Print Friendly 

Improving Treatment Response in Panic Disorder

Mark Vanelli, MD, MHS



 Primary Psychiatry. 2005;12(11):68-73

Faculty Affiliations and Disclosures

Dr. Vanelli is chief medical officer at Adheris Inc. in Burlington, and a lecturer in the Department of Psychiatry at Harvard Medical School in Boston, Massachusetts.

Disclosure: Dr. Vanelli has served as a consultant to and on the speaker’s bureaus of Eli Lilly, Janssen, Novartis, Pfizer, and Schwarz Pharma; and has received research support from Abbott.

Please direct all correspondence to: Mark Vanelli MD, MHS, Chief Medical Officer, Adheris Inc., 1 Van De Graaff Dr, Burlington, MA 01803; Tel: 781-425-6636; Fax: 718-229-8878; E-mail:

Focus Points

• Longitudinal studies of patients with panic and anxiety disorders under routine care suggest that the vast majority of patients will experience ongoing or breakthrough symptoms despite treatment.

• More effective medication strategies, in addition to psychosocial interventions, may help patients self-manage reoccurring symptoms.

• As-needed use of benzodiazepines can help individuals maintain function and quality of life when situational anxiety, avoidance, and panic threaten daily work and social situations.

• Co-therapy of a benzodiazepine and an antidepressant may help improve treatment outcomes by speeding the onset of anti-panic action. Careful follow-up and education at the initiation of therapy may help improve medication compliance.



Although antidepressants and benzodiazepines are effective in reducing the frequency of panic attacks during short-term clinical trials, follow-up studies suggest that under routine care, panic disorder is typically a chronic condition characterized by ongoing or breakthrough symptoms. As a result, panic disorder patients are likely to experience levels of work and role impairment similar to that experienced by patients with arthritis, asthma, and major depression. Selective serotonin reuptake inhibitors and serotonin norepinephrine reuptake inhibitors are often considered first-line drug therapy in treating panic disorder, but their effectiveness may be undermined by rates of premature medication discontinuation that may reach 30% to 55% within 3 months of initiating therapy. Strategies to improve treatment response in panic disorder, and anxiety disorders in general, include careful patient follow-up and education during the first 30 days after therapy is initiated, when the risk of medication discontinuation is greatest. Use of benzodiazepines, particularly on an as-needed or “rescue” basis, may be useful in helping patients better manage breakthrough episodes of anticipatory or situational anxiety, phobic avoidance, or panic that often returns and threatens daily function or quality of life. Recent studies suggest that the risk of significant dose escalation with chronic benzodiazepine use occurs in <2% of patients and that benzodiazepines generally provide a favorable risk-benefit profile.



Panic disorder is common, costly, and often underdiagnosed in primary care settings. Although pharmacotherapy reduces the frequency of panic attacks in panic disorder, particularly in clinical trial settings, studies of patients with panic disorder under routine care suggest that ongoing symptoms and functional impairment are common.

Treatment for patients with panic disorder in primary care may be improved in two ways. First, rates of recognition and diagnosis can be increased; estimates of the prevalence of panic disorder in primary care range from 5% to 8%, but studies suggest that only 34% to 50% of patients with panic disorder are diagnosed and treated.1-3 Second, clinicians can focus on treatment strategies that are likely to help patients with panic disorder improve functioning and quality of life. At 6–10 years after diagnosis, an estimated 70% of patients remain symptomatic with ongoing or breakthrough symptoms.4 The degree of occupational impairment for panic disorder is estimated to be comparable to that associated with other chronic medical conditions, such as arthritis and heart disease.5

With the aim of aiding panic disorder diagnosis and improving the function and quality of life of panic disorder patients, this article will review symptoms of the disorder, diagnostic difficulties, treatment outcomes in clinical trials versus routine care, work and social impairment associated with the disorder, pharmacologic treatment strategies, and recommendations to improve treatment response in clinical practice.

Presenting Symptoms and Diagnostic Difficulties

Panic attacks are characterized by sudden onset of fear or discomfort accompanied by a combination of cardiac, pulmonary, gastrointestinal, neurological, and psychiatric symptoms, which typically peak in 10 minutes. Individuals with unexplained, recurrent panic attacks followed by at least 1 month of ongoing worry about having another attack, are diagnosed with panic disorder, according to Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition-Text Revision (DSM-IV-TR) criteria.6 Although attacks are frightening and disabling in their own right, they are often accompanied by fear of having another attack (anticipatory anxiety) and avoidance of situations in which attacks have previously occurred or in situations from which escape may be difficult should they reoccur (panic disorder with agoraphobia). These avoidance patterns often lead to impaired role and work performance, and, in patients with agoraphobia, may leave individuals housebound.

Panic disorder symptoms mimic potentially life-threatening medical disorders. Therefore, patients often seek medical assistance soon after symptom onset, and present to medical rather than psychiatric providers. In one study that identified the site of care after a self-defined worst panic attack, patients reported seeking care from medical providers 52% of the time and from mental health professionals 17% of the time.7 Underdiagnosis and misdiagnosis may be common; the presentation of symptoms, which include numerous somatic manifestations, often result in referrals to cardiologists, gastroenterologists, or neurologists.3,8 Panic disorder may account for 40% to 60% of patients with atypical chest pain and normal cardiac catheterization findings; 20% to 40% of patients with irritable bowel symptoms; and 15% to 20% of the patients referred for evaluation of dizziness.9 Once patients present with “medical” symptoms, or attribute symptoms to a plausible psychosocial stressor, psychiatric diagnoses become less likely.10 Comorbid conditions can also complicate treatment.

Clinical Course of Panic Disorder During Routine Care

During clinical trials conducted under highly supervised conditions, 30% to 80% of patients with panic disorder achieve complete suppression of panic attacks. However, the trials typically occur over a period of weeks, which cannot determine how well patients fare under conditions of routine care over a period of years.11,12 To determine the degree of symptom relief achieved under conditions of routine care, investigators in the Harvard/Brown Anxiety Research Project (HARP) designed a prospective, longitudinal, observational study in which 711 subjects were evaluated with respect to the degree of symptom remission they experienced. Patients with diagnoses of panic disorder, panic disorder with agoraphobia, generalized anxiety disorder (GAD), or social phobia were enrolled and followed for an 8-year period. Exclusion criteria was kept to a minimum to reflect characteristics of patients in community medical practices.13-20 Recovery was defined as a 2-month period with minimal or no symptoms. Relapse was defined by a full return of symptoms meeting DSM-IV21 criteria for a 2-week period. Patients with ongoing symptoms were those reporting a continuation of moderate symptoms and functional impairment.  

This study found that low rates of recovery and high rates of relapse were the norm.15 The rates for women only are featured in Figure 1. Two-year remission rates for panic disorder were initially good for women (61%) as well as men (58%).16 In contrast, remission for women or men with panic disorder with agoraphobia (23% and 22%, respectively), women and men with GAD (23% and 22%, respectively), and women and men with social phobia (18% and 15%, respectively), were much worse. At 8 years, men with panic disorder had a remission rate of 69% and a relapse rate of 21%. For women, however, the remission rate was 76%, but the relapse rate was 64%, a 3-fold rate of symptom return relative to men (Figure 2).15 Notably, a high percentage of patients who initially experienced remission subsequently relapsed. Although >80% of subjects with panic disorder and GAD received medication management, the dose, duration, and degree of medication adherence achieved at each point in the study were not reported; self reports, not pharmacy records, were used to document medication use. Given the high rates of medication discontinuation observed among outpatients, it is likely that poor medication adherence contributed to the high rates of relapse.

Other studies also document ongoing impairment and disability among patients with panic disorder who are followed over time under routine care. For example, one follow-up study showed that 60% of patients suffered panic attacks and 20% remained agoraphobic 4 years after participating in a clinical trial.22 According to the American Psychiatric Association (APA) guidelines for panic disorder,4 40% to 50% of panic disorder patients in tertiary care settings are likely to be symptomatic but improved, 20% to 30% the same or worse, and only 30% are well at approximately 4–6 years posttreatment.12,22

Role and Work Impairment

Patients with panic disorder are likely to experience better physical functioning but worse role functioning than those with other serious medical or emotional disorders. For example, the Health-Related Quality of Life survey compared the role function and physical and emotional health of 443 patients with panic disorder to that of 9,839 patients with diabetes, hypertension, depression, arthritis, or heart disease. Patients with panic disorder reported good physical functioning, but greater role limitations due to emotional distress.23 The MacArthur Foundation Midlife Development study surveyed 2,074 individuals in the United States on the number of work-loss days (days home due to illness) and partial work days (physically present, but less productive) experienced in the previous month due to any chronic condition.5 On average, individuals with panic disorder reported 5.1 work impairment days, compared to 4.0 for patients with arthritis, 4.3 for major depression, 6.6 for heart disease, and 10.9 for cancer. Moreover, work loss due to poor productivity was as common as sick days, although largely undetectable from an employer perspective.

Clinical Trials Versus Clinical Reality

The low percentage of patients with panic disorder and related anxiety disorders achieving even a brief period of symptom relief is both surprising and sobering given the increasing number of agents available to treat anxiety disorders. This gap, between the outcomes achieved in clinical trials and those achieved during routine care, helps underscore the need for studies that document the clinical effectiveness and cost of routine care. While major funding mandates exist to support basic research (National Institutes of Health) and the development of new medicines (pharmaceutical manufacturers), the same is not true for routine healthcare delivery. Although biomedical research composes 5.6% of US health expenditures, only 1.7% of this is spent on evaluating the effectiveness and cost of care that patients actually receive. More studies like those conducted by HARP investigators are clearly needed.24,25

Poor Medication Adherence

Selective serotonin reuptake inhibitors (SSRIs) and serotonin norepinephrine reuptake inhibitors (SNRIs) are considered first-line treatments for panic disorder, with ongoing maintenance treatment generally recommended for 12–18 months after improvement.4 Therefore, adherence to antidepressant therapy is especially important. Based on their side-effect profiles, SSRIs and SNRIs were anticipated to replace tricyclic antidepressants and improve adherence to antidepressant therapy. However, this may not have occurred. Recent studies suggest that 30% to 55% of patients starting antidepressants may discontinue therapy within 3 months of starting.26 Health system factors that might predict  higher rates of adherence have eluded investigators; for example, rates of antidepressant discontinuation have been found to be similar for patients under psychiatric and generalist care, and for managed care and fee-for-service systems.27 Factors causing patients to discontinue antidepressants are numerous and may include insufficient education regarding the required extended duration of use; impatience with the delayed onset of action; side effects that emerge on antidepressant initiation, such as anxiety and insomnia, or after extended use, such as sexual dysfunction and weight gain; and perception of clinical improvement.28 The cost of medications may increasingly contribute to antidepressant discontinuation. Studies suggest that a doubling of antidepressant copayments may prompt declines in use that range from 8% to 26%.29 

Knowing the point in therapy and the kinds of patients at greatest risk for discontinuation may help clinicians to practically focus their efforts to improve medication compliance. A study of 20,296 patients who initiated therapy on paroxetine (controlled-release formulation), and were followed over a 12-month period after filling an initial outpatient prescription, found that individuals who were new to the medication were at the greatest risk of discontinuation and that this risk was greatest during the first month of therapy (Figure 3).26 In this study, patients were divided into two groups based on the hypothesis that individuals initiating therapy for the first time would be at greatest risk of discontinuing therapy. Patients new to therapy (rookies) were identified as those who had not received antidepressant prescriptions for 180 days prior to the index antidepressant fill; veterans were those continuing therapy. Patients who switched from one antidepressant to another were considered to have continued therapy. Patients were considered to have discontinued therapy when they were 30 days late for a scheduled refill. Notably, only 58% of  9,295 rookies and 80% of 11,071 veterans remained in therapy 60 days after filling their initial prescription. Four months after initiating therapy, only 37% of rookies and 66% of veterans continued with therapy. Similar results were observed with other SSRIs/SNRIs.26

Current Pharmacologic Treatment Strategies in Panic Disorders

High-potency benzodiazepines (eg, alprazolam and clonazepam) can block panic attacks that suddenly emerge and can provide rapid relief of situational and anticipatory anxiety. Drawbacks include sedation, potential disinhibition, and physiologic dependence.9 Initiation at low doses is recommended with increases every 3–4 days until a therapeutic dose is reached.9 Typical dose ranges for alprazolam are 1.5–8.0 mg/day for panic disorder and 0.75–4.0 mg/day for anxiety. Typical dose range for clonazepam are 0.5–4.0 mg/day for panic disorder.30  

The Table provided presents a comparison of selected benzodiazepines and SSRI/SNRIs.31 One advantage of SSRI/SNRIs is their ability to treat both depression and comorbid anxiety disorders. SSRI/SNRIs are also less likely to be associated with physiological dependence or with major impairments of memory or motor coordination.9 A disadvantage of SSRI/SNRIs is latency to onset of effect; it can take up to 4 weeks before a response occurs and from 8–12 weeks for a complete response.4 Recommended daily starting doses for SSRIs and SNRIs are generally half those used for treating depression, to avoid potential antidepressant-induced anxiety.

Current APA practice guidelines recommend 12–18 months of maintenance therapy for panic disorder, with discontinuation attempted if the patient has experienced significant or full improvement. Discontinuation with either antidepressants or benzodiazepines is associated with discontinuation-related symptoms and relapse. Patients who relapse may be offered a new trial on medication along with cognitive-behavioral therapy (CBT).4

Current Trends in the Pharmacotherapy of Panic Disorder

Combination Use of SSRI/SNRIs and Benzodiazepines

In principle, combinations of an SSRI/SNRI and a benzodiazepine could be advantageous because of their different, but potentially complementary, mechanisms of action on serotonin and γ-aminobutyric acid systems.32 Advantages of combination use of antidepressant and benzodiazepines include rapid anxiolysis, improved efficacy in combination, and reduced SSRI/SNRI-related anxiety and agitation upon treatment initiation. Combination medication use may also improve medication compliance and reduce medication dosages, perhaps as a result of improved medication effectiveness, reduced side effects, or some combination of the two. The advantages of such augmentation strategies need to be balanced against the risk of drug-drug interactions.

Reduced Time to Symptom Control

A 12-week randomized, double-blind study in patients with panic disorder suggests that the combination of benzodiazepines and SSRI/SNRIs results in more rapid symptom control than either agent alone (Figure 4).32,33 In this study, patients received sertraline on an open-label basis and then were randomly assigned to receive clonazepam 3 times/day or placebo for 4 weeks, after which the clonazepam or placebo dose was tapered over the course of 3 weeks. At the end of the first week of combination treatment, the proportion of responders was significantly higher (P=.003) among those receiving clonazepam/sertraline (41%) than those receiving placebo/sertraline (4%). The between-group difference was also statistically significant at week 3 in favor of the active combination (63% versus 32%; P=05). Adverse events did not differ between groups. These results suggest that the combination of a benzodiazepine and SSRI/SNRI can provide more rapid anxiolysis of panic symptoms in patients with panic disorder than sertraline alone. In line with these results, the APA treatment recommendations suggest use of benzodiazepines for early control of panic and anticipatory anxiety as the effects of nonbenzodiazepine maintenance medications and psychosocial treatment modalities take effect.4


Reduced SSRI/SNRI-Related Anxiety  

Not uncommonly, patients with anxiety disorder may experience anxiety or agitation when an SSRI/SNRI is first started.34 Use of a benzodiazepine, either on a regular or as-needed basis (PRN), may alleviate such side effects and reduce the risk of medication discontinuation at the start of treatment.

PRN Benzodiazepine Use

PRN use of benzodiazepines appears to be substantial. As part of the longitudinal HARP study, Bruce and colleagues35 found that among patients with panic disorder and panic disorder with agoraphobia, 40% of those on lorazepam and 51% of those on alprazolam used these agents on a PRN basis. PRN use, which stand for the Latin term pro re nata (according to circumstance), allows patients to manage symptoms as they arise; it is best directed by specifying both a frequency and maximal number of daily doses (eg, alprazolam 0.5 mg PRN every 4 hours; maximum of 4 doses in a 24-hour period). Studies of outpatient PRN benzodiazepine use are lacking.  

To better understand the role of PRN benzodiazepine use in managing anxiety and panic symptoms, Fitzgerald36 analyzed self-reported patterns of alprazolam use among 100 patients with anxiety disorders. Overall, patients used PRNs to reduce situational anxiety, anticipatory anxiety, or phobic avoidance when symptoms threatened function in a work or social role. Activities of daily living in which PRN benzodiazepine use proved beneficial for people with panic disorder and panic disorder with agoraphobia, included leaving the house to shop, keeping doctors’ appointments, attending important social and work-related events, participating in community events, traveling by car or plane, or carrying out job obligations. PRN benzodiazepine use often augmented maintenance medications, such as SSRI/SNRIs or long-acting benzodiazepines, to provide “rescue” treatment for breakthrough symptoms. This is illustrated schematically in Figure 5. APA treatment guidelines for panic disorder recommend the preferential use of benzodiazepines in situations in which rapid symptom control is necessary to preserve function or quality of life.4

Although SSRI/SNRIs demonstrate efficacy in decreasing the frequency of panic attacks, they may be less effective than benzodiazepines in treating the anticipatory anxiety and avoidance, which are frequent behavioral responses to panic attacks. It is in this context that PRN benzodiazepine use may be both more beneficial and more common than generally recognized. In clinical practice, physicians may not learn about PRN use from their patients because such usage occurs outside of the office visit. Furthermore, some data suggest that long-term users of benzodiazepines may not discuss their drug usage with their physicians.37 Moreover, many patients may switch to PRN use or supplement regular schedules with PRN use over time.37 Additional study in the outpatient setting is needed to assess the impact of PRN benzodiazepine use in panic disorder, alone and in combination, and its impact on quality of life and functioning.

Benzodiazepines: Low Risk of Dose Escalation

Ongoing concerns about benzodiazepine dependence, abuse, and dose escalation may limit appropriate use of these agents. The APA Task Force on Benzodiazepine Dependence, Toxicity, and Abuse clearly notes that there are no data to suggest that long-term use of benzodiazepines commonly leads to dose escalation or recreational abuse.4

From a clinical perspective, underuse because of unjustified fears of addiction may be a more critical issue. Several recent studies document that dose escalation of alprazolam is not characteristic of long-term use and that the risk of significant dose escalation may be <2%. One study examined changes in dose among 2,440 Medicaid recipients with a history of ≥2 years of continuous benzodiazepine use, converting all doses into diazepam milligram equivalents (DMEs) for comparison.38 The median daily dose was 10 DMEs over 2 years of continuous use (eg, equivalent to 1 mg alprazolam or .5 mg clonazepam). No clinically or statistically significant increases in daily dosages occurred over time. The incidence of dose escalation to a high dose, defined as 20 DMEs for older patients and 40 DMEs for younger patients, was 1.6%. Statistically significant predictors of dose escalation were the use of lorazepam, use of an antidepressant, and filling prescriptions at two different pharmacies in a 7-day period (“pharmacy hopping”). Patients >65 years of age or with disability were significantly less likely to escalate their doses than younger patients. A naturalistic study of patients with panic disorder found that when benzodiazepines were used PRN, patients used far lower average daily doses than typically prescribed.34

Although some evidence suggests that benzodiazepine abuse is more likely to occur among those with known substance abuse disorder, the data are equivocal and do not support an absolute contraindication in such patients.39,40 Benzodiazepines are not primary drugs of abuse, but may be used to enhance or modulate effects of the primary abused substances, such as cocaine, methadone, and alcohol.39  

Together, these data indicate that the likelihood of benzodiazepine dose escalation and abuse is minimal in patients who are using benzodiazepines for appropriate medical uses. Nevertheless, clinical experience suggests that patient concerns about dependence on benzodiazepines are common and may contribute to treatment failure, as patients independently reduce the dose or duration or frequency of use of prescribed benzodiazepines. In panic disorder, patients who completed both a short-term treatment plus a maintenance treatment of 6 months with either alprazolam or imipramine were significantly more likely to be panic free than those who did not complete treatment (85% versus 55%; P<.01) and long-term treatment.31 The maintenance period was not associated with tolerance to either agent, and completion of maintenance was predictive of remission. In summary, inadequate dosing or premature discontinuation of  benzodiazepines may contribute to ongoing symptoms and distress, which in turn may exacerbate medication noncompliance and consequent treatment failure.9

Treatment Recommendations

The natural history and high relapse rate of panic disorder underscores the importance of a well thought out long-term treatment plan. Several steps can increase the probability of successful outcomes:

•    Set appropriate expectations and realistic goals for treatment by discussing the chronic nature of panic disorder and the possibility of ongoing or breakthrough symptoms. An appropriate goal is symptom control, as well as the preservation of function and quality of life rather than cure.

•    Engage the patient in the development of a comprehensive treatment plan involving both pharmacologic and non-pharmacologic options. Consider referral to providers who can help patients learn breathing or relaxation techniques, or begin CBT.

•    Initiate drug therapy with a low dose, titrate slowly, and set appropriate expectations for drug therapy. Inform patients about the slow onset of action of SSRI/SNRIs so they will not discontinue the medication out of frustration or belief that it is not working. Provide a clear review of potential side effects and the transient nature of some of the most common ones. When rapid anxiolysis or coverage for SSRI/SNRIs-related activation is a concern, consider prescribing an SSRI/SNRI plus a benzodiazepine.

•    Provide timely follow-up visits during the first 30 days after a new medication is initiated. This is the period of greatest risk for discontinuing pharmacotherapy, especially for newly diagnosed and treated patients. Ask if prescribed refills have been picked up. Re-emphasize the duration of treatment and the importance of adhering to the recommended dosage schedule. Encourage compliance by connecting medication use to problems patients have identified and are motivated to solve and by teaching patients to link medication use to daily activities (eg, with coffee in morning, at bedtime). Try to uncover barriers to adherence (eg, cost, unexpected side effects, perceived lack of need).   

•    Probe for evidence of symptom breakthrough. Develop a written patient “action plan” for breakthrough symptoms, whether it involves recommendations for PRN medication use, relaxation/cognitive strategies, or a combination of the two.

Ongoing or breakthrough symptoms are likely for the majority of patients with panic disorder, and with anxiety disorders more generally. As a result, clinicians need to provide patients with strategies to maintain function and quality of life in work and social situations when symptoms reoccur.  As high rates of medication discontinuation are common soon after the start of treatment, combination use of a benzodiazepine and SSRIs/SNRIs may improve treatment success by speeding the onset of antipanic action.  PRN benzodiazepine use may help patients cope with recurring anxiety or panic in daily work or social situations. PP


1. Olfson M, Shea S, Feder A, et al. Prevalence of anxiety, depression, and substance use disorders in an urban general medicine practice. Arch Fam Med. 2000;9(9):876-883.

2. Eaton WW, Kessler RC, Wittchen HU, Magee WJ. Panic and panic disorder in the United States. Am J Psychiatry. 1994;151(3):413-420.

3. Wang PS, Berglund P, Kessler RC. Recent care of common mental disorders in the United States: prevalence and conformance with evidence-based recommendations. J Gen Intern Med. 2000;15(5):284-292.

4. American Psychiatric Association Practice Guidelines for the Treatment of Psychiatric Disorders. Compendium 2004. Arlington, VA: American Psychiatric Association; 2004.

5. Kessler RC, Greenberg PE, Mickelson KD, Meneades LM, Wang PS. The effects of chronic medical conditions on work loss and work cutback. J Occup Environ Med. 2001;43(3):218-225.

6. Diagnostic and Statistical Manual of Mental Disorders. 4th ed text rev. Washington, DC: American Psychiatric Association; 2000.

7. Katerndahl DA, Realini JP. Where do panic attack sufferers seek care? J Fam Pract. 1995;40(3):237-243.

8. Barsky AJ, Delamater BA, Orav JE. Panic disorder patients and their medical care. Psychosomatics. 1999;40(1):50-56.

9. Pollack MH, Smoller JW, Lee DK. The anxious patient. In: Stern TA, Herman JB, Slavin PL, eds. Massachusetts General Hospital Guide to Primary Care Psychiatry. 2nd ed. New York, NY: McGraw-Hill; 2003:137-152.

10. Kessler RC, Nelson CB, McGonagle KA, Liu J, Swartz M, Blazer DG. Comorbidity of DSM-III-R major depressive disorder in the general population: results from the US National Comorbidity Survey. Br J Psychiatry Suppl. 1996;(30):17-30.

11. Cowley DS, Ha EH, Roy-Byrne PP. Determinants of pharmacologic treatment failure in panic disorder. J Clin Psychiatry. 1997;58(12):555-561.

12. Roy-Byrne PP, Cowley DS. Course and outcome in panic disorder: a review of recent follow-up studies. Anxiety. 1994;1(4):151-160.

13. Keller MB, Yonkers KA, Warshaw MG, et al. Remission and relapse in subjects with panic disorder and panic with agoraphobia: a prospective short-interval naturalistic follow-up. J Nerv Ment Dis. 1994;182(5):290-296.

14. Yonkers KA, Warshaw MG, Massion AO, Keller MB. Phenomenology and course of generalised anxiety disorder. Br J Psychiatry. 1996;168(3):308-313.

15. Yonkers KA, Bruce SE, Dyck IR, Keller MB. Chronicity, relapse, and illness—course of panic disorder, social phobia, and generalized anxiety disorder: findings in men and women from 8 years of follow-up. Depress Anxiety. 2003;17(3):173-179.

16. Yonkers KA, Ellison JM, Shera DM, et al. Pharmacotherapy observed in a large prospective longitudinal study on anxiety disorders. Psychopharmacol Bull. 1992;28(2):131-137.

17. Yonkers KA, Ellison JM, Shera DM, et al. Description of antipanic therapy in a prospective longitudinal study.  J Clin Psychopharmacol. 1996;16(3):223-232.

18. Yonkers KA, Zlotnick C, Allsworth J, Warshaw M, Shea T, Keller MB. Is the course of panic disorder the same in women and men? Am J Psychiatry. 1998;155(5):596-602.

19. Yonkers KA, Dyck IR, Warshaw M, Keller MB. Factors predicting the clinical course of generalised anxiety disorder. Br J Psychiatry. 2000;176:544-549.

20. Yonkers KA, Dyck IR, Keller MB. An eight-year longitudinal comparison of clinical course and characteristics of social phobia among men and women. Psychiatr Serv. 2001;52(5):637-643.

21. Diagnostic and Statistical Manual of Mental Disorders. 4th ed. Washington, DC: American Psychiatric Association; 1994.

22. Katschnig H, Amering M, Stolk JM, Ballenger JC. Predictors of quality of life in a long-term followup study in panic disorder patients after a clinical drug trial. Psychopharmacol Bull. 1996;32(1):149-155.

23. Sherbourne CD, Wells KB, Judd LL. Functioning and well-being of patients with panic disorder. Am J Psychiatry. 1996;153(2):213-218.

24. Tunis SR, Stryer DB, Clancy CM. Practical clinical trials: increasing the value of clinical research for decision making in clinical and health policy. JAMA. 2003;290(12):1624-1632.

25. Moses H 3rd, Dorsey ER, Matheson DH, Thier SO. Financial anatomy of biomedical research. JAMA. 2005;294(11):1333-1342.

26. Vanelli M, Coca M. The role of patient experience in antidepressant adherence. Poster presented at: 158th Annual Meeting of the American Psychiatric Association; May 21-26 2005; Atlanta, Georgia.

27. Simon GE, Von Korff M, Rutter CM, Peterson DA. Treatment process and outcomes for managed care patients receiving new antidepressant prescriptions from psychiatrists and primary care physicians. Arch Gen Psychiatry. 2001;58(4):395-401.

28. Bull SA, Hu XH, Hunkeler EM, et al. Discontinuation of use and switching of antidepressants: influence of patient-physician communication. JAMA. 2002;288(11):1403-1409.

29. Goldman DP, Joyce GF, Escarce JJ, et al. Pharmacy benefits and the use of drugs by the chronically ill. JAMA. 2004;291(19):2344-2350.

30.  Shader, RI. Manual of Psychiatric Therapeutics. 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2003.

31. Rickels K, Schweizer E. Panic disorder: long-term pharmacotherapy and discontinuation. J Clin Psychopharmacol. 1998;18(6 suppl 2):12S-18S.

32. Stahl SM. Essential Psychopharmacology:  Neuroscientific Basis and Practical Applications. 2nd ed. New York, NY: Cambridge University Press; 2000.

33. Goddard AW, Brouette T, Almai A, Jetty P, Woods SW, Charney D. Early coadministration of clonazepam with sertraline for panic disorder. Arch Gen Psychiatry. 2001;58(7):681-686.

34. Smith WT, Londborg PD, Glaudin V, Painter JR. Short-term augmentation of fluoxetine with clonazepam in the treatment of depression: a double-blind study. Am J Psychiatry. 1998;155(10):1339-1345.

35. Bruce SE, Vasile RG, Goisman RM, et al. Are benzodiazepines still the medication of choice for patients with panic disorder with or without agoraphobia? Am J Psychiatry. 2003;160(8):1432-1438.

36. Fitzgerald M. Utilization of non-traditional dosage forms in treating anxiety—the patient perspective. Innovative drug delivery systems in the management of anxiety. Presentation at: 25th Annual Meeting of the Anxiety Disorders Association of America; March 19, 2005; Seattle, WA.

37. Romach MK, Somer GR, Sobell LC, Sobell MB, Kaplan HL, Sellers EM. Characteristics of long-term alprazolam users in the community. J Clin Psychopharmacol. 1992;12(5):316-321.

38. Soumerai SB, Simoni-Wastila L, Singer C, et al. Lack of relationship between long-term use of benzodiazepines and escalation to high dosages. Psychiatr Serv. 2003;54(7):1006-1011.

39. Sellers EM, Ciraulo DA, DuPont RL, et al. Alprazolam and benzodiazepine dependence. J Clin Psychiatry. 1993;(54 suppl):64-75.

40. Posternak MA, Mueller TI. Assessing the risks and benefits of benzodiazepines for anxiety disorders in patients with a history of substance abuse or dependence. Am J Addict. 2001;10(1):48-68.