Can pill placebo augment cognitive-behavior therapy for panic disorder?
© Furukawa et al; licensee BioMed Central Ltd. 2007
Received: 02 August 2007
Accepted: 20 December 2007
Published: 20 December 2007
In a number of drug and psychotherapy comparative trials, psychotherapy-placebo combination has been assumed to represent psychotherapy. Whether psychotherapy plus pill placebo is the same as psychotherapy alone is an empirical question which however has to date never been examined systematically.
We conducted a systematic review and meta-analysis of randomised controlled trials (RCTs) that directly compared cognitive-behavior therapy (CBT) alone against CBT plus pill placebo in the treatment of panic disorder.
Extensive literature search was able to identify three relevant RCTs. At the end of the acute phase treatment, patients who received CBT plus placebo had 26% (95%CI: 2 to 55%) increased chances of responding than those who received CBT alone. At follow-up the difference was no longer statistically significant (22%, 95%CI: -10% to 64%).
The act of taking a pill placebo may enhance the placebo effect already contained in the effective psychotherapeutic intervention during the acute phase treatment. Theoretically this is an argument against the recently claimed null hypothesis of placebo effect in general and clinically it may point to some further room for enhancing the psychotherapeutic approach for panic disorder.
The observed treatment effect, i.e. the change from baseline till endpoint, is traditionally thought to be due to four factors: regression towards the mean, natural course of disease, placebo effect i.e. non-specific effects of the therapist and the setting in which therapy takes place, and specific effects of physical or psychological intervention on the target condition .
When a specific therapy for a certain disorder exists, we therefore assume that it realizes all these four components through its administration. Cognitive-behavior therapy (CBT) for panic disorder is one such instance, because it has demonstrated its superiority in the rigorously internally calibrated, drug-sensitive  group of patients vis-à-vis the pill placebo arm  and also against a non-specific psychological intervention .
Whether adding pill-placebo to specific psychotherapy enhances its effectiveness is an empirically and theoretically interesting question . Some may assume that a specifically effective treatment has already realized all the above-mentioned four components of treatment effect and therefore addition of a pill placebo cannot enhance its effect. On the other hand, some may suspect that addition of pill placebo may subtract from psychological treatment because it may undermine the active commitment of the patient to follow psychological interventions. Or, some may argue that psychotherapy placebo and pharmacotherapy placebo work through different psychological mechanisms and can therefore be additive or synergistic.
While conducting a comprehensive systematic review of randomized controlled trials (RCT) of the combined psychological and drug treatment for panic disorder [6, 7], we had a unique opportunity to compare the psychotherapy alone versus the psychotherapy plus pill placebo arms and would like to repot the results.
Identification of trials
In order to identify all randomized controlled trials that compared psychotherapy against psychotherapy plus placebo, we looked for all trials that examined the combination of psychotherapy and pharmacotherapy by antidepressant or benzodiazepines, two types of drugs known to be effective for panic disorder.
Both individual and group formats of the following psychological treatments were included: behavior therapy involving some kind of exposure, cognitive therapy which uses some kind of cognitive restructuring, cognitive-behavior therapy involving elements of both cognitive and behavioral therapies, and other psychological interventions. All commonly prescribed antidepressants and benzodiazepines were eligible.
We electronically searched the Cochrane Collaboration Depression, Anxiety and Neurosis Controlled Trials Register (CCDANCTR) with keywords antidepressant OR benzodiazepines and panic in April 2003 and October 2005. The CCDANCTR is a study-based register of randomized trials incorporating results of group searches of MEDLINE (1966-), EMBASE (1980-), CINAHL (1982-), PsycINFO (1974-), PSYNDEX (1977-) and LILACS (1982–1999) and handsearches of major psychiatric and medical journals. Two complementary searches for additional relevant trials were conducted with the Cochrane Central Register of Controlled Trials (CENTRAL) and with MEDLINE. No language restriction was imposed.
Two reviewers examined titles and abstracts of studies identified by the electronic search strategies and then checked full articles for eligibility. To identify further trials, references of these selected studies and of other review papers were also checked, representative studies were subjected to SciSearch, and experts were contacted.
Quality assessment and data extraction
Two independent reviewers (NW and TAF) assessed the methodological quality of the selected studies. The criteria for quality assessment were based on recommendations in the Cochrane Handbook for Systematic Reviews of Interventions , which focused on the quality of allocation concealment. We also rated whether at least one outcome measure was assessed by an independent assessor blind to treatment allocation.
Two reviewers independently extracted data from the original reports using data extraction forms. For studies where exact numbers of responders were not presented but only their means and standard deviations (SDs) of the global severity measure, we imputed response rates by using a validated imputation method  in order to conduct intention-to-treat (ITT) analyses as described below. Any disagreement was resolved by consensus between the first three reviewers.
Our primary outcome was "response," i.e. substantial improvement from baseline as defined by the original investigators. Examples would be "very much or much improved" according to the Clinical Global Impression Change Scale , and more than 40% reduction in the Panic Disorder Severity Scale  score or in panic frequency. The total number of dropouts for any reason was regarded as a proxy measure of treatment acceptability. Adverse effects were evaluated by looking at the number of dropouts due to adverse effects.
For dichotomous outcomes, ITT analysis was adopted. When dropouts were excluded from any assessment in the primary studies (for example, those who never returned for assessment after randomization), they were considered non-responders. Relative risks (RR) and their 95% confidence intervals were calculated using random effects model rather than a fixed effects model because of its generalizability .
Heterogeneity, which refers to variability among studies in a systematic review and generally derives from clinical, methodological or statistical diversity , was assessed by Chi-squared statistics and I-squared statistics .
Description of studies
Characteristics of the included studies
Barlow et al (2000)
DIAGNOSIS: DSM-III-R panic disorder with mild to no agoraphobia
AGE: mean = 34.1 to 37.8 years
SEX: 62% women
12 weeks of:
1. imipramine (mean = 214 to 239 mg/d by week 12) + CBT (11 sessions during 3 months)
2. imipramine alone
3. CBT alone
4. CBT + placebo
5. placebo alone
RESPONSE: > = 40% reduction on PDSS
De Beurs et al (1995)
DIAGNOSIS: DSM-III-R panic disorder with moderate to severe agoraphobia
AGE: mean = 38.8
SEX: 75% women
12 weeks of:
1. fluvoxamine (100–150 mg/day) + exposure (12 weekly sessions)
2. exposure alone
3. placebo + exposure
4. psychological panic management plus exposure
RESPONSE: > = 40% reduction in panic frequency [imputed from its mean & SD]
Sharp et al (1996)
DIAGNOSIS: DSM-III-R panic disorder with or without agoraphobia (% agoraphobia unclear; however, the average score for FQ-Ag was around 15, indicating most had at least some agoraphobia)
AGE: mean = 33.2 to 38.8
SEX: 78% women
12 weeks of:
1. fluvoxamine (100–150 mg/day) + CBT (12 sessions)
2. fluvoxamine alone
3. CBT alone
4. CBT + placebo
5. placebo alone
RESPONSE: "Very much" or "Much improved" on CGI Change
Acute phase treatment
Treatment acceptability, as measured by the total number of dropouts for any reasons, did not differ between the two groups (RR = 0.83, 0.40 to 1.72). There were no dropouts due to side effects in psychotherapy alone or psychotherapy plus placebo arms of the three trials.
Follow-up after trial termination
The present report represents the first systematic empirical examination of the effect of adding pill placebo to psychotherapy. It found that, for panic disorder with or without agoraphobia, pill placebo may enhance CBT during the acute phase treatment. There was no difference in treatment acceptability between these two treatment arms. The difference in effectiveness, however, seemed to wane in 6–24 months after end of the acute phase treatment.
Twenty-five years ago, Hollon and DeRubeis argued that placebo-psychotherapy combination cannot represent psychotherapy in drug-psychotherapy comparative trials but could not determine from the literature review then whether the former would overestimate or underestimate the latter . The current study has assembled newer data and suggests that placebo-psychotherapy combination overestimates the latter and that therefore Hollon and DeRubeis's concerns proved justified.
The fact that pill placebo can add to psychotherapy is intriguing. Theoretical implications of the present findings may be two. Firstly, in contrast to some old claims , taking a pill placebo did not undermine the effectiveness of psychological intervention. Secondly, much to the contrary, pill placebo enhanced the effectiveness of CBT. The act of taking an inactive drug can enhance the placebo effect already contained in the effective psychotherapeutic intervention. This is a strong argument against the recently claimed null hypothesis of placebo effect .
The observed pill placebo effect over and above the proven psychotherapy may be explainable from both of the main theoretical models of the placebo effect, namely the classical conditioning  and expectancy theory  In the life of a modern man the beneficial experience of a pill is almost inescapable and this may have contributed to the placebo effect through pill taking in addition to the same through talk therapy. In other words, the pill has become a conditioned stimulus, eliciting a conditioned response which is placebo effect. Or alternatively, according to the expectancy theory, patients expect more from pill and talk therapy so that they respond more to pill and talk therapy. The fact that the benefit of the pill placebo appeared to wane after the treatment termination is compatible with both of these theoretical models. In fact, although they are often presented as competing perspectives, the two theories are not necessarily incompatible with each other .
Pragmatic implications of the present findings may not be as straightforward. We have formerly established that antidepressant drug plus CBT was more effective than placebo plus CBT at the end of the acute phase treatment . It appears that psychotherapy plus pill placebo would come in-between the two, but in clinical practice prescribing a pill placebo will be very difficult if not unethical. It is important to note, however, that there is some further room to enhance placebo effect when administering CBT to panic disorder.
There are some possible weaknesses to the present study. In the first place, despite our intensive and extensive literature search, we were able to identify only three trials that made a head-to-head comparison between psychotherapy plus placebo against psychotherapy alone. The present findings therefore may lack statistical power and should be regarded exploratory and preliminary. Secondly, in two of the three included studies, the primary outcomes were rated by raters who were not blind to the psychotherapy status but who were blind only to the drug status (active drug vs placebo). However, we must note that the absolute degree of heterogeneity, as expressed by I-squared values and as visually observable in forest plots (Figures 1 and 2), was close to zero among the one blind and two non-blind studies. The open nature of the assessment can therefore not explain away the observed difference. Thirdly, the response rate had to be imputed in one of the three studies  based on the panic frequency, which likely not reflect the whole panic disorder psychopathology. Deleting this study, however, did not affect the outcomes (RR = 1.24, 95%CI: 1.00 to 1.54 at the end of acute phase treatment, and RR = 1.28, 95%CI: 0.91 to 1.80 at 6–24 months after treatment discontinuation). We were also unable to conduct detailed meta-analyses of continuous variables presented in the identified studies because each rated some different aspects of panic disorder. For example, one RCT reported frequency of panic attacks, two reported on phobic avoidance, two reported on general anxiety, one reported depression and one reported social dysfunction on various continuous scales. The field would certainly benefit from wider-spread use of a more common metric of overall panic disorder severity, such as the Panic Disorder Severity Scale . Lastly, our results concern panic disorder only and generalizability beyond this disorder is not guaranteed. More systematic reviews on this topic in other disorders are warranted.
The act of taking a pill placebo may enhance the placebo effect already contained in the effective psychotherapeutic intervention during the acute phase treatment. There may be some further room for enhancing the psychotherapeutic approach for panic disorder.
- Kleijnen J, de Craen AJ, van Everdingen J, Krol L: Placebo effect in double-blind clinical trials: a review of interactions with medications. Lancet. 1994, 344 (8933): 1347-1349. 10.1016/S0140-6736(94)90699-8.View ArticlePubMedGoogle Scholar
- Klein DF: Preventing hung juries about therapy studies. J Consult Clin Psychol. 1996, 64 (1): 81-87. 10.1037/0022-006X.64.1.81.View ArticlePubMedGoogle Scholar
- Barlow DH, Gorman JM, Shear MK, Woods SW: Cognitive-behavioral therapy, imipramine, or their combination for panic disorder: A randomized controlled trial. JAMA. 2000, 283 (19): 2529-2536. 10.1001/jama.283.19.2529.View ArticlePubMedGoogle Scholar
- Shear MK, Houck P, Greeno C, Masters S: Emotion-focused psychotherapy for patients with panic disorder. Am J Psychiatry. 2001, 158 (12): 1993-1998. 10.1176/appi.ajp.158.12.1993.View ArticlePubMedGoogle Scholar
- Hollon SD, DeRubeis RJ: Placebo-psychotherapy combinations: inappropriate representations of psychotherapy in drug-psychotherapy comparative trials. Psychol Bull. 1981, 90 (3): 467-477. 10.1037/0033-2909.90.3.467.View ArticlePubMedGoogle Scholar
- Furukawa TA, Watanabe N, Churchill R: Psychotherapy plus antidepressant for panic disorder with or without agoraphobia: systematic review. Br J Psychiatry. 2006, 188: 305-312. 10.1192/bjp.188.4.305.View ArticlePubMedGoogle Scholar
- Watanabe N, Churchill R, Furukawa TA: Combination of psychotherapy and benzodiazepines versus either therapy alone for panic disorder: a systematic review. BMC Psychiatry. 2007, 7 (1): 18-10.1186/1471-244X-7-18.View ArticlePubMedPubMed CentralGoogle Scholar
- Higgins JP, Green S: Cochrane Handbook for Systematic Reviews of Interventions 4.2.5 [updated May 2005]. 2005, Chichester, UK , John Wiley & Sons, Ltd.Google Scholar
- Furukawa TA, Cipriani A, Barbui C, Brambilla P, Watanabe N: Imputing response rates from means and standard deviations in meta-analysis. Int Clin Psychopharmacol. 2005, 20 (1): 49-52. 10.1097/00004850-200501000-00010.View ArticlePubMedGoogle Scholar
- Guy W: ECDEU Assessment Manual for Psychopharmacology. 1976, Rockville, Maryland , U. S. Department of Health and Human ServicesGoogle Scholar
- Shear MK, Brown TA, Barlow DH, Money R, Sholomskas DE, Woods SW, Gorman JM, Papp LA: Multicenter collaborative panic disorder severity scale. Am J Psychiatry. 1997, 154 (11): 1571-1575.View ArticlePubMedGoogle Scholar
- Furukawa TA, Guyatt GH, Griffith LE: Can we individualize the 'number needed to treat'? An empirical study of summary effect measures in meta-analyses. Int J Epidemiol. 2002, 31 (1): 72-76. 10.1093/ije/31.1.72.View ArticlePubMedGoogle Scholar
- Higgins JP, Thompson SG: Quantifying heterogeneity in a meta-analysis. Stat Med. 2002, 21 (11): 1539-1558. 10.1002/sim.1186.View ArticlePubMedGoogle Scholar
- de Beurs E, van Balkom AJ, Lange A, Koele P, van Dyck R: Treatment of panic disorder with agoraphobia: comparison of fluvoxamine, placebo, and psychological panic management combined with exposure and of exposure in vivo alone. Am J Psychiatry. 1995, 152 (5): 683-691.View ArticlePubMedGoogle Scholar
- Sharp DM, Power KG, Simpson RJ, Swanson V: Fluvoxamine, placebo, and cognitive behaviour therapy used alone and in combination in the treatment of panic disorder and agoraphobia. J Anxiety Disord. 1996, 10 (4): 219-242. 10.1016/0887-6185(96)00008-4.View ArticleGoogle Scholar
- Uhlenhuth EH, Lipman RS, Covi L: Combined pharmacotherapy and psychotherapy. J Nerv Ment Dis. 1969, 148 (1): 52-64.View ArticlePubMedGoogle Scholar
- Hrobjartsson A, Gotzsche PC: Is the placebo powerless? Update of a systematic review with 52 new randomized trials comparing placebo with no treatment. J Intern Med. 2004, 256 (2): 91-100. 10.1111/j.1365-2796.2004.01355.x.View ArticlePubMedGoogle Scholar
- Siegel S: Explanatory mechanisms for placebo effect: Pavlovian conditioning. The Science of the Placebo: Toward an interdisciplinary research agenda. Edited by: Guess HA, Kleinman A, Kusek JW, Engel LW. 2002, London , BMJ Books, 133-157.Google Scholar
- Bootzin RR, Caspi O: Explanatory mechanisms for placebo effect: cognition, personality and social learning. The Science of the Placebo: Toward an interdisciplinary research agenda. Edited by: Guess HA, Kleinman A, Kusek JW, Engel LW. 2002, London , BMJ Books, 108-132.Google Scholar
- Stewart-Williams S, Podd J: The placebo effect: dissolving the expectancy versus conditioning debate. Psychol Bull. 2004, 130 (2): 324-340. 10.1037/0033-2909.130.2.324.View ArticlePubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-244X/7/73/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.