Send email for updates
Tryon G.S., Birch S.E., Verkuilen J.
Psychotherapy: 2018, 55(4), p. 372–383.
Unable to obtain a copy by clicking title? Try asking the author for a reprint by adapting this prepared e-mail or by writing to Dr Tryon at firstname.lastname@example.org.
Research findings amalgamated for the American Psychological Association show that outcomes are substantially better the more clients and therapists agree on goals and methods and form collaborative working relationships to implement those agreements. The findings support engaging patients as partners in setting treatment goals and methods.
Summary [Though not specific to clients with drug and alcohol problems, the principles derived from this review of psychotherapy studies are likely to be applicable, partly because severe substance use problems generally form part of a broader complex of psychosocial problems. Whilst addictions work may not necessarily best conceptualised as psychotherapy, there is a therapeutic element to it which makes these findings relevant to keyworkers and other clinical staff. This review updates an earlier version with the same lead author.]
Research findings amalgamated for the American Psychological Association related outcomes to the collaborative relationship and agreement on goals between clients and therapists in individual psychotherapy.
The links between outcomes and collaboration or goal consensus were moderate and statistically significant. Assuming a causal connection, in the context of other influences these would be relatively important determinants of patient progress.
Though causality could not be established by the types of studies included in the analyses, it is probable, and the safest stance is to presume that how the therapist is and behaves affects how well their patients do partly via the quality of their collaborative relationships.
The featured review is one of several in a special issue of the journal Psychotherapy devoted to features of the therapist–client relationship related to effectiveness, based on the work of a task force established by the American Psychological Association. This particular review synthesised findings on the links between outcomes of therapy and the degree to which therapists and clients agree on objectives and ways of working and the degree to which they function as a team, working together to achieve those goals. These features of the therapeutic relationship – respectively termed ‘goal consensus’ and ‘collaboration’ – are ‘pantheoretical’ in that they are involved in all types of psychotherapy.
The authors defined goal consensus as consisting of:
• patient–therapist agreement on and commitment to goals and the methods by which they will be achieved;
• patient–therapist agreement on the patient’s problem;
• the extent to which goals are discussed and clearly specified.
Collaboration was seen as an active process whereby patient and clinician reach agreement on goals and work together to achieve them. Specifically, the authors defined it as the “mutual involvement of psychotherapist and patient in a helping relationship”, experienced as respect and a mutually cooperative stance.
Measures ( panel right) of collaboration are highly correlated with those of goal consensus, supporting the decision to consider these concepts together. As is clear from the way they are measured, agreement on goals and collaborative working are important components of the overall working alliance between therapist and client, the subject of other articles in the same special issue of Psychotherapy in respect of adults in individual therapy, children and adolescents, couples and families, and group therapy.
Completed by patients, therapists or observers, the Goals and Tasks subscales of the Working Alliance Inventory assess patient–practitioner goal consensus. Sample statements below are taken from a version for clients, who respond by choosing options ranging from “seldom” to “always”.
• “We agree on what is important for me to work on.”
• “I wish my therapist and I could clarify the goals of our sessions.” (reverse-scored)
The California Psychotherapy Alliance Scale (CALPAS) assesses the working alliance as a whole, but includes a Working Strategy Consensus subscale which assesses patient–therapist goal consensus. In its version for clients (the process is similar for therapists) they rate statements describing their experience during the session on a seven-point scale ranging from “Not at all” to “Very much so”; sample statements below.
• “Did you feel that you were working together with your therapist, that the two of you were joined in a struggle to overcome your problems?”
• “Do you feel that you disagree with your therapist about changes you would like to make in your therapy?” (reverse-scored)
Via its Therapist Understanding and Involvement subscale, CALPAS was also used by most of the relevant studies to assess collaboration. Typical statements below.
• “Did you feel pressured by your therapist to make changes before you were ready?” (patient version)
• “Were your interventions tactful and well-timed?” (therapist version)
In the analysed studies, homework completion was the most common measure of the patient’s collaboration with therapy.
To test whether these dimensions of the working alliance really are associated with better outcomes for clients, analysts sought studies of individual psychotherapy with adults which related patient outcomes to goal consensus or collaboration, and reported these relationships in the English language in a way which enabled their findings to be aggregated with those from other studies. The result was 84 studies, of which 54 involving 7,278 participants documented links with goal consensus, 53 involving 5,286 participants, links with collaboration between therapist and patient, and 21 studies involving 2,081 participants, links with the collaborative behaviour specifically of the therapist.
Findings from the three sets of studies were separately amalgamated in three meta-analyses to provide estimates of the overall strength of the link between outcomes and goal consensus or one of the two aspects of collaboration, and to explore possible influences on the strength of these links. Strength was calculated as a correlation coefficient, an expression of the degree to which outcomes co-varied with consensus or collaboration. The chosen metric ranged from -1 (perfect negative co-variation, meaning that as one side of the link gets larger the other diminishes to the same degree) to +1 (perfect positive co-variation, meaning that as one side of the link gets larger so does the other, and to the same degree). Correlation coefficients were also converted to effect sizes. Effectively these metrics indicate how influential consensus or collaboration had been if causally linked to outcomes.
With effect sizes ranging narrowly from 0.49 to 0.61, the three aspects of consensus/collaboration were similarly strongly and positively related to outcomes. In each case there was no indication that studies missed by the searches would have materially altered the finding, but there was appreciable variation in the strength of the link between different studies. When this was explored, no factors were identified which could reliably be said to have accounted for this variation. Detailed findings follow.
Goal consensus Across the relevant studies the strength of the link between goal consensus and therapy outcomes amalgamated to a correlation of 0.24, equating to a medium effect size of 0.49 – a substantial and statistically significant relationship indicating that better outcomes can be expected when patient and therapist agree on therapeutic goals and how to achieve them.
There was no indication that studies missed by the search would have materially altered this finding, but there was considerable variation in the strength of the link between different studies. Of the 20 investigated factors which might account for this variation, only one was found statistically significant: therapists’ perceptions of goal consensus were more strongly related to outcomes than were those of patients or observers. However, with so many factors tested, this may have been a chance finding.
Collaboration between patient and therapist Across the relevant studies the strength of the link between patient–therapist collaboration and outcomes amalgamated to a correlation of 0.29, equating to a medium effect size of 0.61 – a statistically significant relationship indicating that patient experience and wellbeing are considerably improved if there has been a better quality collaborative relationship. Homework completion was the primary measure of collaboration. With a correlation of 0.23 across 26 studies, this particular element was itself significantly and positively related to outcomes.
There was no indication that studies missed by the search would have materially altered this finding, but there was appreciable variation in the strength of the link between different studies. Of the many factors the analysis investigated which might account for this variation, only one was found statistically significant (collaboration ratings taken at unspecified times during therapy differed from those taken at specified times), and with so many factors tested, this may have been a chance finding.
Therapist collaboration Findings were similar to those on patient–therapist collaboration. Across the relevant studies the strength of the link between therapist collaboration and outcomes amalgamated to a correlation of 0.26, equating to a medium effect size of 0.54 – a statistically significant relationship indicating that patient experience and wellbeing are considerably better if the therapist has been more collaborative.
There was no indication that studies missed by the search would have materially altered this finding, but there was appreciable variation in the strength of the link between different studies. Too few studies were analysed to be able to adequately probe for reasons for this variation. However, similar to goal consensus findings, collaboration ratings completed by the therapist tended to be more highly related to outcomes than ratings made by others, and similar to patient–therapist collaboration results, ratings taken at unspecified times were more highly related to outcomes than ratings taken at specified times.
The analyses demonstrated positive links between psychotherapy outcomes and goal consensus or collaboration. Clinically and theoretically, it is not possible to separate patient and practitioner contributions to goal consensus, which involves both agreeing on psychotherapy goals. The meta-analytic result that goal consensus is related to favourable outcomes indicates the importance of cooperation in establishing treatment goals.
Most of the studies in the analyses assessed goal consensus and collaboration at a single time during treatment. In practice, goals change during therapy, meaning goal consensus is ongoing. Similarly, patients and practitioners collaborate throughout the course of treatment. Psychotherapy is a dynamic, ongoing process, and research on its relationship elements should reflect this reality by assessing them frequently during therapy.
The vast majority of the studies reported an association between outcomes and consensus/collaboration, but were unable to investigate whether this arose from a causal link.
These results from over 80 studies and thousands of patients suggest practices that psychotherapists can use to enhance patient outcomes.
• Begin work on the client’s problems only after you agree on treatment goals and the ways you will together set about reaching them.
• Share your knowledge with patients by educating them about assessment and treatment procedures.
• Listen to what patients say and seek their inputs in the formulation of treatment plans. Do not push your own agenda at the expense of the patient’s.
• Seek frequent feedback and reactions from patients regarding their functioning, assessment of psychotherapy, life situation, and desire to change.
• Provide patients with regular feedback about their progress.
• Develop homework assignments that address treatment goals in collaboration with patients. Start with small, easy-to-accomplish tasks and build to larger ones. Provide homework instructions to patients clearly and perhaps in written form. Get patient feedback on homework and incorporate it in your work.
• Encourage homework completion. Review homework with patients and discuss their experiences working through it. If indicated, adapt assignments to ensure patients can complete and benefit from them.
• Be ‘on the same page’ with patients. Get their feedback to ensure you are working toward the same goals and understand each other throughout treatment.
• Modify your treatment stance and methods if ethically and clinically indicated in response to patient feedback.
• Share with patients the results of the featured study linking their active collaboration to successful outcomes.
commentary At a correlation of 0.29, not surprisingly the strength of the link between patient–therapist collaboration and outcomes was almost exactly the same as the 0.28 found for the broader working relationship between patient and therapist, of which collaboration forms a major part (the other element is the emotional bond between patient and therapist). Close also to this mark were goal consensus and the degree of collaboration shown by the therapist.
In respect of the alliance as a whole, its relationship to outcomes was still statistically significant but substantially weaker for substance use clients. This possibility was not examined in the featured review, but it would not be a surprise if especially in the treatment of dependence on illegal drugs, collaboration and consensus goal setting were less salient elements of the therapeutic relationship relatively weakly linked to outcomes, which in turn weakened the overall relationship’s association with outcomes.
Public policy, professional, safety and social acceptance considerations, and the host service’s obligations and aims, limit the degree to which gaols and methods can be individualised to the patient’s preferences. Many patients would, for example, prefer not to have to attend a service several times a week to take their medication under the eyes of a clinician, but guidelines which strongly determine what constitutes acceptable practice have recommended this, and services may feel safety considerations – not just for the patient but for others too – demand it. Some patients, too, may want to be prescribed injectable heroin, an option rarely available and which has been deprecated, and to restrain or control their consumption of substitute opioid medications so they can still experience a ‘high’ from illegal heroin – a literally illegitimate goal.
After a survey of patient desires and preferences, at one English clinic treating heroin-dependent patients it became apparent that “how far the clinic could/would move to satisfy clients’ desires was constrained by national guidelines and professional standards on issues such as supervised consumption, daily pick-up of prescriptions, and the prescribing of injectables. Agencies and doctors are not prepared to risk being pilloried for transgressing these ‘guidelines’ if something goes wrong. Though they can express an opinion on whatever they like, the clients’ influence is effectively relegated to the elements of the service that for most probably matter least. For the rest, a central authority has already decided (perhaps rightly) what is best for them.”
In the treatment of drinking problems too, it has not been unusual for services to insist on abstinence as the treatment goal and to have their own non-negotiable methods for reaching that goal.
Not only may there be conflicting goals for patient and service, but also within the patient ambivalence about changing from a drug- or alcohol-focused life. Resolving ambivalence in a healthier direction without overtly directing the patient is the domain par excellence of motivational interviewing, a major reason why this counselling style has become so popular in the substance use sector. It can help even when the agendas of patient and service seem far apart, as with offenders coerced into treatment, though one risk is that is that its ‘It’s up to you stance’ can undermine progress because to both patient and practitioner it feels less than genuine; seeming genuine is, as another review commissioned for the American Psychological Association concluded, itself an important factor in therapy.
Within the restricted space available for client influence, in England the ITEP programme was an attempt to improve collaboration and in particular collaborative goal-setting via visual flow-chart style ‘maps’ developed or completed and shared by client and keyworker.
The reviewers’ practice recommendations are based on the likelihood of a causal link between collaboration/consensus and patients’ progress, which can be leveraged by the therapist to augment that progress. In other words, the assumption is that how the therapist ‘is’ and behaves affects how well a patient does, and does so partly via the collaboration they help develop between them.
Though a causal link between collaboration and outcomes is plausible, such a link could not be established by the types of studies included in the featured analyses. Rather than randomly assigning clients to therapies/therapists which differ only in their generation of a collaborative relationship, generally these simply measured collaboration or consensus at one point in therapy and related those measures to later outcomes. Such studies are generally unable to eliminate the possibility that (for example) patients who were going to do well in any event were more likely to actively cooperate with their therapists, or be easier for therapists to reach agreement with and to collaborate with. In these scenarios, collaboration would remain associated with better outcomes, but not because it helped cause them. As causality theorists have explained, “Thunder correlates with power outages, but thunder does not cause power outages. To distinguish causal from noncausal correlations, it is important to control for alternative causes.” Without effectively random allocation of patients to high- and low-collaboration therapies, these “alternative causes” cannot completely be eliminated.
Despite not being provable, for at least two reasons, a causal link between collaboration and outcomes seems likely. First is the consistency of the association. Across the 84 studies there were 128 assessments of the link between outcomes and either patient–therapist collaboration, therapist collaboration, or goal consensus. In just six assessments was the link negative; none of these were statistically significant and five were negligible in size. Second is the plausibility of the proposition that establishing a collaborative working relationship, within which therapist and patient are ‘pulling together’ towards the same goals, will help reach those goals. Posing the opposite scenario reinforces this view. If patient and therapist seek incompatibly different things from therapy and one tries to use methods the other finds inappropriate, unconvincing or uncomfortable, premature departure seems more likely than successful completion. The safest stance for trainers, supervisors, therapists, counsellors, patients and clients, is to presume that a collaborative stance is an important determinant of treatment success, and that establishing, maintaining, and as needed, re-establishing collaboration, are core tasks.
Listed below are analyses of the other reviews commissioned by the American Psychological Association task force.
Overall conclusions based on the other 16 reviews
The ‘real’, person-to-person relationship
Alliance in couple and family therapy
Alliance in child and adolescent therapy
Cohesion in group therapy
Therapist self-disclosure and ‘immediacy’
Therapist and client emotional expression
Repairing ruptured alliances between therapists and clients
Treatment outcome expectations
Feeding back client progress data therapists
Thanks for their comments on this entry in draft to research author Georgiana Shick Tryon of The Graduate Center of the City University of New York, USA, and Luke Mitcheson, Consultant Clinical Psychologist at the South London and Maudsley NHS Trust in London, England. Commentators bear no responsibility for the text including the interpretations and any remaining errors.
Last revised 28 March 2019. First uploaded 16 March 2019
Give us your feedback on the site (two-minute survey)
Open Effectiveness Bank home page
Add your name to the mailing list to be alerted to new studies and other site updates
REVIEW 2018 Psychotherapy relationships that work III
REVIEW 2018 Cohesion in group therapy: a meta-analysis