Reflections on the Therapeutic Alliance in Clinical Practice

Many mental health trainees come to psychiatry from medicine, where a doctor-patient relationship is often based on compliance. Such a view in psychiatry can be an unfortunate potential obstacle to learning and practicing dynamic psychotherapy. Implied also is the more authoritative role of the doctor and the more passive, acted-upon role of the patient, as opposed to being "acted with." Some approaches to psychopharmacology and behavioral therapies may also take on this infelicitous form. In contrast, psychotherapy is based on the more collaborative relationship, sometimes called "the rational alliance," which has proven useful in a number of forms of psychotherapy, not just analysis.¹ Clinical and supervisory experience further demonstrate that the therapeutic alliance is also an aid in helping to retain patients in therapy and to diminish dissatisfaction (even liability) in clinical practice. Conversely, alliance threats or failures may lead to rejection and interruption of therapy. An extensive, taxonomic 1979 article attempted to sort through various definitions and uses of the term "alliance" in contemporary psychoanalytic literature.¹ In contrast to that academic complexity, the simplified model of the alliance provided below has proven useful to trainees and consultees for many years.

Moreover, the psychiatric field has become dominated by psychopharmacology, cognitive behavioral therapy, and treatment limited by managed care. These trends hay have turned attention away from something so basic as an alliance-based psychotherapy. The time may thus be opportune to return to basic psychotherapy principles.

A Working Definition: The Therapeutic "We"

In simple terms, the therapeutic alliance in practice may be defined as follows: in the therapy relationship, the healthy side of the therapist collaborates with the healthy side of the patient to confront and help to resolve the illness, symptoms, or conflicts of the patient. This model views the patient and therapist as divided, or of a divided mind. The healthy side of the therapist might include empathy, absence of bias, and freedom from problematic countertransference and exploitation. Note that the healthy side of the patient may be hard to connect with, eclipsed by the severity of a disorder, or unavailable during early stages of the relationship; the therapist may need to seek out the alliance in this case. Some techniques later outlined, together with patience, may evoke an alliance over time.

One surprising implication of this model is that the therapist is not actually working for the patient as would be expected. Instead, the therapist is working for a part of the patient and against another part. Failure to understand this view may lead to an impasse conveyed in 1 patient's complaint, "You are not helping me, you are not doing or prescribing what I want, and you're supposed to be my therapist." One of several responses might be: "I am your therapist, but I work for the side of you that doesn't want to die or kill someone else, doesn't want to use substances, doesn't want to keep up the circle you feel trapped in; we need to work together on this."

One of the hallmarks of an alliance-based approach is the therapeutic "we." This may appear in various forms, such as these examples: "How should we view this dream; what is it trying to tell us? What are we to make of this behavior at this time? How should we proceed? What seems to be happening with us in this room now? Let us go back to basics [when an impasse arises]: do you recall what we are here to do?" Many opportunities typically arise for the use of this helpful pronoun.

Alliance Threats

In practice, most problems in breakdown of the alliance¾and subsequent flight from therapy or liability claims¾arise from what are generally called alliance threats: experiences that threaten the sense of collaboration essential to therapeutic work. Understanding such threats should be part of the training of residents and others; as supervisors, we need to help trainees remain alert to threat events. Some of those commonly seen are incidents of a judgmental attitude toward the patient or patient reports of judgments from family or peers, failures of empathy in the therapist, countertransference, unchallenged distortion of the therapist's role and the like (note also that transference distortions, in the room or in displacement, are perhaps the most frequently occurring alliance threats). Still other forms involve evidence of the treater’s biases in several possible ways:

1. A resident dealing with a kleptomaniac patient brought in a magazine cartoon satirizing this disorder, and the resident assumed that this would further the alliance; however, the patient never returned. While certain uses of humor may foster an alliance, it is a double-edged sword—a patient may feel mocked or dismissed.

2. A resident came to believe that one of his patients was involved in a local crime wave; he went so far as to fingerprint the patient himself, in effect becoming a law enforcement officer, acting on his patient, for which he was censured. This example epitomizes several situations where a clinical treater may slip into forensic mode and pursue a prosecutorial exploration instead of a clinical one.

3. The phrases heard from patients at times—"You're going to think I'm crazy if I tell you this" or "I know you think what I did was wrong”—are similar invitations to judge, thus risking the alliance. As a rule, these should be promptly addressed by reaffirming the nonjudgmental position of the treater: "I am not here to judge, only to understand you."

4. A therapist's personal biases may threaten the alliance. For example, a patient may make a racial or ethnic slur about the therapist or others and seem to invite the therapist to share that view; returning to the nonjudgmental "contract" is called for here, and indicating that the nonjudgmental view will apply to others being discussed (eg, the patient’s family members).

5. Patients may comment: "You are just trying to get me to..." Statements that attribute to the treater a single specific intent, true or not, imply a goal of the therapist that is not collaborative with the patient's. This transference-based view, an alliance threat, commonly appears in patient populations.

A special case of this issue may surround an inpatient's submission of a formal written request to be discharged. This unilateral action may yet offer a chance to repair the alliance by pointing out its absence in the patient's decision. For example: the therapist may say "You have the legal right to sign out, but why aren't we planning together for a discharge in the best and safest way possible? Why use this legal approach?"²

Non-Alliance Interventions

In an emergency, where self-harm or harm to others is at risk, the ideal resolution to this situation would be an alliance-based treatment plan. Should this prove unattainable, the therapist may need to fall back on a unilateral intervention: an ultimatum (eg, deciding not to discharge a dangerous patient despite the latter's wishes and/or petitioning for commitment). Interestingly, even in such a single sided intervention as involuntary commitment, some alliance elements may still be salvaged: doctor and patient are joined in not knowing how the commitment judge in question will rule. This leaves the dyad free to share speculation together about that outcome.³

As a general principle, an impasse is best resolved by falling back to a more basic level of interaction. For example, renegotiating the contract, reviewing the goal, explaining psychotherapy at a more basic level, or revisiting the chief complaint. Any of these approaches may serve to repair an alliance that has failed or begun to fail.⁴

Alliance Failures

In some cases, despite best efforts by the treater, the alliance may fail completely. Patients with paranoid conditions are most likely to be in this group, seeing nefarious goals and conspiracies in the treatment relationship¾a view often resulting in flight from therapy. Such failures may not be remediable, but if the patient in question is not dangerous, the prudent clinician will accept this flight, while emphasizing the clinician's availability for return should the patient change their mind. It is not unusual for a patient to flee therapy, only to return later, perhaps because of a decompensation.

Concluding Thoughts

The therapeutic alliance is a cornerstone of successful therapy. Helping trainees understand this core value is an essential part of therapy training, which should include raising the trainee's consciousness to potential alliance threats. Training should also address approaches to restoring an alliance relationship if alliance threats appear.

Dr Gutheil is a professor of psychiatry and cofounder of the Program in Psychiatry and Law, Beth Israel Deaconess Medical Center Department of Psychiatry at Harvard Medical School in Boston, Massachusetts. He is the first professor of psychiatry in the history of Harvard Medical School to be board certified in both clinical and forensic psychiatry.

References

1. Gutheil TG, Havens LL. The therapeutic alliance: contemporary meanings and confusions. Int R Psycho-Anal. 1979;6:467-481.

2. Gutheil TG, Bursztajn H, Brodsky A. Malpractice prevention through the sharing of uncertainty: informed consent and the therapeutic alliance. N Eng J Med. 1984;311:49-51.

3. Eisenberg GC, Barnes BM, Gutheil TG. Involuntary commitment and the treatment process: a clinical perspective.Bull Am Acad Psychiatry Law. 1980;8:44-54.

4. Gutheil TG. On the therapy in clinical administration, part II: the administrative contract, alliance, ultimatum and goal. Psychiat Q. 1982;54:11-17.