PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of bmjclinevidLink to Publisher's site
 
BMJ Clin Evid. 2008; 2008: 1010.
Published online 2008 December 16.
PMCID: PMC2907935

Panic disorder

Shailesh Kumar# and Darren Malone, Consultant Psychiatrist for Older People, Mental Health Services for Older People#

Abstract

Introduction

Panic disorder occurs in up to 3% of the adult population at some time, and is associated with other psychiatric and personality disorders, and with drug and alcohol abuse. The risk of suicide and attempted suicide has been found to be higher in people with panic disorder than in people with other psychiatric illness, including depression.

Methods and outcomes

We conducted a systematic review and aimed to answer the following clinical questions: What are the effects of non-drug treatments for panic disorder? What are the effects of drug treatments for panic disorder? What are the effects of combined drug and psychological treatments for panic disorder? We searched: Medline, Embase, The Cochrane Library, and other important databases up to June 2007 (Clinical Evidence reviews are updated periodically; please check our website for the most up-to-date version of this review). We included harms alerts from relevant organisations such as the US Food and Drug Administration (FDA) and the UK Medicines and Healthcare products Regulatory Agency (MHRA).

Results

We found 36 systematic reviews, RCTs, or observational studies that met our inclusion criteria. We performed a GRADE evaluation of the quality of evidence for interventions.

Conclusions

In this systematic review, we present information relating to the effectiveness and safety of the following interventions: applied relaxation; benzodiazepines; breathing retraining; brief dynamic psychotherapy; buspirone; client-centred therapy; cognitive behavioural therapy (CBT) (alone or plus drug treatments); cognitive restructuring; couple therapy; exposure (external or interoceptive); insight-orientated therapy; monoamine oxidase inhibitors (MAOIs); psychoeducation; selective serotonin reuptake inhibitors (SSRIs); self-help; and tricyclic antidepressants (imipramine).

Key Points

Panic disorder is characterised by recurrent, unpredictable panic attacks, making people worry about or change their behaviour to avert subsequent panic attacks or their consequences.

  • Panic disorder occurs in up to 3% of the adult population at some time, and is associated with other psychiatric and personality disorders, and with drug and alcohol abuse.
  • The risk of suicide and attempted suicide has been found to be higher in people with panic disorder than in people with other psychiatric illness, including depression.

CBT is effective in reducing symptoms of panic disorder over 6 months or longer, but we don't know whether it is more effective than other psychological treatments.

Other forms of psychotherapy can also be beneficial in reducing symptoms associated with panic disorder, with or without drug treatments.

SSRIs and tricyclic antidepressants are also effective at reducing the symptoms of panic disorder.

  • Benzodiazepines can be effective in reducing symptoms in panic disorder, but their adverse-effect profile makes them unsuitable for long-term treatment.
  • We don't know whether buspirone or MAOIs are effective.

About this condition

Definition

A panic attack is a period in which there is sudden onset of intense apprehension, fear, or terror, often associated with feelings of impending doom. Panic disorder is classified by the DSM-IV as recurrent, unpredictable panic attacks followed by at least 1 month of persistent concern about having another panic attack, worry about the possible implications or consequences of the panic attacks, or a significant behavioural change related to the attacks. The term “panic disorder” excludes panic attacks attributable to the direct physiological effects of a general medical condition, a substance, or another mental disorder. The ICD-10 classifies panic disorder as recurrent, unpredictable panic attacks, with sudden onset of palpitations, chest pain, choking sensations, dizziness, and feelings of unreality, often with associated fear of dying, losing control, or going mad, but without the requirement for the symptoms to have persisted for 1 month or longer. The DSM-IV classifies these conditions as primarily panic disorder with or without agoraphobia, whereas the ICD-10 classifies them as primarily agoraphobia with or without panic disorder. The diagnosis should not be made in people with co-morbid depression, when the panic is considered to be secondary to depression. Diagnosis: Although panic attacks are a necessary feature of panic disorder, panic attacks on their own are not enough to make the diagnosis. Panic attacks may happen in the context of specific situations such as social or specific phobia which are different from panic disorder. A diagnosis of panic disorder is made in the presence of recurrent unexpected panic attacks followed by at least 1 month of persistent concern about having another panic attack.

Incidence/ Prevalence

Panic disorder often starts at about 20 years of age (between late adolescence and the mid-30s). Lifetime prevalence is 1% to 3%, and panic disorder is more common in women than in men. An Australian community study found 1-month prevalence rates for panic disorder (with or without agoraphobia) of 0.4% using ICD-10 diagnostic criteria, and of 0.5% using DSM-IV diagnostic criteria. One systematic review of observational data estimated the prevalence rate of panic disorder during the perinatal period at between 1.3% to 2.0%, and that, although the symptoms of panic during pregnancy may be identical to at other periods, they are often interpreted in the context of the perinatal state (e.g., a woman may interpret panic attacks during pregnancy as an indication that something is wrong with the pregnancy).

Aetiology/ Risk factors

The onset of panic disorder tends to be preceded by stressful life events, although a negative interpretation of these events, in addition to their occurrence, has been suggested as an important causal factor. Panic disorder is associated with major depression, social phobia, generalised anxiety disorder, obsessive compulsive disorder, and a substantial risk of drug and alcohol misuse. It is also associated with avoidant, histrionic, and dependent personality disorders.

Prognosis

The severity of symptoms in people with panic disorder fluctuates considerably, and people commonly have periods of no attacks, or only mild attacks with few symptoms. There is often a long delay between the initial onset of symptoms and presentation for treatment. Recurrent attacks may continue for several years, especially if associated with agoraphobia. Reduced social or occupational functioning varies among people with panic disorder, and is worse in people with associated agoraphobia. Panic disorder is also associated with an increased rate of attempted suicide, with one study finding that it occurred in 20% of people with panic disorder, compared with 12% of people with panic attacks alone, 6% of those with other psychiatric disorder, and 1% of those with no disorders. The odds ratio for attempted suicide was increased if there were co-morbid conditions. One study analysing data from RCTs and systematic reviews found that co-existence of anxiety and depressive features adversely affected treatment response at 12 years compared with treatment of panic disorder alone.

Aims of intervention

To reduce the severity and frequency of panic attacks, phobic avoidance, and anticipatory anxiety; to improve social and occupational functioning, with minimal adverse effects of treatment.

Outcomes

Symptom severity measures of panic attacks, agoraphobia, and associated disability (self-reported and clinician-rated, before and after treatment, and longer term) using general or specific scales for panic disorder (e.g., the Panic and Agoraphobia Scale, the Mobility Inventory for Agoraphobia), relapse rates; quality of life; and adverse effects.

Methods

Clinical Evidence search and appraisal June 2007. The following databases were used to identify studies for this systematic review: Medline 1966 to June 2007, Embase 1980 to June 2007, and The Cochrane Database of Systematic Reviews and Cochrane Central Register of Controlled Clinical Trials 2007, Issue 2. Additional searches were carried out using these websites: NHS Centre for Reviews and Dissemination (CRD) — for Database of Abstracts of Reviews of Effects (DARE) and Health Technology Assessment (HTA), Turning Research into Practice (TRIP), and NICE. We also searched for retractions of studies included in the review. Abstracts of the studies retrieved from the initial search were assessed by an information specialist. Selected studies were then sent to the author for additional assessment, using pre-determined criteria to identify relevant studies. Study design criteria for assessment in this review were: published systematic reviews and RCTs in any language, at least single blinded, and containing more than 20 people of whom more than 80% were followed up for a minimum of 6 months. We excluded all studies described as “open”, “open label”, or not blinded unless blinding was not possible. In addition, we use a regular surveillance protocol to capture harms alerts from organisations such as the FDA and the UK Medicines and Healthcare products Regulatory Agency (MHRA), which are added to the reviews as required. This review includes RCTs in people aged 18 to 65 years old, and excludes RCTs in people solely with the co-morbidity of head injury and organic brain disorder, or people solely with phobic avoidance of social phobia (i.e., social phobia without panic disorder). To aid readability of the numerical data in our reviews, we round many percentages to the nearest whole number. Readers should be aware of this when relating percentages to summary statistics such as RRs and ORs. We have performed a GRADE evaluation of the quality of evidence for interventions included in this review (see table). The categorisation of the quality of the evidence (high, moderate, low, or very low) reflects the quality of evidence available for our chosen outcomes in our defined populations of interest. These categorisations are not necessarily a reflection of the overall methodological quality of any individual study, because the Clinical Evidence population and outcome of choice may represent only a small subset of the total outcomes reported, and population included, in any individual trial. For further details of how we perform the GRADE evaluation and the scoring system we use, please see our website (www.clinicalevidence.com).

Table
GRADE Evaluation of interventions for Panic disorder.

Glossary

Applied relaxation
A technique involving training in relaxation techniques and self-monitoring of symptoms without challenging beliefs.
Beck Depression Inventory
Standardised scale to assess depression. This instrument consists of 21 items to assess the intensity of depression. Each item is a list of 4 statements (rated 0, 1, 2, or 3), arranged in increasing severity, about a particular symptom of depression. The range of scores possible are 0 = least severe depression to 63 = most severe depression. It is recommended for people aged 13 to 80 years. Scores of more than 12 or 13 indicate the presence of depression.
Behavioural therapy
is a type of psychotherapy which is aimed at modifying the behaviour causing distress or impairment, by focusing on the environment and context in which the behaviour occurs. Unlike other psychotherapies, behavioural therapy doesn’t endorse mind–body split or the impact of past developmental issues, and treats the person as a unit.
Client-centred therapy
A system of psychotherapy based on the view that the client has the internal resources to improve and is in the best position to resolve his or her own personality dysfunction. It has roots in psychoanalysis and sees clients as taking a more active role in their treatment.
Clinical Global Impression Scale
A one-item, observer-rated scale for measuring the severity of a condition. It has been investigated for validity and reliability. The scale is scored from 0 (not ill at all) to 7 (severely ill).
Cognitive behavioural therapy (CBT)
Brief structured treatment using relaxation and exposure procedures, and aimed at changing dysfunctional beliefs and negative automatic thoughts (typically 20 sessions over 12–16 weeks).
Cognitive restructuring
An intervention that involves asking questions to help people challenge the stereotyped and repetitive thoughts and images that enhance fear.
Couple therapy
An intervention that involves using significant relationships to help change previous persistent and inflexible patterns of behaviour.
Exposure
A type of behavioural therapy involving deliberate exposure to previously avoided situation or feared stimuli (including thoughts). It can be done by either asking the person to imagine being in the previously avoided situation, especially when direct exposure is impractical or difficult (such as when a person has a fear of flying) — termed in vitro, interoceptive, or imaginal exposure. Alternatively, exposure can be in vivo or exteroceptive, in which exposure is to real life situations or stimuli.
Hamilton Anxiety Rating Scale
A 14-item observer-rated scale for measuring the severity of anxiety. It has been investigated for validity and reliability. Each item is rated on a 5-point scale from 0 (no symptoms) to 4 (severe or grossly disabling symptoms). Total score ranges from 0 to 56, with 14 or higher indicating clinically significant anxiety.
Hamilton Depression Rating Scale
a measure of depressive symptoms using 17 items, with total scores from 0 to 54 (higher scores indicate increased severity of depression).
Low-quality evidence
Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Psychoeducation
An intervention aimed at educating the person with psychiatric disorder in subject areas that serve the goals of treatment and rehabilitation.
Very low-quality evidence
Any estimate of effect is very uncertain.

Notes

Disclaimer

The information contained in this publication is intended for medical professionals. Categories presented in Clinical Evidence indicate a judgement about the strength of the evidence available to our contributors prior to publication and the relevant importance of benefit and harms. We rely on our contributors to confirm the accuracy of the information presented and to adhere to describe accepted practices. Readers should be aware that professionals in the field may have different opinions. Because of this and regular advances in medical research we strongly recommend that readers' independently verify specified treatments and drugs including manufacturers' guidance. Also, the categories do not indicate whether a particular treatment is generally appropriate or whether it is suitable for a particular individual. Ultimately it is the readers' responsibility to make their own professional judgements, so to appropriately advise and treat their patients. To the fullest extent permitted by law, BMJ Publishing Group Limited and its editors are not responsible for any losses, injury or damage caused to any person or property (including under contract, by negligence, products liability or otherwise) whether they be direct or indirect, special, incidental or consequential, resulting from the application of the information in this publication.

Contributor Information

Shailesh Kumar, Division of Psychiatry, Auckland Medical School, Auckland, New Zealand.

Darren Malone, Rotorua Hospital, Rotorua, New Zealand.

References

1. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 4th ed. Washington, DC: American Psychiatric Association, 1994.
2. World Health Organization. The ICD-10 classification of mental and behavioural disorders. Geneva: World Health Organization, 1992.
3. Robins LN, Regier DA, eds. Psychiatric disorders in America: the epidemiologic catchment area study. New York, NY: Free Press, 1991.
4. Weissman MM, Bland MB, Canino GJ, et al. The cross-national epidemiology of panic disorder. Arch Gen Psychiatry 1997;54:305–309. [PubMed]
5. Andrews G, Henderson S, Hall W. Prevalence, comorbidity, disability and service utilisation. Overview of the Australian National Mental Health Survey. Br J Psychiatry 2001;178:145–153. [PubMed]
6. Ross LE, McLean LM. Anxiety disorders during pregnancy and the postpartum period: A systematic review. J Clin Psychiatry 2006;67:1285–1298 [PubMed]
7. Last CG, Barlow DH, O'Brien GT. Precipitants of agoraphobia: role of stressful life events. Psychol Rep 1984;54:567–570. [PubMed]
8. De Loof C, Zandbergen H, Lousberg T, et al. The role of life events in the onset of panic disorder. Behav Res Ther 1989;27:461–463. [PubMed]
9. Rapee RM, Mattick RP, Murrell E. Impact of life events on subjects with panic disorder and on comparison subjects. Am J Psychiatry 1990;147:640–644. [PubMed]
10. Hirschfield RMA. Panic disorder: diagnosis, epidemiology and clinical course. J Clin Psychiatry 1996;57:3–8. [PubMed]
11. Andrews G, Creamer M, Crino R, et al. The treatment of anxiety disorders. Cambridge: Cambridge University Press, 1994.
12. Page AC, Andrews G. Do specific anxiety disorders show specific drug problems? Aust N Z J Psychiatry 1996;30:410–414. [PubMed]
13. Gorman JM, Coplan JD. Comorbidity of depression and panic disorder. J Clin Psychiatry 1996;57:34–41. [PubMed]
14. Tyrer P, Seivewright H, Simmonds S, et al. Prospective studies of cothymia (mixed anxiety-depression): how do they inform clinical practice? Eur Arch Psychiatry Clin Neurosci 2001;251:1153–1156. [PubMed]
15. Oei TPS, Llamas M, Devilly GJ. The efficacy and cognitive processes of cognitive behaviour therapy in the treatment of panic disorder with agoraphobia. Behav Cogn Psychotherapy 1999;27:63–88.
16. Gould RA, Otto MW, Pollack MH. A meta-analysis of treatment outcome for panic disorder. Clin Psychol Rev 1995;15:819–844.
17. Butler AC, Chapman JE, Forman EM, et al. The empirical status of cognitive-behavioral therapy: A review of meta-analyses. Clin Psychol Rev 2006;26:17–31. Search date 2004. [PubMed]
18. Mitte K. A meta-analysis of the efficacy of psycho- and pharmacotherapy in panic disorder with and without agoraphobia. J Affect Dis 2005;88:27–45. Search date 2002. [PubMed]
19. von Knorring L, Thelander S, Pettersson, A. Treatment of anxiety syndrome. A systematic literature review. Summary and conclusions by the SBU. Lakartidningen 2005;102:3561–3562, 3565–3566, 3569. [PubMed]
20. Kenardy JA, Dow MG, Johnston DW, et al. A comparison of delivery methods of cognitive-behavioural therapy for panic disorder: an international multicenter trial. J Consult Clin Psychol 2003;71:1068–1075. [PubMed]
21. Sharp DM, Power KG, Swanson V. A comparison of the efficacy and acceptability of group versus individual cognitive behaviour therapy in the treatment of panic disorder and agoraphobia in primary care. Clin Psychol Psychother 2004;11:73–82.
22. Telch MJ, Lucas AJ, Schmidt NB, et al. Group cognitive-behavioural treatment of panic disorder. Behav Res Ther 1993;31:279–287. [PubMed]
23. Lidren DM, Watkins PL, Gould RA, et al. A comparison of bibliotherapy and group therapy in the treatment of panic disorder. J Consult Clin Psychol 1994;62:865–869. [PubMed]
24. Schmidt NB, Woolaway-Bickel K, Trakowski J, et al. Dismantling cognitive-behavioral treatment for panic disorder: questioning the utility of breathing retraining. J Consult Clin Psychol 2000;68:417–424. [PubMed]
25. Bowen RC, D'Arcy C, Keegan D, et al. A controlled trial of cognitive behavioral treatment of panic in alcoholic inpatients with comorbid panic disorder. Addict Behav 2000;25:593–597. [PubMed]
26. Ost LG, Thulin U, Ramnero J. Cognitive behavior therapy vs exposure in vivo in the treatment of panic disorder with agrophobia. Behav Res Ther 2004;42:1105–1127. [PubMed]
27. Carlbring P, Nilsson IhrfeltE, Waara J, et al. Treatment of panic disorder: live therapy vs. self-help via the Internet. Behav Res Ther 2005;43:1321–1333. [PubMed]
28. Beck JG, Stanley MA, Baldwin LE, et al. Comparison of cognitive therapy and relaxation training for panic disorder. J Consult Clin Psychol 1994;62:818–826. [PubMed]
29. Clark DM, Salkovskis PM, Hackmann A, et al. A comparison of cognitive therapy, applied relaxation and imipramine in the treatment of panic disorder. Br J Psychiatry 1994;164:759–769. [PubMed]
30. Arntz A, van den Hout M. Psychological treatments of panic disorder without agoraphobia: cognitive therapy versus applied relaxation. Behav Res Ther 1996;34:113–121. [PubMed]
31. Ost LG, Westling BE, Hellstrom K. Applied relaxation, exposure in vivo and cognitive methods in the treatment of panic disorder with agoraphobia. Behav Res Ther 1993;31:383–394. [PubMed]
32. Ost LG, Westling BE. Applied relaxation vs cognitive behavior therapy in the treatment of panic disorder. Behav Res Ther 1995;33:145–158. [PubMed]
33. Milrod B, Leon AC, Busch F, et al. A randomized controlled clinical trial of psychoanalytic psychotherapy for panic disorder. [erratum appears in Am J Psychiatry 2007;164:529. Am J Psychiatry 2007;164:265–272. [PubMed]
34. Teusch L, Bohme H, Gastpar M. The benefit of an insight-oriented and experiential approach on panic and agoraphobia symptoms. Results of a controlled comparison of client-centered therapy alone and in combination with behavioral exposure. Psychother Psychosom 1997;66:293–301. [PubMed]
35. Teusch L, Bohme H, Finke J. Conflict-centered individual therapy or integration of psychotherapy methods. Process of change in client-centered psychotherapy with and without behavioral exposure therapy in agoraphobia with panic disorder. Nervenarzt 2001;72:31–39. [In German] [PubMed]
36. Bouchard S, Gauthier J, Laberge B, et al. Exposure versus cognitive restructuring in the treatment of panic disorder with agoraphobia. Behav Res Ther 1996;34:213–224. [PubMed]
37. Cox BJ, Endler NS, Lee PS, et al. A meta-analysis of treatments for panic disorder with agoraphobia: imipramine, alprazolam, and in vivo exposure. J Behav Ther Exp Psychiatry 1992;23:175–182. [PubMed]
38. Ito LM, de Araujo LA, Tess VL, et al. Self-exposure therapy for panic disorder with agoraphobia: randomised controlled study of external v. interoceptive self-exposure. Br J Psychiatry 2001;178:331–336. [PubMed]
39. Marks IM, Swinson RP, Basoglu M, et al. Alprazolam and exposure alone and combined in panic disorder with agoraphobia. A controlled study in London and Toronto. Br J Psychiatry 1993;162:776–787. [PubMed]
40. Barlow JH, Ellard DR, Hainsworth JM, et al. A review of self-management interventions for panic disorders, phobias and obsessive-compulsive disorders. Acta Psychiatr Scand 2005;111:272–285. Search date 2003. [PubMed]
41. Hecker JE, Losee MC, Roberson-Nay R, et al. Mastery of your anxiety and panic and brief therapist contact in the treatment of panic disorder. J Anxiety Disord 2004;18:111–126. [PubMed]
42. Meuret AE, Wilhelm FH, Ritz T, et al. Breathing training for treating panic disorder: useful intervention or impediment? Behav Modif 2003;27:731–754. Search date not reported. [PubMed]
43. Wiborg IM, Dahl AA. Does brief dynamic psychotherapy reduce the relapse rate of panic disorder? Arch Gen Psychiatry 1996;53:689–694. [PubMed]
44. Byrne M, Carr A, Clark M. The efficacy of couples-based interventions for panic disorder with agoraphobia. J Fam Ther 2004;26:105–125. Search date 2001.
45. Oatley K, Hodgson D. Influence of husbands on the outcome of their agoraphobic wives' therapy. Br J Psychiatry 1987;150:380–386. [PubMed]
46. Arnow BA, Taylor CB, Agras WS, et al. Enhancing agoraphobia treatment outcome by changing couple communication patterns. Behav Ther 1985;16:452–467.
47. Jannoun L, Munby M, Catalan J, et al. A home-based treatment program for agoraphobia: replication and controlled evaluation. Behav Ther 1980;11:294–305.
48. Daiuto AD, Baucom DH, Epstein N, et al. The application of behavioral couples therapy to the assessment and treatment of agoraphobia: implications of empirical research. Clin Psychol Rev 1998;18:663–687. [PubMed]
49. Boyer W. Serotonin uptake inhibitors are superior to imipramine and alprazolam in alleviating panic attacks: a meta-analysis. Int Clin Psychopharmacol 1995;10:45–49. Search date not reported; primary sources Medline, Embase, Psychlit, and sponsoring agencies of two trials contacted for supplementary statistical information. [PubMed]
50. Otto M, Tuby K, Gould R, et al. An effect-size analysis of the relative efficacy and tolerability of serotonin selective reuptake inhibitors for panic disorder. Am J Psychiatry 2001;158:1989–1992. Search date not reported; primary sources Medline, Psychlit, and hand searches of references. [PubMed]
51. Lepola UM, Wade AG, Leinonen EV, et al. A controlled, prospective, 1-year trial of citalopram in the treatment of panic disorder. J Clin Psychiatry 1998;59:528–534. [PubMed]
52. Rapaport M, Wolkow R, Rubin A, et al. Sertraline treatment of panic disorder: results of a long term study. Acta Psych Scand 2001;104:289–298. [PubMed]
53. US Food and Drug Administration. Antidepressant use in children, adolescents, and adults. Available online at www.fda.gov/cder/drug/antidepressants/default.htm (last accessed 03 August 2010).
54. US Food and Drug Administration. Paroxetine hydrochloride (marketed as Paxil) information. Available online at: www.fda.gov/cder/drug/infopage/paroxetine/default.htm (last accessed 03 August 2010).
55. US Food and Drug Administration. Paxil (paroxetine hydrochloride) tablets and oral suspension. Paxil CR (paroxetine hydrochloride) controlled-release tablets. Available online at: www.fda.gov/medwatch/safety/2006/safety06.htm#paxil (last accessed 03 August 2010).
56. US Food and Drug Administration. 2005 Safety alerts for drugs, biologics, medical devices, and dietary supplements. Available at: http://www.fda.gov/medwatch/safety/2005/safety05.htm#Paxi12 (last accessed 14 November 2008).
57. Curtis GC, Massana J, Udina C, et al. Maintenance drug therapy of panic disorder. J Psychiatr Res 1993;27:127–142. [PubMed]
58. Mavissakalian MR, Perel JM. Long-term maintenance and discontinuation of imipramine therapy in panic disorder with agoraphobia. Arch Gen Psychiatry 1999;56:821–827. [PubMed]
59. Bakker A, van Balkom AJLM, Spinhoven P. SSRIs vs TCAs in the treatment of panic disorder: a meta-analysis. Acta Psychiatr Scand 2002;106:163–167. Search date 1999; primary sources Medline, Embase, Psychinfo, and hand searches of reference lists of articles obtained. [PubMed]
60. Wilkinson G, Balestrieri M, Ruggeri M, et al. Meta-analysis of double-blind placebo-controlled trials of antidepressants and benzodiazepines for patients with panic disorders. Psychol Med 1991;21:991–998. Search date 1990. [PubMed]
61. Feijo De Mello M. Effectiveness of alprazolam in the treatment of panic disorder: A systematic review. Rev Bras Med 2006;63:606–610.
62. Posternak M, Mueller T. Assessing the risks and benefits of benzodiazepines for anxiety disorders in patients with a history of substance abuse or dependence. Am J Addict 2001;10:48–68. [PubMed]
63. Cottraux J, Note ID, Cungi C, et al. A controlled study of cognitive behaviour therapy with buspirone or placebo in panic disorder with agoraphobia. Br J Psychiatry 1995;167:635–641. [PubMed]
64. Bouvard M, Mollard E, Guerin J, et al. Study and course of the psychological profile in 77 patients expressing panic disorder with agoraphobia after cognitive behaviour therapy with or without buspirone. Psychother Psychosom 1997;66:27–32. [PubMed]
65. Tiller JW, Bouwer C, Behnke K. Moclobemide for anxiety disorders: a focus on moclobemide for panic disorder. Int Clin Psychopharmacol 1997;12:S27–S30. [PubMed]
66. Furukawa TA, Watanabe N, Churchill R. Combined psychotherapy plus antidepressants for panic disorder with or without agoraphobia. In: The Cochrane Library Issue 2, 2007.Chichester, UK: John Wiley & Sons, Ltd. Search date 2005. [PubMed]
67. Roy-Byrne PP, Craske MG, Stein MB, et al. A randomized effectiveness trial of cognitive-behavioral therapy and medication for primary care panic disorder. Arch Gen Psychiatry 2005;62:290–298. [PMC free article] [PubMed]
68. Katon W, Russo J, Sherbourne C, et al. Incremental cost-effectiveness of a collaborative care intervention for panic disorder. Psychol Med 2006;36:353–363. [PubMed]
69. Roy-Byrne P, Stein MB, Russo J, et al. Medical illness and response to treatment in primary care panic disorder. Gen Hosp Psychiatry 2005;27:237–243. [PubMed]
2008; 2008: 1010.
Published online 2008 December 16.

CBT versus no treatment

Summary

CBT is effective in reducing symptoms of panic disorder over 6 months or longer, but we don't know whether it is more effective than other psychological treatments.

CBT is more effective than waiting list and other controls in reducing symptoms in panic disorder with or without mild to moderate agoraphobia.

Benefits and harms

CBT versus placebo or no treatment:

We found five systematic reviews. The first and second systematic reviews were included in the third and fourth systematic reviews so are not reported further. Two other reviews performed different meta-analyses so are reported below. A fifth review included additional RCTs but did not perform a meta-analysis so results from those RCTs meeting Clinical Evidence inclusion criteria are reported below. For full details of review methods and inclusion criteria, see further information about studies.

Symptom severity

CBT or behavioural therapy compared with placebo or no treatment CBT or behavioural therapy may be more effective at improving anxiety and clinical significance scores, but we don't know whether they are more effective at improving depression scores. We don't know whether CBT with or without exposure to the panic-inducing stimulus is more effective than placebo or waiting list control at improving symptoms in people with panic disorder and severe agoraphobia (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

Systematic review
People with panic disorder with and without agoraphobia Treatment effect
with cognitive behavioural treatments in general
with control (not specified)
Absolute results not reported

Effect size 0.68 (see further information on studies for details on effect size)
Significance not reported

RCT
4-armed trial
186 people with panic disorder
In review
Proportion of people who were panic free at follow-up 6 months
66% with CBT (12 sessions)
65% with brief CBT (6 sessions)
63% with brief CBT plus computer programme
9% with waiting list
Absolute numbers not reported

P >0.05 (among group comparison)

RCT
3-armed trial
97 people with panic disorder with or without agoraphobia in primary care
In review
Proportion of follow-up attenders who completed follow-up without receiving intervening treatment 6 months
20/37 (54%) with individual CBT
12/38 (31%) with group CBT
0/22 (0%) with waiting list

Significance not assessed

RCT
67 people with panic disorder with or without agoraphobia
In review
Proportion of people who were panic free at follow-up 6 months
83% with group CBT (including education, breathing retraining plus interoceptive exposure)
30% with waiting list
Absolute numbers not reported

P value not reported

RCT
3-armed trial
36 people with panic disorder
In review
Proportion of people with 'clinical improvement' in frequency of panic attacks (criteria not reported) end of treatment
83% with group CBT
25% with waiting list
Absolute numbers not reported

P <0.01 (group CBT v waiting list control)
Effect size not calculatedCBT

RCT
3-armed trial
45 people with DSM-IV-diagnosed panic disorder with or without agoraphobia
In review
Proportion of people with high end-state function (defined as panic frequency = 0, anxiety on Sheehan Patient-Rated Anxiety Scale <30, and phobic avoidance on the Mobility Inventory Scale <1.5) end of treatment
38% with CBT (12 sessions over 12 weeks)
0% with control (delayed treatment)
Absolute numbers not reported

P <0.0001 (CBT v control)
Effect size not calculatedCBT
Anxiety

Systematic review
People with panic disorder with or without agoraphobia Anxiety
with CBT or behavioural therapy
with no treatment
Absolute results not reported

Effect size +0.87
95% CI +0.71 to +1.03
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Effect size not calculatedCBT

Systematic review
People with panic disorder with or without agoraphobia Anxiety
with CBT or behavioural therapy
with pill placebo
Absolute results not reported

Effect size +0.51
95% CI +0.30 to +0.72
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Effect size not calculatedCBT

RCT
231 alcoholic inpatients with panic disorder
In review
Anxiety 12 months
with Group CBT plus standard alcohol treatment programme (4 weeks)
with standard alcohol treatment programme alone
Absolute results reported graphically

Reported as not significant
P value not reported
Outcome improved from baseline in both groups
Not significant
Depression/mood

Systematic review
People with panic disorder with or without agoraphobia Depression
with CBT or behavioural therapy
with no treatment
Absolute results not reported

Effect size +0.72
95% CI +0.54 to +0.90
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Effect size not calculatedCBT

Systematic review
People with panic disorder with or without agoraphobia Depression
with CBT or behavioural therapy
with pill placebo
Absolute results not reported

Effect size +0.27
95% CI –0.02 to +0.56
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Not significant

RCT
231 alcoholic inpatients with panic disorder
In review
Mood 12 months
with Group CBT plus standard alcohol treatment programme (4 weeks)
with standard alcohol treatment programme alone
Absolute results reported graphically

Reported as not significant
P value not reported
Outcome improved from baseline in both groups
Not significant
'Clinically significant improvement'

Systematic review
People with panic disorder with or without agoraphobia Clinically significant improvement (not further defined)
with CBT or behavioural therapy
with no treatment
Absolute results not reported

Effect size +1.36
95% CI +1.10 to +1.62
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Effect size not calculatedCBT

Systematic review
People with panic disorder with or without agoraphobia Clinically significant improvement (not further defined)
with CBT or behavioural therapy
with pill placebo
Absolute results not reported

Effect size +0.58
95% CI +0.25 to +0.92
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Effect size not calculatedCBT

Quality of life

CBT or behavioural therapy compared with placebo or no treatment We don't know whether CBT or behavioural therapy are more effective at improving quality-of-life scores (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Quality of life

Systematic review
People with panic disorder with or without agoraphobia Quality of life
with CBT or behavioural therapy
with no treatment
Absolute results not reported

Effect size +0.85
95% CI +0.48 to +1.21
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Effect size not calculatedCBT

Systematic review
People with panic disorder with or without agoraphobia Clinically significant improvement (not further defined)
with CBT or behavioural therapy
with pill placebo
Absolute results not reported

Effect size +0.42
95% CI –0.11 to +0.94
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Not significant

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Further information on studies

The review (search date 2004) included several meta-analyses of CBT for the treatment of different anxiety disorders and depressive disorder. The authors recalculated effect sizes for treatments from these meta-analyses, with an adjustment to account for different control-group response rates, to enable a comparison of different treatments where direct-comparison studies had not been performed. The review reported than an effect size of 1.0 would represent a large treatment effect (indicating that the average person in one group would have an outcome superior to that of 84% of people in the control group), while an effect size of 0.0 would indicate no treatment effect.

The review (search date 2002) identified nine meta-analyses, and included a total of 124 controlled clinical studies of CBT or behavioural therapy, pharmacotherapy, or behavioural therapy plus pharmacotherapy in people with panic disorder, agoraphobia, or panic disorder plus agoraphobia; the review identified 32 controlled clinical studies comparing CBT or behavioural therapy versus no treatment, and 13 controlled studies comparing CBT or behavioural therapy versus placebo. Inclusion criteria of the review were a minimum of four people, and a waiting list, pill placebo, or therapy placebo group: the review did not specify that studies had to be randomised. The review calculated effect sizes to determine the additional benefit from active treatment compared with control. The review found no evidence of publication bias in studies that had compared CBT versus waiting list, placebo, behavioural therapy, or combination treatment. However, these results should be interpreted with caution since response rates tend to be greater in pill placebo control groups as opposed to waiting list control groups, and because few studies of CBT used an intention-to-treat analysis.

This review (search date 2005) identified 298 controlled clinical studies of treatments for panic disorder, but did not perform a meta-analysis. The review identified 30 controlled clinical randomised and non-randomised studies of CBT compared with waiting list or placebo control, including 13 that were also identified by the third review, and five that were not included in any of the other reviews and that met our inclusion criteria. It graded evidence, and defined grade 1 evidence as: the conclusion being supported by at least two studies with a high level of proof or good systematic review; and grade 2 evidence as: the conclusion being supported by one study with a high level of proof and at least two studies with a medium level of proof. The authors of the review concluded from included studies that there was good evidence of benefit from CBT, with or without exposure to the panic-inducing stimulus, in people with panic disorder without agoraphobia or mild to moderate agoraphobia (evidence grade 1), but that the role of CBT in the treatment of panic disorder with severe agoraphobia was not established.

Comment

None.

Substantive changes

No new evidence

2008; 2008: 1010.
Published online 2008 December 16.

CBT versus drug treatments

Summary

We don't know whether CBT alone is more effective than antidepressants alone, but weak evidence suggests that the effects of CBT may last longer.

Combined treatment with CBT plus antidepressants has been shown to be more effective than CBT alone or antidepressants alone in reducing symptoms in the short term.

Benefits and harms

CBT versus antidepressants:

We found three systematic reviews. Two of the reviews performed different meta-analyses so both are reported here. The third review did not perform a meta-analysis and identified no RCTs that met Clinical Evidence inclusion criteria so is not reported further. For full details of the inclusion criteria of the reviews, see further information about studies.

Symptom severity

CBT or cognitive therapy compared with antidepressants We don't know whether CBT or cognitive therapy is more effective than pharmacotherapy (mainly SSRIs and tricyclic antidepressants but also including benzodiazepines) at improving symptoms in people with panic disorder (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

Systematic review
People with panic disorder with and without agoraphobia Treatment effect
Effect size 0.68 with cognitive behavioural treatments in general
Effect size 0.47 with drug treatments (not further defined in review)

Significance not reported
Analyses based on indirect comparisons, and so should be interpreted with extreme caution

Systematic review
People with panic disorder with and without agoraphobia "Slippage" in effect size 1 year
–0.07 with cognitive behavioural treatments in general
–0.46 with drug treatments (not further defined in review)

Significance not reported
Analyses based on indirect comparisons, and so should be interpreted with extreme caution
Anxiety

Systematic review
People with panic disorder with or without agoraphobia Anxiety
with drug treatments
with CBT or behavioural therapy
Absolute results not reported

Eeffect size +0.27
95% CI –0.07 to +0.62
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
The review found evidence of publication bias (see further information on studies for more details)
Not significant
Depression

Systematic review
People with panic disorder with or without agoraphobia Depression
with drug treatments
with CBT or behavioural therapy
Absolute results not reported

Effect size +0.21
95% CI –0.34 to +0.75
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
The review found evidence of publication bias (see further information on studies for more details)
Not significant
'Clinical significant improvement'

Systematic review
People with panic disorder with or without agoraphobia Clinically significant improvement (not further defined)
with drug treatments
with CBT or behavioural therapy
Absolute results not reported

Effect size +0.09
95% CI –0.38 to +0.56
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
The review found evidence of publication bias (see further information on studies for more details)
Not significant

Quality of life

No data from the following reference on this outcome.

Adverse effects

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Withdrawal rate

Systematic review
People with panic disorder with or without agoraphobia Average withdrawal rates
21% with drug treatments
16% with CBT or behavioural therapy
Absolute numbers not reported

Significance not assessed
The review did not report details of method of randomisation
The review found evidence of publication bias (see further information on studies for more details)

No data from the following reference on this outcome.

CBT plus buspirone versus CBT alone:

See option on buspirone.

Further information on studies

The review (search date 2004) included several meta-analyses and recalculated effect sizes for treatments from these meta-analyses, with an adjustment to account for different control-group response rates, to enable a comparison of different treatments where direct-comparison studies had not been performed. The review reported than an effect size of 1.0 would represent a large treatment effect (indicating that the average person in one group would have an outcome superior to that of 84% of people in the control group), while an effect size of 0.0 would indicate no treatment effect.

The review (search date 2002) identified 11 controlled clinical studies directly comparing cognitive therapy or CBT versus pharmacotherapy in people with panic disorder, agoraphobia, or panic disorder plus agoraphobia. Inclusion criteria of the review were a minimum of four people, and a waiting list, pill placebo, or therapy placebo group: review did not specify that studies had to be randomised. Drug classes investigated were mainly SSRIs and tricyclic antidepressants, but benzodiazepines were also included. Effect sizes were calculated to determine the additional benefit from active treatment compared with control. The review found some evidence of publication bias in studies comparing CBT versus drug treatments. The author adjusted the calculated effect sizes to account for publication bias, and found that this increased the effect size for CBT or behavioural therapy compared with drug treatment (P <0.01).

The review (search date 2005) identified 10 controlled clinical studies, of which four were included in the first review. None of the other studies met our inclusion criteria. The review did not pool data, but graded evidence based on included studies. It graded evidence, and defined grade 1 evidence as: the conclusion being supported by at least two studies with a high level of proof or good systematic review; and grade 2 evidence as: the conclusion being supported by one study with a high level of proof and at least two studies with a medium level of proof. The review concluded that the effect of psychotherapy was longer lasting compared with that of drug treatments (reported as grade 2 evidence [defined as the conclusion being supported by one study with a high level of proof and at least two studies with a medium level of proof]); review is not discussed further.

Comment

None.

Substantive changes

CBT versus drug treatments Existing evidence re-evaluated, and reporting in benefits section further enhanced with additional detail. Existing categorisation changed. 'CBT versus antidepressants (unclear which more effective, but weak evidence that effects of CBT may last longer than those of antidepressants)' categorised as Unknown effectiveness.

2008; 2008: 1010.
Published online 2008 December 16.

CBT versus other psychological treatments

Summary

Self-help using CBT techniques and therapist-based CBT may be equally effective at improving symptoms of panic disorder.

Benefits and harms

CBT versus behavioural therapy:

We found two systematic reviews. For full details of the methods and inclusion crteria of the first review, see further information about studies. The second review (search date 2005) did not perform a meta-analysis. It identified seven controlled trials, of which three were also identified by the first review. None of the other trial met our inclusion criteria, and so no data from this review are reported below.

Symptom severity

CBT compared with behavioural therapy CBT may be more effective at improving depression scores, but not anxiety scores or clinical significance scores (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Anxiety

Systematic review
People with panic disorder with or without agoraphobia Anxiety
with CBT
with behavioural therapy
Absolute results not reported

Effect size +0.09 (CBT v behavioural therapy)
95% CI –0.07 to +0.24
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Not significant
Depression

Systematic review
People with panic disorder with or without agoraphobia Depression
with CBT
with behavioural therapy
Absolute results not reported

Effect size +0.18 (CBT v behavioural therapy)
95% CI +0.01 to +0.35
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Effect size not calculatedCBT
'Clinically significant improvement'

Systematic review
People with panic disorder with or without agoraphobia Clinically significant improvement (not further defined)
with CBT
with behavioural therapy
Absolute results not reported

Effect size +0.13 (CBT v behavioural therapy)
95% CI –0.13 to +0.39
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Not significant

Quality of life

CBT compared with behavioural therapy We don't know whether CBT is more effective at improving quality of life (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Quality of life

Systematic review
People with panic disorder with or without agoraphobia Quality of life
with CBT
with behavioural therapy
Absolute results not reported

Effect size –0.11 (CBT v behavioural therapy)
95% CI –0.63 to +0.42
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Not significant

Adverse effects

No data from the following reference on this outcome.

CBT versus exposure:

We found one RCT.

Symptom severity

CBT compared with exposure We don't know whether CBT is more effective than exposure in vivo at improving symptoms (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

RCT
3-armed trial
73 people with DSM-IV diagnosis of panic with agoraphobia Behavioural assessor ratings
with CBT
with exposure in vivo
Absolute results not reported

Reported as not significant (CBT v exposure in vivo)
P value not reported
Complex analysis using analysis of variance (ANOVA)
Not significant

RCT
3-armed trial
73 people with DSM-IV diagnosis of panic with agoraphobia Behavioural approach tests
with CBT
with exposure in vivo
Absolute results not reported

Reported as not significant (CBT v exposure in vivo)
P value not reported
Complex analysis using analysis of variance (ANOVA)
Not significant

RCT
3-armed trial
73 people with DSM-IV diagnosis of panic with agoraphobia Self-monitoring of panic attacks
with CBT
with exposure in vivo
Absolute results not reported

Reported as not significant (CBT v exposure in vivo)
P value not reported
Complex analysis using analysis of variance (ANOVA)
Not significant

RCT
3-armed trial
73 people with DSM-IV diagnosis of panic with agoraphobia Self-report scales of agoraphobia and panic
with CBT
with exposure in vivo
Absolute results not reported

Reported as not significant (CBT v exposure in vivo)
P value not reported
Complex analysis using analysis of variance (ANOVA)
Not significant
'Clinically significant improvement'

RCT
3-armed trial
73 people with DSM-IV diagnosis of panic with agoraphobia Proportion of people with a clinically significant improvement (based on ratings of phobic severity, agoraphobia score, and diagnosis criteria) post-treatment
79% with CBT
67% with exposure in vivo
Absolute numbers not reported

Reported as not significant (CBT v exposure in vivo)
P value not reported
Not significant

RCT
3-armed trial
73 people with DSM-IV diagnosis of panic with agoraphobia Proportion of people with a clinically significant improvement (based on ratings of phobic severity, agoraphobia score, and diagnosis criteria) 1 year
76% with CBT
74% with exposure in vivo
Absolute numbers not reported

Reported as not significant (CBT v exposure in vivo)
P value not reported
Not significant

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

CBT versus applied relaxation:

See option on applied relaxation.

CBT versus breathing retraining:

See option on breathing retraining.

CBT versus self-help methods:

See option on self-help.

Further information on studies

The review (search date 2002) identified 26 controlled studies comparing CBT versus behavioural therapy in people with panic disorder, agoraphobia, or panic disorder plus agoraphobia, with a minimum of four people, including a waiting list, pill placebo, or therapy placebo group. Effect sizes were calculated to determine the additional benefit from active treatment compared with control. The review specified that trials had to have a control group (waiting list, pill placebo, or therapy placebo) but did not specify that they had to be randomised

For the analysis of active treatment versus waiting list control, the RCT combined results from both the CBT and exposure groups together and so we have not reported these results further.

Comment

None.

Substantive changes

CBT versus other psychological treatments One RCT (73 people) added comparing CBT, exposure in vivo, and waiting list control. It found no significant difference between CBT and exposure in vivo for a range of outcomes. One small RCT (49 people) added comparing CBT versus a self-help method supplied over the internet.The RCT found no significant difference between groups for a range of outcome measures. Categorisation unchanged (Unknown effectiveness) as it is unclear how CBT compares with other psychological treatments.

2008; 2008: 1010.
Published online 2008 December 16.

Applied relaxation

Summary

Applied relaxation is likely to be effective in reducing symptoms.

Benefits and harms

Applied relaxation versus waiting list control:

We found two systematic reviews. The reviews identified the same two RCTs comparing applied relaxation versus waiting lost control. Neither review analysed the results of these RCTs separately from those of other psychological treatments so we report the results regarding applied relaxation from the RCTs separately here.

Symptom severity

Applied relaxation compared with waiting list control Applied relaxation may be more effective at improving panic and anxiety symptoms at 10 to 12 weeks, but we don't know whether it is more effective at increasing the proportion of people who are panic free at 10 weeks (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Frequency of panic attack

RCT
3-armed trial
64 people with DSM-III criteria for panic disorder with moderate or mild agoraphobia, 9 without avoidance
In review
Number of panic attacks 10 weeks
with applied relaxation
with minimal contact control
Absolute results not reported

Reported as significant (applied relaxation v minimal contact control)
P value not reported
Effect size not calculatedapplied relaxation
Freedom from panic attacks

RCT
3-armed trial
64 people with DSM-III criteria for panic disorder with moderate or mild agoraphobia, 9 without avoidance
In review
Proportion of people who were panic free 10 weeks
9/19 (47%) with applied relaxation
11/17 (65%) with CBT
8/22 (36%) with minimal contact control

P >0.05 (among-group comparison)
Global symptoms

RCT
3-armed trial
64 people with DSM-III criteria for panic disorder with moderate or mild agoraphobia, 9 without avoidance
In review
Global function (measured by Anxiety Disorders Interview Schedule and Hamilton Anxiety and Depression scales) 10 weeks
with applied relaxation
with minimal contact control
Absolute results not reported

Reported as significant (applied relaxation v minimal contact control)
P value not reported
Effect size not calculatedapplied relaxation

RCT
3-armed trial
64 people with DSM-III criteria for panic disorder with moderate or mild agoraphobia, 9 without avoidance
In review
Panic symptoms 6 months
with applied relaxation
with CBT
with minimal contact control
Absolute results not reported

Reported as not significant (among-group comparison)
P value not reported
Not significant

RCT
4-armed trial
64 people
In review
Improvement in a composite panic/anxiety outcome (consisting of 17 validated panic and anxiety measures) 12 weeks
From +1.12 to –0.01 with applied relaxation
From 1.22 to 0.94 with waiting list control

P <0.05 (applied relaxation v waiting list control)
Effect size not calculatedapplied relaxation
Depression

RCT
4-armed trial
64 people
In review
Improvement in depression (measured by Beck Depression Inventory; scale 0 to 63, change from baseline) 12 weeks
From 18.6 to 9.8 with applied relaxation
From 21.3 to 20.9 with waiting list control

Reported as significant (applied relaxation v waiting list control)
P value not reported
Effect size not calculatedapplied relaxation

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Applied relaxation versus CBT:

We found two systematic reviews. The reviews identified five RCTs comparing applied relaxation versus CBT. Neither review analysed the effects of applied relaxation separately from those of other psychological treatments so we report the results of the individual RCTs below.

Symptom severity

Applied relaxation compared with CBT We don't know whether applied relaxation is more effective at improving symptoms (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Frequency of panic attack

RCT
3-armed trial
36 people with panic disorder and with no or mild agoraphobia
In review
Panic frequency 12 weeks
with applied relaxation
with CBT
Absolute results not reported

P = 0.01 (applied relaxation v CBT)
Effect size not calculatedCBT

RCT
3-armed trial
36 people with panic disorder and with no or mild agoraphobia
In review
Panic frequency 6 months
with applied relaxation
with CBT
Absolute results not reported

P <0.00035 (applied relaxation v CBT)
Effect size not calculatedCBT
Freedom from panic attacks

RCT
3-armed trial
64 people with DSM-III criteria for panic disorder with moderate or mild agoraphobia, 9 without avoidance
In review
Proportion of people who were panic free 10 weeks
9/19 (47%) with applied relaxation
11/17 (65%) with CBT
8/22 (36%) with minimal contact control

P >0.05 (among-group comparison)
Not significant

RCT
3-armed trial
36 people with panic disorder and with no or mild agoraphobia
In review
Proportion of people who were panic free 4 weeks
with applied relaxation
with CBT
Absolute results not reported

P = 0.04 (applied relaxation v CBT)
Effect size not calculatedCBT

RCT
3-armed trial
36 people with panic disorder and with no or mild agoraphobia
In review
Proportion of people who were panic free 6 months
with applied relaxation
with CBT
Absolute results not reported

P = 0.04 (applied relaxation v CBT)
Effect size not calculatedCBT
Global symptoms

RCT
38 people with DSM-III criteria for panic disorder with no or mild avoidance
In review
Independent assessor ratings 1 year
with applied relaxation
with CBT
Absolute results not reported

Reported as no significant difference between groups
P value not reported
Not significant

RCT
38 people with DSM-III criteria for panic disorder with no or mild avoidance
In review
Self-report scales 1 year
with applied relaxation
with CBT
Absolute results not reported

Reported as no significant difference between groups
P value not reported
Not significant

RCT
38 people with DSM-III criteria for panic disorder with no or mild avoidance
In review
Self-observation of panic attacks 1 year
with applied relaxation
with CBT
Absolute results not reported

Reported as no significant difference between groups
P value not reported
Not significant

RCT
3-armed trial
45 people with DSM-III criteria for panic disorder with agoraphobia
In review
Mean score on the therapist-assessed Behavioural Agoraphobia Test post-treatment
with applied relaxation
with CBT
Absolute results reported graphically

P <0.05 (applied relaxation v CBT)
All groups had self-exposure instructions: these have been shown to be independently effective in the treatment of agoraphobia, which makes interpretation of the results difficult
Effect size not calculatedapplied relaxation

RCT
3-armed trial
45 people with DSM-III criteria for panic disorder with agoraphobia
In review
Mean score on the therapist-assessed Behavioural Agoraphobia Test 1 year
with applied relaxation
with CBT
Absolute results reported graphically

Reported as not significant (applied relaxation v CBT)
All groups had self-exposure instructions: these have been shown to be independently effective in the treatment of agoraphobia, which makes interpretation of the results difficult
Not significant

RCT
4-armed trial
64 people
In review
Improvement in a composite panic/anxiety outcome (consisting of 17 validated panic and anxiety measures) 12 weeks
From +1.06 to +0.02 with applied relaxation
From +0.83 to –0.82 with CBT

Reported as significant (applied relaxation v CBT)
P value not reported
Effect size not calculatedCBT

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Applied relaxation versus drug treatments:

We found two systematic reviews. The reviews identified one RCT comparing applied relaxation versus drug treatments. The reviews did not analyse the results of this RCT separately from those of other psychological treatments. The second review identified 10 controlled studies, six of which were included in the first review, including the RCT reported here. Again, the review did not discuss the benefits of applied relaxation separately from other psychological treatments.

Symptom severity

Applied relaxation compared with imipramine Applied relaxation may be less effective at improving symptoms (low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

RCT
4-armed trial
64 people
In review
Improvement in a composite panic/anxiety outcome (consisting of 17 validated panic and anxiety measures) 12 weeks
From 1.06 to 0.02 with applied relaxation
From 1.15 to 0.01 with imipramine

Reported as not significant (applied relaxation v imipramine)
P value not reported
Not significant

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Applied relaxation versus panic focused psychodynamic psychotherapy:

See option on brief dynamic psychotherapy.

Further information on studies

None.

Comment

One RCT found that cognitive measures taken at the end of treatment were significant predictors of outcome at follow-up.

Substantive changes

Applied relaxation One small RCT (49 people) added comparing applied relaxation therapy versus panic-focused psychodynamic psychotherapy.It found that panic-focused psychodynamic psychotherapy increased response compared with applied relaxation. Categorisation of 'applied relaxation' unchanged (Likely to be beneficial).

2008; 2008: 1010.
Published online 2008 December 16.

Client-centred therapy

Summary

Client-centred therapy is likely to be effective in reducing symptoms.

We found no direct evidence from RCTs about whether client-centred therapy is better than no active treatment.

Benefits and harms

Client-centred therapy versus no treatment:

We found no RCTs comparing client-centred therapy with placebo or no treatment.

Client-centred therapy versus client-centred therapy plus exposure:

We found two systematic reviews, which identified two RCTs. The RCTs assessed people for panic, using a variety of scales, on admission, at discharge (10–14 weeks), and at 3, 6, and 12 months' follow-up. The RCTs found that both client-centred therapy and client-centred therapy plus exposure treatment significantly reduced panic and avoidance symptoms compared with baseline measures (results presented graphically, P value not reported).

Symptom severity

Client-centred therapy compared with client-centred therapy plus exposure We don't know whether client-centred therapy is more effective than client-centred therapy plus additional behavioural exposure treatment at improving symptoms (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Agoraphobia

RCT
40 people
In review
Agoraphobia symptoms 3 months
with client-centred therapy plus exposure
with client-centred therapy
Absolute results reported graphically

Reported as significant
P value not reported
Effect size not calculatedclient-centred therapy plus exposure
Anxiety

RCT
40 people
In review
Anxiety 12 months
with client-centred therapy plus exposure
with client-centred therapy
Absolute results reported graphically

Reported as not significant
P value not reported
Not significant
Depression

RCT
40 people
In review
Depressive symptoms 12 months
with client-centred therapy plus exposure
with client-centred therapy
Absolute results reported graphically

Reported as not significant
P value not reported
Not significant
General symptoms

RCT
40 people
In review
Readiness to expose oneself actively to phobic situations 6 months
with client-centred therapy plus exposure
with client-centred therapy
Absolute results reported graphically

Reported as significant
P value not reported
Effect size not calculatedclient-centred therapy plus exposure

RCT
68 people
In review
Dependence on the expectations of others
with client-centred therapy plus exposure
with client-centred therapy
Absolute results reported graphically

Reported as significant
P value not reported
Effect size not calculatedclient-centred therapy

RCT
68 people
In review
Level of stress
with client-centred therapy plus exposure
with client-centred therapy
Absolute results reported graphically

Reported as significant
P value not reported
Effect size not calculatedclient-centred therapy

RCT
68 people
In review
Rate of psychosomatic complaints
with client-centred therapy plus exposure
with client-centred therapy
Absolute results reported graphically

Reported as significant
P value not reported
Effect size not calculatedclient-centred therapy

RCT
68 people
In review
Time taken to feel accepted by social environment
with client-centred therapy plus exposure
with client-centred therapy
Absolute results reported graphically

Reported as significant
P value not reported
Effect size not calculatedclient-centred therapy plus exposure

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Further information on studies

Both RCTs assessed people for panic, using a variety of scales, on admission, at discharge (10–14 weeks), and at 3, 6, and 12 months' follow-up. The RCTs found that both client-centred therapy and client-centred therapy plus exposure treatment significantly reduced panic and avoidance symptoms compared with baseline measures (results presented graphically, P value not reported).

Comment

None.

Substantive changes

No new evidence

2008; 2008: 1010.
Published online 2008 December 16.

Cognitive restructuring

Summary

Cognitive restructuring is likely to be effective in reducing symptoms.

We found no direct information about whether cognitive restructuring alone is better than no active treatment.

Benefits and harms

Cognitive restructuring versus waiting list or placebo:

We found no systematic review or RCTs comparing cognitive restructuring alone versus placebo or waiting list control.

Cognitive restructuring plus interoceptive exposure compared with waiting list, pill placebo, or psychological placebo:

We found one systematic review comparing cognitive restructuring plus interoceptive exposure versus placebo or no treatment, which included a meta-analysis that was included in, and reanalysed by, a second systematic review. We therefore report only the results from the second review below.

Symptom severity

Cognitive restructuring plus interoceptive exposure compared with waiting list, pill placebo, or psychological placebo We don't know whether cognitive restructuring plus interoceptive exposure is more effective at improving symptoms (reported as a larger effect size) than waiting list or pill or psychological placebo; review identified did not report on the significance of differences between groups (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Symptom improvement

Systematic review
People with panic disorder with and without agoraphobia Treatment effect
with cognitive restructuring plus interoceptive exposure
with waiting list control
Absolute results not reported

Effect size +0.91 (see further information on studies for details on effect size)
Significance not reported

Systematic review
People with panic disorder with and without agoraphobia Treatment effect
with cognitive restructuring plus interoceptive exposure
with pill placebo
Absolute results not reported

Effect size +0.65 (see further information on studies for details on effect size)
Significance not reported

Systematic review
People with panic disorder with and without agoraphobia Treatment effect
with cognitive restructuring plus interoceptive exposure
with psychological placebo
Absolute results not reported

Effect size +1.29 (see further information on studies for details on effect size)
Significance not reported

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Cognitive restructuring versus exposure:

We found three systematic reviews. The first systematic review (search date 2002) identified one RCT but did not analyse the results separately from other psychological treatments so we report the results of the single RCT separately below. The second review (search date 2005) identified one controlled study that was also identified by the first review, but again did not analyse the results separately from other psychological treatments so is not reported further. The third systematic review re-analysed the results of another review; we report the re-analysis below.

Symptom severity

Cognitive restructuring compared with exposure We don't know whether cognitive restructuring is more effective at improving symptoms (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Frequency of panic attacks

RCT
28 people with DSM-III-R-diagnosed panic disorder with agoraphobia
In review
Frequency of panic attacks
with interoceptive plus exteroceptive exposure therapy
with cognitive restructuring
Absolute results not reported

Reported as not significant
P value not reported
Not significant
Global symptoms

RCT
28 people with DSM-III-R-diagnosed panic disorder with agoraphobia Rate of change of behavioural and cognitive variables
with interoceptive plus exteroceptive exposure therapy
with cognitive restructuring
Absolute results not reported

Reported as not significant
P value not reported
Not significant

Systematic review
People with panic disorder with and without agoraphobia Treatment effect
with cognitive restructuring
with situational exposure
Absolute results not reported

Effect size –0.95 (see further information on studies for details on effect size)
Significance not reported

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Further information on studies

The review reported than an effect size of 1.0 would represent a large treatment effect (indicating that the average person in one group would have an outcome superior to that of 84% of people in the control group), while an effect size of 0.0 would indicate no treatment effect. The review reported effect sizes in favour of cognitive restructuring plus interoceptive exposure, ranging from 0.65 to 1.29 depending on the control group.

Comment

None.

Substantive changes

No new evidence

2008; 2008: 1010.
Published online 2008 December 16.

Exposure (external or interoceptive)

Summary

Exposure to the panic-inducing stimulus is likely to be effective in reducing symptoms.

Benefits and harms

Exposure versus control:

We found three systematic reviews and one additional RCT of exposure in the treatment of panic disorder. We excluded the first systematic review because it included non-randomised trials and did not provide methodological details of the studies included in the meta-analysis, making the results difficult to interpret. The second review (search date 2002) included two RCTs of exposure, but did not analyse the results separately from other psychological treatments so we report the results of the individual RCTs below. The third systematic review (search date 2005) identified 12 controlled studies of exposure, two of which were included in the second review, and a further three met our inclusion criteria. This review did not perform a meta-analysis so, again, we report the results of the RCTs that met our inclusion criteria below.

Symptom severity

Exposure compared with control Exposure may be more effective than placebo plus relaxation at increasing the proportion of people who remain well without relapse at 43 weeks. External self-exposure, interoceptive self-exposure, and combined external and interoceptive self-exposure may be more effective than a waiting list control at improving panic and agoraphobia measures (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Proportion of people free from panic attacks

RCT
67 people with panic disorder with or without agoraphobia
In review
Proportion of people who were panic free at follow-up 6 months
83% with group CBT (including education, breathing retraining plus interoceptive exposure)
30% with waiting list
Absolute numbers not reported

P value not reported

RCT
4-armed trial
80 people with panic disorder with agoraphobia
In review
Proportion of people who were panic free 12 months
with external plus interoceptive self-exposure
with external self-exposure
with interoceptive self-exposure
with control
Absolute results not reported

Reported by review to be not significant (among-group comparison)
P value not reported

RCT
4-armed trial
154 people with DSM-III-diagnosed panic disorder with agoraphobia Proportion of the people who improved and remained well without relapse 43 weeks
36% with exposure (combined treatment) plus alprazolam
29% with relaxation plus alprazolam
62% with exposure (combined treatment) plus placebo
18% with relaxation plus placebo
Absolute numbers not reported

P <0.0001 (among-group comparison of results from baseline; no between or among group comparisons assessed)
Panic and agoraphobia

RCT
4-armed trial
80 people diagnosed with panic disorder plus agoraphobia
In review
Hamilton Anxiety Scale 10 weeks
with external self-exposure
with interoceptive self-exposure
with combined external and interoceptive self-exposure
with delayed-treatment control
Absolute results reported graphically

P <0.001 (for all between-group comparisons of individual exposure therapy v delayed control)
Effect size not calculatedexposure therapy

RCT
4-armed trial
80 people diagnosed with panic disorder plus agoraphobia Clinical Global Impression Rating Scale 10 weeks
with external self-exposure
with interoceptive self-exposure
with combined external and interoceptive self-exposure
with delayed-treatment control
Absolute results reported graphically

P <0.001 (for all between-group comparisons of individual exposure therapy v delayed control)
Effect size not calculatedexposure therapy

RCT
4-armed trial
80 people diagnosed with panic disorder plus agoraphobia Agoraphobic Cognitions Scale 10 weeks
with external self-exposure
with interoceptive self-exposure
with combined external and interoceptive self-exposure
with delayed-treatment control
Absolute results reported graphically

P <0.001 (for all between-group comparisons of individual exposure therapy v delayed control)
Effect size not calculatedexposure therapy

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Exposure versus cognitive restructuring:

See option on cognitive restructuring.

Exposure versus CBT:

See option on CBT versus other psychological treatments.

Further information on studies

All four groups significantly improved all panic measures compared with baseline measures (panic free at week 8: 62% with alprazolam plus exposure v 47% with alprazolam plus relaxation v 43% with placebo plus exposure v 47% with placebo plus relaxation; reported as significant, P value not reported). Compared with previous poorer-quality trials, the RCT had three new features: an exposure therapy contrast group, a 6-month treatment-free follow-up, and a low rate of early placebo withdrawals (“non-evaluables”). Exposure consisted of initial education, weekly diary keeping, weekly discussion of diaries with a therapist, and 2 hours of exposure to one or more phobic targets each week.

Comment

None.

Substantive changes

Exposure (external or interoceptive) One RCT added (73 people) comparing exposure in vivo, CBT, and waiting list control. It found no significant difference between exposure in vivo and CBT for a range of outcomes. Categorisation of 'Exposure (external or interoceptive)' unchanged (Likely to be beneficial).

2008; 2008: 1010.
Published online 2008 December 16.

Self-help

Summary

Exposure to the panic-inducing stimulus is likely to be effective in reducing symptoms.

Benefits and harms

Self-help methods versus no treatment:

We found one systematic review (search date 2002), which identified eight randomised and non-randomised clinical studies comparing self-help versus no treatment in people with panic disorder, agoraphobia, or panic disorder plus agoraphobia. For full details of review methods, see further information about studies.

Symptom severity

Self-help compared with no treatment Self-help may be more effective than no treatment (also including pill placebo and therapy placebo) at improving symptoms (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Anxiety

Systematic review
People with panic disorder with or without agoraphobia Anxiety
with self-help
with no treatment
Absolute results not reported

Effect size +0.80
95% CI +0.29 to +1.30
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Effect size not calculatedself-help
Depression

Systematic review
People with panic disorder with or without agoraphobia Depression
with self-help
with no treatment
Absolute results not reported

Effect size +0.62
95% CI +0.03 to +1.21
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Effect size not calculatedself-help
'Clinically significant improvement'

Systematic review
People with panic disorder with or without agoraphobia Clinically significant improvement (not further defined)
with self-help
with no treatment
Absolute results not reported

Effect size +0.98
95% CI +0.25 to +1.71
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Effect size not calculatedself-help

RCT
3-armed trial
36 people with panic disorder
In review
Proportion of people with 'clinical improvement' in frequency of panic attacks (criteria not reported) end of treatment
83% with bibliotherapy CBT
25% with waiting list
Absolute numbers not reported

P <0.01 (bibliotherapy CBT v waiting list control)
Effect size not calculatedself-help

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Self-help methods versus CBT:

We found two systematic reviews (search date 2002, search date 2005) The first review was narrative in character; for full details of methods, inclusion criteria, and conclusions of the first review, see further information about studies. The second review (search date 2005) identified eight controlled studies. It did not perform a meta-analysis and only two RCTs met our inclusion criteria so these are reported below. We also found one subsequent RCT, which is reported below.

Symptom severity

Self-help compared with CBT We don't know whether bibliotherapy self-help or Internet-delivered self-help are more effective than CBT at improving symptoms (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

RCT
28 people with panic disorder
In review
Proportion of people who were classed as having high end-state functioning 6 months
36% with bibliotherapy CBT plus 1 meeting with therapist plus telephone support (3 sessions)
24% with bibliotherapy plus group CBT (4 sessions)
Absolute numbers not reported

P >0.4
Not significant

RCT
49 people with panic disorder with or without agoraphobia; diagnosis confirmed by administering a structured clinical interview Body sensations questionnaire 1 year
with Internet-administered (IT) self-help plus minimal therapist contact via email (10 modules)
with CBT (10 individual weekly sessions, termed live therapy)
Absolute results not reported

Reported as not significant
P value not reported
Not significant

RCT
49 people with panic disorder with or without agoraphobia; diagnosis confirmed by administering a structured clinical interview Agoraphobic cognitions questionnaire 1 year
with Internet-administered (IT) self-help plus minimal therapist contact via email (10 modules)
with CBT (10 individual weekly live sessions)
Absolute results not reported

Reported as not significant
P value not reported
Not significant
Anxiety and depression

RCT
49 people with panic disorder with or without agoraphobia; diagnosis confirmed by administering a structured clinical interview Beck anxiety and depression inventory 1 year
with Internet-administered (IT) self-help plus minimal therapist contact via email (10 modules)
with CBT (10 individual weekly live sessions)
Absolute results not reported

Reported as not significant
P value not reported
Not significant
Agoraphobia

RCT
49 people with panic disorder with or without agoraphobia; diagnosis confirmed by administering a structured clinical interview Mobilitory inventory for agoraphobia (alone or accompanied) 1 year
with Internet-administered (IT) self-help plus minimal therapist contact via email (10 modules)
with CBT (10 individual weekly live sessions)
Absolute results not reported

Reported as not significant
P value not reported
Not significant
'Clinically relevant improvement'

RCT
3-armed trial
36 people with panic disorder
In review
Proportion of people with 'clinical improvement' in frequency of panic attacks (criteria not reported) 6 months
75% with bibliotherapy CBT
92% with CBT
Absolute numbers not reported

P >0.1 (bibliotherapy CBT versus CBT)
Not significant
Freedom from panic disorder

RCT
49 people with panic disorder with or without agoraphobia; diagnosis confirmed by administering a structured clinical interview Proportion of people no longer meeting criteria of panic disorder 1 year
92% with Internet-administered (IT) self-help plus minimal therapist contact via email (10 modules)
88% with CBT (10 individual weekly live sessions)
Absolute numbers not reported

Reported as not significant
P value not reported
Not significant

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Further information on studies

Inclusion criteria of the review were a minimum of four people, and a waiting list, pill placebo, or therapy placebo group: review did not specify that studies had to be randomised. The review calculated effect sizes to determine the additional benefit from active treatment compared with control. The review found no evidence of publication bias in studies that had compared CBT versus waiting list, placebo, behavioural therapy, or combination treatment. However, these results should be interpreted with caution since response rates tend to be greater in pill placebo control groups as opposed to waiting list control groups, and because few studies of CBT used an intention-to-treat analysis.

The review identified 5 RCTs involving a total of 275 people with panic disorder, with or without agoraphobia, of which only three RCTs had at least 6 months' follow-up). It assessed self-management interventions for panic disorders, phobias, and obsessive compulsive disorder. Two of these RCTs (63 people, assessed post-treatment but with no long-term follow-up) used a home-based, internet-delivered self-help programme with minimal therapist input by email, although the other RCT used individual or group CBT with a therapist plus self-study. The review concluded that CBT and self-exposure to panic-provoking stimuli were effective in reducing the frequency of panic attacks, panic-related cognitions, agoraphobic avoidance, anxiety, and depression at follow-up, by allowing subjects to develop skills and coping strategies (data not reported here). The review also suggested that the use of self-help techniques reduced direct contact time with therapists without reducing the efficacy of treatment.

IT self-help modules in this RCT contained components of psychoeducation, socialisation breathing retraining and hyperventilation tests, cognitive restructuring, interoceptive exposure, exposure in vivo and relapse prevention and assertiveness training. Live therapy consisted of 10 weekly individual sessions of 45 to 60 minutes with inter-sessional homework

Comment

The authors of the narrative systematic review looking at self-management commented that, because the RCTs used small sample sizes, significant differences between treatment groups would be difficult to find, as would be the precise intervention responsible for improvement. Furthermore, because self-management interventions require a high level of motivation and engagement, they may be less suitable for people with greater levels of distress.

Substantive changes

Self-help One small RCT added (49 people) which compared a self-help method supplied over the internet versus CBT given in person. The RCT found no significant difference between groups for a range of outcome measures. Categorisation of 'self-help (may be as effective as other forms of CBT)' unchanged (Likely to be beneficial).

2008; 2008: 1010.
Published online 2008 December 16.

Breathing retraining

Summary

Breathing retraining may be beneficial, but we found insufficient evidence to be sure.

Benefits and harms

Breathing retraining alone versus no treatment:

We found no RCTs of sufficient quality.

Breathing retraining plus CBT versus control:

We found three systematic reviews (search date 2002, 2005, and not reported), which between them identified one RCT on the effects of breathing retraining.

Symptom severity

Breathing retraining plus CBT compared with control Breathing retraining plus CBT may be more effective than a delayed-treatment control at increasing the proportion of people with high end-state function (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

RCT
3-armed trial
45 people with DSM-IV-diagnosed panic disorder with or without agoraphobia
In review
Proportion of people with high end-state function (defined as panic frequency = 0, anxiety on Sheehan Patient-Rated Anxiety Scale <30, and phobic avoidance on the Mobility Inventory Scale <1.5) end of treatment
21% with CBT plus breathing retraining (diaphragmatic breathing instruction plus practise homework in sessions 4 and 5)
0% with control (delayed treatment)
Absolute numbers not reported

P <0.01 (CBT plus breathing retraining v control)
Effect size not calculatedCBT plus breathing retraining

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Breathing retraining plus CBT versus CBT alone:

We found three systematic reviews (search date 2002, 2005, and not reported), which identified one RCT on the effects of breathing retraining.

Symptom severity

Breathing retraining plus CBT compared CBT alone We don't know whether breathing retraining plus CBT is more effective than CBT alone at increasing the proportion of people with high end-state function at end of treatment or at 1 year (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

RCT
3-armed trial
45 people with DSM-IV-diagnosed panic disorder with or without agoraphobia
In review
Proportion of people with high end-state function (defined as panic frequency = 0, anxiety on Sheehan Patient-Rated Anxiety Scale <30, and phobic avoidance on the Mobility Inventory Scale <1.5) end of treatment
21% with CBT plus breathing retraining (diaphragmatic breathing instruction plus practise homework in sessions 4 and 5)
38% with CBT alone
Absolute numbers not reported

P <0.10 (CBT plus breathing retraining v CBT alone)
Not significant

RCT
3-armed trial
45 people with DSM-IV-diagnosed panic disorder with or without agoraphobia
In review
Proportion of people with high end-state function (defined as panic frequency = 0, anxiety on Sheehan Patient-Rated Anxiety Scale <30, and phobic avoidance on the Mobility Inventory Scale <1.5) 1 year
37% with CBT plus breathing retraining (diaphragmatic breathing instruction plus practise homework in sessions 4 and 5)
57% with CBT alone
Absolute numbers not reported

P <0.10 (CBT plus breathing retraining v CBT alone)
Not significant

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Further information on studies

None.

Comment

Breathing retraining is based on the rationale that hypocapnia and respiratory irregularities are underlying factors in the development of panic. The systematic review recommended that these factors should be monitored physiologically throughout treatment and that techniques taught in breathing retraining must take account of respiration rate and tidal volume in the regulation of blood gases (partial pressure of carbon dioxide [pCO2]).

Substantive changes

No new evidence

2008; 2008: 1010.
Published online 2008 December 16.

Brief dynamic psychotherapy

Summary

Brief dynamic psychotherapy may be beneficial, but we found insufficient evidence to be sure.

We found no direct information about whether brief dynamic psychotherapy alone is better than no active treatment.

Benefits and harms

Brief dynamic psychotherapy alone versus no treatment:

We found no systematic review or RCTs.

Brief dynamic psychotherapy plus clomipramine versus clomipramine alone:

We found one RCT.

Symptom severity

Brief dynamic psychotherapy plus clomipramine compared with clomipramine alone Brief dynamic psychotherapy plus clomipramine may be more effective at increasing global improvement and at improving panic scores at 18 months, and at decreasing the proportion of people with relapse 9 months after the end of treatment (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

RCT
40 people Global improvement on Clinical Global Impression 18 months
with brief dynamic psychotherapy plus clomipramine
with clomipramine
Absolute results not reported

P = 0.001
See further information on studies for discussion of potential effects of adding clomipramine to BDP
Effect size not calculatedbrief dynamic psychotherapy plus clomipramine
Freedom from panic attacks

RCT
40 people Proportion of people who were free from panic attacks end of treatment
100% with brief dynamic psychotherapy plus clomipramine
75% with clomipramine
Absolute numbers not reported

Significance not assessed
See further information on studies for discussion of potential effects of adding clomipramine to BDP

RCT
40 people Proportion of people who were free from panic attacks 6 months
20/20 (100%) with brief dynamic psychotherapy plus clomipramine
20/20 (100%) with clomipramine

Significance not assessed
See further information on studies for discussion of potential effects of adding clomipramine to BDP
Panic

RCT
40 people Improvement in panic subscale of Clinical Global Impression 18 months
with brief dynamic psychotherapy plus clomipramine
with clomipramine
Absolute results not reported

P <0.001
See further information on studies for discussion of potential effects of adding clomipramine to BDP
Effect size not calculatedbrief dynamic psychotherapy plus clomipramine
Relapse

RCT
40 people Proportion of people who relapsed 9 months after the end of treatment
20% with brief dynamic psychotherapy plus clomipramine
75% with clomipramine
Absolute numbers not reported

Reported as significant
P value not reported
See further information on studies for discussion of potential effects of adding clomipramine to BDP
Effect size not calculatedbrief dynamic psychotherapy plus clomipramine

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Panic-focused psychodynamic psychotherapy versus applied relaxation:

We found one small RCT.

Symptom severity

Panic-focused psychodynamic psychotherapy versus applied relaxation therapy Panic-focused psychodynamic psychotherapy may be more effective at reducing symptom severity and may be more effective at increasing the proportion of people who respond to treatment (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

RCT
49 adults with primary diagnosis DSM-IV diagnosis panic disorder Mean Panic Disorder Severity Scale score (change from baseline)
From 13.2 to 5.1 with panic-focused psychodynamic psychotherapy
From 12.2 to 9.0 with applied relaxation therapy

P = 0.002
Effect size not calculatedPFPP

RCT
49 adults with primary diagnosis DSM-IV diagnosis panic disorder Proportion of people classed as a responder (response defined as 40% reduction from baseline in Panic Disorder Severity Scale score)
19/26 (73%) with panic-focused psychodynamic psychotherapy
9/23 (39%) with applied relaxation therapy

P = 0.016
Effect size not calculatedPFPP

Quality of life

No data from the following reference on this outcome.

Adverse effects

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Withdrawal

RCT
49 adults with primary diagnosis DSM-IV diagnosis panic disorder Proportion of people withdrawing from RCT
2/26 (7%) with panic-focused psychodynamic psychotherapy
8/23 (34%) with applied relaxation therapy

P = 0.03
Effect size not calculatedPFPP

Further information on studies

The authors of the RCT suggested that the addition of clomipramine may, by reducing the frequency and intensity of the panic attacks, have reduced the level of psychological distress sufficiently for the people to then be able to work on the issues addressed by BDP. The combined treatment could have maximised the patients' confidence by both ameliorating their panic attacks and providing the opportunity to change their maladaptive interpersonal patterns.

Comment

None.

Substantive changes

Brief dynamic psychotherapy One small RCT (49 people) added comparing panic-focused psychodynamic psychotherapy versus applied relaxation therapy. It found that panic-focused psychodynamic psychotherapy increased response compared with applied relaxation. Categorisation of 'Brief dynamic psychotherapy' unchanged (Unknown effectiveness).

2008; 2008: 1010.
Published online 2008 December 16.

Couple therapy

Summary

Couple therapy may be beneficial, but we found insufficient evidence to be sure.

We found no direct information about whether couple therapy is better than no active treatment.

Benefits and harms

Couple therapy versus no treatment:

We found no systematic review or RCTs.

Different forms of couple therapy versus each other:

We found one systematic review (search date 2001), which included three RCTs of sufficient quality.

Symptom severity

Different forms of couple therapy compared with each other Couples communication skills training may be more effective than couples relaxation training at increasing the proportion of people taking unaccompanied excursions and at improving behavioural approach test scores at 8 months. We don't know whether behavioural therapy with husband as co-therapist is more effective than behavioural therapy with a female friend as co-therapist, or whether graded exposure with friends or spouses is more effective than problem solving with friends or spouses at 4 weeks (low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

RCT
30 married women with DSM-III-diagnosed panic disorder with agoraphobia
In review
Mean Behavioural Items score 6 months
with behavioural therapy at home with a female friend
with couple therapy (behavioural therapy at home with husband as co-therapist)
Absolute results reported graphically

Reported as not significant
P value not reported
Not significant

RCT
24 women with DSM-III-diagnosed agoraphobia with panic attacks
In review
Mean Behavioural Approach Test score 8 months
with couples relaxation training
with couples communication skills training
Absolute results not reported

P <0.02
Effect size not calculatedcouples communication skills training

RCT
24 women with DSM-III-diagnosed agoraphobia with panic attacks
In review
Proportion of people taking unaccompanied excursions 8 months
with couples relaxation training
with couples communication skills training
Absolute results not reported

P <0.01
Effect size not calculatedcouples communication skills training
Anxiety

RCT
30 married women with DSM-III-diagnosed panic disorder with agoraphobia
In review
Mean Leeds Anxiety score 6 months
with behavioural therapy at home with a female friend
with couple therapy (behavioural therapy at home with husband as co-therapist)
Absolute results reported graphically

Reported as not significant
P value not reported
Not significant

RCT
28 women with agoraphobia whose main complaint was fear of leaving home and entering public places
In review
Change in physician-assessed ratings of phobic anxiety 6 months
with programmed practise (graded exposure) at home
with problem solving at home
Absolute results reported graphically

Reported as not significant
P value not reported
Not significant

RCT
28 women with agoraphobia whose main complaint was fear of leaving home and entering public places
In review
Change in participant-assessed ratings of phobic anxiety 6 months
with programmed practise (graded exposure) at home
with problem solving at home
Absolute results reported graphically

Reported as not significant
P value not reported
Not significant

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Further information on studies

None.

Comment

Some cognitive behavioural therapists have encouraged spouse participation on the grounds that the person's adherence with exposure homework assignment will improve. However, the evidence is conflicting. One systematic review (that included no studies that met Clinical Evidence inclusion criteria) found that many studies reviewed had small samples, that psychometric data of all measures were not published, and that there was a wide range of variability in parameters used by different studies.

Substantive changes

No new evidence

2008; 2008: 1010.
Published online 2008 December 16.

Insight-orientated therapy

Summary

We found no direct information from RCTs about whether insight-orientated therapy is better than no active treatment.

Benefits and harms

Insight-orientated therapy:

We found one systematic review (search date 2002), which identified no RCTs of sufficient quality.

Further information on studies

None.

Comment

RCTs are needed. There is currently widespread scepticism about the usefulness of insight-orientated therapy in panic disorder.

Substantive changes

No new evidence

2008; 2008: 1010.
Published online 2008 December 16.

Psychoeducation

Summary

We found no direct information from RCTsa about whether psychoeducation alone is better than no active treatment.

Benefits and harms

Psychoeducation:

We found no systematic review or RCTs of psychoeducation as a sole intervention in the treatment of panic disorder.

Further information on studies

None.

Comment

We found no RCTs evaluating psychoeducation as the sole intervention. Most CBT interventions generally started with educational/informational session(s) providing information on the nature of symptoms experienced during panic attack, and on the roles played by fears, avoidance, and catastrophic misinterpretation in the onset and maintenance of panic symptoms. Such information formed the basis of developing a disorder model of panic disorder, and rationale for the specific intervention to be used in the study.

Substantive changes

No new evidence

2008; 2008: 1010.
Published online 2008 December 16.

SSRIs

Summary

SSRIs are effective at reducing the symptoms of panic disorder.

Benefits and harms

SSRIs versus placebo:

We found four systematic reviews (search date 2002, 2005, not reported not reported). Two of the reviews systematic reviews were each included in the third review (search date 2002) so are not reported further. The meta-analysis of the third review, which assessessed anxiety, is reported below. The fourth systematic review identified 22 placebo-controlled studies, 10 of which were included in the third review. It also performed a meta-analysis assessing global symptom improvement which is reported below. For full details of inclusion criteria and methods of the reviews, see further information about studies. We also found two additional RCTs.

Symptom severity

SSRIs compared with placebo SSRIs may be more effective at improving symptoms of panic disorder (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

RCT
5-armed trial
279 people Proportion of people who responded (defined as no panic attacks and either no episodic increases in anxiety or only slight increases in anxiety precipitated by definite events or activities) 12 months
with oral citalopram 10 or 15 mg daily
with oral citalopram 20 or 30 mg daily
with oral citalopram 40 or 60 mg daily
with placebo
Absolute results reported graphically

Citalopram 10 or 15 mg/day v placebo; P = 0.05
Citalopram 20 or 30 mg/day v placebo; P = 0.001
Citalopram 40 or 60 mg/day v placebo; P = 0.003
Only 28/54 (52%) people completed the trial; analysis was by intention to treat, and people who withdrew from the trial were counted as treatment failures
Effect size not calculatedcitalopram

RCT
182 people who had responded to open label sertraline for 52 weeks Proportion of people who had exacerbation of symptoms 28 weeks
13% with sertraline
33% with placebo (discontinuation of sertraline)
Absolute numbers not reported

P = 0.005
The use of treatment responders was likely to bias results in favour of the drug
Effect size not calculatedsertraline

Systematic review
People with panic disorder with or without agoraphobia Symptom improvement
with SSRIs for 8 to 12 weeks
with placebo
Absolute results not reported

NNT 8
95% CI 6 to 11
Effect size not calculatedSSRIs

Systematic review
People with panic disorder with or without agoraphobia Symptom improvement
with paroxetine for 8 to 12 weeks
with placebo
Absolute results not reported

NNT 5
95% CI 3 to 7
Effect size not calculatedparoxetine

Systematic review
People with panic disorder with or without agoraphobia Symptom improvement
with citalopram for 8 to 12 weeks
with placebo
Absolute results not reported

NNT 5
95% CI 3 to 11
Effect size not calculatedcitalopram
People with panic disorder with or without agoraphobia Symptom improvement
with sertraline for 8 to 12 weeks
with placebo
Absolute results not reported

NNT 8
95% CI 5 to 20
Effect size not calculatedsertraline
Anxiety

Systematic review
People with panic disorder with or without agoraphobia Anxiety
with SSRIs
with placebo
Absolute results not reported

Effect size 0.41
Significance and P value not reported
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Depression

Systematic review
People with panic disorder with or without agoraphobia Depression
with SSRIs
with placebo
Absolute results not reported

Effect size 0.50
Significance and P value not reported
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation

Quality of life

No data from the following reference on this outcome.

Adverse effects

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Adverse effects

RCT
5-armed trial
279 people Adverse effects
with oral citalopram 10 or 15 mg daily
with oral citalopram 20 or 30 mg daily
with oral citalopram 40 or 60 mg daily
with placebo
Absolute results not reported

Only 28/54 (52%) people completed the trial; analysis was by intention to treat, and people who withdrew from the trial were counted as treatment failures

RCT
182 people who had responded to open label sertraline for 52 weeks Adverse effects 28 weeks
with sertraline
with placebo (discontinuation of sertraline)
Absolute results not reported

No data from the following reference on this outcome.

SSRIs versus MAOIs:

See option on MAOIs.

SSRIs versus CBT:

See option on CBT versus drug treatment.

Further information on studies

The review identified and performed a meta-analysis on results from 78 controlled studies identified by these and by other meta-analyses. The review stated that included studies had to have a control group, but did not specify that they must be randomised.

The review reported only the number needed to treat (NNT) to improve symptoms in one person for all SSRIs and for each drug individually.

Comment

The second systematic review found that smaller RCTs were associated with larger effect sizes, suggesting the possibility of publication bias.

SSRIs can cause initial increased anxiety, which can exacerbate a tendency to focus on internal sensations, and to avoid situations that trigger these sensations (catastrophisation of somatic sensations). Education about this is likely to improve adherence with medication. The FDA and other regulatory bodies have issued several alerts and revised prescribing information regarding the use of SSRIs, on the increased risk of self-harm and suicide, on increased risk of neonatal persistent pulmonary hypertension in women who had taken SSRIs during pregnancy, on the risk of congenital malformations in women taking paroxetine during early pregnancy, and on the potential for SSRIs to cause hyponatraemia. See harms of prescription antidepressant drugs in review on depression in adults (drug and other physical treatments)

Tricyclic antidepressants versus SSRIs:

See comment on tricyclic antidepressants.

Substantive changes

No new evidence

2008; 2008: 1010.
Published online 2008 December 16.

Tricyclic antidepressants

Summary

Tricyclic antidepressants are effective at reducing the symptoms of panic disorder.

Benefits and harms

Tricyclic antidepressants versus placebo:

We found three systematic reviews, and two additional RCTs. The first systematic review was included in the second review (search date 2002) so is not reported further. The meta-analysis of the second review, assessing anxiety and depression, is reported below. The third systematic review (search date 2005) identified 21 placebo-controlled studies, 10 of which were included in the second review. It performed a meta-analysis assessing global symptom improvement, which is reported below.

Symptom severity

Tricyclic antidepressants compared with placebo Tricyclic antidepressants may be more effective at improving symptoms of panic disorder (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

Systematic review
People with panic disorder with or without agoraphobia Symptom improvement
with tricyclic antidepressants for 6 to 12 weeks
with placebo
Absolute results not reported

NNT 6
95% CI 5 to 8

Systematic review
People with panic disorder with or without agoraphobia Symptom improvement
with imipramine for 6 to 12 weeks
with placebo
Absolute results not reported

NNT 6
95% CI 4 to 8

Systematic review
People with panic disorder with or without agoraphobia Symptom improvement
with clomipramine for 6 to 12 weeks
with placebo
Absolute results not reported

NNT 7
95% CI 4 to 17
Anxiety

Systematic review
People with panic disorder with or without agoraphobia Anxiety
with tricyclic antidepressants
with placebo
Absolute results not reported

Effect size 0.41
Significance and P value not reported
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Depression

Systematic review
People with panic disorder with or without agoraphobia Depression
with tricyclic antidepressants
with placebo
Absolute results not reported

Effect size 0.34
Significance and P value not reported
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Frequency of panic attack

RCT
3-armed trial
181 people with panic disorder with or without agoraphobia Frequency of panic attack 8 months
with oral imipramine (maximum dose 225 mg)
with placebo
Absolute results reported graphically

Significance not assessed
Results favoured imipramine
Relapse

RCT
56 adults with panic disorder and agoraphobia in stable remission after 24 weeks' treatment with oral imipramine Proportion of people relapsing 12 months
1/29 (3%) with oral imipramine 2.25 mg/kg daily
10/27 (37%) with placebo

RR 0.09
95% CI 0.01 to 0.68
NNT 5
95% CI 3 to 14
Large effect sizeimipramine

Quality of life

No data from the following reference on this outcome.

Adverse effects

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Adverse effects

RCT
3-armed trial
181 people with panic disorder with or without agoraphobia Adverse effects
with oral imipramine (maximum dose 225 mg)
with placebo

No data from the following reference on this outcome.

Tricyclic antidepressants versus CBT:

See option on CBT versus drug treatment.

Further information on studies

The review identified and performed a meta-analysis on results from 78 controlled studies identified by this and by other meta-analyses. The review stated that included studies had to have a control group but did not specify that they had to be randomised.

The review performed a meta-analysis and calculated the number needed to treat (NNT) to improve symptoms in one person for each type of drug intervention compared with placebo.

Comment

Short-term effects:

We found one systematic review (search date 1999, 43 studies including 34 RCTs, 2367 people, withdrawal rate 24%, analysis based on completers) that compared the short-term efficacy of SSRIs (fluoxetine, fluvoxamine, paroxetine, citalopram, and sertraline) versus tricyclic antidepressants (imipramine, desipramine, nortriptyline, and clomipramine) and analysed effect size within treatment group rather than within studies. It found no significant difference between treatments in the proportion of people free of panic attacks at 6 to 10 weeks (60% with tricyclic antidepressants v 55% with SSRIs; P value not reported). It found that tricyclic antidepressants significantly increased withdrawal rates (31% with tricyclic antidepressants v 18% with SSRIs; P <0.001). These results should be interpreted with caution because nine of the RCTs were open label and there was no indication of the length of follow-up for any of the RCTs.

Substantive changes

No new evidence

2008; 2008: 1010.
Published online 2008 December 16.

Benzodiazepines

Summary

Benzodiazepines can be effective in reducing symptoms in panic disorder, but their adverse-effect profile makes them unsuitable for long-term treatment.

Benzodiazepines are associated with a wide range of well-recognised adverse effects, both during and after treatment.

Benefits and harms

Benzodiazepines versus placebo:

We found five systematic reviews, The first systematic review was included in the second review (search date 2002) so is not reported further. The second review performed a meta-analysis anxiety and depression, which is reported below. The third systematic review (search date 2005) identified 27 placebo-controlled studies, 17 of which were included in the second review. It performed a meta-analysis assessing global symptoms, which is reported below. For full details of inclusion criteria and methods of these two reviews, see further information about studies. The fourth systematic review was excluded as the RCTs included did not meet Clinical Evidence inclusion criteria. The fifth review (search date 2006, 16 RCTs published from 1986–1999) assessed alprazolam versus placebo and performed a meta-analysis assessing global symptom improvement, which is reported below.

Symptom severity

Benzodiazepines compared with placebo Benzodiazepines may be more effective at improving symptoms of panic disorder (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

Systematic review
People with panic disorder with or without agoraphobia Symptom improvement
with alprazolam for 5 to 8 weeks
with placebo
Absolute results not reported

NNT 5
95% CI 4 to 7

Systematic review
People with panic disorder with or without agoraphobia Symptom improvement
with clonazepam for 5 to 8 weeks
with placebo
Absolute results not reported

NNT 5
95% CI 4 to 7
Anxiety

Systematic review
People with panic disorder with or without agoraphobia Anxiety
with benzodiazepines
with placebo
Absolute results not reported

Effect size 0.40
Significance and P value not reported
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Depression

Systematic review
People with panic disorder with or without agoraphobia Depression
with benzodiazepines
with placebo
Absolute results not reported

Effect size 0.28
Significance and P value not reported
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Freedom from panic attacks

Systematic review
1669 people
8 RCTs in this analysis
Proportion of people free from panic attacks
64% with alprazolam
41% with placebo
Absolute numbers not reported

RR 0.61
95% CI 0.52 to 0.71
Small effect sizebenzodiazepine

Quality of life

No data from the following reference on this outcome.

Adverse effects

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Adverse effects
Adverse effects
with benzodiazepines
with placebo

Non-systematic review
People with a history of substance abuse or dependence and anxiety disorder Adverse effects
with benzodiazepines
with placebo
Withdrawal

Systematic review
2284 people
14 RCTs in this analysis
Proportion of people withdrawing (reasons for withdrawal not further defined)
15% with alprazolam
44% with placebo
Absolute numbers not reported

RR 0.22
95% CI 0.18 to 0.27
Moderate effect sizebenzodiazepine

No data from the following reference on this outcome.

Further information on studies

The review identified and performed a meta-analysis on results from 78 controlled studies identified by this and by other meta-analyses. Mean effect sizes for benzodiazepines were similar to those of SSRIs and of tricyclic antidepressants for reduction of anxiety and depression scores. The review stated that included studies had to have a control group but did not specify that they had to be randomised.

The review performed a meta-analysis and calculated the number needed to treat (NNT) to improve symptoms in one person for each type of drug intervention compared with placebo.

Comment

Many RCTs of psychological and pharmacological treatments (even those not involving benzodiazepines) allowed people to receive small amounts of anxiolytic drugs during the study, because benzodiazepine use and abuse is quite prevalent in people who suffer from panic disorder. We found one systematic review with meta-analysis of placebo-controlled RCTs of antidepressants and benzodiazepines for the treatment of panic disorders (search date 1990, 1276 people, 13 antidepressant trials, mean duration 16 weeks; 6 benzodiazepine trials, mean duration 7 weeks) that found that antidepressants and benzodiazepines were likely to be equally effective in the short-term treatment of panic disorder. However, longer-term follow-up was not performed in any of the RCTs of benzodiazepine treatment.

Substantive changes

Benzodiazepines One systematic review (search date 2006) added which compared alprazolam versus placebo, and found that alprazolam significantly increased the proportion of people free from panic attacks compared with placebo. Categorisation of benzodiazepines unchanged (Trade-off between benefits and harms).

2008; 2008: 1010.
Published online 2008 December 16.

Buspirone

Summary

We don't know whether buspirone is effective in the treatment of panic disorder.

We found no direct information from RCTs about whether buspirone is better than no active treatment.

Benefits and harms

Oral buspirone alone versus placebo:

We found no RCTs.

Oral buspirone plus CBT versus placebo plus CBT:

We found two systematic reviews which identified one RCT, and we found one additional RCT. The RCTs gave no information on adverse effects. For information on adverse effects of buspirone, see harms of buspirone in review on generalised anxiety disorder.

Symptom severity

Oral buspirone plus CBT versus placebo plus CBT We don't know whether oral buspirone plus CBT is more effective at improving symptoms (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Panic and agoraphobia

RCT
41 people with panic disorder and agoraphobia
In review
Proportion of people with a reduction of at least 50% in agoraphobic symptoms 68 weeks
44% with oral buspirone 30 mg daily plus CBT
68% with placebo plus CBT
Absolute numbers not reported

Reported as not significant
P value not reported
Not significant

RCT
48 people Improvement in self-rated panic and agoraphobia scores (using a 90-point symptom scale where each symptom was graded from 0 = not present to 4 = severe) 1 year
with oral buspirone (maximum 60 mg/day) plus CBT
with placebo plus CBT
Absolute results not reported

P = 0.03
Effect size not calculatedoral buspirone plus CBT

Quality of life

No data from the following reference on this outcome.

Adverse effects

No data from the following reference on this outcome.

Further information on studies

None.

Comment

None.

Substantive changes

No new evidence

2008; 2008: 1010.
Published online 2008 December 16.

MAOIs

Summary

We don't know whether MAOIs are effective.

Benefits and harms

MAOIs versus control or placebo:

We found one systematic review (search date 2005), which identified no controlled studies of MAOIs meeting our inclusion criteria for this comparison.

MAOIs versus SSRIs:

We found one systematic review (search date 2005), which identified four controlled studies of MAOIs, one of which met our inclusion criteria.

Symptom severity

MAOIs compared with SSRIs We don't know whether moclobemide is more effective than fluoxetine at reducing the proportion of people who are panic free at 1 year (low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Freedom from panic attacks

RCT
366 people with panic disorder Proportion of people who were panic free 1 year
60% with moclobemide 300 to 600 mg daily
65% with fluoxetine 10 to 30 mg daily
Absolute numbers not reported

Reported as not significant
P value not reported
Not significant

Quality of life

No data from the following reference on this outcome.

Adverse effects

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Adverse effects

RCT
366 people Adverse effects
with moclobemide 300 to 600 mg daily
with fluoxetine 10 to 30 mg daily

Further information on studies

The safety database found that, compared with placebo, moclobemide was associated with higher rates of insomnia (24% with moclobemide v 13% with placebo), and dizziness (11% with moclobemide v 7% with placebo; P values not reported). There was no significant difference in blood pressure changes with moclobemide at doses less than 300 mg daily, 300 mg to 599 mg daily, or 600 mg or more daily, compared with placebo.

Comment

None.

Substantive changes

No new evidence

2008; 2008: 1010.
Published online 2008 December 16.

CBT plus drug treatments versus CBT alone

Summary

Combined treatment with CBT plus antidepressants has been shown to be more effective than CBT alone in reducing symptoms in the short term.

Benefits and harms

CBT plus antidepressants versus CBT alone:

We found two systematic reviews (search date 2002,search date 2005), which pooled data and performed slightly different analysis so both are reported here. For full details of methods and inclusion criteria of reviews, see further information about studies.

Symptom severity

CBT plus antidepressants compared with CBT alone CBT plus antidepressants may be more effective at improving the proportion of people in remission and at decreasing global severity of symptoms in the acute phase at 2 to 4 months, but not at improving the proportion of people with response at 2 to 4 months. We don't know whether CBT plus antidepressants is more effective at improving response/remission with continued treatment after the acute phase, or at improving response/remission 6 to 24 months after treatment discontinuation (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

Systematic review
363 people
3 RCTs in this analysis
Global severity of symptoms in the acute phase 2 to 4 months
with CBT plus antidepressant
with CBT or CBT plus placebo
Absolute results not reported

SMD –0.31
95% CI –0.54 to –0.08
Control group was a combined analysis of CBT alone and CBT plus placebo (see further information on studies for comments on this and generalisability)
Effect size not calculatedCBT plus antidepressant

Systematic review
709 people
9 RCTs in this analysis
Proportion of people with response to treatment in the acute phase 2 to 4 months
187/310 (60%) with CBT plus antidepressant
211/399 (53%) with CBT or CBT plus placebo

RR 1.13
95% CI 0.96 to 1.33
Control group was a combined analysis of CBT alone and CBT plus placebo (see further information on studies for comments on this and generalisability)
Not significant

Systematic review
205 people
Data from 1 RCT
Proportion of people with response/remission after acute phase > 4 months
37/65 (60%) with CBT plus antidepressant
64/140 (46%) with CBT or CBT plus placebo

RR 1.23
95% CI 0.93 to 1.63
Control group was a combined analysis of CBT alone and CBT plus placebo (see further information on studies for comments on this and generalisability)
Not significant

Systematic review
339 people
3 RCTs in this analysis
Proportion of people with response/remission after acute phase 6 to 24 months after treatment discontinuation
41/111 (37%) with CBT plus antidepressant
90/228 (40%) with CBT or CBT plus placebo

RR 0.91
95% CI 0.69 to 1.21
Control group was a combined analysis of CBT alone and CBT plus placebo (see further information on studies for comments on this and generalisability)
Not significant
Anxiety

Systematic review
People with panic disorder with or without agoraphobia Anxiety median follow-up of 16.8 months
with CBT plus antidepressant
with CBT
Absolute results not reported

Effect size +0.23
95% CI +0.09 to +0.37
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Effect size not calculatedCBT plus antidepressant
Depression

Systematic review
People with panic disorder with or without agoraphobia Depression median follow-up of 16.8 months
with CBT plus antidepressant
with CBT
Absolute results not reported

Effect size +0.29
95% CI +0.09 to +0.49
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Effect size not calculatedCBT plus antidepressant
'Clinically significant improvement'

Systematic review
People with panic disorder with or without agoraphobia Clinically significant improvement (not further defined) median follow-up of 16.8 months
with CBT plus antidepressant
with CBT
Absolute results not reported

Effect size +0.40
95% CI +0.23 to +0.56
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Effect size not calculatedCBT plus antidepressant
Remission

Systematic review
625 people
7 RCTs in this analysis
Proportion of people in remission 2 to 4 months
160/268 (60%) with CBT plus antidepressant
178/357 (50%) with CBT or CBT plus placebo

RR 1.23
95% CI 1.02 to 1.47
Control group was a combined analysis of CBT alone and CBT plus placebo (see further information on studies for comments on this and generalisability)
Small effect sizeCBT plus antidepressant

Quality of life

CBT plus antidepressants compared with CBT alone We don't know whether CBT plus drug treatment (mainly antidepressants but also included other drugs) is more effective (analysis also included behavioural therapy) at improving quality of life (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Quality of life

Systematic review
People with panic disorder with or without agoraphobia Quality of life median follow-up of 16.8 months
with CBT plus antidepressant
with CBT
Absolute results not reported

Effect size +0.25
95% CI –0.18 to +0.68
Positive value for effect size means first intervention more effective than comparator; larger value means greater effect
The review did not report details of method of randomisation
Results should be interpreted with caution (see further information on studies for more details)
Not significant

Adverse effects

CBT plus antidepressants compared with CBT alone CBT plus antidepressants may increase the proportion of people discontinuing treatment due to adverse effects compared with CBT alone (low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Withdrawal

Systematic review
604 people
6 RCTs in this analysis
Proportion of people who withdrew because of adverse effects 2 to 4 months
25/258 (10%) with CBT plus antidepressant
2/346 (1%) with CBT or CBT plus placebo

RR 5.00
95% CI 1.96 to 12.72
Control group was a combined analysis of CBT alone and CBT plus placebo (see further information on studies for comments on this and generalisability)
Large effect sizeCBT alone

No data from the following reference on this outcome.

CBT plus buspirone versus CBT alone:

See option on buspirone.

Further information on studies

The review (search date 2002) identified 20 studies comparing psychotherapy (CBT or behavioural therapy) with combined psychotherapy plus pharmacotherapy (mainly antidepressants). The drugs investigated were mainly SSRIs and tricyclic antidepressants. Effect sizes were calculated to determine the additional benefit from active treatment compared with control. The review stated that included studies had to have a control group but did not specify that they had to be randomised. The review found some evidence of publication bias in studies that had compared CBT with drug treatment. Adjustment for this led to an increased calculated effect size for CBT compared with drug treatment, but the authors did not report the change in effect size for combined treatment. These results should therefore be interpreted with caution, as pill placebo response rates tend to be greater than those for waiting list control groups, and because few studies of CBT used an intention-to-treat analysis.

The review (search date 2005) compared psychotherapy plus antidepressant versus psychotherapy alone or psychotherapy plus placebo in people with panic disorder with or without agoraphobia. It reported both short- and long-term outcomes and included only RCTs. The review included any type of psychotherapy (including 9 RCTs of CBT consisting of both behavioural and cognitive therapy elements; 12 RCTs of behavioural therapy [including exposure and/or breathing retraining and/or relaxation]; and 2 RCTs of "psychodynamic and others"). It presented pooled results for all psychotherapies, but also presented a subgroup analysis for RCTs of CBT alone which we have reported here. The review included RCTs in which benzodiazepines were used irregularly. The combination of CBT alone and CBT plus placebo as the control comparison was a post hoc analysis, and the review also reported CBT plus antidepressants versus CBT alone, and CBT plus antidepressants versus CBT plus placebo, separately. Overall, the results were broadly similar in all the analyses, but the significance of some results were sensitive to the method of analysis used. The review noted that the generalisability of the findings beyond specialist psychiatric settings was not straightforward, as only one RCT was undertaken in a primary care setting. It also noted the comparability of treatment arms after the acute phase or continuation phase may have been compromised, as people were free to have other treatments before the final follow-up assessment, and 30% to 77% of people did so.

Comment

None.

Substantive changes

CBT plus drug treatments versus CBT alone One systematic review (search date 2005) added including a meta-analysis for short- and long-term outcomes. Benefits and harms data enhanced. Categorisation of 'CBT plus antidepressants versus CBT alone (combination may be more effective in acute phase; unclear which is more effective with continued treatment, or 6 to 24 months after treatment discontinuation)' unchanged (Likely to be beneficial).

2008; 2008: 1010.
Published online 2008 December 16.

CBT plus drug treatments versus drugs alone

Summary

Combined treatment with CBT plus antidepressants has been shown to be more effective than antidepressants alone in reducing symptoms in the short term.

Benefits and harms

CBT plus antidepressants versus antidepressants alone:

We found one systematic review (search date 2005) and one subsequent RCT with two subsequent analyses. For full details of inclusion criteria and methods of review, see further information on studies.

Symptom severity

CBT plus antidepressants compared with antidepressants alone CBT plus antidepressants may be more effective at increasing the proportion of people who respond and at decreasing the global severity of symptoms in the acute phase at 2 to 4 months, but not at increasing remission at 2 to 4 months. CBT plus antidepressants may be more effective at improving response/remission with continued treatment after the acute phase, or when assessed at 6 to 24 months after discontinuation of treatment, although results at 6 to 24 months are of borderline significance (very low-quality evidence).

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Global symptoms

Systematic review
336 people
5 RCTs in this analysis
Proportion of people who responded in the acute phase 2 to 4 months
94/159 (59%) with CBT plus antidepressant
74/177 (42%) with antidepressants alone

RR 1.46
95% CI 1.05 to 2.02
The review noted that the results may not be generalisable beyond specialist psychiatric settings (see further information on studies for more details)
Small effect sizeCBT plus antidepressant

Systematic review
206 people
2 RCTs in this analysis
Global severity of symptoms 2 to 4 months
with CBT plus antidepressant
with antidepressants alone
Absolute results not reported

SMD –0.30
95% CI –0.57 to –0.02
The review noted that the results may not be generalisable beyond specialist psychiatric settings (see further information on studies for more details)
Effect size not calculatedCBT plus antidepressant

Systematic review
148 people
Data from 1 RCT
Proportion of people responding or in remission after acute phase >4 months
37/65 (57%) with CBT plus antidepressant
31/83 (37%) with antidepressants alone

RR 1.52
95% CI 1.07 to 2.16
The review noted that the results may not be generalisable beyond specialist psychiatric settings (see further information on studies for more details)
Small effect sizeCBT plus antidepressant

Systematic review
240 people
3 RCTs in this analysis
Proportion of people responding or in remission 6 to 24 months after treatment discontinuation
41/111 (37%) with CBT plus antidepressant
32/129 (25%) with antidepressants alone

RR 1.46
95% CI 1.00 to 2.11
P = 0.05
Result is of borderline significance
The review noted that the results may not be generalisable beyond specialist psychiatric settings (see further information on studies for more details)
Not significant
Anxiety

RCT
232 people in primary-care, meeting DSM-IV criteria for panic disorder, with or without co-morbid mental and physical disorders Proportion of people responding (achieving a score of <20 on the Anxiety Sensitivity Index scale) 12 months
63% with CBT plus pharmacotherapy
38% with usual care
Absolute numbers not reported

P <0.001
Effect size not calculatedCBT plus pharmacotherapy

RCT
232 people in primary-care, meeting DSM-IV criteria for panic disorder, with or without co-morbid mental and physical disorders
Further report of reference
Number of anxiety-free days over a 12 month period
with CBT plus pharmacotherapy
with usual care
Absolute results not reported

Difference between groups: +60.4 anxiety-free days
95% CI 42.9 anxiety-free days to 77.9 anxiety-free days
See further information for details of subgroup analysis based on burden of chronic medical illness at baseline
Effect size not calculatedCBT plus pharmacotherapy
Remission

Systematic review
252 people
3 RCTs in this analysis
Proportion of people in remission 2 to 4 months
75/117 (69%) with CBT plus antidepressant
62/135 (46%) with antidepressants alone

RR 1.40
95% CI 0.92 to 2.14
The review noted that the results may not be generalisable beyond specialist psychiatric settings (see further information on studies for more details)
Not significant

RCT
232 people in primary-care, meeting DSM-IV criteria for panic disorder, with or without co-morbid mental and physical disorders Proportion of people achieving remission of panic disorder (no panic attacks in the past month, minimal anticipatory anxiety about panic and agoraphobia subscale score of 10 or lower) 12 months
29% with CBT plus pharmacotherapy
16% with usual care
Absolute numbers not reported

P <0.001
See further information for details of subgroup analysis based on burden of chronic medical illness at baseline
Effect size not calculatedCBT plus pharmacotherapy

Quality of life

No data from the following reference on this outcome.

Adverse effects

Ref (type)PopulationOutcome, InterventionsResults and statistical analysisEffect sizeFavours
Withdrawal

Systematic review
262 people
3 RCTs in this analysis
Proportion of people who withdrew because of adverse effects 2 to 4 months
11/123 (9%) with CBT plus antidepressant
17/139 (12%) with antidepressants alone

RR 0.81
95% CI 0.25 to 2.68
The review noted that the results may not be generalisable beyond specialist psychiatric settings (see further information on studies for more details)
Not significant

No data from the following reference on this outcome.

Further information on studies

The systematic review (search date 2005) compared psychotherapy plus antidepressant treatment versus antidepressant treatment alone in people with panic disorder with or without agoraphobia, and reported short- and long-term outcomes. The review only included RCTs. It included any type of psychotherapy (including 9 RCTs of CBT consisting of both behavioural and cognitive therapy elements, 12 RCTs of behavioural therapy [including exposure and/or breathing retraining and/or relaxation], and two RCTs of "psychodynamic and others"). It presented pooled results for all psychotherapies, but also presented a subgroup analysis for RCTs of CBT alone which we have reported here. The review included RCTs in which benzodiazepines were used irregularly. The review noted that the generalisability of the findings beyond specialist psychiatric settings was not, straightforward as only one RCT was done in a primary-care setting. It also noted the comparability of treatment arms after the acute phase or continuation phase may have been compromised, as people were free to have other treatments before the final follow-up assessment, and 30% to 77% of people did so.

The subgroup analysis compared outcomes in people above (125 people) versus below (107 people) the median for burden of chronic medical illness at baseline, as assessed by self-reported chronic illness and prescribing data. Those above the median for medical illness were more likely to be older, female, and poorer, and had significantly more psychiatric morbidity at baseline. Both morbidity groups responded to combined CBT plus psychotherapy to a similar extent, but the higher baseline scores in the higher morbidity group meant that this group had higher levels of residual symptoms after treatment. The authors concluded that combined CBT plus pharmacotherapy for panic disorder worked equally well regardless of medical illness co-morbidity, but suggested that more intensive treatment may be required for people with co-morbid medical illnesses.

Comment

None.

Substantive changes

CBT plus drug treatments versus drugs alone One systematic review (search date 2005) added which includes a meta-analysis for short- and long-term results. Benefits and harms data enhanced. Categorisation changed. CBT plus antidepressants categorised as Likely to be beneficial compared with antidepressants alone as combination treatment may be more effective.


Articles from BMJ Clinical Evidence are provided here courtesy of BMJ Publishing Group