Review Article

Cognitive-Behavioral Therapy for Panic

Disorder: A Review ofTreatment Elements,
Strategies, and Outcomes
By Nadine Recker Rayburn, MA, and Michael W. Otto, PhD

FOCUS POINTS model for the disorder, corresponding interventions, and the
• Discusses the cognitive-behavioral conceptualization of range of applications for CBT for panic disorder, including
panic disorder. initial intervention, as a strategy for pharmacotherapy nonre-
• Summarizes the most important cognitive-behavioral sponders, and as a replacement strategy for patients who
therapy (CBT) intervention strategies for panic disorder. wish to discontinue their medication. In addition, we will
• Provides a review of the treatment-outcome literature discuss some of the complex issues that underlie combined
regarding CBT for panic disorder. pharmacologic and cognitive-behavioral treatment strategies.
• Discusses some of the complex issues that underlie com-
bined pharmacologic and CBT strategies. A COGNITIVE-BEHAVIORAL MODEL
OF PANIC DISORDER
ABSTRACT According to cognitive-behavioral models of the disor-
This article provides an overview of cognitive-behavioral der,1 5 panic attacks represent intense anxiety responses that
therapy (CBT) for panic disorder. CBT is currently considered are cued by subjective rather than objective danger. The
a first-line treatment for panic disorder. It offers benefit after subjective danger is defined by fears of panic sensations
short-term intervention, typically consisting of 12—15 sessions and their perceived consequences, including fears of
conducted in either an individual or a group format. The treat- embarrassment ("everyone will notice"), loss of control ("I
ment focuses on the elimination of the patterns that underlie will not be able to continue driving"), and death ("what if I
and perpetuate the disorder. Through CBT, patients learn have a heart attack") associated with panic episodes. Fears
about the nature of the disorder and acquire a set of strategies of future panic attacks help ensure vigilance to feared sen-
that counter the fears ofpanic attacks themselves, and break sations as well as a high level of anticipatory anxiety. Once
the recurring cycle of anticipatory anxiety, panic, and agora- symptoms are detected, they are misinterpreted within the
phobic avoidance. The collaborativeformat of treatment, and a context of fears of these sensations, expectations of cata-
focus on elimination of core fears may be factors in enhancing strophic outcomes,6 as well as memories of past attacks; the
longer-term outcome. In this article, we review the efficacy of result is a cascade of rising anxiety, apprehension, and
CBT as a first-line treatment, a strategyfor medication nonre- panic characterizing a self-perpetuating cycle (Figure).
sponders, a replacement strategy for patients who wish to dis- Avoidance of situations where panic attacks are
continue pharmacotherapy, and a potential preventive strategy feared—elevators, subway trains, grocery stores, etc.—
for at-risk individuals. We also discuss some of the complex typifies the agoraphobic subtype, but also may include
issues involved with combination-treatment strategies. avoidance of internal stimuli that are reminiscent of
CNS Spectr 2003;8(5):356-362 feared sensations of anxiety and panic.7'' For example,
patients with panic disorder may stop exercising because
INTRODUCTION they experience the physiological arousal as too frighten-
Over the past 2 decades, cognitive-behavioral interven- ing due to its resemblance to a panic attack. One conse-
tions for the anxiety disorders have become increasingly quence of avoidance is that it prevents individuals from
specialized. Cognitive-behavioral therapies (CBTs) for panic re-establishing a sense of safety in feared situations.
disorder reflect this specialization, with a honing of treat- Cognitive-behavioral models differ on the degree to
ment to focus on the fears of anxiety sensations that which fears of anxiety sensations are assumed to be learned
are thought to enhance vulnerability to and maintain panic as a result of panic attacks. Nevertheless, there is compelling
disorder. In this paper, we will discuss a cognitive-behavioral evidence that these fears occur in individuals without a
Ms. Rayburn is clincal psychology fellow in the Cognitive-Behavior Program in the Department of Psychology at Massachusetts General Hospital and
Harvard Medical School in Boston. Dr. Otto is associate professor of psychology and director of the Cognitive-Behavior Therapy Program in the
Department of Psychology at Massachusetts General Hospital and Harvard Medical School.
Disclosures: This work was supported by grants from GlaxoSmithKline, Pfizer Inc., and Eli Lilly and Company. Dr. Otto is a consultant from Pfizer Inc.,
Janssen Pharmaceutica, and Wyeth. Ms. Rayburn has no other ajfliations.
Please direct all correspondence to: Michael Otto, PhD, Cognitive-Behavior Therapy Program, Massachusetts General Hospital, 55 Fruit St. Boston, MA
02114; Tel: 617-726-2714, Fax: 617-726-7541; E-mail: motto@partners.org.
Downloaded from https:/www.cambridge.org/core. Centre Universitaire, on 30 Apr 2017 at 19:53:58, subject to the Cambridge Core terms of use, available at
Volume 8-Number 5
https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S1092852900018617 356 CNS Spectrums - May 2003

history of panic. There is also are a vulnerability factor for
the development of panic attacks or the reemergence of
panic disorder."1 This evidence ranges from longitudinal
studies of large cohorts undergoing stress to studies of the
biological provocation of panic. Fears of anxiety sensations
serve as an excellent predictor of panic provocation, includ-
ing provocation rates in individuals without a history of panic
disorder. Moreover, effective treatment leads to significant
reductions in fears of anxiety sensations, and residual eleva-
tions in these fears are associated with risk of relapse."12
Attention to fears of anxiety sensation as the core fears
maintaining panic disorder has been important for the
development of current treatment packages for panic disor-
der and the emergence of panic disorder prevention pro-
grams. Panic prevention is a noteworthy test of this model of
the etiology of the disorder. For example, using the presence
of at least moderate fears of anxiety sensations (ie, anxiety
sensitivity) and occasional unexpected panic attacks as a
marker of risk, Gardenswartz and Craske 1 ' examined the
efficacy of a 5-hour workshop to prevent the onset of panic
disorder. Elements of treatment were similar to those in full
CBT programs and included education about the nature and
etiology of panic and agoraphobia, cognitive restructuring,
exposure to feared somatic sensations (interoceptive expo-
sure), and instructions for in vivo exposure to avoided situa-
tions. Six-month follow-up data was available for
121 participants. Consistent with rates for other studies of
at-risk individuals, 13.6% of the wait-list group developed
panic disorder relative to only 1.8% of those attending the
prevention workshop. Given the emotional, social, and eco-
nomic costs of panic disorder and its treatment, further
investigation of such preventive strategies is encouraged.
Preventive and treatment programs are increasingly turn-
ing attention away from symptom management skills, such
as relaxation training or diaphragmatic breathing retraining,
in favor of a focus on eliminating fears of anxiety sensations
with a combination of cognitive restructuring and exposure
interventions. This shift draws attention to what may be a
fundamental difference between the elimination of fears and
the use of anxiety coping strategies; the latter may actually
reduce the efficacy of exposure-based procedures when
included in treatment packages.1415 Accordingly, the model
of treatment described in the next section focuses most
strongly on fear elimination rather than anxiety coping.
COMPONENTS OF COGNITIVE-BEHAVIORAL
THERAPY FOR PANIC DISORDER
CBT for panic disorder consists of a combination of psy-
choeducational, exposure, and cognitive interventions.
These components are typically delivered in 12—15 sessions
conducted in either individual or group formats.1019
Psychoeducation
At the beginning of treatment, therapists provide patients
with a strong informational background about the nature of
the disorder from a cognitive-behavioral perspective.
Downloaded from https:/www.cambridge.org/core. Centre Universitaire, on 30 Apr 2017 at 19:53:58, subject to the Cambridge Core terms of use, available at
Volume 8 - Number 5
https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S1092852900018617
Review Article
The purpose of these interventions is two-fold. First, the
information helps break the experience of panic attacks into
identifiable elements and provides a rationale for the inter-
ventions to follow. Second, informational interventions are
used to help the patient become a more active collaborator
in treatment. Active collaboration is essential because all
interventions are to be applied, ultimately, in the absence of
the therapist. Indeed, relapse prevention efforts often rely
on patients being able to apply elements of treatment in the
future should new problems be encountered.2"-2'
Information interventions commonly include a discussion
of the role of thoughts and avoidance patterns in helping cue
and maintain panic episodes (Figure), and review of the phys-
iologic source of panic symptoms. To try to enhance indepen-
dent learning in CBT, informational interventions may
include home reading assignments"'1719 that further explain
the CBT conceptualization and treatment components.
Cognitive Interventions
Cognitive interventions aim to correct patients' thoughts
and beliefs about the meaning and consequences of the
somatic symptoms of panic and anxiety. Common targets
include both the tendency to overestimate the probability of
negative outcomes (eg, "I will faint") as well as the degree of
catastrophe of these outcomes (eg, "I will not be able to
stand it"). Consistent with the application of cognitive inter-
ventions to other disorders, 2224 a variety of strategies are
used to help patients recognize the content of their thinking,
evaluate its accuracy, and guide themselves more effec-
tively. As a central feature to these interventions, patients
are asked to treat thoughts as hypotheses, and to evaluate
the validity of these hypotheses. Socratic questioning (ie,
open-ended questions guiding self-discovery) may be
adopted to help patients learn from their own history by
becoming better at attending to actual outcomes relative to
their catastrophic expectations. In reviewing past evidence,
for example, patients may realize that their catastrophic
Stress
Biological Diathesis
Alarm Reaction
Rapid heart rate, heart palpitations,
shortness of breath, smothering sensations,
chest pain or discomfort, numbness or tingling
Conditioned Catastrophic
Increased anxiety
fear of misinterpretations
and fear
somatic sensations of symptoms
Hypervigilance to symptoms
Anticipatory anxiety
Memory of past attacks
FIGURE. A COGNITIVE-BEHAVIORAL MODEL OF
PANIC DISORDER.
Adapted from Otto MW, Pollack MH, Meltzer-Brody S, Rosenbaum JF.
Cognitive-behavioral therapy for benzodiazepine discontinuation in panic
disorder patients. Psychopharm Bull. 1992; 28 123-130.
Rayburn NR n Otto MW. CNSSpectr. Vol 8, No 5. 2003.
357 CNS Spectrums - May 2003
 Review Article
predictions have never occurred (eg, they realize that they
have never fainted on the subway).
Regular thought monitoring in response to situational,
emotional, and physiologic cues may also be used to help
patients become better observers of their current thoughts,
and to evaluate systematically the objective evidence for
and against the veracity of these thoughts. For example,
patients may be asked to note the situation (eg, being in a
crowded elevator), the emotional and physiologic sensations
(eg, anxiety-increased heart rate, sweating, dizziness), and
the cognitions ("I am going to have a heart attack," "I am
going crazy"). Patients then learn to challenge these cata-
strophic thoughts and adopt alternative and more accurate
cognitive responses (eg, "I have never fainted in these situa-
tions in the past and chances are that I am not going to faint
now"). More active reality testing may also be adopted in
the context of behavioral experiments—a cross between
cognitive and exposure interventions where patients are
asked to systematically test whether feared outcomes occur
in the context of an exposure assignment (eg, going to the
subway station and checking out the prediction that one is
going to faint). Based on the evidence obtained, patients are
asked to form more accurate thoughts. This kind of "collab-
orative empiricism" is the hallmark of cognitive interven-
tions,25 and may be complemented stylistically by a variety
of rich metaphors or stories that are designed to aid reten-
tion and subsequent application of session material.26
Exposure Interventions
Systematic exposure exercises are typically offered in
conjunction with cognitive techniques. Interoceptive expo-
sure techniques target patients' fears of their own physio-
logic anxiety sensations. With the help of specific
procedures (eg, hyperventilation to induce dizziness, paras-
thesias, flushes, etc.), patients systematically confront
feared somatic sensations.6'19 For example, patients who are
afraid of being dizzy because they believe they will faint are
asked to hyperventilate for one minute to deliberately
induce dizziness. This exposure allows them to test cata-
strophic fears about these sensations (ie, they learn that they
will not lose consciousness), and also provides them with an
opportunity to rehearse alternative responses. Central to
these alternative responses is learning to simply "note the
sensations" while doing nothing to control or avoid them.
The result is that patients are provided with an opportunity
to examine the outcome of the symptoms apart from their
catastrophic expectations and (often anxiogenic) attempts to
control the symptoms. With repeated exposure, patients find
that they no longer fear the symptoms; and consequently the
symptoms lose their ability to elicit panic episodes.
In addition, patients engage in situational (in vivo) expo-
sure exercises. These interventions are focused on breaking
and modifying patterns of agoraphobic avoidance. In the
early stages of the treatment, the patient and therapist
together establish a hierarchy of situations the patient
avoids due to the situation's ability to elicit panic. Common
Downloaded from https:/www.cambridge.org/core. Centre Universitaire, on 30 Apr 2017 at 19:53:58, subject to the Cambridge Core terms of use, available at
Volume 8 - Number 5
https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S1092852900018617
examples of in vivo exposures are riding in an elevator,
going to the grocery store, standing in a crowd of people, or
riding on a subway train. Throughout the treatment, patients
confront these situations in a stepwise manner, starting with
moderately anxiety-provoking situations. The majority of the
in vivo exposures are typically performed as homework
assignments, however, it is also possible to engage in some
of them in the context of therapy sessions. Again, in situa-
tional exposures patients are encouraged to resist the urge
to leave when they start feeling anxious. Instead, they are
instructed to stay in the feared situation until the fear sub-
sides. That way, patients gradually overcome their agora-
phobic avoidance and relearn that the once-feared places
and circumstances are safe.
Relapse Prevention
CBT for panic disorder also includes a relapse prevention
component. The last two or three therapy sessions are devoted
to relapse prevention. The therapist encourages the patient to
identify potential problem areas that he or she may encounter
and think of possible solutions based on knowledge gained
from therapy (eg, to engage in interoceptive exposures in case
panic symptoms return and to implement in vivo exposures if
agoraphobic avoidance recurs). During these final sessions,
therapist and patient may also establish a written relapse pre-
vention contract.20 This contract includes an outline of spe-
cific skills learned in therapy and concrete plans for using
these skills under certain circumstances. A formal contract
may be helpful in terms of motivating patients to apply cogni-
tive-behavioral tools in the future.
OUTCOME FINDINGS
There is a wealth of data documenting the efficacy of
CBT for the treatment of panic disorder both with and with-
out agoraphobia.27-28 Panic-free rates following short-term
treatment often range from 74% to 85%, with evidence that
similar rates can be achieved when this treatment is trans-
ported from research clinics to community mental health
centers.29 However, panic-free rates give an overly opti-
mistic view of treatment response. It is clear that attainment
of remission status (high endstate functioning) occurs at a
lower rate (50% to 70%) than panic cessation. Nonetheless,
CBT results in some of the most promising outcomes in the
literature, especially with regard to the reduction of agora-
phobic avoidance.
Treatment Outcomes of Cognitive-behavioral
Therapy Versus Pharmacotherapy
Meta-analytic review of the panic treatment-outcome lit-
erature indicates that treatment effect sizes for CBT are
equal to or surpass those for antidepressant or benzodi-
azepine treatments.272S i() This conclusion is also supported
by many comparative trials," :i! including the recent, four-
site, comparison of CBT and imipramine treatment.1*
Although these studies suggest a subtle edge for CBT over
pharmacologic alternatives, a conservative conclusion is
358 CNS Spectrums - May 2003

that approximately equal outcome is achieved by these two
modalities of treatment during the acute treatment phase.
However, efficacy estimates for the two modalities of
treatment differ dramatically when longer-term outcome is
considered. It is generally accepted that to maintain treat-
ment gains from pharmacotherapy, medications need to be
continued over the long term.3' Indeed, when medications
are discontinued after acute treatment, relapse rates in the
range of 54% to 70% are common.36 39 Even with continua-
tion treatment, there is some evidence that ongoing slippage
in treatment gains may occur regardless of the use of single
agent or combined pharmacologic strategies.40
In contrast, there is good evidence that treatment gains
from CBT tend to be maintained over time, with up to 81% of
patients remaining panic free over follow-up periods of
1-2 years.1541 This supports the idea that patients may be able
to fundamentally shift fears maintaining the disorder.
Nonetheless, it is important to remember whereas there is
strong overall evidence of continued benefit in cross-sectional
studies, many panic patients continue to have a waxing and
waning course of symptoms, albeit at a lower level of severity.42
Cost-efficacy and Acceptability
of Cognitive-behavioral Therapy
The apparent ability of CBT to confer longer-term advan-
tages with brief treatment is a central strength of the
approach for efficacy estimates. Certainly, this perspective
is supported by the overall performance of patients as
assessed in meta-analytic studies.2" It is also supported by
direct analysis of costs of and outcome in a specialty anxiety
clinic offering treatment with either CBT or psychopharma-
cology specialists.13 Consistent with the controlled treat-
ment-outcome literature, Otto and colleagues43 found that
clinic treatment by these specialists resulted in approxi-
mately equal outcome. Examination of the costs of the treat-
ments indicated that individual CBT was somewhat more
costly than pharmacotherapy during the acute 4-month
treatment phase, but then quickly makes up its initial costs
over follow-up, so that by 1 year it was found to be roughly
half as costly as pharmacotherapy. Group treatment pro-
vided a more extreme example, with cost savings relative to
pharmacotherapy starting in the acute treatment phase and
extending forward, so that it was less than 25% of the costs
of pharmacotherapy by the end of the 1-year study.
Given that both CBT and pharmacotherapy can offer
promising treatment outcome (albeit with different profiles
of maintenance strategies), on what basis should one modal-
ity of treatment be selected over the other? So far, there is
an absence of data indicating how patients should be
matched to either treatment modality based on characteris-
tics of their symptom profile. Moreover, predictors of poor
outcome appear to be similar for CBT and psychopharma-
cology. Greater baseline panic and agoraphobic avoidance,
depressed mood, personality psychopathology, and marital
dissatisfactions, as well as lower motivation for treatment
have all been linked to poorer treatment outcome.44"17 In the
Downloaded from https:/www.cambridge.org/core. Centre Universitaire, on 30 Apr 2017 at 19:53:58, subject to the Cambridge Core terms of use, available at
Volume 8 - Number 5
https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S1092852900018617
Review Article
absence of any empirically validated matching criteria, what
other considerations should guide the decision process?
In a study of treatment preference and acceptability,
Hofmann and colleagues48 examined reasons potential par-
ticipants refused randomization in a large multicenter trial
at sites known for their specialization in CBT or pharma-
cotherapy. Whereas 34% refused participation due to con-
cerns over imipramine treatment, <1% refused participation
due to concerns about CBT. Examination of treatment pref-
erence among clinic patients also supports the notion that
CBT compares well with pharmacotherapy for treatment
preference, with evidence of approximately equal rates of
preference of CBT and pharmacotherapy.3''
CBT also appears to be an especially tolerable treatment,
at least as estimated by dropout rates in clinical trials.
In Gould and colleagues'28 meta-analytic review of well-
controlled treatment trials, CBT had an average 5.6% dropout
rate, compared with 13.1% for benzodiazepine treatment, and
25.4% for antidepressant treatments (excluding selective
serotonin reuptake inhibitor [SSRI] treatment). Subsequent
examination of dropout rates in SSRI trials suggested that, at
least in the acute treatment phase, dropout rates (19.9%)
were not substantially different from older agents.3"
Part of the acceptability of CBT may come from its rela-
tively quick onset of action, well within the time frame of
antidepressant treatments for panic. For example, studies of
CBT indicate improvements as early as the second session,
with evidence of incremental improvement thereafter.1"
Cognitive-behavioral Therapy for Pharmacotherapy
Nonresponders and Medication Discontinuation
CBT also appears to be an effective strategy when nonre-
sponse to pharmacotherapy is encountered. Examination of
outcome for patients who are either partial responders or
nonresponders to an adequate dose and duration of pharma-
cotherapy indicates that brief CBT offers an efficacious
alternative.50"'3 Moreover, there is good evidence that
patients can use CBT as a strategy to discontinue their phar-
macotherapy while maintaining or extending treatment
gains. Studies in support of this application originally
focused on the difficult job of helping patients discontinue
their benzodiazepine medication. Three studies indicate
that brief CBT, typically offered during and after the taper
process, helps patients tolerate benzodiazepine withdrawal
while further treating panic disorder and extending treat-
ment gains.53"55 Furthermore, preliminary evidence from an
ongoing, two-site study of benzodiazepine discontinuation
indicates that nonspecific effects of treatment cannot
account for the success of the CBT program.'6 Specifically,
manualized CBT for benzodiazepine discontinuation19
appears to offer a significant advantage over an adjunctive
treatment program offering relaxation skills alone.
Examination of the nature of symptom changes across ben-
zodiazepine discontinuation, suggests that the ability of
CBT to further reduce fears of anxiety sensations is one
important element of treatment.57
359 CNS Spectrums - May 2003
 Review Article
More recent evidence has also emerged documenting the
value of CBT for SSRI discontinuation. In a case series,
Whittal and colleagues58 documented successful SSRI dis-
continuation in the context of improved clinical outcome
with a brief program of group CBT, with maintenance of
improved outcomes at 3-month follow-up. In a second study,
Schmidt and colleagues59 randomly assigned patients with
panic disorder to either continue or discontinue their SSRI
treatment in the context of group CBT. Both groups of
patients improved dramatically, with over 75% of patients
meeting criteria for high endstate functioning, without sig-
nificant differences in the magnitude of improvement
between groups. Together, these early studies suggest that
the success utilizing CBT for medication discontinuation in
panic disorder can be extended to SSRIs, and underscores
the important role CBT can play in replacing medication
treatment, particularly in individuals who have not fully
responded to pharmacotherapy.
COMBINATION TREATMENT STRATEGIES
Thus far, we have discussed the issue of combination
treatment primarily from the perspective of sequential treat-
ment: adding CBT as an adjunctive or replacement strategy
for patients already on medications. What about the issue of
regular combined treatment strategies?
Considering the evidence that most patients will enjoy acute
improvement with either CBT or pharmacotherapy, it is tempt-
ing to consider combination treatment as a strategy to boost
efficacy. Although, there is some evidence for synergistic
effects, this relationship appears to be especially complex.
There is good evidence that CBT can extend the gains from
pharmacotherapy, when added sequentially or as an initial
strategy,5"'52"' and there is limited evidence that pharmacother-
apy holds promise for CBT nonresponders.61 However, for the
addition of pharmacotherapy to CBT, there are some indica-
tions that any additive effects achieved during the acute treat-
ment period may come at a cost to longer-term outcome. Two
large multicenter trials""'4 provide evidence of this cost; each
suggests that some of the benefits of CBT provided during med-
ication treatment may be lost when medication is discontinued.
The first investigation examined33 the relative and com-
bined efficacy of exposure therapy and treatment with alpra-
zolam. There was evidence of marginal gains in efficacy with
combined treatment, but this treatment strategy had poorer
longer-term outcome. After discontinuation of the alprazolam
the combined treatment group fared less well than the expo-
sure alone treatment condition. The second study34 examined
the individual and combined efficacy of CBT and
imipramine at each of four treatment centers,34 utilizing five
different treatment conditions: CBT alone, imipramine alone,
placebo alone, CBT plus imipramine, and CBT plus placebo.
The last condition was especially important since it allowed
the assessment of the nonspecific effects of pill taking in
combination with an active treatment like CBT.
Consistent with prior research, this study28 found that both
CBT and imipramine alone were superior to placebo alone.
Downloaded from https:/www.cambridge.org/core. Centre Universitaire, on 30 Apr 2017 at 19:53:58, subject to the Cambridge Core terms of use, available at
Volume 8 - Number 5
https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S1092852900018617
Patients who received CBT or imipramine had responded
equally well to the treatments at the end of the acute trial. In
the short-term, combined treatment with CBT and
imipramine turned out to be superior to either monotherapy
on several measures. However, it is important to note that
this combination treatment did not outperform CBT plus
placebo. This suggests that nonspecific effects associated
with pill-taking account for many of the advantages of adding
imipramine to CBT. Moreover, when patients were re-
assessed at long-term follow-up (after discontinuation of
medication use), CBT alone and CBT plus placebo both
showed an advantage over combined treatment, suggesting
that the addition of imipramine to CBT may have under-
mined the long-term stability of treatment gains from CBT.
These results are consistent with findings from studies of
other anxiety disorders,62 and also consistent with data from
the animal laboratory, indicating that extinction of fear
learning may be strongly dependent on context. Bouton63
details findings from his animal laboratory indicating that
changes in internal state, such as anxiety reduction from a
benzodiazepine, may be a powerful enough context such
that adequate safety learning from exposure (extinction) is
achieved only in that context. When the drug state is with-
drawn, so is the learned safety. This principle is also consis-
tent with the maintenance of treatment gains seen with
medication discontinuation programs utilizing CBT. In these
programs, CBT is offered during the process of medication
taper, allowing patients to learn safety in response to cues
present during and after the withdrawal of medication
effects. This timing of treatment may provide the crucial
experience in a medication-free state that may be required
to help maintain fear-extinction effects from exposure.
In addition, completion of CBT in the absence of medica-
tion may also provide patients with more complete learning
of safety, especially in relation to emotional cues of anxiety.
When CBT is conducted without medications, patients have
the experience of completing exposure and cognitive
restructuring in the context of anxiety, at least in the initial
phase of treatment. Anticipatory anxiety and occasional
panic episodes become part of the context of exposure treat-
ment. As documented by Bouton,03 inclusion of occasional
aversive outcomes (in this example, a panic attack) during
exposure (extinction trials) may confer greater resistance to
relapse, assuming that at some point in the future the
patient will again confront contextual cues of anxiety. This
effect may help explain indications of deleterious effects of
medication on the maintenance of treatment gains in CBT,
even when the medications are continued.''4 This may be
manifested clinically in the difference between a recovered
patient who can respond with resilience to a subsequent
episode of anxiety compared with the patient who responds
with a dramatic increase in fear, reporting that "it feels like
I have gone back to square one.65"
Given these data, it appears that any routine application
of combination treatment should proceed cautiously. Indeed,
from the perspective of long-term cost-benefit, it appears
360 CNS Spectrums - May 2003

that, when CBT is available, it should be considered as a
monotherapy for patients who are new to treatment.
INTEGRATING COGNITIVE-BEHAVIORAL
THERAPY ELEMENTS INTO STANDARD
PHARMACOTHERAPY PRACTICE
In the context of the many strengths of CBT for panic dis-
order, perhaps the greatest limitation is the availability of
CBT providers. That is, many communities and clinics may
not have access to state-of-the-art CBT providers.
Accordingly, pharmacotherapy applied alone or in conjunc-
tion with dynamic treatment will continue to be a common
treatment."' To try to enhance gains from pharmacotherapy,
several elements of CBT may be added to standard pharma-
cologic care. Upon entering treatment, we recommend that
all patients receive education about the role of negative
thoughts, fears of anxiety sensations, and avoidance pat-
terns in maintaining panic disorder. There is nothing about
this model that is inconsistent with the use of medications to
help control anticipatory anxiety, negative affect, and panic.
This information may offer patients immediate benefit,67 and
it provides a model for the adjunctive or subsequent addi-
tion of CBT strategies.
In particular, stepwise exposure to avoided situations also
presents an uncomplicated addition to pharmacotherapy. As
noted, controlled trials indicate that exposure extends treat-
ment gains from pharmacotherapy.5<)'60"8'69 In addition, simple
exposure assignments, relative to assurance alone, are pre-
dictive of better outcomes for patients with panic disorder
presenting to hospital emergency departments.70
Exposure treatment should begin with a discussion of the
range of situations the patient avoids. Then, therapists
should encourage patients to confront these avoided situa-
tions in a gradual manner, starting out with moderately diffi-
cult situations. Patients should remain in the situations until
their anxiety decreases. In addition to instructing patients in
the role of negative thoughts during exposure, clinicians
may also consider utilizing self-help books to help patients
expand their independent role in treatment. Similarly, many
therapy manuals are available that can aid therapists with
integrating elements of CBT into their practice.1617 For clini-
cians considering referral to a CBT specialist, professional
associations, such as the Association for the Advancement
of Behavior Therapy (www.aabt.org) or the Anxiety
Disorders Association of American (www.adaa.org), can
help identify CBT specialists in a given area.
CONCLUSION
In summary, in the past 2 decades, a great deal of
research has emerged in support of the efficacy of CBT for
panic disorder. Treatment packages commonly include psy-
choeducation, cognitive restructuring, exposure to feared
somatic sensations (interoceptive exposures), and situa-
tional (in vivo) exposures to reduce agoraphobic avoidance.
Evidence suggests that in terms of long-term outcome, CBT
applied as a monotherapy appears to result in more stable
Downloaded from https:/www.cambridge.org/core. Centre Universitaire, on 30 Apr 2017 at 19:53:58, subject to the Cambridge Core terms of use, available at
Volume 8 - Number 5
https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S1092852900018617
Review Article
gains than treatment combining CBT and medication. CBT
offers beneficial outcome when applied during any of a
number of phases of the disorder. It should be considered as
an effective first-line treatment for the disorder, and for
patients who have not responded to, or who have exhibited
only partial improvement with pharmacotherapy. CBT can
also be of particular benefit during the medication discon-
tinuation phase. Finally, brief CBT interventions have
shown promise in stemming the tide of new cases of panic
disorder in vulnerable individuals. ISfiEl
REFERENCES
1. Harvey JM, Richards JC, Dziadosz T, Swindell A. Misinterpretation of ambigu-
ous stimuli in panic disorder. Cogn Ther Res. 1993;17:235-248.
2. McNally RJ, Foa EB. Cognition and agoraphobia: bias in the interpretation of
threat. Cogn Ther Res. 1987;11:567-581.
3. Barlow DH. Anxiety and Its Disorders: The Nature and Treatment ofAnxiety and
Panic. New York, NY: Guilford Press; 2001.
4. Clark DM. A cognitive approach to panic. Behav Res Ther. 1986;24:461-470.
5. McNally RJ. Psychological approaches to panic disorder: a review. Psychol
Bull. 1990;108:403-419.
6. Barlow DH, Craske MG. Mastery of Your Anxiety and Panic. Albany, NY:
Graywind; 1989.
7. Salkovskis PM, Clark DM, Gelder MG. Cognitive-behavior links in the persis-
tence of panic. Behav Res Ther. 1996;34:453-458.
8. Salkovskis PM. The importance of behavior in the maintenance of anxiety and
panic: a cognitive account. Behav Psychother. 1991;19:6-19.
9. Salkovskis PM. Avoidance behavior is motivated by threat beliefs: a possible
resolution of the cognitive-behavior debate. In: Salkovskis PM, ed. Key Trends
in Cognitive-Behavior Therapy. New York, NY: Wiley; 1996.
10. McNally RH. Anxiety sensitivity and panic disorder. Bid Psych. In press.
11. Otto MW, Reilly-Harrington N. The impact of treatment on anxiety sensitivity.
In: Taylor S, ed: Anxiety Sensitivity: Theory, Research and Treatment of the Fear
ofAnxiety. Mahwah, NJ: Lawrence Erlbaum; 1999.
12. Hoffart, A. Cognitive and guided mastery therapy of agoraphobia: long-term
outcome and mechanisms of change. Cogn Ther Res. 1998;22:195-207.
13. Gardenschwartz CA, Craske MG. Prevention of panic disorder. Behav Ther.
2001;32:725-737.
14. 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.
15. Craske MG, Brown TA, Barlow DH. Behavioral treatment of panic: a two year
follow-up. Behav Ther. 1991;22:289-304.
16. Craske MG, Meadows E, Barlow DH. Mastery of Your Anxiety and Panic and
Agoraphobia Supplement (Therapist Guide). 2nd ed. San Antonio, TX:
Psychological Corporation; 1994.
17. Margraf J, Schneider S. Panik: Angstanfaelle und Hire Behandlung. Berlin,
Germany: Springer; 1990.
18. Otto MW, Jones JC, Craske MG, Barlow DH. Stopping Anxiety Medication:
Panic Control Therapy for Bezodiazepine Discontinuation (Therapist Guide).
San Antonio, TX: Psychological Corporation; 1996.
19. Otto MW, Pollack MH, Barlow DH. Stopping Anxiety Medication: A Workbook
for Patients Wanting to Discontinue Benzodiazepine Treatment for Panic
Disorder. Albany, NY: Graywind Publications; 1995.
20. Ost LG. A maintenance program for behavioural treatment of anxiety disor-
ders. Behav Res Ther. 1980;27:123-130.
21. Jansson L, Jerremalm A, Ost LG. Follow-up of agoraphobia patients treated with
exposure, in vivo, or applied relaxation. Br J Psychiatry. 1986;149:486-490.
22. Beck AT. Cognitive Therapy and the Emotional Disorders. New York, NY:
International University Press; 1976.
23. Ellis A. Changing rational-emotive therapy (RET) to rational emotional behav-
ior therapy (REBT). Behav Ther. 1993;16:257-258.
361 CNS Spectrums - May 2003
 Review Article
24. Ellis A. Fundamentals of rational-emotive therapy forthe 1990s. In: Dryden W, Hill
L, eds. Innovations in Rational-Emotive Therapy. Newbury Park, CA: Sage; 1993. ative treatment outcome study on panic disorder. Am J Psychiatry.
25. Young JE, Weinberger AD, Beck AT. Cognitive therapy for depression. In:
Barlow DH, ed. Clinical Handbook of Psychological Disorders. 3rd ed. New
York, NY: Guilford Press; 2001.
26. Otto MW. Stories and metaphors in cognitive-behavior therapy. Cogn Behav
Pract. 2000;7:166-172.
27. Clum GA, Clum G, Surls R. A meta-analysis of treatments for panic disorder.
J Consult ClinPsychol. 1993;61:317-326.
28. Gould RA, Otto MW, Pollack MH. A meta-analysis of treatment outcome for
panic disorder. ClinPsychol Rev. 1995;15:819-844.
29. Wade WA, Treat TA, Stuart GL. Transporting an empirically supported treat-
ment for panic disorder to a service clinic setting: a benchmarking strategy. J
Consult Clin Psychol 1998;66:231-239.
30. Otto MW, Tuby KS, Gould RA, McLean RYS, Pollack MH. An effect-size
analysis of the relative efficacy and tolerability of serotonin selective reuptake 53. Hegel MT, Ravaris CL, Ahles TA. Combined cognitive-behavioral and time-
inhibitors for panic disorder. Am J Psychiatry. 2001;158:1989-1992.
31. Clark DM, Salkovskis PM, Hackmann A, Middleton H, Pavlos A, Gelder M. A
comparison of cognitive therapy, applied relaxation and imipramine in the
treatment of panic disorder. Br J Psychiatry. 1994;164:759-769.
32. Klosko JS, Barlow DH, Tassinari RB, Cerny JA. Comparison of alprazolam
and behavior therapy in the treatment of panic disorder. J Consult Clin Psychol
1990;58:77-84.
33. Marks IM, Swinson RP, Basaglu 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.
34. Barlow DH, Gorman JM, Shear MK, Woods SW. Cognitive-behavioral therapy,
imipramine, or their combination for panic disorder. JAMA 2000;283:2529-2536.
35. American Psychiatric Association. Practice guidelines for the treatment of
patients with panic disorder. In: Practice Guidelines for the Treatment of
Psychiatric Disorders. Washington, DC: American Psychiatric Association; 1998.
36. Mavissikalian M, Perel JM. Clinical experience in maintenance and discon-
tinuation of imipramine therapy in panic disorder with agoraphobia. Arch Gen
Psychiatry. 1993;49:318-323.
37. Noyes R, Garvey MJ, Cook BL, Samuelson L. Problems with tricyclic antide-
pressant use in patients with panic disorder or agoraphobia: results of a natu-
ralistic follow-up study. J Clin Psychiatry. 1989;50:163-169.
38. Noyes R, Garvey MJ, Cook B, Suelzer M. Controlled discontinuation of benzo-
diazepine treatment for patients with panic disorder. Am J Psychiatry.
1991;148:517-523.
39. Pollack MH, Smoller JW. Pharmacologic approaches to treatment resistant
panic disorder. In: Pollack MH, Otto MW, eds. Challenges in Clinical Practice:
Pharmacological and Psychosocial Strategies. New York, NY: Guildford,
Press; 19%.
40. Simon NM, Safren SA, Otto MW, Sharma SG, Lanka GD, Pollack MH.
Longitudinal outcome with pharmacotherapy in a naturalistic study of panic
disorder. J Affect Disord. 2002; 69: 201-208.
41. Beck AT, Sokol L, Clark DA, Berchick R, Wright F. A cross-over study of focused
cognitive therapy for panic disorder. Am J Psychiatrty. 1992;149:778-783.
42. Brown TA, Barlow DH. Long-term outcome in cognitive-behavioral treatment
of panic disorder: Clinical predictors and alternative strategies for assessment.
J Consult Clin Psychol. 1992;63:754-765.
43. Otto MW, Pollack MH, Maki KM. Empirically supported treatments for panic dis-
order: costs, benefits, and stepped care. J Consult Clin Psychol. 2000;68:556-563.
44. Mavissakalian M. The relationship between panic disorder/agoraphobia and
personality disorders. Psychiatr Clin North Am. 1990;13:661-684.
45. Pollack MH, Otto MW, Sachs GS, et al. Anxiety and psychopathology predic-
tive of outcome in patients with panic disorder and depression treated with
imipramine, alprazolam and placebo. J Affect Disord. 1994;30:273-281.
46. Pollack MH, Otto MW, Tesar GE, Cohen LS, Meltzer-Brody S, Rosenbaum JF.
Long-term outcome after acute treatment with clonazepam and alprazolam for
panic disorder. J Clin Psychopharmacol. 1993;13:257-263.
47. Keijsers GPJ, Hoogduin CAL, Schaap CPDR. Prognostic factors in the behav-
ioral treatment of panic disorder with and without agoraphobia. Behav Ther.
1994;25:689-708.
Downloaded from https:/www.cambridge.org/core. Centre Universitaire, on 30 Apr 2017 at 19:53:58, subject to the Cambridge Core terms of use, available at
Volume 8 - Number 5
https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S1092852900018617 362
48.Hofmann SG, Barlow DH, Papp LA, et al. Pretreatment attrition in a compar-
1998;155:43-47.
49.
Penava SJ, Otto M W, Maki KM, Pollack MH. Rate of improvement during cog-
nitive-behavioral group treatment for panic disorder. Behav Res Ther.
1998;36:665-673.
50.Mavissakalian M. Sequential combination of imipramine and self-directed
exposure in the treatment of panic with agoraphobia. J Clin Psychiatry.
1990;51:184-188.
51. Otto MW, Pollack MH, Penava SJ, Zucker BG. Group cognitive-behavior ther-
apy for patients failing to respond to pharmacotherapy for panic disorder: a
clinical case series. Behav Res Ther. 1999;7:763-770.
52. Pollack MH, Otto MW, Kaspi SP, Hammerness PG, Rosenbaum JF. Cognitive-
behavior therapy for treatment-refractory panic disorder. J Clin Psychiatry.
1994;55:200-205.
limited alprazolam treatment of panic disorder. Behav Ther. ] 994;25:183-195.
54. Otto MW, Pollack MH, Sachs GS, Reiter SR, Meltzer-Brody S, Rosenbaum JF.
Discontinuation of benzodiazepine treatment: efficacy of cognitive-behavior
therapy for patients with panic disorder. Am J Psychiatry. 1993;150:1485-1490.
55. Spiegel DA, Bruce TA, Gregg SF, Nuzzarello A. Does cognitive behavior ther-
apy assist slow-taper alprazolam discontinuation in panic disorder? Am J
Psychiatry. 1994;151:876-881.
56. Otto MW, Barlow DH, Pollack MH, Spiegel DA. Further evaluation of panic
control therapy for benzodiazepine discontinuation. Paper presented at:
Annual Convention of the Association for the Advancement of Behavior
Therapy; November 11-14,1999; Toronto, Ontario.
57. Bruce TJ, Spiegel DA, Gregg SF, Nuzzarello A. Predictors of alprazolam dis-
continuation with and without cognitive behavior therapy in panic disorder.
Am J Psychiatry. 1995;152:1156-1160.
58.
Whittal ML, Otto M W, Hong JJ. Cognitive-behavior therapy for discontinua-
tion of SSRI treatment of panic disorder: a case series. Behav Res Ther.
2001;39:939-945.
59.Schmidt NB, Wollaway-Bickel K, Trakowski JH, Santiago HT, Vasey M.
Antidepressant discontinuation in the context of cognitive behavioral treat-
ment for panic disorder. Behav Res Ther. 2002;40:67-73.
60.Telch MJ, Lucas RA. Combined pharmacological and psychological treatment
of panic disorder; current status and future directions. In: Wolfe BE, Maser JD,
eds. Treatment of Panic Disorder. Washington, DC: American Psychiatric
Association Press; 1994.
61. Kampman M, Keijsers GP, Hoogduin CA, Hendriks GJ . A randomized, dou-
ble-blind, placebo-controlled study of the effects of adjunctive paroxetine in
panic disorder patients unsuccessfully treated with cognitive-behavioral ther-
apy alone. J Clin Psychiatry. 2002;63:772-777.
62. Foa EB, Franklin ME, Moser J. Context in the clinic: How well do CBT and
medications work in combination. Biol Psychiatry. In press.
63.Bouton ME. Context, ambiguity, and unlearning: Sources of relapse alter
behavioral extinction. Biol Psychiatry. In press
64.Otto MW, Pollack MH, Sabatino SA. Maintenance of remission following cog-
nitive behavior therapy for panic disorder: possible deleterious effects of con-
current medication treatment. Behav Ther. 1996;27:473-482.
Otto MW. Learning and "unlearning" fears: preparedness, neural pathways,
65.
and patients. Biol Psychiatry. In press.
66.
Goisman RM, Warshaw MG, Keller MB. Psychosocial treatment prescriptions
for generalized anxiety disorder, panic disorder, and social phobia 1991-19%.
Am J Psychiatry. 1999;156:1819-21.
67. Shear KS, Pilkonis PA, Cloitre M, Leon AC. Cognitive behavioral treatment
compared with nonprescriptive treatment of panic disorder. Arch Gen
Psychiatry. 1994;51:395-401.
68. Mavissakalian M, Michelson L Two-year follow-up of exposure in imipramine
treatment of agoraphobia. Am J Psychiatry. 1986;143:1106-1112.
Telch MJ, Agras WS, Taylor CB, Roth WT, Gallen C. Combined pharmacological
69.
and behavioral treatment for agoraphobia. Behav Res Ther. 1985;23:325-335.
70. Swinson RP, Soulios C, Cox BJ, Kuch K. Brief treatment of emergency room
patients with panic attacks. Am J Psychiatry. 1992; 149:944-946.
CNS Spectrums - May 2003