Richard G.

Heimberg

Current Status of Psychotherapeutic

Interventions for Social Phobia
Richard G. Heimberg, Ph.D.

Psychotherapeutic interventions, especially the cognitive-behavioral psychotherapies, have been

well studied as treatments for social phobia. The purposes of this article are to (1) enumerate and de-
©

scribe the varieties of cognitive-behavioral therapy (CBT) that have been applied to the treatment of
social phobia, (2) provide a meta-analytic overview of the efficacy of these approaches, (3) examine
Co

the relative utility of CBT versus that of pharmacotherapy for social phobia, (4) examine the potential
utility of multidisciplinary approaches to treatment, and (5) discuss possible future directions in the
py

development of psychotherapeutic strategies for the treatment of social phobia, including the use of
computers as adjunctive tools. (J Clin Psychiatry 2001;62[suppl 1]:36–42)
rig
ht

T wenty years have passed since social phobia became ders with an interpersonal component such as bulimia ner-
20

an official diagnostic category in the Third Edition vosa,6 no other studies have been conducted of IPT in so-
of the Diagnostic and Statistical Manual of Mental Disor- cial phobia. Similarly, although psychodynamic theorists7
01 ne p

ders.1 Over the years, we have accumulated a great deal of have offered accounts of the development of social pho-
O

knowledge about the nature of this disorder and how it can bia, no specific recommendations for treatment have been
Pherso

be treated. In this article, I shall attempt to summarize the put forth, and no studies of psychodynamic treatment for
ys nal c

current state of knowledge about the use of specific social phobia have yet been conducted. Given this state of
psychosocial interventions for the treatment of social pho- affairs, I shall focus my attention in the remainder of this
ic op

bia. Thereafter, I shall examine the evidence for the rela- article on the cognitive-behavioral therapies (CBTs) for
ia y m

tive efficacy of psychotherapeutic and pharmacotherapeu- social phobia.

ns ay

tic approaches to the treatment of social phobia as well as Rather than a single school of psychotherapy, CBT is
the potential utility of pharmacotherapy-psychotherapy best considered a collection of techniques held together by
Po be

combinations. Future directions for the study of psychoso- a series of philosophical, clinical, and empirical emphases.
cial interventions for social phobia will also be discussed. The CBTs are time-limited, present-oriented approaches to
stgprin

There are many varieties of psychotherapy, but few psychotherapy that attempt to teach patients the cognitive
ra ted

have been studied in controlled trials in patients with so- and behavioral competencies needed to function adaptively
cial phobia. For instance, a recent uncontrolled trial of in their interpersonal and intrapersonal worlds. Cognitive-
du

interpersonal psychotherapy (IPT),2 a time-limited ap- behavioral therapy is a collaborative effort between thera-
a

proach to psychotherapy based on the assumption that pist and patient, who form a working team to address the
te

psychiatric disorders occur and are maintained within a patient’s presenting concerns. The various CBTs are also
psychosocial and interpersonal context, provided promis- bound together by an emphasis on the empirical demon-
Pr

ing results. After treatment, 78% of patients were classi- stration of efficacy in controlled research.
es

fied by independent evaluators as responders, and im- The major classes of CBT that have been applied to the
s,

provement was noted on several other measures. Although treatment of social phobia include (1) exposure, a series of
IPT has been shown to be efficacious in the treatment of techniques designed to help patients face situations they
I nc

major depression,3,4 dysthymic disorder,5 and other disor- fear and stay psychologically engaged with the situation
so that habituation and extinction processes can take ef-
.

fect; (2) cognitive restructuring, a series of techniques de-

From the Department of Psychology and the Adult Anxiety
Clinic, Temple University, Philadelphia, Pa. signed to help the patients view their world in unbiased,
This article is based on “Psychotherapy for Social Phobia,” a presumably more accurate, ways that will enable them to
presentation made by Dr. Heimberg at a symposium entitled
Advances and Emerging Treatments in Social Phobia. The move through the world less preoccupied with the dangers
symposium took place on January 10, 2000, in Atlanta, Georgia. that are perceived to lurk around every turn; (3) relaxation
Advances and Emerging Treatments in Social Phobia was training techniques, which help patients learn to attend to,
sponsored by Duke University School of Medicine and supported
through an unrestricted educational grant from Pfizer Inc. and control, the degree of physiologic arousal experienced
Reprint requests to: Richard G. Heimberg, Ph.D., Adult during or in anticipation of feared events; and (4) social
Anxiety Clinic, Temple University, 1701 North 13th St.,
Philadelphia, PA 19122-6085 skills training, or the specific training of patients in the be-
(e-mail: rheimber@nimbus.ocis.temple.edu). havioral skills of social interaction, a treatment approach

36 J Clin Psychiatry 2001;62 (suppl 1)

 Psychotherapeutic Interventions for Social Phobia

that follows logically from an assumption that individuals Just as the agoraphobic patient is prone to say that she sur-
who are socially anxious lack the necessary skills to derive vived the last panic attack only because she ran out of the
positive outcomes from social interaction. These classes of room before it got too bad, the patient with social phobia
techniques can be, and often are, combined in the treat- may falsely attribute successful outcomes in feared situa-
ment of a specific patient. In the next section, I shall de- tions to the fact that he carefully rehearsed every line before
scribe the basics of each of these approaches to CBT. speaking, that she stood at the periphery of the circle so as
to avoid becoming the center of attention, or that he was able
COGNITIVE-BEHAVIORAL TREATMENTS to avoid spilling the drink only because he clasped it hard
FOR SOCIAL PHOBIA with both hands at all times. Wells and colleagues11 have
demonstrated that analyzing a patient’s safety behaviors and
Exposure instructing the patient to drop them during exposure exer-
©

In exposure treatments, patient and therapist first col- cises facilitate the outcome of exposure treatment.
Co

laborate on the development of a “fear and avoidance hier-

archy,” that is, a rank-ordered list of situations that are Cognitive Restructuring
py

problematic and anxiety-provoking for the patient. This The findings of studies by Wells and colleagues,9,11 as
list is typically rated by the patient on scales assessing the well as a large body of experimental psychopathology re-
rig

degree of anxiety experienced and the degree to which the search, suggest the importance of working with patients
ht

patient avoids the situation in everyday life. To keep anxi- to examine their thoughts about specific situations and the
ety in a tolerable range, the patient then starts working on beliefs that may underlie them. In fact, recent cognitive-
20

the least fearsome situation and gradually approaches behavioral models of social phobia10,12 suggest that social
more and more difficult situations as a sense of mastery of phobia arises from inaccurate beliefs about the potential
01 ne p

the lesser situations is achieved. Either in imagination (as dangers posed by social situations, negative predictions
O

the therapist narrates scenes for the patient to imagine), by about the outcomes of these situations, and biased process-
Pherso

role-playing with the therapist or therapy assistants, or by ing of events that transpire during social situations.
ys nal c

confronting feared situations in the patient’s life outside of Cognitive restructuring is a set of interventions originat-
session (or typically in a combination of all these modali- ing from the cognitive therapy of Beck and Emery13 and
ic op

ties), the patient is asked to engage the situation and con- from the rational-emotive therapy of Ellis.14 In cognitive re-
ia y m

tinue to do so until anxiety naturally begins to subside. structuring, individuals are taught (1) to identify negative
ns ay

Most CBTs for the anxiety disorders feature exposure thoughts that occur during stressful or anxiety-provoking
as a prominent part of the overall treatment effort. For ex- situations, (2) to evaluate the accuracy of those thoughts as
Po be

posure techniques to be maximally effective, the patients compared with objective information derived via repeated
should allow themselves to be fully engaged in the feared questioning or as a result of planned “behavioral experi-
stgprin

situation, that is, to pay full attention to the situation, to ex- ments,” and (3) to derive rational alternative thoughts based
ra ted

perience it completely, and to allow the inevitable cascade on the acquired information. As implied by this description,
of subjective fear and physiologic arousal to occur.8 How- cognitive restructuring techniques contain a substantial ex-
du

ever, anxiety disorder patients may find this a frightening posure component, although this exposure may not be as
a

proposition and may engage in efforts to manage their anx- systematic and graduated as described above. Furthermore,
te

iety that will be counterproductive and may reduce the ef- the purpose of exposure in this treatment approach differs
ficacy of exposure treatments. For instance, these patients from that described in the previous section. In cognitive re-
Pr

may attempt to distract themselves to avoid paying full at- structuring, exposure is less about habituation to anxiety or
es

tention to the situation in which they find themselves. Al- remaining in the situation until no more anxiety is experi-
s,

ternatively, they may focus inward on themselves and pay enced, and more about the opportunity for patients to col-
attention to their negative thoughts about themselves lect information that will enable them to “restructure” their
I nc

rather than examine what is actually going on in the situa- view of these situations and to revise their judgments about
tion. Wells and Papageorgiou9 have demonstrated that effi- the degree of risk to which they are exposed. Patients are
.

cacy is increased when the standard directions are aug- given assignments that are intended to undermine their
mented with instructions that help the patient focus on belief(s)—that they are unacceptable to others, that they
what is occurring in the situation. will not know how to behave in particular situations, that
Another maladaptive strategy that may undermine the other people will be harsh and critical, or that they will be
effectiveness of exposure treatments is the tendency of pa- overwhelmed by their anxiety in a way that will be humili-
tients to engage in “safety behaviors.”10 Safety behaviors ating and embarrassing. For example, a patient who be-
are actions that patients falsely believe will enable them to lieved that he had to be profoundly witty in conversation in
manage the feared situation successfully. However, these order to be accepted by others was asked to eavesdrop on
behaviors may prevent patients from learning that they the (typically mundane) conversations of others in the com-
might have survived and flourished with no special effort. pany cafeteria. Another patient who believed that he would

J Clin Psychiatry 2001;62 (suppl 1) 37

Richard G. Heimberg

not be able to break silences if they occurred in conversa- poor eye contact, poor conversation skills). These behav-
tion was asked to artificially create these very silences at an ioral deficiencies may elicit negative reactions from oth-
upcoming social event. A person with a fear of drinking in ers, thereby causing social interactions to be punishing
public was asked to go out and do so, making sure to fill his and anxiety-provoking for the patient.18 The studies exam-
glass to the brim and carry it around with him, rather than ining the social skills of socially anxious individuals have
leave it sitting on the table. Furthermore, he was asked to come to different conclusions, with some suggesting be-
spill a drink on purpose at some point during the evening. havioral deficiencies19,20 and others not.21–23 Furthermore,
The outcomes of these behavioral experiments were then even when behavioral deficits are observed, it is unclear
juxtaposed with the patients’ negative predictions about the whether they are a function of a lack of social knowledge
social catastrophes that they believed were certain to occur. or skill, behavioral inhibition or avoidance produced by
Repeated efforts of this nature encourage patients to be- anxiety, or a combination of these and other factors. In our
©

come “scientists” for whom the “experimental hypotheses” clinic, it is not uncommon for patients to complain that
Co

are their own dire predictions and the “data” derived from they do not know how to behave in various situations.
these behavioral experiments will require their hypotheses However, observation of their performance during expo-
py

to be revised in the best tradition of the scientific method. sures often reveals their behavior to be within acceptable
limits. This observation is consistent with research sug-
rig

Relaxation Training gesting that individuals with social phobia often underesti-
mate the adequacy of their social performance.23
ht

As noted above, relaxation training techniques help the

patient learn to attend to and control the degree of physi- Techniques commonly used in social skills training in-
20

ologic arousal experienced during or in anticipation of clude therapist modeling, behavioral rehearsal, corrective
feared events. There are a large number of different ap- feedback, social reinforcement, and homework assign-
01 ne p

proaches to relaxation training, although most are derived ments. Notably, if these techniques effectively reduce anx-
O

in some way from the pioneering work of Wolpe15 and iety for some individuals with social phobia, this reduction
Pherso

Bernstein and Borkovec.16 In the typical application, pa- does not necessarily occur because deficiencies in the
ys nal c

tients learn to relax through exercises involving different patient’s repertoire of social skills have been remediated
muscle groups. These exercises are practiced in session (although this certainly may be the case). Social skills
ic op

with a therapist and at home alone. Patients focus on a par- training may provide benefits because of the training (e.g.,
ia y m

ticular muscle group, tense the muscle, hold the tension for repeated practice of feared social behaviors), the exposure
ns ay

5 to 10 seconds, and next focus on the sensations accompa- (e.g., confrontation of feared situations), or the cognitive
nying the tension. They then release the tension, notice the elements (e.g., corrective feedback about the adequacy of
Po be

difference between the feelings of tension and of relax- one’s social behavior) inherent in the procedures. Social
ation, and focus on the sensations accompanying relaxation skills training may also be easily combined with other
stgprin

(e.g., warmth, heaviness). Patients begin by working with techniques such as cognitive restructuring or exposure.
For instance, Social Effectiveness Training24 is a multi-
ra ted

16 muscle groups, but, over time, practice relaxing larger

groups of muscles to achieve more rapid relaxation. The component treatment package combining exposure with
du

next step involves relaxation by recall, in which patients social skills training and education in a mixture of group
a

scan their bodies for muscle tension and release any tension and individual formats.
te

by recalling how these muscles felt when relaxed. Patients

are also taught cue-controlled relaxation, in which a word EFFICACY OF COGNITIVE-BEHAVIORAL
Pr

such as relax is repeatedly paired with a relaxed state and TREATMENTS FOR SOCIAL PHOBIA
es

then used as a cue to begin the process of rapidly relaxing

s,

during daily activities. The number of studies examining the efficacy of CBTs
Relaxation strategies for social phobia are typically not for social phobia increased dramatically in the 1990s, and
I nc

effective unless they are “applied.” Applied relaxation con- this rate of growth shows little sign of slowing down.
sists of training in 3 skills. Patients learn to (1) attend to the Qualitative reviews of the literature become increasingly
.

physiologic sensations of anxiety, (2) relax quickly while difficult to do and even more difficult to digest. Readers
engaging in everyday activities, and (3) apply relaxation interested in pursuing such reviews are referred to other
skills in anxiety-provoking situations. Applied relaxation articles previously published by my research group.25–27
thus combines relaxation and exposure to help individuals Here, I shall focus instead on quantitative reviews that use
cope with anxiety-provoking situations.17 the techniques of meta-analysis. Meta-analysis is a method
for examining the outcomes of several studies simulta-
Social Skills Training neously by reducing the results of each study to a common
The use of social skills training in the treatment of social metric, the effect size.28
phobia is predicated on the assumption that socially anx- Three meta-analytic reviews have been conducted ex-
ious patients often exhibit behavioral deficiencies (e.g., amining the relative efficacy of various classes of CBTs

38 J Clin Psychiatry 2001;62 (suppl 1)

 Psychotherapeutic Interventions for Social Phobia

Table 1. Percent Attrition for Various Classes of Table 3. Follow-Up Effect Sizes for Various Classes of
Cognitive-Behavioral Psychotherapy for Social Phobiaa Cognitive-Behavioral Psychotherapy for Social Phobiaa
No. of No. of
Treatment Category Mean SD Trials Treatment Category Mean SD Trials
Waiting list 5.7 4.6 6 Waiting list ... ... ...
Placebo 7.5 7.5 6 Placebo ... ... ...
Cognitive restructuring 12.2 10.9 5 Cognitive restructuring 0.96 0.47 5
Social skills training 16.6 8.2 5 Social skills training 0.99 0.64 3
Exposure 16.4 7.4 8 Exposure 0.93 0.25 8
Exposure and cognitive restructuring 18.0 11.0 12 Exposure and cognitive restructuring 1.08 0.41 9
a
Adapted, with permission, from Taylor.31 a
Adapted, with permission, from Taylor.31
©

Table 2. Posttreatment Effect Sizes for Various Classes of restructuring. These were compared with both waiting-list
Co

Cognitive-Behavioral Psychotherapy for Social Phobiaa control and placebo control conditions (a combination of
No. of psychological and pill-placebo conditions, which were
py

Treatment Category Mean SD Trials equivalent to each other in effect size). Table 1 shows that
Waiting list –0.13 0.15 5 the CBTs did not differ from each other in percent attrition,
rig

Placebo 0.48 0.26 5

Cognitive restructuring 0.63 0.32 5 and although CBTs were associated with somewhat higher
ht

Social skills training 0.65 0.46 4 withdrawal rates than either of the control conditions, the
Exposure 0.82 0.25 8 difference was not significant.
20

Exposure and cognitive restructuring 1.06 0.32 11

a The effect sizes for the different CBTs and control con-
Adapted, with permission, from Taylor.31
ditions at the end of acute treatment appear in Table 2.
01 ne p

Several points are evident. All CBT variations were more

O

for social phobia.29–31 Each meta-analysis has many merits, effective than the waiting-list control conditions, and the
Pherso

but for the sake of simplicity, I have selected the meta- CBT variations did not differ from each other. However,
analysis by Taylor31 for the current discussion. Before an
ys nal c

only the combination of exposure and cognitive restructur-

examination of the results of Taylor’s efforts, certain tech- ing was statistically superior to the placebo controls. Al-
ic op

nical aspects of his meta-analysis should be described. He though not indicated in the table, the effect sizes were
ia y m

calculated effect sizes for each study on the basis of the similar for group and individual interventions.
ns ay

formula for Cohen’s28 d ([Mpre – Mpost]/SDpooled), where Many of the studies in Taylor’s meta-analysis included
Mpre is the pretreatment mean, Mpost is the posttreatment a follow-up period averaging about 3 months after discon-
Po be

mean, and SDpooled is the standard deviation calculated tinuation of acute treatment. The effect sizes for the CBT
across all relevant observations. This method defines an variants appear in Table 3 (effect sizes for the control con-
stgprin

effect size as the number of standard deviation units of im- ditions were not available because control patients in most
ra ted

provement made by patients receiving a particular treat- studies had received other treatments by that time). These
ment in a particular study. An average effect size of 1 for a effect sizes were not different from each other, but they are,
du

specific treatment indicates that, on average, the patients as a group, significantly larger than they were at the end of
a

in all the studies who received that treatment improved by treatment. Thus, the patients who received CBTs continued
te

1 standard deviation. Generally, half that effect size is to improve after the discontinuation of treatment.
thought to be a meaningful level of response. Taylor ex-
Pr

amined the results of 42 outcome trials and calculated ef- COMPARISON OF COGNITIVE-BEHAVIORAL
es

fect sizes on the basis of validated self-report measures THERAPY AND PHARMACOTHERAPY
s,

that were available in most studies. Because clinician- FOR SOCIAL PHOBIA
rated measures tend to yield larger effects than self-report
I nc

measures, this is a conservative approach. The question of the relative efficacy of cognitive-
A potential problem with meta-analyses is that effect behavioral and medication approaches to the treatment of
.

sizes may be inflated by failure to publish negative trials. social phobia has not received the attention it deserves.
To guard against this, calculations are made of the number Only a handful of studies have been conducted, and 2 of
of articles that would need to have a null result to render a these32,33 examined medications that have not been suc-
finding trivial. This calculation is called the “fail-safe N.” cessfully differentiated from placebo in double-blind tri-
A large fail-safe N suggests that the effect size for a par- als. Another study34 compared group CBT with the mono-
ticular treatment is robust. Fail-safe Ns in Taylor’s meta- amine oxidase inhibitor phenelzine, the high-potency
analysis ranged from 48 to 223, numbers that should in- benzodiazepine alprazolam, and placebo. However, pa-
spire confidence in his results. tients in the medication and placebo arms of the study re-
Taylor examined cognitive restructuring, social skills ceived instructions to engage in exposure to feared situa-
training, exposure, and exposure combined with cognitive tions between sessions, thus making the results of this

J Clin Psychiatry 2001;62 (suppl 1) 39

Richard G. Heimberg

study difficult to interpret. Yet another study35 compared pleters and 20 (77%) of 26 phenelzine completers as hav-
group CBT with the benzodiazepine clonazepam, but in- ing a clinically significant response (intent-to-treat analy-
cluded no placebo condition. A potentially excellent study sis: CBGT = 58%, phenelzine = 65%). CBGT and phenel-
comparing group CBT with fluoxetine and their combina- zine produced response rates higher than rates associated
tion, conducted by Jonathan R. T. Davidson, M.D., at Duke with pill placebo and ES, but not different from each other.
University and Edna B. Foa, Ph.D., at the University of Many phenelzine patients who were classified as respond-
Pennsylvania, is still underway (unpublished study). The ers after 12 weeks of treatment had achieved gains by the
only other published comparative study of medications and midtreatment (6-week) assessment; however, this was less
CBT for social phobia is the collaborative study conducted common among CBGT patients. Phenelzine patients were
by myself and Michael Liebowitz, M.D., of Columbia Uni- also more improved than CBGT patients on a subset of
versity and the New York State Psychiatric Institute Anxi- measures after 12 weeks.
©

ety Disorders Clinic, in which we examined the efficacy of In the second phase of this study, patients who re-
Co

group CBT and phenelzine in patients with social pho- sponded to CBGT or phenelzine were continued through 6
bia.36,37 A description of that study follows a brief explana- months of maintenance treatment and a 6-month treat-
py

tion of our cognitive-behavioral program, which uses a ment-free follow-up period.37 After the follow-up period,
therapeutic approach known as cognitive-behavioral group 50% of previously responding phenelzine patients re-
rig

therapy (CBGT). lapsed, compared with only 17% of CBGT patients. The
Cognitive-behavioral group therapy38,39 integrates cog-
ht

difference in relapse between treatments was especially

nitive restructuring techniques and exposure in the treat- pronounced for patients with generalized social phobia.
20

ment of social phobia. Most commonly, CBGT is admin- The overall pattern of results suggests that phenelzine
istered to groups of 6 patients in 12 weekly sessions of might have slightly greater immediate efficacy, but that
01 ne p

approximately 2.5 hours each. Ideally, male and female CBGT may confer greater protection against relapse.
O

cotherapists lead the groups to allow for maximum flexibil-

Pherso

ity in constructing within-session exposure exercises. In COMBINING COGNITIVE-BEHAVIORAL

ys nal c

the first and second sessions, patients are presented with AND PHARMACOLOGIC TREATMENTS
the rationale and instructions for exposure, cognitive re- FOR SOCIAL PHOBIA
ic op

structuring, and homework assignments and are given

ia y m

opportunities to practice cognitive restructuring skills. Few studies have examined the efficacy of combining
ns ay

Thereafter, therapists lead patients through individualized cognitive-behavioral and pharmacologic treatments for so-
exposures (most often in the form of role-played simula- cial phobia, even though combined treatments are common
Po be

tions of each patient’s feared situations) that are preceded in clinical practice. There are only 2 published trials,32,44
and followed by therapist-directed cognitive restructuring and, as was the case for drug-CBT comparisons, these
stgprin

exercises. Patients are also coached in rational thinking studies did not examine medications that surpassed place-
ra ted

during the exposure itself. At the end of each session, thera- bo in other trials. The only other studies of combination
pists work individually with patients to develop homework treatment are the Davidson-Foa collaboration (group CBT
du

assignments for completion during the upcoming week. and fluoxetine; J. R. T. Davidson, M.D., E. B. Foa, Ph.D.,
a

Homework typically consists of exposures to real-life situ- unpublished study) and a follow-up to the Liebowitz-
te

ations and patient-directed preexposure and postexposure Heimberg collaboration that is currently investigating
cognitive restructuring, with the goal of teaching patients combined CBGT and phenelzine (M. R. Liebowitz, M.D.,
Pr

to become their own cognitive-behavioral therapists over R. G. Heimberg, Ph.D., unpublished study). However, nei-
es

the long term. ther of these studies is yet complete. Right now, the utility
s,

Several controlled studies have evaluated the efficacy of of combined treatment is an open question; however, it is
CBGT. Prior to our collaborative study, CBGT had pro- interesting to consider the possibilities of this approach.
I nc

duced outcomes superior to waiting-list control conditions40 There are 3 potential outcomes if drugs and CBT are
and psychological placebo treatment.41 CBGT was also su- combined. The therapies may synergize each other, thereby
.

perior to individual CBT programs in cost-effectiveness.42 producing a better outcome than either treatment alone.
Most importantly, patients treated with CBGT maintained This is, of course, the desired result, and it may occur if
their gains at follow-up assessments 4 to 6 years after treat- the 2 treatments increase the chances of response for a spe-
ment had been discontinued.43 cific individual or if they increase the magnitude of the
In our collaborative study,36 133 patients were randomly individual’s response. However, it is also possible that
assigned to CBGT, phenelzine, pill placebo, or educational medication and CBT may add little to each other. This
supportive group psychotherapy (ES), a credible psycho- would be the case if the medication and the psychotherapy
logical placebo treatment. One hundred seven patients were found to work on the same aspects of the target prob-
completed 12 weeks of acute treatment. At posttest, inde- lem, or if the first treatment was sufficiently powerful so
pendent assessors classified 21 (75%) of 28 CBGT com- the second treatment would have little to contribute. It is

40 J Clin Psychiatry 2001;62 (suppl 1)

 Psychotherapeutic Interventions for Social Phobia

also possible that one treatment may detract from the effi- American Psychiatric Association; 1980:225–239
2. Lipsitz JD, Markowitz JC, Cherry S, et al. Open trial of interpersonal psy-
cacy of the other. For example, medication might detract chotherapy for the treatment of social phobia. Am J Psychiatry 1999;156:
from CBT if a patient believes that the medication is pro- 1814–1816
viding the entire benefit. The patient, therefore, might not 3. Elkin I, Shea MT, Watkins JT, et al. National Institute of Mental Health
Treatment of Depression Collaborative Research Program: general effec-
invest in the activities of CBT or might use the medication tiveness of treatments. Arch Gen Psychiatry 1989;46:971–982
as a safety net. Of course, this is an example of a potentially 4. Weissman MM, Prusoff BA, Dimascio A, et al. The efficacy of drugs and
negative belief that might be tested in CBT. psychotherapy in the treatment of acute depressive episodes. Am J Psychi-
atry 1979;136:555–558
There is much that we need to study and learn about how 5. Markowitz JC. Psychotherapy of dysthymia. Am J Psychiatry 1994;151:
various CBTs and medications work together or against 1114–1121
each other and whether there are differences in patients that 6. Wilfley DE, Agras WS, Telch CF, et al. Group cognitive-behavioral therapy
and group interpersonal psychotherapy for the nonpurging bulimic indi-
suggest better responsiveness to one or the other. To under-
©

vidual: a controlled comparison. J Consult Clin Psychol 1993;61:296–305

stand the relative efficacy of different methods of starting 7. Gabbard GO. Psychodynamics of panic disorder and social phobia. Bull
Co

and sequencing the treatment components would also be of Menninger Clin 1992;56(2, suppl A):A3–A13
8. Foa EB, Kozak MJ. Emotional processing of fear: exposure to corrective
great interest. For instance, is it best to start a medication information. Psychol Bull 1986;99:20–35
py

first to take the edge off the patient’s fears and promote 9. Wells A, Papageorgiou C. Social phobia: effects of external attention in anx-
quicker entry into feared situations? Might the medication iety, negative beliefs, and perspective taking. Behav Ther 1998;29:357–370
rig

10. Clark DM, Wells A. A cognitive model of social phobia. In: Heimberg RG,
be phased out as CBT takes hold? It is also important to Liebowitz MR, Hope DA, et al, eds. Social Phobia: Diagnosis, Assessment
ht

consider whether cognitive-behavioral interventions might and Treatment. New York, NY: Guilford Press; 1995:69–93
be used to help patients with social phobia discontinue 11. Wells A, Clark DM, Salkovskis P, et al. Social phobia: the role of in-situation
20

safety behaviors in maintaining anxiety and negative beliefs. Behav Ther

medications on which they have become psychologically 1995;26:153–161
or physically dependent. This strategy has been used quite 12. Rapee RM, Heimberg RG. A cognitive-behavioral model of anxiety in social
01 ne p

successfully with panic disorder patients.45 phobia. Behav Res Ther 1997;35:741–756
O

13. Beck AT, Emery G. Anxiety Disorders and Phobias: A Cognitive Perspec-
Pherso

tive. New York, NY: Basic Books; 1985

FUTURE DIRECTIONS IN 14. Ellis A. Reason and Emotion in Psychotherapy. New York, NY: Lyle Stuart;
ys nal c

THE PSYCHOTHERAPY OF SOCIAL PHOBIA 1962

15. Wolpe J. Psychotherapy by Reciprocal Inhibition. Oxford, UK: Pergamon
ic op

Press; 1958
Clearly, there is much to do in the further development 16. Bernstein DA, Borkovec TD. Progressive Relaxation Training. Champaign,
ia y m

and evaluation of CBTs for social phobia. However, rather Ill: Research Press; 1973
ns ay

17. Öst L-G. Applied relaxation: description of a coping technique and review of
than praise our successes and criticize our failures, I should controlled studies. Behav Res Ther 1987;25:397–409
like to look toward the future. This is a time when we can 18. Lucock MP, Salkovskis PM. Cognitive factors in social anxiety and its treat-
Po be

become extremely creative with our psychotherapeutic or ment. Behav Res Ther 1988;26:297–302
19. Stopa L, Clark DM. Cognitive processes in social phobia. Behav Res Ther
psychoeducational interventions. Our present-day tech-
stgprin

1993;31:255–267
nologies can help us in ways heretofore unavailable. Pa- 20. Halford K, Foddy M. Cognitive and social skills correlates of social anxiety.
ra ted

tients can use hand-held computers that will lead them step Br J Clin Psychol 1982;21:17–28
21. Clark JV, Arkowitz H. Social anxiety and self-evaluation of interpersonal
by step through actual exposures to feared situations and
du

performance. Psychol Rep 1975;36:211–221

help them to do their cognitive restructuring work.46 With 22. Glasgow RE, Arkowitz H. The behavioral assessment of male and female
a

the creation of “virtual exposures,” patients can face their social competence in dyadic heterosexual interactions. Behav Ther 1975;6:
te

488–498
fears by becoming immersed in environments that simulate 23. Rapee RM, Lim L. Discrepancy between self- and observer ratings of per-
reality. The therapeutic applications of these technologies
Pr

formance in social phobics. J Abnorm Psychol 1992;101:728–731

have already been studied for the fear of heights,47 claustro- 24. Turner SM, Beidel DC, Cooley MR, et al. A multi-component behavioral
es

treatment for social phobia: social effectiveness therapy. Behav Res Ther
phobia,48 and the fear of flying,49 and their use for overcom- 1994;32:381–390
s,

ing the fear of public speaking is just around the corner. Pa- 25. Heimberg RG, Juster HR. Cognitive-behavioral treatments: literature re-
tients can have access to CD-ROMs or Web pages, and view. In: Heimberg RG, Liebowitz MR, Hope DA, et al, eds. Social Phobia:
I

Diagnosis, Assessment and Treatment. New York, NY: Guilford Press;

nc

computer-to-computer interactive treatments can allow 1995:261–309

therapists in the technologically developed world to reach 26. Juster HR, Heimberg RG. Social phobia. In: Hersen M, Bellack AS, eds.
.

persons with Internet access in distant areas for “face- Comprehensive Clinical Psychology, vol. 6: Adults: Clinical Formulation
and Treatment. New York, NY: Elsevier; 1998:475–498
to-face” therapy that would be otherwise unavailable. The 27. Turk CL, Fresco DM, Heimberg RG. Social phobia: cognitive behavior
new century should be an exciting one. therapy. In: Hersen M, Bellack AS, eds. Handbook of Comparative Treat-
ments of Adult Disorders. 2nd ed. New York, NY: John Wiley & Sons;
Drug names: alprazolam (Xanax and others), clonazepam (Klonopin and 1999:287–316
others), fluoxetine (Prozac), phenelzine (Nardil). 28. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed.
New York, NY: Academic Press; 1977
29. Gould RA, Buckminster S, Pollack MH, et al. Cognitive-behavioral and
REFERENCES pharmacological treatment for social phobia: a meta-analysis. Clin Psychol
Sci Prac 1997;4:291–306
1. Anxiety disorders. In: American Psychiatric Association. Diagnostic and 30. Feske U, Chambless DL. Cognitive behavioral versus exposure only treat-
Statistical Manual of Mental Disorders, Third Edition. Washington, DC: ment for social phobia: a meta-analysis. Behav Ther 1995;26:695–720

J Clin Psychiatry 2001;62 (suppl 1) 41

Richard G. Heimberg

31. Taylor S. Meta-analysis of cognitive-behavioral treatments for social pho- 41. Heimberg RG, Dodge CS, Hope DA, et al. Cognitive behavioral group
bia. J Behav Ther Exp Psychiatry 1996;27:1–9 treatment for social phobia: comparison with a credible placebo control.
32. Clark DB, Agras WS. The assessment and treatment of performance anxi- Cognit Ther Res 1990;14:1–23
ety in musicians. Am J Psychiatry 1991;148:598–605 42. Lucas RA, Telch MJ. Group versus individual treatment of social phobia.
33. Turner SM, Beidel DC, Jacob RG. Social phobia: a comparison of behavior Presented at the 27th annual meeting of the Association for Advancement of
therapy and atenolol. J Consult Clin Psychol 1994;62:350–358 Behavior Therapy; Nov 19, 1993; Atlanta, Ga
34. Gelernter CS, Uhde TW, Cimbolic P, et al. Cognitive-behavioral and phar- 43. Heimberg RG, Salzman DG, Holt CS, et al. Cognitive-behavioral group
macological treatments of social phobia: a controlled study. Arch Gen Psy- treatment for social phobia: effectiveness at five-year follow-up. Cognit
chiatry 1991;48:938–945 Ther Res 1993;17:325–339
35. Otto MW, Pollack MH, Gould RA, et al. A comparison of the efficacy of 44. Falloon IRH, Lloyd GG, Harpin RE. The treatment of social phobia: real-
clonazepam and cognitive-behavioral group therapy for the treatment of life rehearsal with nonprofessional therapists. J Nerv Ment Dis 1981;169:
social phobia. J Anxiety Disord 2000;14:345–358 180–184
36. Heimberg RG, Liebowitz MR, Hope DA, et al. Cognitive-behavioral group 45. Otto MW, Pollack MH, Sachs GS, et al. Discontinuation of benzodiazepine
therapy versus phenelzine in social phobia: 12-week outcome. Arch Gen treatment: efficacy of cognitive-behavioral therapy for patients with panic
©

Psychiatry 1998;55:1133–1141 disorder. Am J Psychiatry 1993;150:1485–1490

37. Liebowitz MR, Heimberg RG, Schneier FR, et al. Cognitive-behavioral 46. Newman MG, Kenardy J, Herman S, et al. Comparison of palmtop-computer-
Co

group therapy versus phenelzine in social phobia: long-term outcome. De- assisted brief cognitive-behavioral treatment to cognitive-behavioral treat-
press Anxiety 1999;10:89–98 ment for panic disorder. J Consult Clin Psychol 1997;65:178–183
38. Heimberg RG, Becker RE. Treatment of social fears and phobias. New 47. Rothbaum BO, Hodges LF, Kooper R, et al. Effectiveness of computer-
py

York, NY: Guilford Press. In press generated (virtual reality) graded exposure in the treatment of acrophobia.
39. Heimberg RG, Juster HR, Hope DA, et al. Cognitive behavioral group Am J Psychiatry 1995;152:626–628
rig

treatment for social phobia: description, case presentation, and empirical 48. Botella C, Banos RM, Perpina C, et al. Virtual reality treatment of claustro-
support. In: Stein MB, ed. Social Phobia: Clinical and Research Perspec- phobia: a case report. Behav Res Ther 1998;36:239–246
ht

tives. Washington, DC: American Psychiatric Press; 1995:293–321 49. Rothbaum BO, Hodges L, Watson BA, et al. Virtual reality exposure
40. Hope DA, Heimberg RG, Bruch MA. Dismantling cognitive-behavioral therapy in the treatment of fear of flying: a case report. Behav Res Ther
20

group therapy for social phobia. Behav Res Ther 1995;33:637–650 1996;34:477–481
01 ne p

Question and Answer Session

O

Pherso

Question: Do psychotherapeutic techniques work in so- disorder and those who did not. The response to CBT was
ys nal c

cial phobia for patients with comorbidities? the same in both subgroups.
Dr. Heimberg: Yes. Cognitive-behavioral therapy Question: When do you find that the combination of
ic op

(CBT) is effective in the treatment of social phobia, even if CBT and pharmacologic therapy is most called for in the
ia y m

concomitant disorders are present. In a study that my col- treatment of social phobia?
ns ay

leagues and I have just finished,1,2 for instance, 3 distinct Dr. Heimberg: When patients are frightened and resis-
groups of patients were treated for social anxiety with tant to CBT, you can start them on medication—perhaps a
Po be

CBT: patients with social phobia only; patients with social benzodiazepine—to get them to a point where they are
phobia and a comorbid anxiety disorder, mainly general- comfortable beginning psychotherapy. And some patients
stgprin

ized anxiety disorder; and those with social phobia plus de- might first present to my office stabilized on medication;
ra ted

pression. In this setting, patients with social phobia alone they might continue their drug regimen, or, as they begin
and those with social phobia and comorbid anxiety re- to benefit from CBT, they might do well even if weaned
du

sponded equally well to psychotherapy. The presence of a from the medication.

a

second anxiety disorder did not seem to affect response to As a psychologist, when I see patients with social pho-
te

CBT. Experience in clinical practice may be a little more bia, they often have been through multiple medication tri-
complicated, with more varied outcomes and a need for als, frequently at inadequate doses. But even if they have
Pr

several treatment techniques to improve the dual disorders. been adequately medicated, they generally are not having
es

However, a favorable overall response can be expected. an adequate response or they would not be coming to see
s,

On the other hand, the group with comorbid depression me. So there are many patients and many situations in
did not fare as well. The patients did benefit from CBT, but which psychotherapy and pharmacotherapy can work
I nc

the response was somewhat less satisfying because their hand in hand.
clinical state at the end was not what we had hoped for. On
.

social phobia measures, they were more impaired at base- REFERENCES

line than the patients in the other treatment groups. They
1. Erwin B, Heimberg RG, Juster HR, et al. Comorbid mood and anxiety dis-
improved on these measures by about the same amount, orders among patients with social phobia: effect on treatment outcome. In:
but remained more impaired after treatment. Our data did Program and Abstracts of the 19th National Conference of the Anxiety Dis-
not enable us to study the effect of more treatment. We also orders Association of America; March 25–28, 1999; San Diego, Calif. Ab-
stract NR-15:56
used the Beck Depression Inventory to measure outcome in 2. Erwin BA, Heimberg RG, Juster HR, et al. Predictors of treatment outcome
the patients with comorbid depression. According to this among patients with social phobia: comorbid mood and anxiety disorders.
scale, they were no longer depressed after treatment. In this Presented as part of the symposium “Social Phobia: What Can We Learn
From Predictors of Treatment Response?” at the 33rd annual convention of
group with social phobia plus depression, there were 2 sub- the Association for Advancement of Behavior Therapy; Nov 1999; Toron-
groups: individuals who also had another comorbid anxiety to, Ontario, Canada

42 J Clin Psychiatry 2001;62 (suppl 1)