Predictors of outcome in residential cognitive and interpersonal treatment for social phobia: Do cognitive and social dysfunction moderate treatment outcome?

https://doi.org/10.1016/j.jbtep.2010.01.005Get rights and content

Abstract

Background

: The predictors of residential cognitive (RCT) and residential interpersonal Treatment (RIPT) for social phobia were explored. (1) Sotsky et al. (1991) found differential effects of CT and IPT for depression, suggesting that the level of cognitive or social dysfunction predicted differential outcome. We examined whether an analogous effect could be demonstrated in 10 weeks of residential treatment of 80 social phobia subjects. (2) We also included expectations, age of onset, severity of illness, concurrent anxiety, mood, avoidant personality disorder, and body dysmorphic disorder as predictors in this exploratory study.

Method

: Main outcome was the social phobia subscale of Social Phobia and Anxiety Inventory (SPAI SP). DSM-IV axis I and II interviews were completed.

Results

: (1) Sotsky et al. (1991) findings were not reproduced. However, RIPT subjects with poor general functioning were less improved following treatment. Subjects with concurrent agoraphobia responded better with RCT than subjects without agoraphobia. (2) Age of onset and expectations were the most powerful predictors of post treatment outcome.

Conclusion

: Some patient characteristics appear to impact outcome with RIPT and RCT differentially. The findings are discussed.

Introduction

In the last two decades effective psychological treatments for social phobia have been developed and empirically validated (Clark et al., 2003, Heimberg and Juster, 1995). We have previously reported that two residential treatments using either a cognitive or an interpersonal model for treatment of social phobia have been associated with equal improvements (Borge et al., 2008). However, not all patients fully profit from empirically supported treatments. If the predictors of outcome for a specific treatment can be identified, then those who probably would benefit from that treatment could likely be identified as well. Predictors are potential moderators of treatment. Kraemer, Wilson, Fairburn, and Agras (2002) postulate, that for a factor to be a moderator of treatment, this factor must be a pre-randomization characteristic that has an interactive effect with the treatment on outcome.

Sotsky et al. (1991) found different patient predictors of outcome in cognitive and interpersonal therapy for depression, although the overall treatment outcomes were equivalent in the two therapy forms. Low pre treatment cognitive dysfunction predicted a superior response to cognitive therapy, whereas low social dysfunction predicted a superior response to interpersonal therapy. Sotsky et al. suggested that each therapeutic modality relies on specific and different learning techniques. Thus, each modality requires an adequate capacity in the function domain corresponding to the modality, to facilitate recovery. According to this principle, patients with adequate social function would gain most from interpersonal therapy, whereas patients with adequate cognitive function might benefit more from cognitive therapy. If such model specific effects can be demonstrated in other disorders than depression, one could argue that a model that builds on the patient's resources can lead to therapeutic change.

Although model specific therapeutic interventions have shown efficacy, other factors may also account for improvement. Few consistent outcome predictors have been found in the literature on social phobia treatment. In fact, a recent study (Chen et al., 2007) found no pre treatment predictors of the outcomes. However, placebo effects are linked to expectations (Kirsch, 2004), and several studies have demonstrated that patients' expectations are powerful predictors of outcome in psychotherapy (Chambless et al., 1997, Kirsch, 1999, Meyer et al., 2002, Safren et al., 1997, Sotsky et al., 1991, Weinberger and Eig, 1999). Delsignore, Carraro, Mathier, Znoj, and Schnyder (2008) found high expectations of the therapist powerfulness to be the strongest predictor for clinical improvement. The age of onset of social phobia has seldom been included as a predictor in studies of predictors, but when reported, it has usually shown that early onset predicted poorer outcome (Mersch, Emmelkamp, & Lips, 1991) and lower recovery rates (Davidson et al., 1993, DeWit et al., 1999). A recent study (Van Ameringen, Oakman, Mancini, Pipe, & Chung, 2004) confirms these earlier reports demonstrating better treatment outcome for those who had a later onset. The age of onset could be an epiphenomenon of the severity or duration of illness, but remained as a significant predictor in the Van Ameringen et al. study when these factors were controlled. We also wanted to examine some other patient characteristics that have predicted outcome in studies of social phobia. Co-occurring personality disorders are frequent among patients with social phobia. Avoidant personality disorder (AvPD) has been negatively related to short term outcome (Chambless et al., 1997, Feske et al., 1996), whereas others (Hofmann, Newman, Becker, Taylor, & Roth, 1995; Scholing and Emmelcamp, 1999, Van Velzen et al., 1997) or reviews (Dreessen & Arntz, 1998) have not noted this. Other anxiety disorders (e.g. agoraphobia; Angst, 1993) and mood disorders frequently co-occur with social phobia (Brown et al., 2001, Chartier et al., 2003, Erwin et al., 2002, Kessler et al., 1994, Kessler et al., 1999). Chambless et al. (1997) and Lincoln et al. (2005) noted that pre treatment depression predicted poorer post treatment outcome. Scholing and Emmelcamp (1999) only partly replicated the finding in the Chambless et al. trial, noted that depression was a weak predictor of post treatment outcome, and to have no predictive power of 18-month follow-up. Erwin et al. (2002) found no relationship between pre treatment depression and outcome. Thus we chose to explore whether these characteristics and disorders were predictors of treatment outcome in this study. Also, in the current sample, a number of patients presented with symptoms associated with dysmorphic disorder (Borge et al., 2008). Clinically it appeared that co-occurring body dysmorphic disorder might hinder improvement as it represents an additional reason for social avoidance. In addition, it tended to interfere with video-feedback, which is seen as an important technique in the cognitive manual. Short term (Otto et al., 2000) and longer term outcome (Scholing & Emmelcamp, 1999) has also been related to the initial severity of illness in some studies, whereas unrelated in others (Chambless et al., 1997). However, the inconclusive status of these pre treatment patient characteristics as outcome predictors does not seem to justify specific hypotheses.

The purpose of this explorative study was to examine potential short term and longer term predictors and moderators of the main outcome in this trial of two equally effective residential treatments, RCT and RIPT for social phobia. There were primary questions. First, (1) could we find analogous moderator effects of cognitive and interpersonal treatment for social phobia as did Sotsky et al. (1991) for depression? That is, did more cognitive dysfunction relate to poorer outcome in RCT than RIPT, and greater social dysfunction predict poorer outcome in RIPT than RCT? Second, (2) do later age of onset and more positive treatment expectations be positive predictors of treatment outcome? Third, (3) the empirical status of the other factors we investigated (concurrent panic disorder with agoraphobia, level of depression, body dysmorphic disorder, avoidant PD and dimensional measures of avoidant PD, have previously been inconsistent. Consequently we examined to what extent do these factors predicted outcome.

Section snippets

Participants

The participants were 80 individuals who were referred for treatment of social phobia to a residential psychiatric serving the entire country, from February 2002 to April 2003. They were selected among one hundred and thirty consecutive referrals for possible inclusion. Individuals who had not responded adequately to local outpatient services or for whom local services were not available were referred to the hospital for residential treatment. They were assessed by the therapists using the

Results

The overall intercorrelations of the dependent outcome variables and the predictor variables are presented in Table 1.

Discussion

The primary aim of this study was to examine whether a treatment that built upon patients' particular strengths was associated with greater improvement as Sotsky et al. (1991) had noted in a study of depression treated with either CBT or IPT. Specifically we examined whether higher levels of pre treatment cognitive dysfunction would be associated with poorer outcome from cognitive than from interpersonal therapy, and pre treatment social dysfunction would be associated with poorer outcome from

References (55)

  • C.J.M. Van Velzen et al.

    The impact of personality disorders on behavioural treatment outcome for social phobia

    Behaviour Research and Therapy

    (1997)
  • L.E. Alden et al.

    Construction of circumplex scales for the inventory of interpersonal problems

    Journal of Personality Assessment

    (1990)
  • American Psychiatric Association

    Diagnostic and statistical manual of mental disorders

    (1994)
  • J. Angst

    Comorbidity of anxiety, phobia, compulsion and depression

    International Clinincal Psychopharmacology

    (1993)
  • A.T. Beck et al.

    Beck depression inventory – II manual

    (1995)
  • Borge, F. -M., Hoffart, A., Sexton, H., Martinsen, E., Gude, T., Hedley, M., et al. Pre-treatment predictors and...
  • T.A. Brown et al.

    Current and lifetime comorbidity of the DSM-IV anxiety and mood disorders in a large clinical sample

    Journal of Abnormal Psychology

    (2001)
  • T.A. Brown et al.

    Anxiety disorders interview schedule for DSM-IV

    (1994)
  • M.J. Chartier et al.

    Considering comorbidity in social phobia

    Social Psychiatry and Psychiatric Epidemiology

    (2003)
  • J. Chen et al.

    Group cognitive behavior therapy for jananese patients with social anxiety disorder: preliminary outcomes and their predictors

    BMC Psychiatry

    (2007)
  • Clark, D. M. (1997). Cognitive therapy for social phobia: Some notes for therapists. Unpublished manuscript: Oxford:...
  • D.M. Clark et al.

    Cognitive therapy versus exposure & relaxation in social phobia: a randomized controlled trial

    Journal of Consulting and Clinical Psychology

    (2006)
  • D.M. Clark et al.

    Cognitive therapy versus fluoxetine in generalized social phobia: a randomized placebo-controlled trial

    Journal of Consulting and Clinical Psychology

    (2003)
  • Clark, D. M., Wells, A., Hackmann, A., Butler, G., & Fennell, M. J. U. (1994). Measures of social phobia. Unpublished....
  • J.R.T. Davidson et al.

    The epidemiology of social phobia: findings from the duke epidemiological catchment area study

    Psychological Medicine

    (1993)
  • A. Delsignore et al.

    Perceived responsibility for change as an outcome predictor in cognitive-behavioural therapy

    British Journal of Clinical Psychology

    (2008)
  • D.J. DeWit et al.

    Antecedents of the risk of recovery from DSM-III-R social phobia

    Psychological Medicine

    (1999)
  • Cited by (18)

    • Elements of the therapeutic relationship in CBT for anxiety disorders: A systematic review

      2020, Journal of Anxiety Disorders
      Citation Excerpt :

      Across nine credibility-outcome studies, only two revealed significant associations, with both finding a small effect of credibility on outcomes. In a study of 80 individuals randomized to either CT or IPT for SAD, researchers found higher levels of credibility to significantly predict improved outcomes regardless of treatment condition (Borge, Hoffart, & Sexton, 2010). This finding was supported by a subsequent larger study of 126 individuals enrolled in either group or internet-based CBT for SAD, where results showed that greater credibility predicted improved symptoms and diagnostic status at six-month follow up regardless of treatment type (Hedman et al., 2012).

    • Stepped care versus face-to-face cognitive behavior therapy for panic disorder and social anxiety disorder: Predictors and moderators of outcome

      2015, Behaviour Research and Therapy
      Citation Excerpt :

      This trend appears to be similar for ICBT of these conditions. At present, only a few studies have investigated potential moderators of outcome comparing two psychological treatments of anxiety disorders (Borge, Hoffart, & Sexton, 2010; Dow et al., 2007b; Hedman et al., 2012; Wolitzky-Taylor, Arch, Rosenfield, & Craske, 2012), of which two have compared different formats of CBT. Dow et al. (2007b) reported that higher severity, disability and comorbidity at baseline was associated with a worse outcome from brief (six weeks) but not standard (12 weeks) CBT of PD, and Hedman et al. (2012) reported that lower levels of anxiety and depression symptoms was associated with a better outcome from ICBT but not group CBT of SAD.

    • Treatment of Adult Social Anxiety Disorder: A Treatments-by-Dimensions Review

      2014, Social Anxiety: Clinical, Developmental, and Social Perspectives
    • Development of a measure quantifying adverse psychotherapeutic ingredients: The experiences of therapy questionnaire (ETQ)

      2013, Psychiatry Research
      Citation Excerpt :

      Thus, it would be important for studies using our derived measure to determine whether scores have any capacity to capture the possible division between ‘unsuccessful’ and ‘bad’ therapy. As the evidence base grows for certain psychotherapy approaches being effective for specific disorders, there is a call for more refined assessment of factors that may render the same approach beneficial, unhelpful or even harmful to differing clients (O'Donohue and Fisher, 2009; Colom et al., 2010; Borge et al., 2010), and the current measure may assist identifying risk factors for adverse outcomes. The Experiences of Therapy Questionnaire (ETQ) may be used to evaluate differing psychotherapies and psychotherapists, particularly in terms of contribution to unsuccessful outcome; to determine actual adverse outcome risks associated with each factor, under what circumstances and during which time points during the course of therapy; to explore contributions to such negative therapist ingredients (and corrective strategies); and to identify what individual patient characteristics shape the therapist's interactive style and, conversely, shape differing patient responses.

    View all citing articles on Scopus
    View full text