Specific Techniques Vs. Common Factors? Psychotherapy Integration and its Role in Ethical Practice

Introduction

Providing the best evidence-based practice is becoming increasingly important when considering the ethics of informed consent, professional competence, and continuing professional development. The American Psychological Association (2012) requires psychologists to provide information to their clients on “. . . the nature and anticipated course of therapy” (p 13) at the earliest opportunity in treatment. In the complex world of therapy, therapists re admonished to make a choice in treating clients by “. . . basing . . . (their) . . . service on the established knowledge of the discipline and profession of psychology (Australian Psychological Society, 2012, p 18). Psychologists are encouraged to use “. . . [methods] based upon established scientific and professional knowledge of the discipline generally recognized techniques and procedures” (American Psychological Association, 2012, Standard 2, http://www.apa.org/ethics/code/). One of the most direct references to evidence-based practice is in the policy statement accepted by the American Psychological Association, which defined evidence-based practice as using the best research together with clinical experience in the context of individual clients and their wishes and needs (APA Presidential Task Force on Evidence-Based Practice, 2006). The guidelines for continuing professional development (CPD) in Australia, which were developed with the formation of a nation-wide accreditation and registration body for psychologists, makes explicit reference for therapists “to provide optimal evidence-based services to clients” (Psychology Board of Australia, 2012, p. 2). Thus, there is increasing emphasis on evidence-based practice as an ethical imperative. Perhaps the somewhat softer expression, evidence-supported practice, is a more accurate description of what we all try to do.

For professionals considering the role of specific techniques and common factors, the literature on evidence-supported therapies is often confusing. There have been attempts to simplify the search for the most up-to-date evidence-based practice strategies for individual clinicians (e.g. Falzon, Davidson, & Bruns, 2010). Other works described the process of an entire organization moving from an eclectic approach to an evidence-based practice, as being driven by a number of factors, including funding arrangements (e.g. Williams, Rogers, Carson, Sherer, & Hudson, 2011). Then there are others who argue that some cases are so complex that a single evidence-based approach is not feasible (e.g. Magnavita, Levy, Critchfield, & Lebow, 2010). There are useful strategies for determining the best evidence-based practice with specific diagnoses (Falzon et al., 2010)—though this approach relies upon an assumption that each client has only one relevant diagnosis. There is also considerable evidence that the commonalities among therapies may outweigh the differences (e.g. Laska, Gurman, & Wampold, 2014; Wampold, 2000). This paper summarizes literature related to evidence-based practice in the interests of advocating an approach that pragmatically embraces ideas related to specific techniques and common factors. The discussion is then related to the psychotherapy integration movement, which is highlighted in the case study of integration within a range of therapies that have developed an impressive widely accepted literature demonstrating an evidence base, the three waves of behavior therapy. The discussion then relates to the psychotherapy integration movement (highlighted in the case study of integration) within a range of therapies that have developed a widely accepted literature demonstrating an evidence base: the three waves of behavior therapy. Finally, within this context, implications are drawn within the current literature on evidence-based practice.

Evidence-Supported Therapies: Specific Techniques

A brief review of the literature comparing the two major forms of therapy related to specific techniques will be presented. Clear criteria have been documented as Wachtel (2010) who highlighted studies that only include subjects with a single diagnostic category to eliminate confounding factors, studies that have a manual or a well-outlined ritual to validate the specific techniques being used, and studies that conform to randomized controlled trials (RCTs) to ensure that there are no systemic differences between subjects placed in the experimental and control groups. If therapies comply with these criteria, and can repeatedly demonstrate effectiveness in pre-and post-tests, then the therapy is likely to be deemed an EST. Meyer and Scott (2008) in defining an EST, used the criteria that “. . . efficacy had been established in two or more carefully designed and methodologically reliable randomized controlled trials that evaluate the treatment of a specific disorder” (p 685). While they noted the concerns related to using criteria similar to drug studies, they justified their rationale by noting that these standards were, in cognitive behavior therapy (CBT) more than by other therapies, acceptable as an evidence base.

While many therapies have some evidence supporting them, one key difference is based on the decision to rely upon cognitive or affective theories and practices. This basic decision results in two major ways of working with clients that often conceptualized as cognitive-behavioral (CB) or psychodynamic-interpersonal (PI) treatments (Blagys & Hilsenroth, 2000). Thus, in comparisons of specific therapies, these two theoretically different ways of working with clients are often pitted against each other. In a comparison study of interpersonal psychotherapy (IPT) and CBT, Lambert and Ogles (2004) concluded that there were no differences in the effectiveness of the two treatments for clients with both less and more severe pathologies. In comparing a wide range of studies, Shedler (2010) concluded that the effect sizes for therapeutic outcome are the same for psychodynamic therapy as for other forms of therapy, including CBT. Furthermore, when Ahn and Wampold (2001) conducted an analysis of the various components of therapy during a more than 18-year period, they found that removing or adding specific components of therapy did not change the overall effectiveness of the treatment. Thus it seems that even when comparing the overall effectiveness of specific techniques, there are very few differences between a cognitive and an affective focus.

Blagys and Hilsenroth (2000) further studied this issue of differences between cognitive and affective strategies and identified seven activities used by therapists: PI therapists attended to affect or emotion and its expression in the session, concentrated on the process in the therapy room that was related to factors facilitating or hindering progress, explored client’s experiences and relationships, focused on past experiences, attended to the client’s interpersonal experiences, explored the client-therapist alliance, and attended to dreams, fantasies and wishes of the client. In a subsequent study, they (Blagys & Hilsenroth, 2000) found that CB was distinguished from PI in the following ways: a use of homework and outside activities, an active direction of the session by the therapist, the teaching of skills in the session to help clients cope with presenting problems, an emphasis on client’s future experience, the therapists presenting information about the client’s disorder, and a focus on cognitive interpersonal factors. However, straightforward this may appear, experimental evidence suggests that it is not quite so simple. The process of therapy was examined in the National Institutes of Mental Health study in which CBT and IPT were compared. The researchers reported that despite adherence checks for the two separate therapies, there was substantial overlap between the two treatment types that had been described as different (Ablon & Jones, 2002). They concluded that all therapists were more closely aligned in their process to CBT, even though the content of what was discussed was quite different. They all tended to offer advice for how clients should operate between sessions, and were encouraging, rarely talking about the interaction between themselves and their clients. Thus, the evidence seems to suggest both approaches are linked to cognition and affect, noting considerable overlap between them, which could be accounted for as common factors.

Evidence-Supported Therapies: Common Factors

Common factors between therapies have frequently been implied through study results that do not find much difference in effectiveness when two or more therapies are compared. While Lamabert and Ogles’ (2004) substantial review of a range of therapies concluded that all therapies seem to be effective regardless of the specific techniques used (a good proportion of clients improve after 10 sessions), 75% of clients only met more conservative standards of improvement after 50 sessions, and limiting treatment to 20 sessions meant that 50% of the clients did not make substantial gains. They also reported that about 5% to 10% of clients became worse following their involvement in therapy. While calling into question the effectiveness of limited clinical trials in many effectiveness studies, this conclusion supports the argument for the effectiveness of common factors that appear to exist among therapies.

For major comparisons among effectiveness studies, meta-analyses have been used to compare results. Each study is accepted into the analysis based on a pre-determined set of criteria as to what must be present in the study design for it to be included in the statistical analysis. The results in meta-analytic studies that indicate little difference among models are seen as evidence for common factors. There have been a number of extensive reviews of the effectiveness of psychotherapy that have used this process of meta-analysis (e.g. Weston, et al., 2004; Kazdin, 2007; Lambert & Ogles, 2004), including a consideration of the comparative processes of diverse therapies (Blagys & Hilsenroth, 2000). One of the first major studies involving meta-analysis reported an effect size of .85 on 475 psychotherapy effectiveness studies using a range of therapies (Smith, Glass, & Miller, 1980). Effect size is the difference between treatment and control groups expressed as standard units of deviation. An effect size of .5 is considered a moderate effect size (Cohen, 1988). Lipsey and Wilson (1993) reported on 302 meta-analyses of behavioral, educational, and psychological treatments, which indicated extremely positive results. Refining their study to a smaller sample of studies with more rigorous criteria, they reported an effect size of .47 on a sample of 156 therapy outcome studies. Matt and Navarro (1997) suggested that some of the problems with meta-analytic studies included a limited range of people engaging in therapy and very few studies involving therapy in applied settings rather than laboratory settings. Overall, therapy appears to be effective according to most reviews of psychotherapy.

While these reviews generally state that therapy is effective, the conclusions do not seem to indicate why this is the case. However, many therapies give similar procedures different names (Marks, 2000). Furthermore, many of the reviews looked for differences among therapies rather than for the commonalities in asserting claims of effectiveness. In an age where there are hundreds of types of therapies, and our instruments are very blunt in measuring their effectiveness, it is important to concentrate on the commonalities that may account for similarities.

Common Factors in Therapies: Process and Content Variables

One of the most salient factors linking therapies at the process level is the client-therapist relationship. The therapeutic alliance has been an important ingredient in the process of change and has consistently been reported as a direct relationship in that the stronger the alliance, the more effective the therapy (Horvath & Bedi, 2002; Norcross, 2002; Orlinsky, Ronnestad, & Willutzki, 2004). The quality of the client-therapist relationship established very early in the treatment was shown to be very important in outcome, regardless of whether the focus of treatment was CB or PI (Blatt & Zuroff, 2005). However, there could be confounding variables related to the client-therapist relationship, as the more effective the therapy, the more likely it is that the client will view the therapist favourably (Blatt & Zuroff, 2005).

One attempt at thinking more objectively about the client-therapist relationship has been the work on motivational interviewing (MI). The method drew on Rogers (1959) empathic style, while consciously moving in the direction of resolving the ambivalence (Miller & Rollnick, 2009). Asking questions that allowed clients to confront their own ambivalence, invite clients to defend arguments for change, were thought to increase the client motivation towards a greater commitment to action. Thus, the process appeared to be aligned with improving the client-therapist relationship as an important element. Motivational interviewing focussed on readiness for change through a careful process of not challenging clients (Hettema, Steele, & Miller, 2005). It is interesting to note that Hettema et al. (2005) found that in studies where there was no manual for MI, the effect size was double when compared to studies where the treatments were manualized. In further analysis, Hettema et al. (2005) identified that the manual was clearly focussed on bringing clients to the point of change within one session, whether or not these clients were ready for the change which therapists hoped to implement.

The client-therapist relationship is also influenced by the level of directness of the therapist, and this has an impact on therapeutic compliance. Resistance was the focus of a study that noted differences in therapist behavior related to teaching and confronting (Chamberlain, Patterson, Reid, Kavanaugh, & Forgatch, 1984). In their study, they reported that facilitative and supportive therapist behavior led to less client resistance than teaching and confronting therapist behavior. Miller and Rolnick (2009) were quite clear that MI was not a school of therapy or in any way a comprehensive approach to treatment. Thus, MI was originally defined as a common factor that could be aligned with any therapeutic strategy that appears to be focussed on the therapeutic alliance.

There is also a common factor in therapy related to the content of therapy, that is the principle of exposure. This is drawn from Marks (2000) who observed that many different therapies have different names for similar procedures. Carey (2011) highlighted treatment principles rather than treatment strategies as a way of discussing active ingredients of change, and he suggested that exposure is a principle common to many therapies. Although exposure originally was used as a treatment for anxiety, Carey believed it was possible to extend its use with any uncomfortable thought, feeling, or situation that clients confronted. While we may not know why therapy is effective (Kazdin, 2009), we know that the common principle of exposure seems to appear in many therapies, including paradoxical intention (Frankel, 1975; Watzlawick, Weakland, & Fisch, 1974), systematic desensitization (Wolpe & Wolpe, 1988), emotionally focussed therapy (Greenberg & Paivio, 1997), self-psychology (Kohut, 1971), and interpersonal psychotherapy (Gillies, 2002), just to name a few.

There is a direct relationship between mental health and emotion regulation (Davidson, 2000; Gross, 1998). The strategy of avoiding emotion can be useful at times in the short term, but when this strategy is continued, it creates greater life problems, as it may lead to resignation, suppression, repression, substance abuse, dissociation, self-harm, and disengagement (Ottenbreit & Dobson, 2004). However, emotion regulation can also lead to preoccupation with the emotion, which includes rumination, obsession, and compulsive behavior. Thus, both over-and under-involvement with emotion can lead to problems. Therapy involves recognizing, labelling, and differentiating emotions through a process that includes beliefs about control of emotion, expression of emotion, and validation of these emotions (Leahy, 2007). These ways of dealing with emotion are similar to the processing of emotions in therapy, which leads to creating new meaning rather than becoming overwhelmed by the emotion (Hayes, Beevers, Feldman, Laurenceau, & Perlman, 2005). Hayes, Beevers, et al. (2005) also suggested that emotional processing is a central variable for change, which is independent of particular models of therapy. Various theorists use a range of terms for this process, including emotional processing (Foa & Kozak, 1986; Teasdale, 1999), cognitive-emotional processing (Rachman, 2001), and experiencing (Greenberg, 2002). It has been suggested that an emotional processing approach may provide integration for theories that have appeared to be incompatible in the past (Westen, 1994). Hayes et al. (2004) also noted the importance of dealing with emotion when they suggested that acceptance and commitment therapy (ACT) encouraged exposure to actively experiencing emotion moment by moment, as can be achieved through mindfulness. The processes of mindfulness involved labelling of the feeling, which was associated with higher satisfaction with life and greater emotional regulation (Baer, Smith, & Allen, 2004). Common factors appear to account for significant amounts of the change, and can be considered a form of integration. The psychotherapy integration movement has long been instrumental in the search for commonality between therapies, and is perhaps best illustrated through the three waves of beahvior therapy.

The Three Waves of Behavior Therapy: A Movement Towards Integration

Psychotherapy integration has been an ambitious objective, starting with an attempt to integrate Freud and Pavlov at a conference of the American Psychiatric Association in 1932 (French, 1933). At that early stage of identifying differential treatments, the idea of integration was extraordinary and clearly had very little support. Key ideas on integration were not seriously considered until the 1980s (Norcross & Goldfried, 2005). Three principle components of integration have been delineated, which are the therapeutic relationship, meaning-making, and change promotion (Castonguay & Beutler, 2006).

The first wave of behavior therapy (BT) was associated with the scientific respectability of learning theory and the body of research that supported this theory (Hayes, 2004). This therapy relied heavily upon operant conditioning, creating change by considering behavior within the context of its antecedents and consequences. Thus, it moved well beyond the Freudian inspired therapy of the day, which lacked scientific respectability. Consequently, to provide scientific bases for BT, change strategies were implemented to act directly upon symptoms, thus creating a more obvious link between therapeutic strategies and client change in a bid for evidence-based practices. These approaches included operant conditioning, exposure therapy, arousal reduction interventions, skills training, and behavioral activation (Taylor, & Steel, 2008).

The second wave of behavior therapy was associated with client cognition through emphasis on CBT. Private thoughts were defined as objective, observable behavior, providing the required scientific bases for altering these thoughts through direct challenge of some thoughts and substitution of other more appropriate or helpful thoughts. This very quickly spread into a range of cognitive therapies associated with Beck (1976), Ellis and Harper (1975), and Meichenbaum (1977). The scientific respectability of CBT was related to its base in learning theory, with an emphasis on techniques to directly alter client beliefs. Cognitive behavior therapy challenged the validity of client thoughts, rather than attempting to understand them within the context of these thoughts. Thus, emotional processing was not necessarily the first line of action, as attempts were made to modify the beliefs associated with the emotion, thereby changing the antecedent emotional regulation strategies (Hofmann & Asmundson, 2008). The strategies included behavioral activation, cognitive restructuring, psycho education, and relaxation training. Using questions was particularly important in CBT, and reflected a change in therapist attitude towards confrontation (James & Morse, 2007). Socratic questioning was used to help clients achieve insights that were derived from their answers to these questions (James, Morse, & Howard, 2010). Thus, these questions helped clients discover inconsistencies in their own thinking, and provided them with new insights (Overholser, 1993). Consequently, rather than simply eliciting information from a client so that the therapist could directly challenge thoughts, these questions provided information to the client (Brown, 1997; James, Morse, & Howarth, 2010). This form of questioning, drawn from a broad range of therapeutic approaches, demonstrated some movement towards integration and also allowed for a less directive form of therapy because clients challenged their own thinking through the answers they provided to questions asked.

The third wave of behavior therapy was associated with Hayes’ (2004) acceptance and commitment therapy (ACT), and made even more moves toward psychotherapy integration. However, there was some controversy as to whether ACT was beginning a new third wave or, according to Zettle (2005), was another approach of a group of therapies that superseded the second-wave CBT therapies, such as functional analytic psychotherapy (Kohlenberg & Tsai, 1991), dialectical behavior therapy (Lineham, 1993), integrative behavioral couples therapy (Jacobson & Christensen, 1996) and mindfulness-based cognitive therapy (Segal, Williams, & Teasdale, 2002). ACT focused on the contextual and functional meaning of behavior through its emphasis on functional contextualism and associated research on language and cognition (Hayes, 2004). Hayes (2004) stated for example, that anxiety was not the problem, but the tendency to fight against that anxiety, emphasizing the contextual nature of the problem, much in line with many of the systemic therapies in the 1970s and 1980s (e.g. Watzlawick, Weakland, & Fisch, 1974). As Hayes (2004) stated: “. . . Thus ACT relies heavily on paradox, metaphors, stories, exercises, behavioral tasks, and experiential processes, while logical analysis has a relatively limited role” (p 652). He also suggested that ACT was breaking down some of distinctions between BT and some older, less researched therapies. With regard to emotion regulation, it was suggested that ACT relied upon response-focussed emotion regulation strategies (Hofmann & Asmundson, 2008). The list of techniques were drawn from experiential therapies, mindfulness, gestalt therapy, and a range of other sources including psycho-education, problem-solving, behavioral homework, skills building, exposure and other interventions developed in the first and second waves of behavior therapy (Hayes, Strosahi, Bunting, Twohig, & Wilson, 2004; Singh, Lancioni, Wahler, Winston, & Singh, 2008). The three waves of behavior therapy are important in the development of significant advances in psychotherapy integration. While the first wave differentiated itself from Freudian therapy with an emphasis on behavior change, the second wave reconnected itself with a more insight-oriented therapy by defining and exploring cognitions such as behavior and also represented an integration between the first and second waves. Finally, by the third wave, there was a recognition that therapy could not reside with only one perspective. These ideas coexist with ideas associated with the target of intervention, which encapsulates therapies past and present where therapeutic strategies move between changing the targeted behavior and the context around the targeted behavior. Thus, a form of integration is becoming more apparent in the prime example of an EST. The three waves of BT have moved from a focus on the specific behavior to the context of the behavior, which was fundamental in the change strategies that evolved in the systemic therapies, where first-and second-order change highlighted an important distinction in explaining the mechanism of therapeutic change. The target of intervention: the actual behavior or the context of the behavior, has also highlighted an important principle of change through the distinction between first and second order change.

First-and Second-Order Change

Ideas related to first-and second-order change have been emphasized in the therapeutic literature for many years, probably initiated through psychological treatment strategies related to therapeutic paradox (Asher, 1989). The first reported use of this strategy was by Duncan (1928) when he described a process by which a patient could be rid of an unwanted behavior by practicing it in a particular prescribed manner to bring it under control. The technique was also advocated by Frankel (1955), and further elaborated upon in a paper on paradoxical intention (Frankel, 1975). These strategies were popularized by work associated with the Bateson project at Palo Alto, California that was designed to study the communication patterns in families of schizophrenics. These strategies were documented in one book (among others) by Watzlawick, Weakland, and Fisch (1974) in the early days of the family systemic therapy movement. This movement, along with BT were radical attempts to deal with the elaborate and lengthy psychodynamic therapy of the day.

Watzlawick et al. (1974) described first-order change as the change that occurred when direct approaches were used to deal with a problem by focussing directly on that problem in a linear manner. For example, if a client could not sleep, the therapist, in a straightforward manner, would advocate direct approaches and suggest techniques that would facilitate more sleep. Watzlawick et al. (1974) also documented second-order change, which was change that considered context rather than the particular behavior as in first-order change. For example, with a client who couldn’t sleep, a more indirect approach was used, with suggestions that the client should try to remain awake instead of trying to get to sleep. This approach emphasized the context in which anxiety about sleeping was probably a limiting factor in getting to sleep, which is very closely aligned to principles presented in ACT. By removing the added anxiety that was incompatible with sleep, and substituting a task that could more easily be accomplished, a new context around the sleep problem was created that would be easier to achieve, thus decreasing the anxiety (Brown, & Slee, 1988). Thus, the concepts of first-and second-order change were applied in many therapies related to systemic therapy practiced during the 1970s and 1980s, and which retained the second-order perspective of psychodynamic therapy, but provided much more efficient change strategies. Therapies that rely upon first-order change are much easier to evaluate using RCTs than are therapies that focus on second-order change as the contexts of many problems are difficult to measure adequately. Yet, there seems to be something very important about including the context in explorations of client problems. Of course, one of the most prominent examples of the context in therapy is the client-therapist relationship, particularly where it is seen as an active ingredient of change, as discussed above.

Implications for Clinicians

There are a number of implications that emerge from this analysis for the clinician who wants to ensure the most effective evidence-based treatment:

At first glance, these implications appear to be at odds with an evidence-based approach. However, it is clear that all of the above principles have been linked to studies that support an evidence-based approach. It is only when we limit consideration of EST’s to those that examine specific techniques that we fail to notice the impressive evidence for commonalities among therapies. The general strategies listed above can certainly be called evidence-based and may be used in conjunction with most therapies that include specific problems that can provide additional evidence through appropriate searches (Falzon et al., 2010), in organizations where specific techniques have been adopted (Williams et al., 2011), and certainly with very complex cases (Magnavita et al, 2010). The general strategies listed above can certainly be called evidence-based and may be used 1) in conjunction with most therapies that include specific problems that can provide additional evidence through appropriate searches (Falzon et al., 2010), 2) in organizations where specific techniques have been adopted (Williams et al., 2011), and 3) certainly with very complex cases (Magnavita et al, 2010). Thus, these principles link with a range of current evidence-based practices and they also satisfy the ethical imperatives required of psychologists and other therapists.