Therapists know that the therapeutic relationship (or “alliance”) is crucial in achieving successful outcomes in psychotherapy. It is often considered the most important ingredient. But have you ever wondered what the research on this says? In this post, we will explore the many facets of the therapeutic alliance, from its definition and measurement to its role in different therapeutic approaches. I will delve into some of the latest research on this topic and share sources for getting science-backed practical tips on how to strengthen the therapeutic relationship.
The basics
As so often in psychology, measuring alliance is a matter of questionnaires. The Working Alliance Inventory (short: WAI) is the most commonly used, developed by Adam Horvath. It can be downloaded for free on his homepage. Among all the putative mechanisms of action of psychotherapy, the alliance is the one on which the most research has been done. A seminal review of almost 300 studies including more than 30,000 patients led by Christoph Flückiger from the University of Kassel (Germany) found that patients who report a better therapeutic alliance also tend to have better treatment outcomes. They were also less likely to quit therapy prematurely. This was the case in all theoretical orientations they studied, be it cognitive behavioral therapy, psychodynamic therapy, or other approaches. The effect even was there even if psychotherapy was conducted over the Internet! But it was small. Differences in the therapeutic alliance explain 7-8% of the differences in therapy outcomes. In other words: there is only a 59% chance to observe an above-average therapy outcome if the therapeutic alliance is also rated above-average.
The devil is in the details
While there is an impressive amount of research on the therapeutic alliance, there are still many unanswered questions. One of the biggest challenges in studying the mechanisms of therapy is determining causality. For instance, is it accurate to say that a strong therapeutic alliance causes a reduction in a patient's problems? Or could it be the other way around - that patients who improve significantly during their therapy are more likely to rate the alliance higher because they appreciate the help they received? After all, it's natural to feel positively towards someone who has helped you. While Flückiger and his team acknowledge that there is certainly a correlation between the therapeutic alliance and treatment outcomes, it's not possible to definitively conclude that a better alliance causes better outcomes.
To determine the causal relationship between the therapeutic alliance and treatment outcomes, researchers have looked at whether patients report fewer symptoms after an improvement in the alliance. However, if the alliance is simply a result of a patient feeling better, the relationship should be reversed. Recently, a top-notch team of psychotherapy researchers investigated this issue by analyzing data from over 5,000 patients who rated their alliance and current symptom severity after their first seven therapy sessions. Their findings supported both directions - if patients felt worse before starting therapy, they rated the alliance less positively after the session, but if they rated the alliance more positively after the session, they tended to report less severe symptoms before their next session. This suggests that there might be an "upward spiral" between the therapeutic alliance and treatment outcomes. However, the effect the study found was small - even smaller than in Flückiger’s meta-analysis.
Just because one event occurs after another, it doesn't necessarily mean they're causally related. To determine if altering the therapeutic alliance affects treatment outcomes, experimental studies are necessary. However, it's currently unclear how to manipulate the alliance in a specific way. One suggestion is for therapists to use different communication styles, such as being neutral or warm. However, the first studies on this show that both styles receive equally high alliance ratings. As a result, new technical developments in psychotherapy research are still needed to determine how to directly modify the therapeutic alliance. Without a proven method to modify the alliance, it's impossible to definitively conclude that it leads to greater success in therapy.
Making the alliance useful in therapy
While some therapeutic orientations, such as psychodynamic methods, directly use the relationship between patient and therapist for therapeutic work, newer approaches like CBASP or Cognitive-Analytic Therapy have integrated similar techniques. However, research has shown that the therapeutic alliance can play a role in any type of therapy, making it important to keep it in mind regardless of the therapeutic approach.
A team of researchers, Jeremy Safran, Christopher Muran, and Catherine Eubanks approached the alliance and therapy outcome question from a different perspective. Instead of examining how strong the relationship between alliance and outcome is, they focused their attention on breaks in the therapeutic relationship and how they are repaired during therapy. These temporary deteriorations in the alliance are called “alliance ruptures”. Even in the best therapies, disagreements, and misunderstandings can occur. Recognizing and repairing them is associated with better therapy outcomes. Training therapists in recognizing alliance ruptures and teaching them effective repair strategies significantly improved therapy outcomes, but only for cognitive behavioral therapists. This video gives a short example of how to deal with alliance ruptures.
The most important factor?
All the encouraging study results suggest that the therapeutic relationship plays an important role in helping patients. However, the effects are rather small, and problems in establishing causality warrant healthy skepticism. From the available research, we can’t really conclude that the alliance determines success or failure of psychotherapy Moreover, it is not without risk to single out the alliance as the "most important factor". This could lead to a situation where the most effective method is not chosen for the treatment of severe mental disorders. For example, in the case of post-traumatic stress disorder, methods that are trauma-focused and involve exposure were shown to be more effective than less focused methods, but these are often used in routine practice. Thus, although the alliance could certainly play its role in therapeutic success, it often comes down to technique. Depending on the problem to be worked on, other mechanisms and factors may be important for a good outcome, so it is important to look closely.
By the way: some treatment approaches even explicitly do not want to allow a long-term alliance between patient and therapist. In the Dutch PSYTREC clinics that treat thousands of post-traumatic stress disorder patients every year, the therapist is changed after each treatment session. Nevertheless, the treatment seems to work well. Therefore, it seems doubtful whether an alliance that goes beyond single sessions is a necessary condition for effective treatment.
What now?
In summary, alliance plays a small role in all therapeutic methods. There is no reason to overemphasize its role, but it can be quite understandable if therapists want to keep a special eye on it. Based on the research I presented here, there are some things you can do as a practitioner to explicitly include the alliance in ongoing therapies.
Measure the alliance regularly during therapies. The WAI is available in many languages and it is free. If you do this regularly, you can notice alliance ruptures and address them right away. Of course, because of your training, you have some sense of how the alliance is pronounced, but the patient perspective can often be quite different - and that's what counts in the end.
Apply tried and tested rupture-resolution strategies. These can be learned in specialized workshops called “alliance-focused training”. These are provided in many areas in the United States or Germany and can be valuable for therapists from all orientations. These trainings seem to work especially well if you practice cognitive behavioral therapy.
Don't worry if your therapeutic approach is not as relationship-oriented as others. Therapy can be effective even if the alliance is just "good enough".