3 CBT clinicians can rightly feel pleased to be part of a therapy tradition where clinicians do not just “believe” we do good work. Instead, CBT clinicians are expected to remain open and nondefensively interested in reflecting and revising our interventions as informed by research and immediate client or group feedback. This stance is not unique to CBT clinicians but does tend to especially define them. Research shows that therapists have grossly inflated self-assessments about their personal effectiveness (Walfish, McAlister, O’Donnell, & Lambert, 2012), including findings showing that the least effective therapists rated themselves as being on par with the most effective therapists (Hiatt & Hargrave, 1995). This praise for CBT does of course not mean that all CBT therapists are equally effective in delivering an empirically supported CBT intervention. Research has identified a number of therapist factors that enhance outcomes for all kinds of psychotherapy including CBT (e.g., ability to make clients feel understood, ability to adjust our interpersonal behavior to the needs of our clients, as well as ability to encourage our clients to attempt new solutions to their problems and abandon maladaptive patterns; Lambert, 2013). For CBT, technical interventions are not likely to be successful unless the therapeutic relationship, or group process, is characterized by trust, acceptance, empathy, and warmth. The empirically supported psychotherapist is as important as the empirically supported intervention (Lambert, 2013). CBT is a comfortable professional home for many clinicians because of our willingness to subject clinical questions and hunches to further scrutiny, whether we become involved in conducting clinical research or consume research by doing literature reviews. Either way, we become prepared to make practice changes based on new evidence. Evidence-based practice is thus a guiding principle for CBT therapists and means that we turn to the literature when choosing a treatment program for a particular disorder or problem. We are aware that research evidence does not imply absolute truths as the science of psychology and psychotherapy slowly evolves and new findings become translated into new guidelines for clinical practice. CBT therapists whose training involved courses on research methods are aware that research findings are only as good as the particular research method used—including the chosen statistical analyses. An ability to critically evaluate any research design and method is important. However, this skill is not consistently taught in mental health therapy training programs. A problem with many treatment outcome studies, for which CBT has especially been famous, is that they tell us a lot about how people did on average, in other words which treatment approach “won” in the horse race comparing different treatments. But they hardly tell us anything about for whom this treatment did not work well and why. This group is sometimes referred to as the alternative minority. Instead of the horse race-type outcome studies, this minority would benefit from a “true horses for courses” approach (White, 2010). It is thus important to not interpret evidence-informed practice in a narrow sense—to mean only treatment outcome studies of the randomized controlled kind. CBT clinicians benefit from being informed by additional kinds of evidence. Evidence comes in many forms including case studies on individuals or groups in community clinic settings without control groups, pre- and posttreatment outcome evaluations, simple wait-list control designs, and research as well as theory on the importance of the therapeutic relationship and group cohesion as a necessary condition for more technical interventions. Indeed, with extensive research showing CBT produces positive long-term outcomes for between 50% and 80% of people receiving it (highest success for panic disorder and lowest for GAD), CBT research is turning more to questions pertaining to the alternative minority (Lambert, 2013). This chapter will look at key literature findings on the effectiveness of CBGT for the mental health problems covered in this book. Of special interest are studies directly evaluating group CBT against individual CBT for the same disorder. Each subsequent chapter contains more detailed literature reviews for the various topics covered. When clinicians and program managers seek to become familiar with the CBGT research, they are naturally interested in both evidence of cost-effectiveness and improved client functioning. As will become clear, the majority of studies for all the reviewed disorders support the equivalence hypothesis. In other words, research ranging from randomized controlled trials to nonrandomized community clinic effectiveness studies has not found significant differences between individual and group CBT. One important caveat to keep in mind when reviewing group treatment outcome studies is that not all aspects of improved well-being are measured. There may be other benefits not captured in traditional measurements of CBGT, such as increased self-esteem, decreased social isolation, and improved quality of life. For example, some members in a group for panic disorder became encouraged to sign up for a community class on yoga because a member spoke about how she enjoyed it. The benefits to her were apparent to the other members. On the other hand, although the literature indicates that for some problems a group format is highly promising or even superior to individual, this may not fully match clinicians’ experiences. Our groups often feel less successful than what the outcome literature shows. This is due to the fact that group clients in research studies are carefully selected and screened, a luxury group therapists in community clinics do not enjoy to the same extent. Lastly, treatment outcome evaluations naturally only reflect the kind of treatment offered. CBGT is a relatively new approach for a number of disorders. Many group therapists continue to feel they are experimenting to some degree with how to best tailor individual protocols to a group setting. Any evaluation of CBGT is thus limited to its present mode of delivery and is just one step in a continuous dynamic cycle of evaluating revised forms of interventions based on feedback from both research and frontline group therapists. The chapter closes with suggestions for research-informed improvements in CBGT practice. Depression was the first disorder for which a group CBT format was created (Beck, Rush, Shaw, & Emery, 1979). The past three decades has seen an increase in research support for CBGT for depression. The steadily better outcomes are likely a reflection of CBT clinicians becoming more familiar and better trained with this format. As we see below, clinician researchers predict that CBGT for depression will become even more successful as clinicians begin to work consistently with the group process factors. In their review, Tucker and Oei (2007) concluded that group CBT is as effective as individual and certainly less costly in terms of the ratio of paid clinician hours per client. Several other researchers draw a similar conclusion, stating that group CBT is effective and performs at roughly the same level as individual treatment (Burlingame, Strauss, & Joyce, 2013; Craigie & Nathan, 2009; Cramer, Salisbury, Conrad, Eldred, & Araya, 2011; DeRubeis & Crits-Cristoph, 1998; Oei & Dingle, 2008; Scott & Stradling, 1990). Most recently, a meta-analysis of 34 effectiveness studies on outpatient individual and group CBT for depression concluded that CBGT for depression was equally as effective as individual CBT (Hans & Hiller, 2013a). The most encouraging aspect of this is that CBGT required fewer sessions (about 11) to achieve the same improvement in depression compared to individual CBT (about 21 sessions). And CBGT resulted in fewer dropouts. Although the average individual CBT dropout rate is not reported, the overall dropout rate for both group and individual was 24.6%, compared to 21.4% for CBGT. These dropout rates are, however, considered unacceptably high by the authors (Hans & Hiller, 2013a). I agree and offer suggestions for how to improve dropouts from CBGT for depression in Chapter 6. The aforementioned research was done on mid-age adults, usually from ages 18 to 65, but the same conclusions seem to hold for both older and younger depressed people. For older people, the few studies to date show that group CBT can be considered just as helpful as individual. Group CBT can in some ways even be viewed as superior, when one considers that social isolation is a greater risk factor for depression in elderly people compared to other age groups (Kennedy & Tanenbaum, 2000; Krishna et al., 2010). Krishna and colleagues dampen enthusiasm, however, when they point out that although they found a statistically significant effect (six qualifying trials) favoring CBGT for depression in the elderly, these differences between CBGT and other forms of treatment were “at best modest,” suggesting that CBGT for the elderly could be improved (Chapter 11 discusses CBGT for depressed elderly people). For children and adolescents, we see the same picture of CBGT gaining in popularity as clinicians increasingly recognize it as at least equal to individual (Clarke, Rohde, Lewinsohn, Hops, & Seeley, 1999; Chapter 12 resumes this topic of CBGT for depressed children). From the aforementioned summary, group therapists will hopefully feel supported in their commitment to preferring a group CBT format for depression. But they may also feel some frustration that, when subjected to quantitative scrutiny, this format does not look as effective as we would like to believe. Bieling and colleagues (Bieling, McCabe, & Antony, 2006) are convincing when they, in their discussion of the literature on CBGT for depression, point out that group process factors do not appear to have been formally taken into account, which may explain why many outcomes are not better. This speculation is consistent with the spirit of this book, namely, that we CBGT clinicians can capitalize even more on the group processes. Oei and Dingle (2008) echo this sentiment when they challenge clinicians to an “urgent need to develop and evaluate a coherent group CBT theory, in particular the roles of group processes, before further major advancement in this area can be made.” The research is mixed on whether an individual or a group CBT approach is better, but, overall, it is reasonable to assume equivalence. Despite a couple of trials pointing to the superiority of a group format, the majority of studies—including meta-analyses (Fedoroff & Taylor, 2001; Gould, Buckminster, Pollack, Otto, & Yap, 1997)—have failed to find significant differences between formats. It makes clinical intuitive sense that a problem so clearly rooted in social situations may be best addressed in a group setting. As we will see, however, this is not always the case, and there is room for improvement in how we design and deliver CBGT for social anxiety. While Hope, Heimberg, and Bruch (1995) found socially anxious people benefited from being in a group compared to those who were on a list waiting to go into a group, Stangier and colleagues (Stangier, Heidenreich, Peitz, Lauterbach, & Clark, 2003) did not find a group format especially effective. When Stangier and colleagues did a direct comparison of individual and group cognitive therapy over 15 weekly sessions, they found that individual outperformed group (50% no longer met criteria after individual treatment compared to only 13.6% after group). The protocol included the full range of standard treatment components, such as shifting attention to external cues, stopping safety behaviors, video feedback to correct distorted self-imagery, behavioral experiments, and cognitive restructuring. Although the treatment protocol seems comprehensive, Scott (2011) offers a helpful observation by pointing out that only half of the treatment sessions included actual in-session behavioral exposures. A further argument for the importance of exposure can be advanced from the Hope et al. (1995) study, where the exposure alone was more effective than the overall CBGT package, at least in the shorter term. According to Scott’s analysis of the Stangier study, with which I concur, it is possible that the clients involved were deprived of some unique opportunities provided by a group setting to “test out” and overcome their social fears. Those opportunities include creating more in-session exposures—such as role-playing and speaking in front of a “mock” audience, not to mention exposure to internal feared sensations, such as feeling hot and sensing sweaty palms and a dry mouth. Interestingly, one other study has found that a group format was superior to an individual when the key treatment ingredient was precisely in vivo exposures that rehearsed real-life social interactions (Wlazlo, Schroeder-Hartwig, Hand, Kaiser, & Mϋnchau, 1990). Despite studies showing CBGT to be effective, especially if exposure is included, many CBGT therapists admit with some frustration that their social anxiety groups do not seem to help their clients as much as the research claims. A recent study comparing four potential mediators in CBT for social anxiety (avoidance, self-focused attention, cognitive processing related to anticipation of social encounters, and postevent processing) found that individual CBT had larger positive effects on each of the four mediators (Hedman et al., 2013). But most interestingly, and contrary to expectations, the individual format resulted in larger improvements in avoidance behavior, despite exposure exercises being more emphasized both implicitly and explicitly in the group format. Hedman and colleagues offer various speculations on this somewhat puzzling finding. They wonder about the possibility that individual CBT allows for more idiosyncratically tailored designs of exposure exercises. They also suggest that the behavioral experiments used in individual CBT aimed at testing specific negative assumption have a more generalizable effect across feared situations compared to in-session exposure limited to the group room. Lastly, a recent meta-analysis on CBT for anxiety disorders shows no difference in dropout rates between individual CBT and CBGT for social anxiety (Hans & Hiller, 2013b). These research findings on CBGT for social anxiety can help clinicians consider possible improvements to our groups. It appears that including behavioral exposures within the group is critical but that persistent practice of individual exposure between group sessions is equally important. There is no doubt that other creative solutions to better CBGT for social anxiety are needed. It is our experience—similar to other clinicians including Scott (2011)—that mixing panic disorder clients with socially anxious ones in the same group is one possible step toward making the group format more effective. All group members have in common an intense fear of unexpected panic attacks. But the mix of people with panic disorder and social anxiety creates a lighter atmosphere. The socially anxious people often enjoy the more forward and easy interactions that are typical of people with panic disorder. Even though it is not explicitly encouraged by the facilitators, group members with panic disorder become helpful models for social interactions. Chapter 7 offers more information on this transdiagnostic option.
Effectiveness of CBGT Compared to Individual CBT: Research Review
Depression
Social Anxiety Disorder (SAD)
Obsessive–Compulsive Disorder (OCD)
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
