Using Motivational Interviewing to Manage Process Markers of Ambivalence and Resistance in Cognitive Behavioral Therapy
- 794 Downloads
Resistance can be a substantive problem that limits treatment efficacy in cognitive behavioral therapy (CBT). Resistance often arises in a context of client ambivalence about change and has been consistently found to be exacerbated by directive responses and reduced by supportive ones. When not properly managed, resistance can have detrimental consequences for the therapeutic alliance and both proximal (e.g., engagement, homework compliance) and distal (e.g., symptom severity) treatment outcomes. Given its impact, resistance should be considered a key process marker in CBT. Motivational interviewing (MI) provides strategies for ‘rolling with resistance’ and there is mounting evidence that it can be successfully integrated with CBT to improve overall treatment response. This paper will review the research on resistance and ambivalence, particularly research conducted in the context of CBT. MI and its ability to successfully address these impasses will be outlined in detail, together with a clinical illustration.
KeywordsCognitive behavioral therapy Psychotherapy Process markers Resistance Ambivalence Homework noncompliance Client engagement Therapeutic alliance
Motivational interviewing (MI) is a brief client-centered, directive method for enhancing intrinsic motivation for change (Miller and Rollnick 2012). While MI is an empirically supported treatment for substance abuse, extending it to the treatment of other major mental health problems is clearly appealing to clinicians and researchers, and conceptual and empirical work in these areas is advancing quickly (e.g., Arkowitz et al. 2008; Westra 2012). MI is likely appealing since it addresses important clinical problems (i.e., resistance, ambivalence, lack of engagement in treatment), and can complement, rather than replace, existing treatments. For example, MI is rapidly gaining momentum as an adjunct to, or integrated with, cognitive behavioral therapy (CBT), most recently in the area of anxiety treatment.
Change is a turbulent process that is often fraught with competing and opposing feelings. Although clients come to treatment because they desire change, they often also simultaneously fear and resist change. This ambivalence can set the stage for tensions and breakdowns in the working alliance (i.e., resistance, alliance ruptures). Such tensions often manifest as homework noncompliance, debates and disagreements, high levels of expressed emotion toward the therapist, ignoring and sidetracking, or disengaging behaviors such as passivity, withdrawal or frequent use of “I don’t know”, among other forms. The risk of encountering these impasses is arguably higher in more directive treatments, such as CBT, and this has important clinical implications.
In surveys of CBT practitioners, therapists consistently identify resistance and related variables (e.g., noncompliance) as key barriers responsible for limited treatment efficacy (Kennard et al. 2005; Szkodny et al. 2014). As well, CBT therapists have suggested that resistance is one of the main reasons for the effectiveness gap between research and clinical practice (Amodeo et al. 2011; McAleavey et al. 2014). And, in qualitative studies of alliance, clients identify disagreement with the therapist as one of the most important events in therapy (Viklund et al. 2010). Overall, both clients and therapists acknowledge that resistance can be a substantive obstacle to successful treatment.
A major contribution of motivational interviewing (MI) to clinical practice resides in its specific consideration of resistance and ambivalence, and the provision of accessible and effective means of understanding and working with these important events. Indeed, there is mounting empirical evidence from process research and clinical outcome trials that adding and integrating MI into CBT results in the creation of a more harmonious, engaging, and productive therapy environment (Aviram and Westra 2011; Constantino et al. 2016; Westra et al. 2016). Thus, MI is beginning to occupy an important place in the practice of CBT in particular, specifically for the management of resistance and low motivation/ambivalence which can often derail therapy. Looking forward, MI may also provide a way to facilitate the training of CBT practitioners by strengthening their ability to maintain the therapeutic alliance in the face of resistance and ambivalence.
This paper will focus specifically on the process markers of ambivalence about change and the resistance to which it can give rise. We refer to them as process markers given that their effective identification on a moment-to-moment basis (i.e., ongoing attunement) in therapy is necessary as a precursor to effective responses for processing ambivalence and reducing resistance. We begin by discussing what resistance is and the context (client ambivalence) that gives rise to it. Next, we review the evidence supporting the importance of ambivalence and resistance to subsequent treatment processes and outcomes. Much of the research on extrapolating the study of motivational language and resistance to a CBT context has been conducted by our research group and so we will focus on this work, which occurs in the context of CBT for generalized anxiety disorder (GAD). Finally, we outline the use of MI to navigate ambivalence and resistance, and summarize the current evidence for the effectiveness of adding or integrating MI with CBT.
What is Resistance?
Figure 1 provides an illustration of how resistance typically arises. In a nutshell, resistance is the product of two interacting forces: the client’s ambivalence about change and how the therapist responds to that ambivalence (Moyers and Rollnick 2002). In particular, a therapist being directive in the context of client ambivalence very predictably results in opposition and tension such as “Yes-but…”, “I can’t”, or counterargument. For example, the therapist makes a suggestion for homework and the client responds with lack of enthusiasm or disagreement. Or, the therapist might ask a Socratic question to prompt an alternative way of viewing a situation and the client ignores, interrupts or sidetracks. Alternatively, the client might roll their eyes or be silent, dismissive, or overtly critical of the therapist.
While CBT strives to be collaborative, in the presence of client resistance, therapist reactions to noncompliance, disagreement etc., can often take the form of “convincing or persuading”. Unfortunately, resistance dynamics are typically perpetuated and sustained by the therapist continuing to be directive and advocating more strongly for change or ‘problem solving’, in an effort to get the client to comply or agree with the therapist’s preferred direction or line of thinking. In MI terms, the client and therapist caught in this cycle end up ‘acting out’ the client’s ambivalence (each taking a side) rather than working productively through it. And here, the therapist ends up being the one who ‘takes all the good lines’ (Rollnick et al. 2007) by making the arguments for change, while the client is unfortunately put in the position of arguing against change. In other words, the therapist, inadvertently, ends up causing the client to repeatedly articulate their objections to change and why they cannot change. That is, the client is placed in the position of making “counter-change talk” statements (arguments against change), rather than “change talk” (arguments in favor of change) statements.
Thus, a higher frequency of signals of resistance needs to be seen as feedback that reflects a tension or breakdown in the working alliance (that the therapist then needs to take action to correct in order to reestablish collaboration). This is why sustained client resistance in MI is actually considered a clinician skill error. Interestingly, supportive (not directive) therapist responses at these moments, such as those in MI (e.g., empathy, underscoring autonomy), end up being the way out of the trap. For example, Aspland et al. (2008) found that alliance strains often arise in CBT in the context of therapist demand (e.g., convincing or persuading) and are only corrected when the therapist realigns to understand the client’s viewpoint. This concept of ‘rolling with resistance’ will be elaborated upon later on in this paper.
Ambivalence and Resistance as Empirically Supported Process Markers
As we will review below, there is mounting and consistent evidence that both ambivalence, and the resistance to which it gives rise, are strong predictors of both proximal and distal psychotherapy outcomes. Process research in MI (typically in the addictions domain) supports the importance of client motivational language within MI (Miller and Rollnick 2012). Interestingly, recent research extrapolating the investigation of client motivational language to a CBT for anxiety context also suggests that early client in-session statements regarding change have strong predictive value. In these studies, while early change-talk (CT; arguments in favour of change) has been found to be unrelated to outcomes in CBT for GAD, early client arguments against change (or counter-change talk: CCT) are highly significant predictors. Namely, a greater frequency of arguments against change in the first session of CBT for generalized anxiety disorder (GAD) has been found to be a substantive predictor of lower homework compliance and poorer treatment outcomes across two randomized controlled trials (Button et al. 2014; Lombardi et al. 2014; Goodwin et al. 2015). Moreover, a higher frequency of early arguments against change differentiated those who went on to experience an alliance rupture later in therapy (major disruption in the therapeutic alliance as defined by significant drops in client alliance ratings), from those who did not experience an alliance rupture (Hunter et al. 2014).
In addition, recent research on early CCT has found it to be a significant moderator of outcomes in CBT for GAD. In particular, for those high in ambivalence (high CCT in session 1 of treatment), motivational interviewing integrated with CBT (MI-CBT) is superior to CBT alone in reducing worry up to 1 year post-treatment (Button et al. 2016). In contrast, CBT alone was equivalent or even slightly better to MI-CBT for those who were highly motivated at the outset (low in CCT). Thus, individual differences in client ambivalence are important early markers indicating a need for different treatment approaches. In particular, as outlined earlier, being directive, or pushing for change or action with clients who are ambivalent (‘stuck’) is counter-productive.
While ambivalence in and of itself is not problematic, it becomes problematic when not approached with supportive, empathic responses. That is, the mismanagement of ambivalence gives rise to the toxic dynamic of resistance. In support of this, Sijercic et al. (2016) partitioned client CCT into statements against change meant to oppose the therapist (i.e., CCT statements made in the context of resistance or non-collaboration) and those that were mere expressions of ambivalence (and occurred in a harmonious interpersonal context and thus represented mere disclosure). Interestingly, CCT only predicted outcome when it was uttered in order to oppose the therapist. In contrast, arguments against change that represented mere disclosures of ambivalence (and which did not occur in a context of opposition) were unrelated to outcomes. This again suggests that ambivalence (stuckness) is not a problem in and of itself and that whether or not it turns into resistance (argument, disharmony) depends on how it is managed by the therapist.
Resistance & Outcomes
Resistance essentially reflects a lack of collaboration between client and therapist, and is a robust predictor of both early termination and negative treatment outcomes (Beutler et al. 2011; Piper et al. 1999; Strupp 1980). For example, higher levels of resistance occurring as early as the first session of therapy have been shown to strongly predict poorer client outcomes, even up to 1-year post-treatment (Aviram and Westra 2011). Notably, there is also evidence from mediational analyses that resistance plays a direct role in influencing treatment outcomes. Specifically, a clinical trial comparing the efficacy of CBT alone to MI-CBT for GAD, found substantially lower levels of resistance in MI-CBT at mid-treatment, which in turn fully accounted for group differences in treatment outcomes at 1-year follow-up (Constantino et al. 2016).
In terms of proximal outcomes, there is strong evidence that resistance is associated with reduced subsequent engagement, such as lower levels of in-session task involvement in CBT (Jungbluth and Shirk 2009) and of later homework compliance (Aviram and Westra 2011; Hara et al. 2015). Perhaps most importantly, there is recent evidence that higher levels of resistance are associated with a drop in client outcome expectations (Mamedova et al. 2016). This reduction in optimism about treatment, in turn, has been found to account for the deleterious impact of resistance on CBT outcomes. In other words, resistance (arguing, debating, etc.,) is demoralizing. Interestingly, Mamedova and colleagues found that resistance is demoralizing for both clients and therapists, but that only the lowering of client’s outcome expectations adversely influences treatment outcomes. In short, experiencing such alliance tensions (arguing, debating, ignoring, etc.) with the therapist negatively impacts the willingness of clients to continue to believe that treatment will be helpful to them, and maintaining hope or positive outcome expectations is an important common ingredient for therapy success (Constantino et al. 2011).
Resistance & Alliance
Resistance has also been shown to have a damaging effect on the therapeutic alliance. For example, a study by Watson and McMullen (2005) found that increased resistance within a session was associated with lower client-ratings of the working alliance for that session. Similarly, a study based on data from a clinical trial of CBT for GAD found a significant relationship between higher levels of observed resistance and lower post-session client alliance ratings (Hara et al. 2015). Relatedly, Hara et al. (2016) have found that in-session resistance was negatively correlated with client post-session ratings of therapist empathy. And, this is consistent with the observations of several studies that therapists become markedly less supportive and empathic during moments of resistance (e.g., Castonguay et al. 1996). For example, in an experimental study, Francis et al. (2005) found that manipulating levels of client resistance (by instructing actors portraying patients to behave in a resistant manner) caused clinicians to display significantly more negative confrontational behaviors, such as interrupting or directly disagreeing with clients. Furthermore, qualitative analyses indicated an associated decrease in positive therapist behaviors such as offering praise or encouragement compared to conditions of high motivation (low resistance).
There is also a related body of research regarding a very similar construct in the literature - alliance ruptures. Ruptures, or strains, to the therapeutic alliance are moments of deterioration in the quality of the relationship between therapist and client, such as the tension following a client’s communication that they are dissatisfied with their therapist (Safran and Muran 2000). Like resistance, alliance ruptures predict a wide array of psychotherapy outcomes. For example, Muran et al. (2009) linked the presence and intensity of alliance ruptures to the risk of early termination. In particular, they found that alliance ruptures were more common in dyads in which clients dropped out of therapy prematurely and that intensity of the rupture was positively correlated with the likelihood of dropout. Similarly, Cumming et al. (2011) found that unresolved ruptures directly contribute to the risk of client dropout. As well, research focused on outcomes during therapy has shown that alliance strains are associated with subsequent decreases in client ratings of session quality (Muran et al. 2009) and outcome expectations (Westra et al. 2011).
Apart from their ability to adversely impact client outcomes, such deteriorations or impasses in the therapeutic alliance have also been associated with negative therapist attitudes. Coutinho et al. (2011) observed that, following a rupture, therapists tended to experience negative emotions, such as confusion or guilt regarding their ability to help their client. In the same study, some therapists even reported feeling anger toward their clients. This finding is in line with research indicating that therapists are more likely to blame clients (e.g., for lack of motivation) following moments of conflict or hostility (Binder and Strupp 1997). Though it is often assumed that therapists possess the natural ability to build and maintain an alliance with ease, findings such as these have led researchers like Binder and Strupp (1997) to suggest that this assumption is far too hasty. Altogether, research consistently shows that, especially when left unresolved, alliance ruptures, like resistance, have far-reaching consequences which negatively affect both clients and therapists.
Resistance & Therapist Performance
Moreover, in addition to damaging the alliance, resistance also negatively impacts therapist performance. For example, resistance has been found to have the capacity to derail CBT therapists. Boswell et al. (2013) observed that overt client hostility during sessions was associated with lower observer-rated therapist adherence and competence. And Zickgraf et al. (2016) found that higher levels of resistance predicted reduced therapist adherence and increased reliance on techniques outside those prescribed by the CBT model. Consistent with this, CBT therapists report feeling anger and other negative emotional reactions toward clients that display higher levels of resistance (Westra et al. 2012). These findings have led researchers to suggest that CBT manuals are lacking in explicit guidance for coping with client hostility or resistance (Boswell et al. 2013; Westra 2012; Zickgraf et al. 2016).
In summary, given the strong capacity of interpersonal resistance to predict outcomes and subsequent process, the presence of such resistance should then serve as a critical marker in therapy. Stated differently, not all moments are equally important in the therapy process and the identification of key moments (even if relatively rare) of tension in the therapy alliance and client disengagement seem to be very important. In their review, Orlinsky et al. (1994) identified active client involvement with the process of treatment as among the most critical contributors to treatment outcomes. This result is consistent with other studies finding that alliance ruptures are associated with poorer treatment outcome across a range of therapies, including CBT (Safran et al. 2011; Westra et al. 2011).
Using MI to Roll with Resistance
As articulated by Sheldon et al. (2003), clinicians who are technically proficient and knowledgeable about methods of facilitating change/action, will often find themselves impotent if they are unable to build motivation and help clients work through their conflicting, powerful, and often contradictory feelings about change. And through watching many hours of videotape of resistance in CBT, we have noticed that therapists can often see that there is ambivalence there (that the client is ‘stuck’), and even be able to see that the client has conflicting beliefs about change, but CBT therapists often seem to lack the tools to know how to move the client’s dilemma productively forward. Becoming proficient in MI can equip CBT therapists with such tools for identifying, understanding, and navigating resistance; thereby not allowing ambivalence about change to devolve into toxic impasses and interchanges that derail collaboration, threaten the alliance, and imperil outcomes. Let us first begin with considering what MI is and then discuss how it can be used specifically to address ambivalence and resistance.
What is MI?
Any discussion of MI should begin with the “MI spirit” or client-centered nature of the approach, since this is considered essential to the effective use of the method. MI without the underlying spirit is like words without music and is not considered MI (Rollnick and Miller 1995). MI is an evolution of the client-centered therapy explicated by Carl Rogers (1951, 1965) who emphasized empathic understanding of the client’s internal frame of reference, and therapist communication and provision of core facilitative relational conditions for client growth and change, including accurate empathy, unconditional positive regard, and therapist genuineness or congruence (Rogers 1957).
Like client-centered therapy, MI stresses the essential importance of the development of a safe, collaborative atmosphere in which the client can sort out his or her conflicting and often contradictory views of change. In this sense, MI converges with the client-centered tradition of prioritizing the therapeutic relationship as an essential vehicle in which greater self-awareness can be developed and new meanings generated. Roger’s emphasis on the importance of the therapeutic relationship has since been supported by decades of research on the importance of relatedness, including research on the therapeutic alliance (Constantino et al. 2002; Horvath and Symonds 1991), attachment (Cassidy and Shaver 1999), and the necessity of caring, affection, and interpersonal safety for facilitating exploration and new learning (Gilbert 1993, 2010). In client accounts of their experiences of MI, therapist empathy, and the provision of safety and freedom to explore, emerge as prominent aspects of the approach (Marcus et al. 2011).
Consistent with its client-centered roots, Miller and Rollnick have emphasized that MI is fundamentally a way of being with clients. This emphasis is consistent with Roger’s view of empathy as an attitude or way of being rather than a specific technique per se (Rogers 1980). The attitude one holds toward the client (prizing, unconditional regard, warmth, genuineness, client as expert, etc.) is more pivotal in MI than are the specific techniques. Stated differently, MI is not a set of techniques. MI cannot be distilled into a set of questions or techniques one can memorize and regurgitate in the absence of this fundamental spirit or attitude. In other words, techniques can never be disembedded from their relational context, which is of paramount importance in MI. In fact, MI without a manual tends to be more effective than structured MI with a manual (Hettema et al. 2005). Similarly, MI cannot be equated with any particular method; it is not the sum of its constituent parts. Rather, any technique (e.g., decisional balance, importance and confidence ratings, forward looking, etc.) is merely an expression or instantiation of the underlying spirit and objectives of MI. Even decisional balance, a technique to which MI is often (incorrectly) equated, for example, is not an exercise that one completes and is not even mandatory; rather it is merely a convenient and potentially useful heuristic for advancing therapists’ (and therefore clients’) understanding and exploration of ambivalence about change.
A major component of expertise in therapy is knowing what to reflect and how to steer the conversation. What the therapist watches for is based on their theoretical model, and with experience therapists may develop multiple lenses that guide them and tell them what is important and key in a therapy session. In the case of MI, this involves client motivational language (change-talk, counter-change talk or sustain talk, ambivalence), moments of disharmony or disagreement, and signals of readiness for change. Thus, a major aspect of the ‘directive’ nature of MI is observational capacity for key markers of motivation in language and process, and then responsively navigating these with indicated supportive, MI-consistent strategies.
Observing Ambivalence and Resistance
In order to effectively navigate ambivalence and resistance, one must first be able to accurately observe these important process markers on a moment-to-moment basis. Recognizing resistance both in the unfolding of the interaction (i.e., opposition to therapist direction; disharmony) and in client statements and language about change/treatment (change talk and especially counter-change talk) are two important sources of information about client readiness for change. Such ongoing attentiveness can provide valuable information about a client’s engagement in therapy and level of readiness to change, and consequently, their receptivity to taking action to change.
Important information about readiness can be gleaned from attending carefully to how clients talk about change, particularly objections to and concerns about change. Examples of ambivalent statements that capture conflict or competing motives of approach and avoidance include “I want to, but...” (e.g., “I want to be happy but I feel guilty, like I don`t deserve it”), “I know it doesn’t make sense, but I can’t stop seem to stop” (e.g., “I know worry doesn`t help but I feel like I have to do it”), or “Logically I know it, but emotionally...” (e.g., “I know I`m not worthless, but it feels like I am”). A high frequency of counter-change talk, or statements reflecting arguments against change (referred to as “sustain talk”), is a particularly important indicator of ambivalence and resistance to change. These include both arguments in favor of maintaining the status quo (e.g., “Worrying helps me feel prepared”, or “I’m less anxious if I avoid others”, or “I just feel too anxious if I don’t check”) and arguments against change (e.g., “It would be too hard for me to change”, or “I’m afraid of what would happen if I didn’t check”, etc.). In contrast, examples of arguments for change might include language such as “I can see that this will be helpful (ability & desire to change). Which I really need because this anxiety has gotten way out of control. I know it won’t be easy but if I’ve built up my thinking to be so negative, I can break it down too (ability to change).”
Of the various systems for observing resistance in process, the Client Resistance Code (CRC: Chamberlain et al. 1984, 1985) is not tied to a particular therapy approach. And we have adapted and evaluated this system for a CBT context (and research using this system is reviewed below). While resistance in the CRC can take many forms including confronting, challenging, disagreeing, blaming, complaining, defending self, sidetracking, ignoring, withdrawing, and interrupting, in the adapted system, a gestalt code of ‘opposition’ is used. That is, resistance is considered any behavior (verbal and often nonverbal) that opposes, blocks, diverts, or impedes the direction set by the therapist. Importantly, recent research suggests that accurately observing signs of resistance is easier said than done and that this skill cannot be assumed but rather must be explicitly trained (Hara et al. 2015).
To illustrate, consider the following example of some observations of resistance using the adapted CRC (Westra et al. 2009) in a clinical exchange. The client fears being late, and the therapist is attempting to help her reframe this experience in order to reduce her anxiety. Instances of resistance appear in italics.
T: So even if a person was late once in a while, that’s no good?
C: (pause) I don’t know (Challenging/Disagreeing). Maybe one class in a semester. Things happen, right? (physically restless). But I don’t know (pauses; Withdrawing). I guess that’s the way I see it. (Challenging/Disagreeing; Defending self).
T: And does everyone else see it that way?
C: (sarcastically) It doesn’t seem like it. People come in late all the time.
T: And does it affect their grades or...?
C: But it`s disruptive too!...you know. (Interrupting; Ignoring; Challenging/Disagreeing).
T: Sure it is. I’m not saying being late is a good thing. But is it that bad that you would spend time and energy worrying about it?
C: (very silently & quickly) hmm (Withdrawing).
T: So it`s kind of weighing the pros and the cons.
C: (very silently & quickly) hmm (Withdrawing).
Here the client uses a variety of different strategies (overt and covert) to communicate her unwillingness to go along with the therapist including interrupting, ignoring, disagreeing and defending her position, and, ultimately, passively withdrawing. These examples also illustrate how immediately responsive the interaction is to therapist demands when made in the context of low client readiness for change. Therapist directiveness in this context can very quickly and reliably elicit client resistance. Continually watching for such behaviors thus provides an immediate, readily available source of feedback to the therapist about the client`s level of readiness for change.
Responding to Ambivalence and Resistance
Just as therapists can perpetuate resistance, therapist responses can also be powerful tools for decreasing resistance. That is, ambivalence and resistance are not static qualities of clients, and resistance in particular is highly responsive to therapist inputs. Directiveness has been found to reliably increase client resistance, while supportive responses reduce it. For example, Patterson and Forgatch (1985) had therapists alternate within a session between “teach and confront” on the one hand, and “facilitate and support” on the other hand. The former increased resistance while the latter evoked greater co-operation. A similar finding comes from a study by Miller et al. (1993) in which clients with problem drinking were randomly assigned to therapists who used either a client-centered or directive-confrontational counselling style. The directive style was associated with significantly higher levels of resistance which, in turn, predicted poorer outcomes one-year post-treatment. Finally, Aviram and Westra (2011), as well as Constantino et al. (2016), found that the addition or integration of MI to CBT for GAD was associated with large reductions in observed interpersonal resistance. In other words, clients who received MI in these studies were visibly more engaged with the therapy process in CBT than were individuals who had not received MI. Moreover, post-therapy interviews revealed that clients who received MI prior to CBT themselves reported that they were more actively engaged in CBT and experienced their CBT therapists as more collaborative than did participants who did not receive prior MI (Kertes et al. 2011).
While, directiveness may be helpful in certain contexts, such as when the client is cooperative, it is detrimental in a context of resistance and ambivalence in which supporting the therapeutic alliance should be prioritized (Beutler et al. 2011, 2002a, 2002b; Ilgen et al. 2006). In general, MI strategies for navigating resistance reflect a spirit or attitude of “dancing rather than wrestling” with resistance. This involves reframing or shifting one’s view of resistance. In MI, resistance is not viewed as an obstacle to be defeated but rather as important information to be understood, validated, and integrated. In essence, at these times, the client is sending critical signals that he or she has important concerns that need to be heard and processed. As Miller and Rollnick (2002) discuss, the presence of resistance in the relationship should operate as a type of `stop signal`, indicating that the therapist is working ahead of the client`s level of readiness; placing demands on the client (to do, be, think something) that they are not ready for. That is, client resistance to therapist demands offer critical information and represent client efforts to protect their autonomy and reassert freedom of choice. The onus is on the therapist to recognize and hear such messages, in order to re-establish collaboration and harmony in the relationship.
As such, the ability to hear client resistance and disengagement, and to shift out of being more directive in the presence of doubts about treatment and change, is very important. For example, Burns and Auerbach (1996) note that continuing to use cognitive therapy techniques in the context of client anger or noncompliance can convey the message that the patient’s perceptions are irrational or ridiculous. Rather, they argue that, “When patients are stuck or angry or expressing strong negative affect, therapists need to set their cognitive and behavioral techniques temporarily on the shelf and respond in an empathic manner.” (p. 150). And Arkowitz and Westra (2004) note that if change-oriented strategies are used in the context of strong ambivalence about change, the therapist risks being perceived by the client as an advocate for change (i.e. hearing only one side of their ambivalence), similar to the stance often taken by significant others (e.g., family members, doctors). Finally, Aspland et al. (2008) concluded that, on noticing an alliance rupture, CBT therapists should become more empathic and responsive, switch focus to issues more salient to clients, and encourage clients to express their concerns, rather than continuing with technical interventions.
To further illustrate with the previous clinical vignette, we outline below some possible therapist responses intended to ‘roll with resistance’.1 Notice that using these tools (e.g. getting alongside) can often minimize resistance quite quickly, since resistance is a phenomena that is highly responsive to therapist inputs. Also, note that each of these responses would open up the dialogue/exploration (to help the client decide what she thinks/prefers) as opposed to shutting down the dialogue (because the therapist has decided that being late should be regarded by the client as fine).
T: So even if a person was late once in a while, that’s no good?
C: (pause) I don’t know (Challenging/Disagreeing). Maybe one class in a semester. Things happen, right? (physically restless). But I don’t know (pauses; Withdrawing). I guess that’s the way I see it. (Challenging/Disagreeing; Defending self).
T: It sounds like not being late is very important to you. Say more.
Or T: It’s all well and good for folks like me to say “being late shouldn’t be a big deal” but if I am hearing you right, for you it is a big deal. Talk from that part of you that says “I can’t let go of this”.
Or T: It sounds like there might be some small part of you that thinks it might be okay to worry less about being late, but the bigger part is saying ‘stop right there’. And I really want to reassure you that only you can decide whether it makes sense, or not, to hang on to the belief that being late is terrible. What do you think?
Or T: And I’m guessing that this comes from a good place in you….that there are good reasons that are pushing you to work so hard at being on time. I’m just guessing but I suspect you might have been burned for being late in the past… Would that be right?… Say more about where this comes from.
In summary, it becomes incumbent on the therapist to continually monitor client engagement with the process of therapy, and gauge the level of harmony and collaboration in the process. Therapists also need to become adept at identifying the signs of in-session client disengagement from the process of therapy (disagreeing, ignoring, interrupting, withdrawal, passivity, criticizing, etc.). Moreover, since such process markers (i.e., higher levels of interpersonal resistance) are strong predictors of subsequent engagement (e.g., later homework compliance), CBT therapists do not have to wait until the client fails to complete homework to realize that there is a problem with client engagement. Once identified, the manner in which therapists respond to resistance plays a major role in perpetuating or diminishing it. However, responding to client opposition supportively, rather than directively, is easier said than done and is a difficult skill to master.
Evidence on Adding or Integrating MI into CBT
The diversity of ways in which MI and other related procedures that include elements of MI, often known as motivational enhancement therapy (MET), have been used in the treatment of anxiety disorders is striking (for a review see Westra et al. 2011). Within this growing body of literature, MI has been most commonly used as a prelude to other therapies, as an approach that is integrated into standard assessment and intake procedures, or integrated throughout treatment as one part of a larger, multi-component treatment package. Beyond these uses, MI has also been applied to increase treatment seeking among individuals with social anxiety who are not yet seeking treatment (e.g. Buckner and Schmidt 2009), and for early prevention among individuals deemed at risk for depression (Van Voorhees et al. 2009).
Although preliminary studies investigating the use of MI in the treatment of anxiety have shown that it may be applied flexibly, research has only recently begun to examine the value of adding MI to existing treatments for anxiety. Consistent with the early stages of this work, this research includes uncontrolled case studies and controlled pilot studies, which have generally been supportive of the use of MI. For example, in small randomized controlled studies comparing MI to psychoeducational, supportive, or no treatment control conditions, MI is demonstrating promise in increasing problem-recognition and treatment attendance (Murphy 2008), enhancing receptivity to recommended exposure-based treatments (Merlo et al. 2010), and in improving response to CBT for anxiety (Westra and Dozois 2006).
In addition to these studies, two randomized controlled trials (RCTs) have been conducted on GAD in particular. Westra et al. (2009) randomly assigned GAD clients to an MI-CBT condition of 4 individual MI pretreatment sessions followed by 14 h (8 sessions) of individual CBT. This group was compared to a no pretreatment (NPT) CBT condition in which clients did not receive any pretreatment but did receive the same amount of individual CBT. Findings from 67 treatment completers (34 MI-CBT and 33 NPT-CBT clients) revealed a post-treatment between groups effect size for worry reduction of d = 0.47 in favor of the MI-CBT group. In addition, the MI-CBT group showed greater therapist-rated homework compliance than the NPT-CBT group (post-treatment between groups effect size of d = 0.59). Thus, the addition of MI to CBT substantially increased the effects of treatment on symptoms of worry and on homework compliance. Interestingly, the most pronounced effects were found for severe worriers, with a large between groups post-treatment effect size of d = 0.97, while for the moderate worriers it was small at d = 0.20.
The results of a second RCT are more compelling given the rigorous nature of the experimental design. Westra et al. (2016) restricted the sample to those of high severity GAD and then randomly assigned them to receive either 15 sessions of CBT alone or 4 session of MI followed by 11 sessions of CBT (MI-CBT). As a rigorous test of the value of integrating MI into CBT, experimenter (and therapist) allegiance was controlled in this study by nesting therapists within treatment condition such that each therapist delivered treatment in only one of the two conditions. Therapists also self-selected the condition they wished to deliver treatment in. Finally, they were supervised independently by experts in that treatment. This was done to ensure that the treatment for the CBT alone group in particular was delivered in a manner that was very high quality, with therapists and supervisors who had allegiance to CBT, thus making the ‘control’ treatment as effective as possible.
Findings of the Westra et al. (2016) study indicated no immediate post-treatment differences in outcome. However, over long-term follow-up, MI-CBT clients continued to improve on the trial’s primary outcome measures of worry and distress reduction, while CBT alone clients generally maintained gains. For example, at 1-year follow-up, MI-CBT clients were over five times more likely to no longer meet diagnostic criteria for GAD compared to CBT alone clients. And, 60% of MI-CBT clients no longer met GAD criteria at 1-year compared to 35% in the CBT alone group. Moreover, while both groups had low dropout, the CBT alone group had twice as many dropouts as the MI-CBT group. Finally, there was evidence of strong differences in interpersonal process between the groups, with MI-CBT therapists demonstrating substantially higher levels of empathy and facilitative interpersonal skills (autonomy support, evocation, collaboration). We interpreted these differences in long-term outcome to reflect increased client agency (self-efficacy) in the MI-CBT group, due to the ‘client-as-expert’ style of MI-CBT compared to the ‘therapist-as-expert’ style of CBT alone.
Importantly, in both of these RCTs, the MI-CBT group was found to have substantially lower levels of resistance (Aviram and Westra 2011; Constantino et al. 2015), which in turn mediated the differences in outcome. That is, because there was less resistance (opposition, disharmony) in MI-CBT compared to CBT alone, there were superior outcomes. This suggests that the main advantage of integrating MI into CBT may in fact be that it trains therapists in the crucial skill of recognizing and effectively navigating resistance.
Further evidence in support of the conclusion that the management of resistance is essential in CBT comes from a recent study by Aviram et al. (2016). We examined variability in therapists’ style of responding to client disagreement within CBT. That is, Aviram and colleagues used only the CBT alone dyads (whose therapists were not trained in MI), and then identified precise moments of disagreement (e.g., homework noncompliance, task disagreement). In other words, they examined natural variation in MI skills among CBT therapists. Results indicated that even when accounting for overall CBT competence, those CBT therapists who were ‘naturally’ more empathic and supportive at times of disagreement had substantially better outcomes than therapists who were less empathic and more controlling at these times. Moreover, therapist ability to be MI-like at moments of disagreement was found to be ten times more powerful than being MI-like at other times (randomly selected times). In other words, timing matters, and the ability to be MI-like at key moments (e.g., during disagreement about tasks, homework, etc.) may be a skill differentiating good and poor CBT therapists. Overall, these effects were large, suggesting that learning how to ‘roll with resistance’ would be of substantial benefit in enhancing outcomes. And, rather than allowing this critical source of variance to vary naturally between therapists, these findings suggest that CBT therapists can and should be explicitly trained in this process skill.
Ambivalence and resistance are key process markers which, when not properly managed, have the capacity to derail therapists and limit CBT treatment efficacy. Though the directive nature of CBT tends to be counterproductive in the face of resistance, MI provides a way for clinicians to effectively address it; namely, by rolling with resistance. Importantly, MI and CBT are approaches that can complement each other and there is strong evidence from recent randomized controlled trials that MI can be successfully integrated with CBT and result in improved treatment outcomes. Moreover, recommending the use of MI in response to specific process markers is arguably more truly integrative than approaches that ‘add’ MI to CBT in a more sequential fashion. Thus, there is enormous promise for the continued role of the integration of MI into CBT and, specifically, for its potential to facilitate the effectiveness of CBT clinicians through learning to identify and manage resistance.
Compliance with Ethical Standards
Conflict of Interest
Henny Westra and Nikoo Norouzian declare that they have no conflict of interest.
Human and Animal Participants
This article does not contain any studies with human participants or animals performed by any of the authors.
For this type of study formal consent is not required.
- Amodeo, M., Lundgren, L., Cohen, A., Rose, D., Chassler, D., Beltrame, C., & D’Ippolito, M. (2011). Barriers to implementing evidence-based practices in addiction treatment programs: Comparing staff reports on motivational interviewing, adolescent community reinforcement approach, assertive community treatment, and cognitive-behavioral therapy. Evaluation and Program Planning, 34, 382–389.CrossRefPubMedGoogle Scholar
- Arkowitz, H., Westra, H. A., Miller, W. R., & Rollnick, S. (2008). Motivational interviewing in the treatment of psychological problems. New York, NY: Guilford Press.Google Scholar
- Beutler, L. E., Moleiro, C. M., & Talebi, H. (2002a). Resistance. In J. Norcross (Ed.), Psychotherapy relationships that work: Therapist contributions and responsiveness to patients (pp. 129–144). Oxford: Oxford University Press.Google Scholar
- Binder, J. L., & Strupp, H. H. (1997). “Negative process”: A recurrently discovered and underestimated facet of therapeutic process and outcome in the individual psychotherapy of adults. Clinical Psychology: Science and Practice, 4, 121–139.Google Scholar
- Boswell, J. F., Gallagher, M. W., Sauer-Zavala, S. E., Bullis, J., Gorman, J. M., Shear, M. K., & Barlow, D. H. (2013). Patient characteristics and variability in adherence and competence in cognitive-behavioral therapy for panic disorder. Journal of Consulting and Clinical Psychology, 81, 443–454.CrossRefPubMedGoogle Scholar
- Burns, D. D., & Auerbach, A. (1996). Therapeutic empathy in cognitive-behavioural therapy: Does it really make a difference? In P. M. Salkovskis (Ed.), Frontiers of cognitive therapy (pp. 135–164). New York, NY: The Guilford Press.Google Scholar
- Button, M. L., Westra, H. A., Constantino, M. J., & Antony, M. M. (2016). Client Ambivalence as a Moderator of Treatment Outcomes in MI and CBT for Generalized Anxiety. Paper presented at the annual meeting of the Society for the Exploration of Psychotherapy Integration, Dublin, Ireland.Google Scholar
- Button, M. L., Westra, H. A., & Hara, K. M. (2014). Ambivalence and homework compliance in CBT. Paper presented at the Society for the Exploration of Psychotherapy Integration (SEPI) 30th Annual International Meeting, Montreal, QC.Google Scholar
- Cassidy, J., & Shaver, P. R. (Eds.). (1999). Handbook of attachment: Theory, research, and clinical applications. New York, NY: Guilford Press.Google Scholar
- Chamberlain, P., Davis, J. P., Forgatch, M., Frey, J., Patterson, G. R., Ray, J., Rothschild, A., & Trombley, J. (1985). The Therapy Process Code: A Multidimensional System for Observing Therapist and Client Interactions. Unpublished coding manual.Google Scholar
- Constantino, M. J., Castonguay, L. G., & Schut, A. J. (2002). The working alliance: A flagship for the “scientist-practitioner” model in psychotherapy. In G. S. Tryon (Ed.), Counseling based on process research: Applying what we know (pp. 81–131). Boston, MA: Allyn & Bacon.Google Scholar
- Constantino, M. J., Westra, H. A., & Antony, M. M. (2015). Mediators of treatment effects in a randomized controlled trial testing the efficacy of augmenting CBT with motivational interviewing for severe generalized anxiety disorder. Paper presented at the meeting of the Society for Psychotherapy Research, Philadelphia, PA.Google Scholar
- Constantino, M. J., Westra, H. A., Antony, M. M., & Coyne, A. E. (2016). Integrative motivational interviewing and cognitive-behavioral therapy for generalized anxiety disorder: Secondary outcomes of a randomized controlled trial. Paper presented at the annual meeting of the Society for the Exploration of Psychotherapy Integration, Dublin, Ireland.Google Scholar
- Cumming, J. A., Hayes, A. M., Newman, C. F., & Beck, A. T. (2011). Navigating therapeutic alliance ruptures in cognitive therapy for avoidant and obsessive-compulsive personality disorders and comorbid axis I disorders. International Journal of Cognitive Therapy, 4(4), 397–414.CrossRefGoogle Scholar
- Gilbert, P. (2010). Attachment and the importance of affection. In P. Gilbert (Ed.), Compassion focused therapy (pp. 39–42). New York, NY: Routledge.Google Scholar
- Goodwin, B. J., Constantino, M. J., Westra, H. A., Button, M. L., & Antony, M. M. (2015). Motivational language as a predictor of change and time to response in psychotherapy for generalized anxiety disorder. Paper presented at the Annual Meeting of the Society for Psychotherapy Research, Philadelphia, USA.Google Scholar
- Hara, K. M., Westra, H. A., Constantino, M. J., & Antony, M. M. (2016). Therapist awareness of client-identified alliance ruptures in cognitive-behavioral therapy. Paper presented at the annual meeting of the Society for the Exploration of Psychotherapy Integration, Dublin, Ireland.Google Scholar
- Mamedova, K., Westra, H. A., Constantino, M. J., & Antony, M. M. (2016). Initial Expectations as a Moderator of the Impact of Resistance on Outcome Expectations in Cognitive-Behavioral Therapy. Paper presented at the annual meeting of the Society for the Exploration of Psychotherapy Integration, Dublin, Ireland.Google Scholar
- Merlo, L. J., Storch, E. A., Lehmkuhl, H. D., Jacob, M. L., Muprhy, T. K., Goodman, W. K., & Geffken, G. R. (2010). Cognitive behavioral therapy plus motivational interviewing improves outcome for pediatric obsessive-compulsive disorder: A preliminary study. Cognitive Behavioral Therapy, 39(1), 24–27.CrossRefGoogle Scholar
- Miller, W. R., & Rollnick, S. (2002). Motivational interviewing: Preparing people for change. New York, NY: Guilford Press.Google Scholar
- Miller, W. R., & Rollnick, S. (2012). Motivational interviewing: Helping people change (3rd edn.). New York, NY: Guilford Press.Google Scholar
- Muran, J. C., Safran, J. D., Gorman, B. S., Wallner-Samstag, L., Eubanks-Carter, C., & Winston, A. (2009). The relationship of early alliance ruptures and their resolution to process and outcome in three time-limited psychotherapies for personality disorders. Psychotherapy: Theory, Research, Practice, Training, 46(2), 233–248.CrossRefGoogle Scholar
- Murphy, R. T. (2008). Enhancing combat veterans’ motivation to change posttraumatic stress disorders symptoms and other problem behaviors. In H. Arkowitz, H. A. Westra, W. R. Miller, et al., (Eds.), Motivational interviewing in the treatment of psychological problems (pp. 85–108). New York, NY: Guliford Press.Google Scholar
- Orlinsky, D. E., Grawe, K., & Parks, B. K. (1994). Process and outcome in psychotherapy: Noch einmal. In A. E. Bergin & S. L. Garfield (Eds.), Handbook of psychotherapy and behaviour change (4th edn., pp. 270–376). Oxford: John Wiley & Sons.Google Scholar
- Rogers, C. R. (1951). Client-centered therapy. Boston, MA: Houghton Mifflin.Google Scholar
- Rogers, C. R. (1965). Client-centered therapy: It`s current practice, implications, and theory. Boston, MA: Houghton Mifflin.Google Scholar
- Rogers, C. R. (1980). A way of being. Boston, MA: Houghton Mifflin.Google Scholar
- Rollnick, S., Miller, W. R., & Butler, C. C. (2007). Motivational interviewing in health care: Helping patients change behavior. New York, NY: Guliford Press.Google Scholar
- Safran, J. D., & Muran, J. C. (2000). Negotiating the therapeutic alliance: A relational treatment guide. New York, NY: Guilford Press.Google Scholar
- Van Voorhes, B. W., Fogel, J., Reinecke, M. A., Gladstone, T., Stuart, S., Gollan, J., et al. (2009). Randomized clinical trial of an Internet-based depression prevention program for adolescents (Project CATCH-IT) in primary care: 12-week outcomes. Journal of Developmental and Behavioral Pediatrics, 30(1), 23–37.CrossRefGoogle Scholar
- Westra, H. A. (2012). Motivational Interviewing in the Treatment of Anxiety. New York, NY: Guilford.Google Scholar
- Westra, H. A., Aviram, A., Kertes, A., Ahmed, M., & Connors, L. (2009). Manual for coding resistance in CBT for anxiety. Unpublished coding manual.Google Scholar
- Westra, H. A., Constantino, M. J., & Antony, M. M. (2016). Integrating motivational interviewing with cognitive behavioral therapy for severe generalized anxiety disorder: An allegiance-controlled randomized clinical trial. Journal of Consulting and Clinical Psychology, 84(9), 768–782.CrossRefPubMedGoogle Scholar
- Zickgraf, H. F., Chambless, D. L., McCarthy, K. S., Gallop, R., Sharpless, B. A., Milrod, B. L., & Barber, J. P. (2016). Interpersonal factors are associated with lower therapist adherence in cognitive-behavioural therapy for panic disorder. Clinical Psychology and Psychotherapy, 23(3), 272–284.CrossRefPubMedGoogle Scholar