Blog

Friedlander, M. L., Escudero, V., Welmers-van de Poll, M. J., & Heatherington, L. (2018). Meta-analysis of the alliance–outcome relation in couple and family therapy. Psychotherapy, 55(4), 356-371.

In individual psychotherapy the therapist’s tasks include to develop an alliance with one patient. Goals and tasks of therapy need to be collaboratively negotiated, and therapists need to develop an emotional bond with the patient. The alliance also has to be nurtured continuously throughout treatment. This process is more complicated in couple and family therapy. Only in couple and family therapy (and in group therapy) does a therapist have to develop an alliance with multiple people simultaneously. The challenge is greater when family members are in conflict, or when the therapist’s alliance is stronger with one member than another. Such “split” alliances can be problematic especially when family members view their experiences of the therapist differently. To complicate things more, therapists have to be aware of the alliance within the family or couple system. That is, are the family members allied with each other – do they agree on therapy goals and tasks, and are they able to maintain an emotional connection to each other? In addition, just as therapeutic alliance ruptures can occur in individual therapy, so can they occur in couple and family therapy. An alliance rupture may occur when a there is a “split” alliance or when a patient responds to the therapist or other family members with confrontation or withdrawal behaviors. In this meta-analysis of therapeutic alliance in couple and family therapy, Friedlander and colleagues included 48 studies with a total of 2,568 families and 1,545 couples. The correlation between quality of the alliance and outcome was significant (r = .297, 95% CI [0.223, 0.351], p < .001), indicating that a stronger alliance was related to better outcomes. There was some evidence of publication bias suggesting that this estimate may be over-inflated, but even after adjusting for publication bias the correlation was still significant. The correlation between split alliances and outcome was also significant (r = .316, 95% CI [0.157, 0.458], p < .001), indicating that more split alliances contributed to poorer outcomes. The correlations were similar in strength both in couple and in family therapy, and the alliance was important in all therapeutic orientations. However, correlations were larger when the targeted child in the family was younger, and when families were seeking help and not mandated.

Practice Implications

Like in individual therapy, the therapeutic alliance in couple and family therapy is important to improve the outcomes of patients. Regardless of therapeutic orientation, therapists must spend time and effort developing therapeutic alliances with each member of the system, and must try to maintain relatively equal alliances with each family member to avoid splits in the alliance. Therapists should be particularly aware of any confrontation and withdrawal behaviors towards the therapist or within the family or couple as these may indicate an alliance rupture. In such instances, therapists should emphasize shared goals and feelings, validate the common struggle among family members, and focus on the emotional bond with the disaffected patient. Each person’s alliance matters, and family member alliances are not interchangeable. Assessing the alliance with each member throughout therapy will identify potential problems and facilitate better outcomes.

Karver, M. S., De Nadai, A. S., Monahan, M., & Shirk, S. R. (2018). Meta-analysis of the prospective relation between alliance and outcome in child and adolescent psychotherapy. Psychotherapy, 55(4), 341-355.

Over the past decades there has been increasing research on the efficacy of psychotherapy for children and adolescents, but outcomes have not always been positive. Treatment of children and adolescents comes with challenges that are unique from those experienced in therapy of adults. For example, unlike most adults, children and adolescents may not be the ones to choose to attend therapy - that decision is often made by adults in their lives. Furthermore, psychotherapists must also develop and maintain a collaborative relationship with parents, on whom the therapist and child/adolescent rely in order to be able to engage in treatment. Because of the unique characteristics of working with children and adolescents, negotiating, developing, maintaining, and repairing the therapeutic alliance is potentially complex. The therapeutic alliance is defined as an agreement on tasks of therapy, an agreement on goals of therapy, and the relational bond between therapist and client. In this meta-analysis, Karver and colleagues reviewed 28 studies of psychotherapy with children and adolescents. The mean age was about 12 years, most children/adolescents had internalizing problems, but others had problems with externalizing behaviors, and substance abuse. Almost two thirds of the studies involved a version of behavior or cognitive behavioral therapy. The therapeutic alliance was measured from the perspective of the client, therapist, and/or the parent. The overall mean effect size of the alliance-outcome relationship was small to moderate: r = .19 (p < .01, 95% confidence interval [CI] [0.13, 0.25]). Larger effect sizes were seen in those therapies of children and adolescents with internalizing disorders (r = .19), and when the therapist – parent alliance was measured and correlated with outcomes (r = .30). In other words, a positive alliance was most important for internalizing disorders, and for the relationship between therapist and parent.

Practice Implications

The findings of this meta-analysis indicate that the therapeutic alliance, especially with the parent, is important to the outcomes of children and adolescents in psychotherapy. Clinicians should not only develop an alliance with the youth, but also with the parent/caregiver. Therapists should also consider measuring the alliance regularly during therapy as a means of heading off any ruptures (with the youth or the parent) that might endanger the therapy. The authors recommended using the Therapeutic Alliance Scale for Children – Revised with children/adolescents, and the Working Alliance Inventory with parents.

Eubanks, C. F., Muran, J. C., & Safran, J. D. (2018). Alliance rupture repair: A meta-analysis. Psychotherapy, 55(4), 508-519.

It is difficult to over-state the importance of developing and maintaining a therapeutic alliance in order for patients to experience a good outcome from psychotherapy. The alliance is the collaborative agreement between therapist and patient on the tasks and goals of therapy, and the emotional bond between therapist and patient. A previous meta-analysis found a moderate but highly reliable association between a good alliance and patient outcomes. The alliance is a trans theoretical construct – that is, it is important to all types of therapy regardless of theoretical orientation. Sometimes deteriorations in the alliance occur manifested by a disagreement on the goals, a lack of collaboration on the tasks, or a strain in the relational bond. Other terms for this phenomenon include weakenings, misattunements, challenges, resistances, enactments, and impasses. Such deteriorations can vary from minor tensions to major ruptures in the relationship. Tensions and ruptures in the alliance are common occurrences in therapy with some studies showing 50% of therapy cases experience at least a minor tension within the first six sessions of therapy. There are two main types of alliance tensions/ruptures. (1) Withdrawal tensions/ruptures occur when the patient moves away from the therapist, such as when the patient changes the subject, goes silent, and cancels appointments. These tensions/ruptures are more subtle and harder for therapists to detect. (2) Confrontation tensions/ruptures occur when the patient moves against the therapist, such as when the patient expresses dissatisfaction with or pressures or tries to control the therapist. These tensions/ruptures are more obvious, but also difficult for therapists to manage because of the feelings they evoke. In this meta-analysis, Eubanks and colleagues reviewed 11 studies representing 1,314 patients. They found that the association between rupture repair episodes and patient outcomes was on average moderately large r = .29, d = .62, 95% CI [.10, .47], p = .003.

Practice Implications

The research on alliance tensions/ruptures and repairs is still new but points to some important therapist practices that could improve patient outcomes. Therapists must be attuned to indications of tensions and ruptures in the therapeutic relationship. Therapists immediately need to attend to confrontation tensions/ruptures, in which patients express dissatisfaction or hostility. Similarly, therapists must address more subtle withdrawal tensions/ruptures, in which patients go silent, evade, or appease. Therapists can acknowledge the tension/rupture directly and nondefensively by inviting patients to explore their experience of the rupture. If necessary, therapists might change the tasks or goals of the therapy to better match the patient’s concerns. Therapists should empathize with a patient’s negative feelings about the therapy, and validate the patient for bringing up their concerns. If appropriate, therapists should take responsibility for their part in the tension/rupture and not blame the patient. Also, if the tension/rupture is a repetition of an interpersonal pattern for the patient (e.g., the patient tends to withdraw in relationships), then the therapist might consider carefully exploring the tension/rupture as it occurs in the therapy with the understanding that it is a repetitive pattern. Mainly, therapists need to anticipate that tensions and ruptures will occur in therapy, that they can be destabilizing for the therapist and therapeutic relationship, and so therapists need to recognize and know how to explore their own and their patient’s negative feelings.

Author email: catherine.eubanks@einstein.yu.edu

Tryon, G. S., Birch, S. E., & Verkuilen, J. (2018). Meta-analyses of the relation of goal consensus and collaboration to psychotherapy outcome. Psychotherapy, 55(4), 372-383.

A key element of the therapeutic alliance is for therapists and clients to collaboratively come to a consensus about what they will work on. Goal consensus is part of the agreement between therapist and client, and in part it defines what will be the tasks of therapy. The tasks of therapy (i.e., what a therapist and client do in therapy to alleviate the problems or address issues) follow from the goals and conceptualization of the problems. At times goal consensus is straight forward. The client wants to feel less depressed and the therapist proposes certain therapeutic actions to help the client to be relieved of their depressive symptoms. However at other times, despite an agreement on the main symptoms, the client may not agree with a therapist’s conceptualization and tasks of therapy. For example, a therapist might believe that the client’s history of abuse and/or their current problem with alcohol may underlie the depression, but the client does not want to address these underlying issues. In such an example, the therapist and client only barely agree on a goal, and may not agree on how to go about alleviating the symptoms. In some cases there is outright disagreement, a misunderstanding, or vagueness about the goals, and so there is no consensus and therefore no basis for a collaboration. Collaboration and goal consensus are pan-theoretical processes that apply to all forms of therapy. However, research in the past decade has focused almost exclusively on behavioral or cognitive therapy studies using homework compliance as the index of collaboration. Tyron and colleagues conducted a meta-analysis of 54 studies of the association between goal consensus and client outcomes and found a moderate and significant correlation, r = .24 with 95% CI [.19, .28]. They also reported similar findings from a meta-analysis of 53 studies of therapist and client goal collaboration and client outcomes, in which they found a moderate and significant effect, r = .29, 95% CI [.24, .34].

Practice Implications

These meta-analyses show a positive link between goal consensus and collaboration with psychotherapy outcomes. Therapists should clarify clients’ goals for therapy, and therapists should share their conceptualization of the clients’ issues or symptoms. This conceptualization will determine to some extent the tasks or methods of therapy. For some clients, this process may take time and require revisiting throughout the course of treatment. Collaborative work to establish the goals and focus of therapy may in and of itself be therapeutic for those clients who have long standing interpersonal problems. Therapists should seek input from clients about the formulation and treatment plans, and be prepared to adjust their intentions according to client preferences. Therapists could invite continuous client feedback about the goals and tasks of therapy and monitor client progress. Then therapists can use this feedback to modify their interpersonal stances and treatment methods.

Author email: gtryon@gc.cuny.edu

Burlingame, G. M., McClendon, D. T., & Yang, C. (2018). Cohesion in group therapy: A meta-analysis. Psychotherapy, 55, 384-398.

Many writers consider group cohesion to be one of the most important concepts in group psychotherapy and that is a pre-requisite for positive patient outcomes. That is, patients in a group must feel a bond with the group and its members, must value the relationships in the group, and must see the group experience as a vehicle by which to achieve the change that they want. As Burlingame once noted, over time cohesion has become synonymous with the therapeutic relationship in group therapy. Although an important concept, cohesion has been elusive to define partly because of the complexity of group therapy itself. From the patient’s point of view, relationships in group therapy can take on three structural aspects in the form of member to member, member to group, and member to leader interactions. And so, cohesion may refer to the quality of the member’s relationship: to other members, to the group as a whole, and/or to the group leader. By “quality of group relationships”, clinicians and researchers often mean the positive affective bond (warmth, empathy, attraction, compatibility, trust) and working relationship (consensus on tasks and goals, willingness to work) that members have with other members, the group as a whole, and/or the group leaders. In this meta-analysis, Burlingame and colleagues identified 55 group therapy studies including over 6,000 patients that investigated the cohesion – outcome relationship. The average correlation of cohesion to patient outcomes in the 55 studies was statistically significant, r = .26 (95% CI [.20, .31], p = .01), suggesting a moderate effect. Leaders who had an interpersonal orientation had the highest cohesion – outcome relationship (r = .48), although leaders of other theoretical orientations also posted statistically significant but lower values. A greater group process orientation (r = .36), emphasizing greater interactions among group members (r = .36), composing groups of members with similar diagnoses or problems (r = .23), and groups lasting more than 20 sessions (r = .41) also each produced significantly higher cohesion – outcome correlations.

Practice Implications

The group cohesion – outcome relationship is highly reliable and suggests that clinicians of all theoretical orientations should routinely assess and enhance group cohesion to improve patient outcomes. Ways of increasing cohesion include emphasizing member to member interactions in a group, and discussing group processes as they occur. These processes may be related to members interacting with other members, with the group as a whole, or with the leader. In particular, group therapists should promote a positive emotional climate by handling conflict and avoidance when it arises in the group. This takes particular skills, training, and knowledge in group therapy processes, and so it is important for therapists to be aware of current practice guidelines for group therapy.

Author email: gary_burlingame@byu.edu

Beutler, L. E., Edwards, C., & Someah, K. (2018). Adapting psychotherapy to patient reactance level: A meta‐analytic review. Journal of Clinical Psychology. Advance online publication.

This is another meta-analysis part of the Psychotherapy Relationships That Work series. In this study Beutler and colleagues looked at client resistance and its more extreme form, reactance. Resistance refers to a client avoiding to make changes advocated by the therapist, whereas reactance indicates not only that a client resists but also moves in a direction away from what the therapist is advocating. Social psychologists define resistance as a state of mind aroused by threat to one’s freedom and then attempts to restore one’s freedom. Resistance and reactance are relational concepts – that is, they are not only qualities of the client but defined by the therapeutic relationship. Therapists play a role in resistance by the degree to which they are directive, and by their ability to adjust their level of directiveness or control to the client’s characteristics. Therapist directiveness refers to the degree to which a therapist uses suggestion, interpretation, and assignments in therapy, such as: homework, setting topics, and leading the session. One way for a therapist to adjust their interpersonal stance is to reduce their level of directiveness with clients who are more resistant. In this meta-analysis, Beutler and colleagues reviewed 13 studies representing 1,028 clients. The aggregate effect size for the association between client reactance and therapist directiveness with client outcomes was d = 0.78 (SE = 0.1; p < .001; 95% CI: 0.60–0.97), which is large and significant. In other words, if a therapist adjusted their level of control by lowering it in the face of a resistant client, then client outcomes were better. The opposite was also true, if a therapist increased their directiveness for clients who were less resistant then those clients had better outcomes.

Practice Implications

The results indicate that if client resistance or reactance is not met with confrontation and control, but with acceptance and non-defensiveness, the client may have a better outcome. Resistant or reactant clients will likely do better in a therapy that is less directive, whereas clients with lower levels of resistance may do better with more directive interventions. Therapists may do well to assess routinely the level of a client’s resistance, and adjust their interventions accordingly. Highly resistant clients may need a more collaborative approach, and a transparent discussion that focuses on the impact of certain interventions and therapist interpersonal stances on the client’s sense of control and personal freedom in the therapy.