Breaking Up Is Hard to Do: Terminating Therapy Before Things Get Out of Hand

Psychiatric TimesPsychiatric Times Vol 29 No 6
Volume 29
Issue 6

Cases that come to our attention as malpractice claims, ethics claims, or Board of Registration complaints raise the question: why did the treating clinician not terminate the treatment before things got so out of hand?

Cases that come to our attention as malpractice claims, ethics claims, or Board of Registration complaints raise the question: why did the treating clinician not terminate the treatment before things got so out of hand?

The answers to this question embrace a wide range of issues, including rescue fantasies; beginners’ errors; transference/countertransference impasses; the drive or need to complete a task once begun; power issues; conflicts around abandonment; and narcissistic, sadistic, or erotic dynamics in the therapist and/or the patient. Often these difficulties play themselves out through boundary violations. Moreover, patients may, in some cases, employ a variety of strategies aimed at resisting termination, including threats of suicide, violence, or litigation and other forms of complaint.

Termination is the appropriate ending of therapy, but also it challenges the emotions of both parties in the dyad.1-5 We focus here on the therapist’s problem, recognizing that the particular chemistry of the dyad may be the wellspring of the issues leading to the impasse. Therefore, although patients may have difficulty in leaving treatment, this analysis addresses the matter from the therapist’s side.

Long-term treatment is not inappropriate, although it is difficult to achieve in today’s environment. The point is that treatment at impasse may be inappropriately prolonged under certain conditions-conditions that may lead to various forms of medicolegal trouble.

The case presented here is a composite of actual cases encountered by the authors. Although the case features a patient who has a personality disorder, termination problems can occur with a broad spectrum of conditions.

The italicized sections in the case are intended to indicate those ideas and postures that were pivotal tothe bad outcome, but the issues, of course, spread far wider than this one example.


Ms J is a 42-year-old married woman with recurring and persistent depression. She tells her psychiatrist, Dr S, that she terminated treatment with her 2 previous therapists out of frustration because they did not give her enough support and did not understand her depression adequately. She refuses to give Dr S permission to contact any previous treater, anxious lest those records “poison” this treatment. Dr S feels there is no patient he cannot treat, records or not, and decides to rescue her from her current depressed state that has apparently been mishandled by previous treaters. He agrees to treat her in his private home office, accessed by a separate entrance in his house. Ms J frequently passes, and interacts with, Dr S’s children as sheuses this entrance.

Early in the therapy, Ms J tells Dr S he is a miracle worker who, she hopes and prays, will never abandon her. Dr S, willingly accepting this description of himself, repeatedly reassures Ms J that he would never think of abandoning her.

Over time, Ms J demands that Dr S run over sessions if she is late, switch her session time to late in the day, and come to her house for a session-later, for every other session. After refusing all these requests at first, Dr S ultimately agrees to all of them, wishing not to be the disappointing therapist like her previous treaters. On one visit, she offers him wine, which he first refuses, then accepts. When she invites him to swim in her pool, he decides he should leave. He considers ending the therapy but feels that he should see it through, especially because of his patient’s obvious sensitivity to abandonment.

Invited in a demanding manner to visit again on a later occasion, he refuses on ethical grounds. The patient, enraged and arguing that she can only be treated in her home environment, states that, after all, he did come over before and if he really cares, he will come over again. When he does not do so, she leaves multiple messages stating that if he does not immediately call back, she will take her own life. Dr S rejects the idea of consultation because he perceives it as a sign of weakness. When he calls, she denies that she meant her threats and states she is “testing his loyalty.” She reminds him pointedly of all the sessions that ended with mutual hugs and of all the letters and e-mails signed “Love,” as though to say, “You’ve come this far . . .” Dr S decides to consult his insurer’s attorney.

Dr S writes the patient a letter indicating that because her calls and harassment have made the therapy unworkable, he believes it will be in her best interest to move on with another therapist. He offers 3 termination sessions and a 30-day supply of medication. The patient’s return letter is filled with recriminations, threats, reports of consultations with other therapists and plaintiffs’ attorneys, and reminders of his promise never to abandon her. Not so subtly, she threatens him with Board of Registration complaints, civil suits, and claims of ethics violations. She begs him to continue treatment, promising changed behavior according to his rules; if he refuses, he can discharge his obligations by a letter of apology and a full refund.

No specialized training is required to recognize how soon the situation in the case presented here spiraled out of control. Clearly, when the decision to terminate comes very late in the game, it often fails to “take.” We repeatedly encounter factors that delay the decision to terminate-even when that response is the only remaining appropriate one.6

Narcissistic issues and rescue fantasies

Clinicians enter the health field to help others and derive satisfaction from so doing. They have a sense of their own helpfulness and competence. However, these laudable attitudes may go astray into pathological narcissism and rescue fantasies. Informed consent represents the core of therapeutic work and should stand clearly at the outset.

The patient in the case refused to permit her previous records to enter into the treatment. A patient’s refusal to grant permission to obtain past records (or, at a minimum, to speak to the previous treater) should trigger active discussion of this matter. It may ultimately be a deal breaker to continuation of treatment, especially when the patient calls attention to problems in previous treatment. Many ominous pitfalls and therapeutic errors or impasses may be revealed by the records of, or remarks about, even failed previous treatments. Of course, great difficulty may attend locating previous treaters and obtaining records, assuming they kept records.

Grandiose and possibly counterphobic attitudes led Dr S to go ahead with treatment, despite being unable to obtain the patient’s past records. The expressed rescue fantasy (“I will save this patient”) and competitive strivings with devalued past treaters (“they failed, I will not”) also played a role in the decision. The narcissistic view that “I can treat anyone” is as erroneous as “anyone can be treated.”


The patient’s early idealizing transference was met with reciprocal self-idealization by the therapist. This reciprocity may create the so-called magic bubble,7 a conceptual sphere that begins with a mutual admiration society containing super-patient and wonder-doctor. The doctor then becomes impervious to consultation, supervision, good judgment, and common sense.

We can infer some dynamic conflict around countertransference-derived sadistic feelings in Dr S, who bends over backward, as it were, to avoid abandonment-even when responsible termination and referral would have avoided that form of malpractice, although that was not the patient’s subjective feeling about it. Consultee therapists often report fearing the anger of the patient if they were to suggest termination. The unneutralized rage of individuals with personality disorder has the apparent ability to intimidate even seasoned therapists.8

In addition to aggression, libido may play a role in delaying termination; gratifying erotic feelings and fantasies may account for the prolongation of some therapies beyond the appropriate point. A patient’s idealization of the therapist may have an erotic force.

A common precipitant to transference/countertransference problems is a personal crisis in the life of the therapist-for example, serious illness, divorce, or death of a loved one. Such stressors can easily weaken the therapist’s objectivity and ability to set limits early.9

Common pitfalls

You’ve gone this far . . . Some features in the case presented above are commonly encountered and merit particular focus. One is the “you’ve already gone this far . . .” approach, in which the patient points out that some boundaries have already been crossed and thus further transgressions are required or demanded. In this situation, therapist guilt about past boundary transgressions may be the driving force behind agreeing to continue. Conversely, threats of complaint about early boundary issues may frighten or coerce the treater into continuing or expanding the crossings.

Consultee clinicians encountering “you’ve gone this far” report a complementary reaction of “it’s too late to change.” In reality, it is never too late to change one’s approach and terminate and refer as indicated, although in fact that response may not avert unfortunate consequences. The distance gone “too far” all too often leads to very undesirable results.

Countertransference sadism. The ability to say “No” to a patient’s demands, knowing that this response will frustrate, anger, or hurt the feelings of the patient, requires that the treater be comfortable with sadistic feelings, since one will be causing pain to another. We suggest that the widely acknowledged decline in awareness and teaching of dynamic theory has left many therapists unfamiliar and uncomfortable with this common yet problematic countertransference response.

Issues of power and control may also explain difficulty in terminating. Some therapists may be reluctant to give up a position of power over the patient and thus continue to hold on.

Consultation phobia. A significant number of cases on which we have consulted reveal a highly disturbing unwillingness of the treater to obtain consultation, thus violating a sovereign principle of risk management, “never worry alone.” This reluctance empirically stems from 3 main sources: feelings of shame and guilt at what has already happened in the therapy; fear of reporting consequences (to Board or ethics committees) if the details are revealed; and, more regrettably, fear that the consultant will tell the therapist to get out of the relationship.

Failure to maintain termination. While flexibility and willingness to negotiate are hallmarks of good therapy, a necessary and indicated termination should be sustained, even against psychological and other pressures from the patient to retract it, replete with promises to change, as in the case example. Some patients may attempt to exert pressure on the present therapist by refusing to see subsequent therapists, although that is the patient’s choice. Consultation may be helpful in strengthening the treater’s resolve.

Documentation issues. Appropriate documentation is extremely important in the charged situations described, but our experience is that the record can cut two ways. Recording inappropriate behavior by the patient, and the patient’s failure to be able to remedy such behavior after ample time and assistance, can be extremely helpful in refuting claims of negligence or abandonment. Inclusion in the record of all letters and e-mails to and from the patient avoids the appearance of concealing potentially damaging materials. However, e-mails and letters from the therapist that display an inappropriate tone, excessive informality, hints of intimacy, and expressions of endearment are damaging rather than protective.

The need to finish. Even those therapists who are not highly compulsive may express an understandable wish to finish what they started: they resist the idea of terminating before the therapy is “completed.” This view acts as a strong deterrent to necessary termination.

The benefit of the doubt. Efforts should always be made first, in the service of informed consent, to negotiate an alliance-based, boundary-respecting therapeutic relationship. Clear explanations of professional limits should be freely given. Barring physical attack or serious threat by the patient, which may require abrupt cessation of therapy, most terminations should be discussed in advance, negotiated, and enacted in a professional process. A responsible termination with appropriate referral does not constitute abandonment.


Given the complexity of the therapist-patient relationship, this brief review can only point to particular known trouble spots; it cannot cover the entire issue comprehensively. However, the ubiquity of this problem in leading to trouble for therapists requires at least a warning.

All of us who strive to be good therapists want to be helpful, to stick with the patient through the rough times, to finish a job once begun, and to assist the patient in meeting his or her clinical needs. But we cannot help everyone, and in some cases, we may need to stop. Breaking up is, in fact, hard to do, especially in a field that is largely based on sustained relationships. At certain points, however, it is necessary. Not everyone can treat everyone; not everyone can be treated. Awareness of the issues noted here may be helpful to clinicians struggling with essentially irreconcilable differences with their patients. Section 6 of the Principles of Medical Ethics With Annotations Especially Applicable to Psychiatry states10:

A physician shall, in the provision of appropriate care, except in emergencies, be free to choose whom to serve, with whom to associate, and the environment in which to provide care.

Similar provisions are contained in other behavioral health specialties.

One approach has the therapist describe a treatment plan at the outset of therapy that requires such basics as on-time sessions, limited phone calls for emergencies, release of prior records, and a boundary-respecting approach going forward. The patient’s consent to this plan is sought (and usually given). If the plan is violated, the therapist explains that the patient is violating the agreed-on treatment plan, and the therapist cannot continue to provide treatment under those conditions. Such an approach makes informed consent the heart of the contract, as it should be.

The avoidance of consultation is a particular problem. We all have some hesitation in exposing our work to scrutiny, but the importance and clear value of consultation, not only as advice but also as protection, should be stressed for trainees and encouraged for peers among practitioners. Whether or not one consults through fear, one should never fear to consult. A patient who refuses to consult should be strongly urged to follow that advice. (Comparably, a patient should consider terminating with a clinician who refuses to consult.)

A point sometimes lost sight of by clinicians in crisis is this: if you realize you are practicing, by whatever means, below your own standard of care, you cannot defend your practice as being above the needed standard of care. No matter how far you fear you have gone astray, it is never too late to change your ways in the service of the patient’s welfare.

Few things are as valuable in the situations described as advice from an experienced health law attorney-preferably one who is familiar with psychiatric issues. In addition to valuable advice, an attorney can write termination letters to patients and intervene if the patient refuses to stop, threatens, calls excessively, or fills up your answering machine to a degree constituting harassment or other criminal acts.

In sum, clinicians must realize that in some cases, the best therapy is letting go.



1. Levinson H. Termination of psychotherapy: some salient issues. Social Casework. 1977;58:480-489.

2. Hynan DJ. Client reasons and experiences in treatment that influence termination of psychotherapy.J Clin Psychol. 1990;46:891-895.

3. Hiatt H. The problem of termination of psychotherapy. Am J Psychother. 1965;19:607-615.

4. Martin ES, Schurtman R. Termination anxiety asit affects the therapist. Psychotherapy. 1985;22:92-96.

5. Dewald PA. The termination of psychotherapy. Psychiatry Dig. 1967;28:33-43.

6. Gutheil TG, Brodsky A. Preventing Boundary Violations in Clinical Practice. New York: Guilford Press; 2008

7. Gutheil TG, Gabbard GO. The concept of boundaries in clinical practice: theoretical and risk-management dimensions. Am J Psychiatry. 1993;150:188-196.

8. Gutheil TG. Medicolegal pitfalls in the treatmentof borderline patients. Am J Psychiatry. 1985;142:9-14.

9. Norris DM, Gutheil TG, Strasburger LH. This couldn’t happen to me: boundary problems and sexual misconduct in the psychotherapy relationship. Psychiatr Serv. 2003;54:517-522.

10. American Psychiatric Association. The Principles of Medical Ethics With Annotations Especially Applicable to Psychiatry. Arlington, VA: American Psychiatric Publishing; 2010.

Related Videos
suicide prevention
gender and sexuality
© 2024 MJH Life Sciences

All rights reserved.