurately recognized as having panic disorder receive guideline levels of care.
"Collaborative care" strategies that seek to overcome patient, physician, and process-of-care barriers to mental health treatment, along with judicious use of specialty consultation and close and sustained follow-up of patients, have been shown to improve depression outcomes compared with care as usual. Such strategies have demonstrated superior quality of mental health care and greater cost-effectiveness compared with care as usual in the primary care setting. In a randomized controlled study of the clinical effectiveness of panic disorder pharmacotherapy embedded in a disease management framework of "collaborative care," 115 patients with panic disorder from three primary care clinics were randomly assigned to "collaborative care" (in which patients received educational videotapes and pamphlets; pharmacotherapy with a selective serotonin reuptake inhibitor; 2 psychiatrist visits and 2 telephone calls in the first 8 weeks; and up to 5 telephone calls between 3 and 12 months' follow-up) or "usual care" [66]. Patients in collaborative care were more likely to receive adequate (type, dose, duration) medication treatment and improved significantly more over time compared with usual care patients on anxiety, depression, and disability measures over a one-year period.
Treatment duration ? Patients with panic disorder can usually be stabilized (eg, alleviation of panic attacks, avoidance, and agoraphobic behavior) within four months. Pharmacotherapy should be continued for at least one year and then reassessed. Relapse during or after a medication taper is less likely if comorbid psychiatric and medical conditions have been optimally treated and psychosocial stressors have been significantly reduced.
Antidepressants can be tapered over 4 to 8 weeks, while benzodiazepines may need to be tapered over 10 to 16 weeks to minimize withdrawal symptoms [67]. If panic, anxiety, avoidance, or agoraphobic symptoms recur after tapering, indefinite maintenance therapy should be considered. In addition, a referral for cognitive-behavioral treatment may be added if this has not yet been tried.