Developing process sensitivity: Reply to Wilcox (2024) and Boswell (2024).

Wilcox (2024) and Boswell (2024) make a number of important observations about facilitating process sensitivity training, and here, we respond to those suggestions. We postulate that cultivating process sensitivity is complementary, not antithetical, to traditional training in viewing therapy from a theoretical lens, and thus, can serve to enhance, rather than replace one's existing psychotherapy skills. Moreover, we argue that seeing the impact of process adjustments in real time can be a significant motivator for training in process sensitivity since the benefits are more immediately obvious. We further argue that the field can be more thoughtful about the use of simulations in training and the emerging interactive training platforms using video stimuli, since they facilitate exposure to clinical situations in a safe manner. Finally, while learning to identify process markers can play a valuable role in optimizing the timing of interventions, we argue that this pales in comparison to the value of process sensitization as a means of attending to outcome information in the moment ("little outcome"). If successful, this opens up the possibility of developing expertise in psychotherapy, which hitherto was considered not possible. However, these propositions require rigorous testing in studies on training and supervision. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Developing expertise in psychotherapy: The case for process coding as clinical training.

Routine outcome monitoring (ROM) is a major development in the field since it offers likely outcome trajectories and is particularly helpful for failing cases. However, ROM has not led to improved skill development more generally, and it is debatable as to whether expertise is even possible to acquire in psychotherapy. What is missing but crucial to expertise is feedback on the outcome of one's actions in real time, which would enable responsive adjustments and improve outcomes. It is argued in this article that by identifying empirically validated moment-to-moment markers capable of differentiating later clinical outcomes, process researchers have uncovered the possibility of extracting prognostic information in real time, but one must develop the requisite observational skills. Multiple lines of research are reviewed to support the contention that real-time outcome information is available to guide responsivity and improve outcomes. And the typically hidden nature of these important signals further underscores the need for systematic training in process acuity. Given the pressing need to improve training methods, process coding training should not be restricted to research laboratories but should be exported to the clinical setting and tailored to the needs of clinicians for use in real time during therapy sessions. These are testable hypotheses that, if successful, hold the possibility of improving training and reversing the worrying trend of experience in psychotherapy being unrelated to outcome. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Therapist verification of patient self-concepts as a responsive precondition for early alliance development and subsequent introject change.

OBJECTIVE: Social psychological research has indicated that people strive for self-consistent feedback and interactions, even if negative, to preserve the epistemic security of knowing themselves. Without such self-verification, any interpersonal exchange may become frustrated, anxiety-riddled, and at risk for deterioration. Thus, it may be important for therapists to meet patients' self-verification needs as a responsive precondition for early alliance establishment and development. We tested this hypothesis with patients receiving cognitive behavioral therapy for generalized anxiety disorder-a condition that may render one's self-verification needs especially strong. We also tested the hypothesis that better early alliance quality would relate to subsequent adaptive changes in and posttreatment level of patients' self-concepts. METHOD: Eighty-four patients rated their self-concepts at baseline and across treatment and follow-up, their postsession recollection of their therapist's interpersonal behavior toward them during session 2, and their experience of alliance quality rated after sessions 3-6. RESULTS: As predicted, the more therapists verified at session 2 a patient's baseline self-concepts (which trended toward disaffiliative and overcontrolling, on average), the more positively that patient perceived their next-session alliance. Moreover, better session 3 alliance related to more adaptive affiliative and autonomy-granting self-concepts at posttreatment. CONCLUSION: Results are discussed within a therapist responsiveness framework.

Predicting resistance management skill from psychotherapy experience, intellectual humility and emotion regulation.

OBJECTIVE: Resistance management in psychotherapy remains a foundational skill that is associated with positive client outcomes (Westra, H. A., & Norouzian, N. (2018). Using motivational interviewing to manage process markers of ambivalence and resistance in cognitive behavioral therapy. Cognitive Therapy and Research, 42(2), 193-203). However, little is known about which therapist characteristics contribute to successful management of resistance. Research has suggested that psychotherapy performance does not improve with experience (Goldberg, S. B., Rousmaniere, T., Miller, S. D., Whipple, J., Nielsen, S. L., Hoyt, W. T., & Wampold, B. E. (2016). Do psychotherapists improve with time and experience? A longitudinal analysis of outcomes in a clinical setting. Journal of Counseling Psychology, 63(1), 1-11), that psychotherapists lack humility (Macdonald, J., & Mellor-Clark, J. (2015). Correcting psychotherapists' blindsidedness: Formal feedback as a means of overcoming the natural limitations of therapists. Clinical Psychology & Psychotherapy, 22(3), 249-257), and that difficult therapeutic moments may dysregulate therapist emotions (Muran, J. C., & Eubanks, C. F. (2020). Therapist performance under pressure: Negotiating emotion, difference, and rupture. American Psychological Association). This study aimed to 1) identify whether psychotherapy experience (i.e., training versus no training and number of years of psychotherapy experience) was associated with resistance management skill, and 2) identify whether humility and difficulties regulating emotions among trained individuals were each associated with resistance management. METHOD: A sample of 76 trained and 98 untrained participants were recruited for the present study. All participants completed the Comprehensive Intellectual Humility Scale (CIHS, Krumrei-Mancuso, E. J., & Rouse, S. V. (2016). The development and validation of the comprehensive intellectual humility scale. Journal of Personality Assessment, 98(2), 209-221), the Difficulties in Emotion Regulation Scale (DERS; Gratz, K. L., & Roemer, L. (2004). Multidimensional assessment of emotion regulation and dysregulation: Development, factor structure, and initial validation of the difficulties in emotion regulation scale. Journal of Psychopathology and Behavioral Assessment, 26(1), 41-54), and the Resistance Vignette Task (RVT; Westra, H. A., Nourazian, N., Poulin, L., Hara, K., Coyne, A., Constantino, M. J., Olson, D., & Antony, M. M. (2021). Testing a deliberate practice workshop for developing appropriate responsivity to resistance markers: A randomized clinical trial. Psychotherapy, 58, 175-185 ) which was used to assess resistance management skill. RESULTS: Trained individuals performed significantly better on resistance management than untrained individuals; however, years of experience within the trained sample were not associated with resistance management. Conversely, lower humility and greater difficulties regulating emotions were each associated with significantly poorer resistance management in trained individuals. CONCLUSION: These findings suggest the possibility of improving training to focus on key skills, like resistance management, through supporting humility and emotion regulation in training, as opposed to simply acquiring more experience.

The Resistance Vignette Task: Validating a rapid measure of therapist skill at managing resistance.

BACKGROUND: Therapist appropriate responsivity to client ambivalence and resistance is considered an important interpersonal skill to avoid disengagement and ensure a continued collaborative, productive process. The present study examined the predictive validity of the newly developed Resistance Vignette Task (RVT), a 10-item rapidly administered measure of therapist ability to appropriately respond to various presentations of client resistance. METHODS: Following a resistance management workshop, the concurrent and prospective predictive capacity of RVT scores were examined through test interviews with ambivalent simulators and volunteers. RESULTS: Prospectively, in test interviews with ambivalent interviewees, higher RVT scores immediately postworkshop were associated with significantly greater responsivity (appropriate responsivity and fewer responsivity errors) at 4-month follow-up. RVT scores at the 4-month follow-up point were also concurrently associated with significantly greater therapist responsivity and lower levels of interviewee resistance. CONCLUSIONS: These findings provide further validation for the RVT as a measure of therapist responsivity in vivo, in actual interviews by predicting and being concurrently associated with therapist performance in response to client resistance. Thus, the RVT holds promise in advancing therapist training, as well as research on resistance as it represents an efficient measure of this key therapist skill.

Therapist affiliation and hostility in cognitive-behavioral therapy with and without motivational interviewing for severe generalized anxiety disorder.

OBJECTIVE: Although therapist supportive, rather than directive, strategies have been particularly indicated during client resistance, little systematic research has examined how therapists responsively navigate resistance in different therapy approaches and how this responsiveness is related to outcome. METHOD: In the context of disagreement episodes in cognitive-behavioral therapy (CBT) for generalized anxiety disorder (GAD; Westra, H. A., Constantino, M. J., & Antony, M. M. 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. https://doi.org/10.1037/ccp0000098, 2016), the present study examined (1) the degree to which therapist management of resistance differed between therapists trained in CBT integrated with motivational interviewing (MI-CBT; i.e., training centered on the responsive management of resistance) and therapists trained in CBT-alone, and (2) the impact of specific therapist behaviors during disagreement on client worry outcomes immediately posttreatment and 1-year posttreatment. Episodes of disagreement were rated used the Structural Analysis of Social Behavior (Benjamin, L. S. Structural analysis of social behavior. Psychological Review, 81(5), 392-425. https://doi.org/10.1037/h0037024, 1974). RESULTS: Therapists trained in MI-CBT were found to exhibit significantly more affiliative and fewer hostile behaviors during disagreement compared to those trained in CBT-alone; both of these, in turn, were found to mediate client 1-year posttreatment outcomes, such that increased affiliation during disagreement was associated with improved outcomes. CONCLUSION: This study highlights the value of training therapists in the responsive detection and management of resistance, as well as the systematic integration of MI into CBT.

Testing a deliberate practice workshop for developing appropriate responsivity to resistance markers.

Deliberate practice (DP) is an emerging training method for improving individual performance that may be worth adapting and testing for applicability to groups, given the prevalence of group training for continuing education. This study compared an adapted DP workshop to the same traditional, non-DP workshop for managing ambivalence and resistance. The same presenter delivered the workshops to 88 randomly assigned community psychotherapists. The DP workshop involved repeated interaction with multiple recreations of resistance, with consistent group feedback especially on ideal expert performance. The control workshop was more didactic, with fewer opportunities for practice and feedback. We assessed video vignette performance and coded 20-min interviews with ambivalent interviewees from the community. Both workshops produced equivalent trainee satisfaction and significant increases in self-reported skills. However, the DP versus control group demonstrated better observer-rated skill on all performance measures postworkshop. Although skills declined to 4-month retest in both groups, the DP trainees retained their relative advantage over traditional workshop trainees. Moreover, at the 4-month follow-up, DP versus control trainees were rated as more empathic by community interviewees and self-reported practicing the skills at higher rates. These findings support the continued investigation of DP as a means for improving therapist skill in continuing education workshops. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

The impact of a deliberate practice workshop on therapist demand and support behavior with community volunteers and simulators.

The present study used a newly developed simplified coding system, the Therapist Demand and Support Code, to examine specific therapist behaviors in the context of a previously conducted training trial on Deliberate Practice (DP). The parent trial randomized trainees to a DP workshop or its Traditional, more didactic counterpart (Westra et al., 2020). In both groups, trainees were taught to use Support, rather than Demand, for managing ambivalence and resistance, with the DP group having more feedback and practice. In this study, 68 trainees interviewed both an ambivalent community volunteer and an ambivalent simulator 4 month post workshop. The DP group was found to exhibit significantly fewer Demand behaviors than the Traditional group, with the latter also being significantly quicker to use Demand in the interviews. Moreover, the simulator evoked significantly greater Demand from therapists, regardless of the Training group, suggesting the simulators were more resistant. Although therapist use of Support was equal for community volunteers across training groups, Traditional workshop trainees decreased Support when interviewing the more resistant simulators, whereas DP trainees increased their Support with this same group. This is consistent with findings that DP trainees were more appropriately responsive, making fewer Demands following interviewee counterchange talk and using more Support at these times. These results provide some initial validation of the simplified therapist behavior coding system and offer further evidence for the benefits of DP workshop training for managing resistance. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Association between therapist attunement to patient outcome expectation and worry reduction in two therapies for generalized anxiety disorder.

Patients' higher psychotherapy outcome expectation (OE) correlates with improvement. Thus, it seems important that therapists attune to this belief, both in the moment and over time, to capitalize on its value when higher or respond to its potential risk when lower. Conceptually, attunement can have different guises, including the extent to which therapists (a) accurately estimate their patients' momentary OE level (low directional discrepancy), (b) become more accurate in estimating OE over time (convergence), (c) accurately track shifts in their patients' OE (temporal congruence), and (d) become more temporally congruent over time (alignment). To date, though, little is known empirically about therapist attunement to patient OE. Thus, we examined the presence of attunement indices and their relation to posttreatment outcome. Data derived from a randomized trial that compared cognitive-behavioral therapy (CBT; n = 43) to CBT plus motivational interviewing (n = 42) for patients with generalized anxiety disorder. After each session, patients rated their OE, and therapists estimated their patients' OE. Patients rated worry at baseline and posttreatment. Dyadic multilevel modeling revealed that across both treatments, therapists were directionally discrepant in that they underestimated patients' OE (p < .001), which did not change over time (no average convergence/divergence pattern; p = .43). Additionally, therapists exhibited temporal congruence with patients' OE (p < .001) and became more aligned with this rating over time (p = .008). Only greater OE convergence, when it occurred, predicted lower worry (p = .04). A therapist's increasingly accurate empathy about their patients' OE may be therapeutic. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Dyadic, longitudinal associations among outcome expectation and alliance, and their indirect effects on patient outcome.

Research indicates that patient outcome expectation (OE) correlates with improvement, and that this association may be mediated by better patient-therapist alliances. However, despite OE and alliance being dyadic and dynamic constructs, most research on these direct and indirect associations has assessed these variables from only one dyad member's perspective and at single time points. Addressing these gaps, we used a longitudinal actor-partner interdependence model to first examine OE-alliance associations. Namely, we assessed "actor" effects (relation between each member's OE at 1 session and his or her own next session alliance) and "partner" effects (relation between each member's partner's OE at 1 session and his or her own next session alliance). Second, we tested whether significant actor or partner effects of OE on alliance translated into better patient outcomes (indirect effects). Analyses were conducted at within- and between-dyad levels. Data derived from a generalized anxiety disorder trial in which 85 patients received 15 sessions of either cognitive-behavioral therapy (CBT) or CBT integrated with motivational interviewing. After every session, patients and therapists rated OE and alliance, and patients rated their worry. At the within-dyad level, there were OE-alliance actor effects for both patients and therapists. There was also a within-dyad partner effect; when patients had greater OE at one session their therapists reported better next-session alliances. Finally, all within-dyad effects in turn related to lower subsequent worry. Results reveal ways in which session-by-session fluctuations in both patient and therapist OE translate into better outcomes through their influence on alliance quality. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Patient motivational language in the prediction of symptom change, clinically significant response, and time to response in psychotherapy for generalized anxiety disorder.

Existing research demonstrates that patient change-talk (CT), or self-arguments for change, associates variably with favorable outcomes, whereas counter change-talk (CCT), or self-arguments against change, associates consistently with poorer outcomes. However, most studies on change language have focused on posttreatment versus more proximal outcomes. Addressing this gap, we examined Session 1 CT and CCT as predictors of during-treatment worry change, likelihood of clinically significant response during treatment, and time to response across cognitive-behavioral therapy (CBT; n = 43) and CBT integrated with motivational interviewing (MI; n = 42) for generalized anxiety disorder. We also explored whether treatment moderated these associations. Multilevel modeling revealed that, across both treatments, more CT associated with lower midtreatment worry level (p = .04), whereas more CCT associated with lower worry level (p = .048) and a slower rate of worry reduction at midtreatment (p = .04). Treatment moderated only the associations between CT and both midtreatment worry level (p = .03) and rate of change (p = .03). CBT patients with higher versus lower CT had lower worry and a faster rate of worry reduction; in MI-CBT, CT was unrelated to these outcomes. Survival analyses revealed that, across both treatments, more CT associated with a greater likelihood of response (p = .004) and approached a faster time to response (p = .05), and more CCT associated with a lower likelihood of response (p = .001) and approached a slower time to response (p = .06). Motivational language predicts proximal outcomes and may be useful in treatment selection. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Baseline Client Interpersonal Agency Moderates the Indirect Effect of Treatment on Long-term Worry in Variants of CBT for Generalized Anxiety Disorder.

In a recent trial for generalized anxiety disorder (GAD), cognitive-behavioral therapy (CBT) integrated with motivational interviewing (MI) promoted more long-term worry reduction than CBT alone (Westra, Constantino, & Antony, 2016). A follow-up analysis found that CBT vs. MI-CBT clients evidenced greater increases in friendly submissiveness (FS) across treatment, which in turn promoted lower long-term worry (Constantino, Romano, Coyne, Westra, & Antony, 2018). It was unsurprising that traditional directive CBT promoted more FS than when person-centered MI was integrated; however, given that problematic low agency characterizes GAD, that greater FS promoted better outcome was unexpected. To further unpack this unexpected result, we tested the following moderated mediation hypothesis: for clients with more vs. less problematic low agency at baseline, CBT would still promote more in-session FS than MI-CBT, but this increase would in turn predict increased worry over follow-up. Clients receiving CBT (n = 43) or MI-CBT (n = 42) rated their interpersonal problems at baseline and their worry after treatment and across 12-month follow-up. Therapists rated clients' in-session FS multiple times. As predicted, multilevel modeling revealed that for clients with more problematic low agency, CBT vs. MI-CBT facilitated greater FS, which in turn related to increased worry across follow-up. For clients with more problematic high agency, CBT's facilitation of greater FS related to reduced worry across follow-up. A baseline interpersonal problem characteristic of GAD may have implications for treatment matching and for appreciating different pathways to long-term improvement, or deterioration, for different GAD subgroups.

Interpersonal change as a mediator of the within- and between-patient alliance-outcome association in two treatments for generalized anxiety disorder.

OBJECTIVE: Although a better therapeutic alliance associates with treatment outcome, it may do so in different ways. For example, alliance quality may promote improvement as it shifts over time (within-patient changes); alternatively, alliance quality may also, or instead, influence outcome when generally higher or lower for some patients relative to others (between-patient differences). Although both components have been linked to patient improvement, the distinct mechanisms of these associations have been rarely examined. Conceptually, it follows that within-patient alliance fluctuations (representing the changing nature of a current relationship intended to be ameliorative) may facilitate other interpersonal improvements that could, in turn, translate into symptom reduction. This path squares with the corrective relational experience notion. In contrast, whereas as patients' average alliances across therapy may generally facilitate or hinder improvement, the mechanism may not be other relationship functioning. This squares with the notion that when people generally experience good alliances with their therapist, it may reflect an existing relational ability that catalyzes the effectiveness of other nonrelational therapeutic means. This study tested these distinct hypotheses. METHOD: Patients (N = 85) with generalized anxiety disorder were randomly assigned to cognitive-behavioral therapy, either alone or integrated with motivational interviewing. They rated alliance quality, interpersonal problems, and outcome repeatedly. RESULTS: Using multilevel structural equation modeling, both within- and between-patient alliances related to subsequent worry reduction. As predicted, change in interpersonal problems mediated the association only at the within-patient level. CONCLUSIONS: Results contribute to the literature on mechanisms of the within- and between-person alliance-outcome association. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Competing indirect effects in a comparative psychotherapy trial for generalized anxiety disorder.

In a randomized trial for generalized anxiety disorder, cognitive-behavioral therapy (CBT) and CBT integrated with motivational interviewing (MI) promoted comparable worry reduction at posttreatment, whereas MI-CBT outperformed CBT over 12-month follow-up (Westra, Constantino, & Antony, 2016). Secondary analyses revealed competing mediators of the long-term treatment effect: MI-CBT related to lower patient resistance to the treatment, which promoted lower follow-up worry, whereas CBT related to greater increases in patient friendly submissiveness (FS), or compliance, which also promoted lower follow-up worry (that suppressed an even greater long-term advantage of MI-CBT). In this study, we tested these competing, though theoretically consistent, variables as mediators of the nonsignificant treatment effect on posttreatment worry, as there could also be treatment-specific means to arriving to these comparable ends. Eighty-five patients received 15 sessions of MI-CBT or CBT. Therapists rated patient FS through treatment, observers rated resistance at midtreatment, and patients rated worry at posttreatment. Bootstrap analyses indicated that MI-CBT patients exhibited less resistance, which promoted lower posttreatment worry, whereas CBT patients had greater increases in FS, which also promoted lower worry. CBT and MI-CBT achieved comparable posttreatment outcomes through separate indirect paths that each conferred an advantage for one treatment over the other (and had canceled out a direct treatment effect immediately after therapy). The composite trial findings have significance for understanding different pathways to personal change in both the short- and long terms, and for the importance of testing indirect effects both when direct treatment effects do and do not emerge. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

The predictive capacity of self-reported motivation vs. early observed motivational language in cognitive behavioural therapy for generalized anxiety disorder.

Client motivation to change is often considered a key factor in psychotherapy. To date, research on this client construct has largely relied on self-report, which is prone to response bias and ceiling effects. Moreover, self-reported motivation has been inconsistently related to treatment outcome. Early observed client in-session language may be a more valid measure of initial motivation and thus a promising predictor of outcome. The predictive ability of motivational factors has been examined in addiction treatment but has been limited in other populations. Addressing this lack, the present study investigated 85 clients undergoing cognitive behavioural therapy (CBT) alone and CBT infused with motivational interviewing (MI-CBT) for severe generalized anxiety disorder. There were two aims: (1) to compare the predictive capacity of motivational language vs. two self-report measures of motivation on worry reduction and (2) to examine the influence of treatment condition on motivational language. Findings indicated that motivational language explained up to 35% of outcome variance, event 1-year post-treatment. Self-reported motivation did not predict treatment outcome. Moreover, MI-CBT was associated with a significant decrease in the most detrimental type of motivational language compared to CBT alone. These findings support the importance of attending to in-session motivational language in CBT and learning to respond to these markers using motivational interviewing.

Can Self-Report Measures of Readiness for Change and Treatment Ambivalence Predict Outcomes in Cognitive Behavioral Therapy for Generalized Anxiety Disorder?

Clients with generalized anxiety disorder (GAD) that demonstrate observer-coded treatment ambivalence benefit from the addition of motivational interviewing (MI) to cognitive behavioral therapy (CBT; Button, Westra, Constantino, & Antony, 2016). While observer-coded assessment of ambivalence and readiness for change is resource-intensive, the present study investigates the use of more efficient self-report measures to predict treatment outcomes. Participants (N = 85) with GAD received CBT or MI-CBT and completed self-report measures of readiness for change (Change Questionnaire, Miller & Johnson, 2008) and ambivalence (Treatment Ambivalence Questionnaire, Purdon, Rowa, Gifford, McCabe, & Antony, 2012). Greater self-reported baseline readiness for change was associated with lower posttreatment worry and symptom severity and faster reduction in worry. Self-reported ambivalence was not associated with outcomes. Patients with less concern about adverse consequences of treatment who received CBT experienced greater increases in readiness for change than those receiving MI-CBT. We discuss implications for using these measures in clinical settings.

Client reflections on confirmation and disconfirmation of expectations in cognitive behavioral therapy for generalized anxiety disorder with and without motivational interviewing.

Objective: Addressing methodological shortcomings of prior work on process expectations, this study examined client process expectations both prospectively and retrospectively following treatment. Differences between clients receiving cognitive behavioral therapy (CBT) versus motivational interviewing integrated with CBT (MI-CBT) were also examined. Method: Grounded theory analysis was used to study narratives of 10 participants (N = 5 CBT, 5 MI-CBT) who completed treatment for severe generalized anxiety disorder as part of a larger randomized controlled trial. Results: Clients in both groups reported and elaborated expectancy disconfirmations more than expectancy confirmations. Compared to CBT clients, MI-CBT clients reported experiencing greater agency in the treatment process than expected (e.g., that they did most of the work) and that therapy provided a corrective experience. Despite nearly all clients achieving recovery status, CBT clients described therapy as not working in some ways (i.e., tasks did not fit, lack of improvement) and that they overcame initial skepticism regarding treatment. Conclusions: Largely converging with MI theory, findings highlight the role of key therapist behaviors (e.g., encouraging client autonomy, validating) in facilitating client experiences of the self as an agentic individual who is actively engaged in the therapy process and capable of effecting change.

Specific and common processes as mediators of the long-term effects of cognitive-behavioral therapy integrated with motivational interviewing for generalized anxiety disorder.

OBJECTIVE: A trial of psychotherapy for generalized anxiety disorder (GAD) demonstrated that motivational interviewing (MI) integrated with cognitive-behavioral therapy (CBT) outperformed CBT alone on clients' worry reduction across a 12-month follow-up. In the present study, we hypothesized and tested that less client resistance and greater client-perceived therapist empathy (specific foci of MI) would account for MI's additive effect. Exploratory analyses assessed whether the common processes of homework completion and therapeutic alliance quality mediated the treatment effect. METHOD: Clients with GAD were randomized to 15 sessions of MI-CBT (n = 42) or CBT alone (n = 43). Worry was assessed throughout treatment and follow-up. Observers rated resistance at midtreatment, and clients reported on perceived therapist empathy, alliance, and homework completion throughout treatment. Mediation was tested with bootstrapping methods. RESULTS: Expectedly, MI-CBT clients evidenced less resistance and perceived greater therapist empathy, each of which related to lower 12-month worry. However, when both variables were tested simultaneously, only resistance remained a significant mediator of treatment. No indirect effects through homework completion or alliance emerged. CONCLUSIONS: Reducing client resistance may be a theory-consistent mechanism through which integrative MI-CBT promotes superior long-term improvement than traditional CBT when treating GAD. Clinical or methodological significance of this article: This study further supports the long-term clinical benefit of integrating MI into CBT when treating the highly prevalent and historically difficult-to-treat condition of GAD. In particular, it points to the theory-specific mechanism of MI (helping to reduce/resolve patients' in-treatment resistance) as accounting for the integrative treatment's additive effect on worry reduction across a follow-up period. Therapists using CBT to treat patients with GAD should be trained to incorporate MI principles (e.g., empathy, collaboration, autonomy support) in general and in response to explicit markers of resistance.

The impact of resistance on empathy in CBT for generalized anxiety disorder.

OBJECTIVE: Client resistance has been shown to relate to poorer therapy outcomes, thus making it important to better understand the mechanisms underlying this association. Given observational research suggesting that therapist empathy decreases during moments of resistance, the present study examined client-rated therapist empathy as a potential mediator of the resistance-outcome association. METHOD: Participants included 44 therapist-client dyads receiving cognitive-behavioral therapy for generalized anxiety disorder. Trained observers rated an early therapy session for the level of client resistance, and clients completed a corresponding postsession measure of therapist empathy. Posttreatment outcome was measured via client-rated worry severity. RESULTS: Higher client resistance was significantly associated with poorer treatment outcome and lower client postsession ratings of therapist empathy; however, therapist empathy was not observed to mediate the relationship between resistance and treatment outcomes. CONCLUSIONS: As empathy did not mediate the association between resistance and outcome, future research is needed to uncover other potential mechanisms of this association. However, the current results underscore an important link between resistance and client perceived therapist empathy. As empathy has been shown to relate positively to therapy outcomes, our result highlights the need to enhance therapist in-session responsivity to resistance in psychotherapy research and training.