Psychiatrist-Patient Boundary Issues Following Treatment Termination

Current discussions of boundary problems often take the stance that matters have been resolved and that everyone knows what the issues are and can take heed. However, the issue of boundaries, particularly in clinical settings, is actually quite a recent one. Gabbard (6) has shown how the history of psychoanalysis is replete with examples of sexual boundary violations. Presumably, the disciplines of psychiatry, clinical psychology, and social work have similar histories.

Confusion exists with respect to boundary problems as clinical discussions have become mixed with ethical discussions and issues being raised in the legal arena. Legislatures of various states have often drafted their own versions of what they believe should be standards of conduct about boundaries in therapy. In addition, licensing boards and professional organizations have begun to develop their own standards and procedures as well. The American Psychiatric Association has revised its position on sexual relationships with former patients several times. Until 1989 such cases were decided on an individual basis, but then APA included in its published standards of conduct that sexual activity with a former patient was “almost always” unethical (7). In 1993, the APA Assembly adopted a position that sexual contact between a psychiatrist and former patient was prohibited, with no time limit specified; the APA Board of Trustees approved this change for inclusion in The Principles of Medical Ethics With Annotations Especially Applicable to Psychiatry(8). APA does not address nonsexual activities with former patients.

In this article we wish to focus on the way that posttermination boundary debates have entered into the legal arena and the confusion that has arisen when courts misuse scientific concepts in their attempts to resolve legal disputes. The goal of this article is not to undo the new standards that have been introduced, such as the prohibition against sexual intimacies between a treating psychiatrist and patient. It is rather to point out the legal confusion that arises when a clinical concept is introduced into the judicial system and then expanded into diverse posttermination situations to create the possibility of legal liability. This requires a brief look at the concept of transference, since in the various situations in which boundary violations have been alleged, the typical testimony has been that the transference was mishandled, thus leading to a former patient being harmed.

The Concept of Transference

Transference and countertransference have had extensive coverage in the psychoanalytic literature. However, even among the different schools of psychoanalysis, there has not been a universally agreed upon theory of the definition and extent of these concepts. The idea of transference originated in Freud’s early attempts to make sense out of what was happening during the course of the “talking cure” (9–12). He had come to realize that there was a need to move beyond catharsis to a clinical theory of what was happening in carrying out psychoanalysis. Some of the irrational behavior of patients was explained by the idea that feelings and other aspects of earlier relationships were “transferred” to the analyst, who represented figures from the patient’s past.

When asked what transferences were, Freud once replied,

Freud went on to point out that the value of the transference is for the patient to experience a sense of conviction about the validity of the connections made. “Psycho-analytic treatment does not create transferences, it merely brings them to light, like so many other hidden psychical factors” (13).

A historical paradox is that Freud initially did not see transference as essential to the treatment relationship. He simply viewed it as a kind of displacement. This idea was an extension of his work on dream interpretation, in which an unconscious wish was masked and became attached to some preconscious thought—a particular example of displacing affect from one idea to another. The analyst, seen almost daily, was simply one available object among several possibilities in daily life on which to carry out displacements. Transference manifestations came to be viewed as no different than any other expressions. In some cases they were viewed as symptomatic but in other cases not. Such an approach was a far cry from viewing the treatment process in psychoanalysis as heavily focusing on the transference relationship. Transference phenomena directed toward the analyst or someone else were not seen as different in nature, and they were to be explicated for purposes of getting rid of them.

Some have argued that the technique conceptually attributed to Freud as part of psychoanalysis was more a product of the 1950s. People analyzed by Freud reported that he was actually directive, sometimes scolding, and that he debated theory with patients (14). Reviewing the evolution and ongoing clinical arguments about transference over time would be a separate undertaking. What can be said is that the ideal of an analyst maintaining strict neutrality in the context of abstinence and fostering a regressive transference came after Freud.

The techniques now used in psychoanalysis and psychotherapy are much more variable, leading to the existence of ambiguities in the interpretations of transference. Some practitioners believe transference exists in all therapies, but many treatment modalities do not use the concept or believe in it. Some point to the ascetic conditions imposed on the participants to elicit unconscious wishes and fantasies during psychoanalysis as evidence that the approach fosters transference, whereas others emphasize that the analytic situation is simply one situation among several in a person’s life in which diverse wishes and reactions are elicited. One viewpoint in psychoanalysis is that transference never ends and that through the power invested in analysts, a valid consent can never be given for posttermination relationships. Others stress that all interpersonal relationships in various settings kindle feelings and fantasies and that such phenomena as a part of everyday interactions were not what unresolved transference in psychoanalysis was supposed to be about.

The psychoanalytic idea of transference should not be confused with the popularized version that has arisen in America from the burgeoning of diverse therapies. In the popular view, transference came to refer to whatever emerged in the relationship between a therapist and patient. Such a broadened, but mistaken, idea of transference is subject to misuse and misunderstanding.

Scientific Concepts as Legal Evidence

It is important to keep the uncertainty regarding the role of transference in mind as the concepts of transference and countertransference become incorporated into legal decision making without cognizance being given to the lack of scientific clarity or agreement. A preliminary question is whether the clinical concept of transference meets legal standards for admissibility as scientific evidence. Such a standard is quite different from clinical discussions about the meaning of transference. When the question is admissibility of testimony in a courtroom, the assumption is that testimony will be based on established scientific knowledge.

The standard courts used until recently was based on the Frye test (15), often called the “general acceptance” standard. As stated in that case:

While this test reigned for several decades, it had the limitation of being based on when a particular hypothesis was generally accepted—a type of majority vote mentality without such a plebiscite having ever taken place.

The general acceptance problem was present for many areas of the hard sciences, let alone for some of the vaguer concepts employed in the social sciences or clinical fields. When it came to ideas such as boundaries to maintain after a professional relationship had ended, there was not only the problem of different mental health fields with diverse standards but also, more cogently, that not all types of therapies relied upon the same concepts. A narrowly defined group, such as that of classical psychoanalysts, might reach a greater degree of consensus, but if the group surveyed was broadened, much disagreement would be apparent. From a legal standpoint the issues were how many practitioners agreed with the idea and how wide a survey should be employed to determine the acceptance rate. In practice, things rarely got beyond clinicians offering opinions.

More recently the federal courts, and some states, shifted to the Daubert rule (16). Rather than asking if most of those in a particular field believe in a concept, the court required an inquiry about the scientific validity of the concept. In fact, specific hearings were required in some jurisdictions to determine scientific validity, and judges could exclude testimony on a concept if they did not deem that it met scientific validity. In summary, the Daubert case listed four factors to determine the validity of a scientific proposition: 1) testability/falsifiability, 2) error rate, 3) peer review and publication, and 4) general acceptance. The last two are not much different from the older test, but the first two inject criteria from the philosophy of science. In fact the Daubert opinion cited Sir Karl Popper to the effect that the scientific status of a statement or theory is its falsifiability, meaning its potential to be tested and refuted. The error rate gets into areas of false positives and false negatives. If the proposition or idea is more in the nature of one side simply asserting its validity and the other disagreeing, scientific status has not been achieved. Testability is taken as the prerequisite for scientific status in contrast to an assertion of belief. At least an amenability to be tested must be present. Without further pursuing this topic, it is noted here because it is germane to the issue of how courts might consider posttermination boundary violation allegations that are based on the idea of mishandling the transference.

As civil litigation cases involving allegations of boundary violations arose during the last 15 years, the courts sought some scientific basis on which to render their opinions. Without such a basis, no standard to assess liability was present, which is where things were for much of the 20th century.

The pattern that developed was that experts of diverse clinical (and sometimes ethical) backgrounds expressed their opinions in court as to whether or not there had been a boundary violation. Those arguing that there had been a violation fell back on the concept of “mishandling the transference” as the basis for judging an alleged boundary violation. One side would offer the opinion that the transference was mishandled in treatment and continued after termination; the other side would deny either that the transference was mishandled or that it continued to any significant degree. Testimony to the effect that a continuation of transference controlled a relationship after termination was usually simply an extension of the viewpoint that treatment was never terminated. There are two implicit assumptions in this argument: that the type of treatment utilized did not matter and that there is really no difference if treatment had ended or not, since the power differential did not end.

The theme that a continuing transference was operating was often cited in trial testimony or in written reports submitted to attorneys, who used them in attempts to coerce a pretrial settlement. In time, this language became incorporated into legal decisions or assimilated uncritically into appellate opinions. This meant that courts assimilated the language of testifying experts rather than creating their own legal framework as a basis for decision making. Thus, courts did not create their own jurisprudence but rather took on the language of testifying clinicians and assumed a scientific validity existed. Innumerable examples could be given, but the following cases were selected to illustrate the confusion resulting from the legal system having not developed its own jurisprudence on posttermination boundary issues.

Posttermination Issues in the Legal Arena

Allegations of “Mishandled” Transference

Another area in which legal cases involve issues of transference is when insurance companies resist coverage for sexual misconduct claims. The same type of testimony arises, and often the question of whether the conduct occurred during treatment or after termination is bypassed. The beginning case on this issue arose in 1977, when a jury determined that a sexual relationship during treatment constituted malpractice, but the insurance company refused to pay damages (20). The legal argument was that because such actions are intentional torts, they should not be covered.

Since then courts have gone in different directions (21). A Georgia case held that the transference and a romantic involvement with a patient created a duty to defend the psychiatrist as distinguished from covering damages (22). A Minnesota Supreme Court case extended this thinking (23):

Other jurisdictions have similarly used such reasoning for finding liability based on the transference being mishandled. In passing it should be noted that some jurisdictions are now coming to different conclusions, using reasoning that does not rely on the transference paradigm. A California Court of Appeals reversed the revocation of a psychologist’s license when a relationship did not occur until after the termination of a professional relationship (24). A recent Florida case provides another example:

Another issue that has arisen involves the nature of reality and how it is to be viewed during intensive courses of psychotherapy. Apart from viewing things in extremes, where everything is viewed as either transference or realistic, it is difficult for experienced clinicians to make such determinations. Nor is raising such a question merely academic. It involves difficult assessments about what reality-based behaviors and feelings might be when compared to transference-based behaviors. The practical implications are how accurate a person’s assessments of another might be or whether misinterpretations and misperceptions are occurring.

Allegations of “Undue Influence”

More posttermination complaints are emerging in the areas of business and financial arrangements as well as in consultation situations. Attacks on such arrangements may arise from former patients or their relatives or business associates who feel deprived or cheated and use the argument that boundaries were not kept clear or “undue influence” was operating. They may argue that since transference continued, the treatment termination lacked legal significance. The undue influence operating in these situations is presented in terms of the impaired capacity of former patients to give or withhold consent to certain transactions—such as funds transferred to nonprofit foundations, charities, trust funds, educational establishments, and religious organizations—which should therefore retroactively be annulled.

When allegations of boundary violations are made, depositions and cross-examinations at trial inquire into a host of business relationships and financial arrangements. One problem encountered with increasing frequency is that of patients whose insurance has run out but who later wish to handle past overdue billings by bartering services. Later, it is alleged retrospectively that they were exploited because their goods or services were undervalued and that dependency on the therapist made them unable to withhold consent for offering services for less than they later believe they were worth.

Even more uncertainty regarding persistent transference exists in the area of professional consultations. Consultees may later claim boundary violations were present without their awareness of them at the time because of “power differences.” The logical fallacy again is to assume that every assessment or treatment relationship, no matter what the diagnosis, is primarily characterized by a power differential. Even if the diagnosis per se is ignored, and it is argued that a continuing transference creates a distortion, it leaves the question as to the degree of distortion unresolved. This question-begging conclusion is, in fact, the question that needs answering.

Discussion

Multiple questions arise from these posttermination situations. Does the former therapist retain control? Are the feelings toward an individual nothing more than what arose during the course of treatment? Are former patients unable to inhibit themselves from pursuing posttermination relationships? To view every aspect of such a relationship as “nothing but” a transference phenomenon goes beyond what is claimed in most psychoanalyses.

What is often bypassed in critiques of the scientific validity of transference and psychoanalytic theory is the fact that psychoanalysis involves different aspects of a therapeutic relationship. One posited level involves transference, but there is also the actual collaborative relationship (sometimes referred to as the working alliance) as well as the relationship between a person and the analyst as a real person (29). The working alliance is seen as the rational and nonneurotic aspect of a relationship between a patient and therapist. Even in intensive analytic treatment, this part of the patient’s rational ego functioning remains intact, and choices continue to be made about such matters as the frequency of visits, fees, and even the continuation of treatment itself. Without such a supposition of rationality, all treatment and the decision to enter or terminate treatment can only be seen as determined and devoid of any meaningful exercise of choice.

A major problem with the use of mishandled transference as the basis for legal liability is the indiscriminate application of the transference concept, which was based originally on intensive treatment situations. The variety of therapeutic approaches currently used are thus ignored. Thus, we see the same concept being introduced in courtroom procedures no matter what treatment modality was once used with a former patient. The dependent nature of the transference may be interpreted in terms of a therapist knowing, or supposedly having reason to know, that all former patients should be seen as not able to withhold or give consent to social relationships, no matter what type of treatment modality was once used. The idea that former patients cannot provide legal informed consent is thus created from the assumption that a continuing, unresolved transference will interfere with their judgment and predispose them to agree to certain relationships. This assumption bypasses an assessment of whether such an incapacity in fact exists. Even if the treatment model used had been psychoanalysis, and the assumption made that transference had continued, that in itself should not invalidate the capacity of the person to exercise some rational choices. Transference triumphalism has some limits.

The question concerning the continuing influence of transference after treatment termination remains. If transference is hypothesized as remaining unresolved, what effect does it have on a person’s capacity to exercise choices later in their life? Given that the issue is debatable for someone who has been in psychoanalytic therapy, it remains much more debatable for those who have been in other types of treatment. Note that important treatment goals are to increase the capacity of individuals to make more rational choices in their lives and to be relatively freer from disabling conflicts. In a developmental sense, treatment seeks to foster such maturation. Yet, to then argue that after treatment the individual remains perpetually dependent amounts to saying that for legal purposes a person can never achieve such a level of maturation or rehabilitation.

A basic question is whether the “power” imbalance alleged in boundary violations holds across diverse psychotherapies. Is the power balance so distorted in therapy that former patients thereafter lack the capacity to exercise choice? If this type of question cannot be answered in a satisfactory manner, it may be better to look elsewhere for the justification of regulating contacts between psychiatrists and their former patients after termination.

Allegations of a boundary violation after termination of the treatment relationship are based on the following theoretical position: a transferential relationship has occurred, the influence of which has continued into the future, thereby changing forever the capacity of the two people to have an undistorted relationship. The question often bypassed is whether there is any scientific merit in using the psychoanalytic concept of transference for assessing legal liability, or even for assessing general decision-making capacity or ability to provide informed consent. Conclusions that turn on allegations of mishandling the transference, or denials of its continuance, presumably have problems in meeting a legal standard of scientific credibility. Since the hypothetical construct of transference appears to be the primary legal rationale in arguments purporting that the transference has been “mishandled,” it leads to questions about the efficacy of such an explanation when used by the judicial system. Does it offer assistance to propose legislation that asserts former patients can only exercise distorted choices in entering into social relationships subsequent to the termination of therapy? If such choices are later viewed as undesirable, should they be primarily attributed to the transference? Such an explanation seems to provide a series of ad hoc explanations grafted onto a position already adopted.

An alternative approach could argue that social policy should shift from focusing on the concepts of transference-countertransference as the sine qua non for assessing legal liability. Rather, if society were to now consider that posttermination social contacts with someone who was once a patient are undesirable, that should simply be stated as the consensus of the community at this time. Such an approach would be in accord with positive law. Ultimately, the proposed rules would be enacted through legislation. However, it would need to be clearly noted that a new policy was being placed into effect, with no implications that it had always been the rule. Were this not announced as a new policy, it would amount to an ex post facto creation of legal liability. Such a clarification would be necessary, since the behavior had not always been viewed as prohibited.

Search for Consensus

Empirically, a question exists as to whether there is a common clinical consensus about posttermination professional relationships. If there is, who has reached it? The consensus has not been evident in such popular images as movies or in the general population. Yet, a critic might say that such sources are not the ones to employ in establishing professional standards. Rather, surveys of the profession should be used. To some extent this has been done. A 1987 survey of 5,574 psychiatrists (to which 1,423 [26%] responded) showed that while 98% said that therapist-patient sexual contact was always inappropriate, 29.6% answered that the prohibition ended with the termination of therapy, and an additional 8.5% had no opinion on the question (30). Such a survey suggests that perhaps one-third of responding psychiatrists at that time did not view such behavior as unethical. In another study, psychologists were asked their opinions as to posttermination sex between a therapist and former patient, and about 50% rated it as unethical (31). While more contemporary surveys might find different percentages, Daubert would presumably raise questions about the scientific validity of surveys on posttermination relationships. The only consensus among mental health professionals seems to be that terminating treatment for the purpose of having sex would be unethical. Schoener (32), on the basis of consultations on several thousand cases, found that the vast majority of complaints arise in situations where there has really not been a termination at all.

Determining Liability Without Scientific Validity

To assume that transference in a broad sense is present in any type of treatment situation as a clinical construct is one thing. To extend the idea and assume it is a controlling factor in all future relationships with a once-upon-a-time treating person becomes an untestable proposition in terms of scientific criteria. While one meaning of transference operates on the level of a theoretical abstraction, it is also used in diverse contexts, such as directly observable interactions. Treatment situations offer the possibility for various levels of meaning. In such situations hypotheses about transference can be proposed, discounted, and treatment approaches modified, etc. However, it is a quite different situation when clinical hypotheses are taken as the standard for assessing penalties (civil or criminal) from the idea that boundaries that were based on a past therapeutic experience have been violated. The open-endedness of the concept and its lack of clear definition are tolerable as part of clinical discussions and proposed hypotheses. Problems arise when loose clinical formulations are applied to assess legal liability where ambiguity is not desirable.

The past life of a person with its complexities of thoughts and residues of feelings exists before and after any type of therapeutic encounter. If one views therapeutic encounters from the past as being recapitulated and then carried into many future relationships, the past history of a person remains pervasive in future relationships. However, feelings that are influenced by past experiences emerge in many situations other than treatment settings, such as in the development of romantic love. Broad conceptions of transference need to be distinguished from the unconscious repetition of thoughts, feelings, and behaviors originating in early relationships. If the treatment is psychoanalysis, these are the phenomena hopefully worked through over the course of several years of treatment. Many nontherapeutic repetitions operate in people’s lives. While they may represent conflict, that is different than attributing the behaviors to a treatment transference.

What is paradigmatic in legal cases that are based on a charge of a posttermination boundary violation is that the treating person continues to possess controlling power over the patient’s free will, meaning the capacity to choose has been lost. In legal jargon the attribution is that a patient has lost his or her capacity to make informed choices. Specifically, it is that one cannot give consent to a social relationship after treatment has ended because of the interference of the past transference influences. A further allegation is sometimes made that undue influence has been present.

The furthest extension is when it is assumed that the capacity to choose remains lost, without limits, for an indefinite period of time. Those holding to such a position—that a capacity to withhold or give consent continues to be lacking because of the continuance of the transference—again seem to be using the paradigm of a person who has been in some type of treatment that encourages a strong transference. If the model is applied to all other types of treatments, including the psychopharmacological, the burden to show a lack of capacity to consent should shift to those arguing that transference has been operating.

For those who use the concept of transference, it is seen as operating in everyday encounters of life. As such, it is simply one more relationship. Transference may be seen in a variety of human relationships, such as the response of an individual to medical specialists, psychotherapists, professional relationships, friendships, and other daily encounters. A pervasive evidentiary problem is the adequacy of evidence that a person was actually reacting on the basis of a particular transference rather than to one of many other past and ongoing occurrences in life. Again, while such open-ended possibilities may not matter in making clinical formulations, it is a different matter when such difficult-to-assess determinations are used to assess legal liability.

Significance of Treatment Variables

All too often what is ignored are distinctions as to what type and degree of transference existed in patients with different diagnoses and given different types of treatment. Diagnosis is a crucial variable in selecting any treatment. If psychoanalysis is selected, transference feelings are encouraged, whereas other approaches may try to minimize them. When such a treatment choice is made by the patient and professional, it requires assessment of factors such as the capacity of the patient to distinguish reality from fantasy, form an intense therapeutic alliance capable of dealing with interpretations, and make a commitment to lengthy treatment. The potential for future boundary problems in such a situation is quite different than if some other type of brief and focused therapy was chosen, such as a behavioral or psychopharmacological approach.

Treatment may be entered for a brief time and then terminated after the person decides not to examine further the basis for certain traits, such as suspiciousness or manipulating others. After leaving treatment the person continues to exhibit the same traits. Cases involving allegations of boundary violations may stem from such traits, which are kindled in a retrospective distortion after treatment has ended but have existed before and following treatment. Legal issues about boundaries arise particularly in cases of patients with personality disorders. As an example, consider the manifestations of a patient with narcissistic personality disorder, in which psychoanalysis or intensive psychotherapy tries to deal with unresolved developmental problems. Distortions of the self or of object representations are prominent, along with disturbances in separation-individuation processes. In the course of treatment, such patients have difficulty experiencing the therapist as separate, since other people are often experienced narcissistically as extensions of themselves. They may expect to control the therapist, much like they try and control their world. Treatment would not try primarily to resolve conflict but rather help development progress. Since the person in the therapist role is only given a “satellite” existence, contingent on the patient, when patients cannot control a situation, their aggression is mobilized to strike out or hurt. Clinicians know not only how difficult such behavior is to treat but that it is a high-risk situation, since such individuals may blame others for their unhappiness and disappointments throughout their lives. It is the perfect setting for complaints of boundary violations to arise if some type of posttreatment situation occurs.

An International Perspective

In 1988, Coleman (33) reviewed how other countries deal with relationships between therapists and former patients and found many that have a different perspective from that seen in the United States. While an update of such a cross-cultural comparison is needed, a few countries included in the review are presented here for illustration. The General Medical Council in England has noted that a great variety of circumstances affect relationships with former patients, and the Council will not deal with situations when the doctor has attended the patient in the distant past. The Canadian Psychiatric Association has not addressed the issue of contact with former patients. In Israel there is no statutory or even case law on sexual relations between psychiatrists and patients. The Jerusalem Ministry of Health has taken the position that a psychiatrist engaging in sex with a patient would be disciplined but that it becomes “more problematic” with a former patient.

In Norway, a distinction is made depending on whether there has been an explicit termination. In Sweden, posttermination situations are viewed as an ethical rather than a legal matter. If necessary, the question would be considered in the context of a continued dependency situation. The policy of West Germany (and presumably continued under a united Germany) is that medical professional standards do not prohibit sex following termination. In Brazil there are no specific laws against sexual relationships between consenting adults, but if an issue arose ethically, it would be decided on the basis of whether the transference was resolved. The approaches of these countries emphasize an ethical rather than legal approach when posttermination issues arise. In addition, issues other than those concerning continuing sexual relations with a former patient are not viewed any differently than dealings between any two adults.

Conclusions

The diverse and unresolved clinical issues concerning the nature of boundaries after termination of treatment, including psychoanalysis, have direct implications for psychiatrists, the legal system, and society. They raise questions about the validity of drawing legal conclusions based on allegations of the continuance of the transference and its being “mishandled.” A new set of problems arises when these psychoanalytic and clinical concepts are injected into the legal arena for legal decision making. Many of the approaches used in attempts to regulate posttermination behavior have relied upon an unproven assumption of transference phenomena continuing for whatever type of therapeutic modality that has been used and for an unlimited duration thereafter. A further implication is that such phenomena mean valid legal consent can never be given for any type of future relationship. In many cases, diagnoses are also relatively ignored. Note that the period under discussion here pertains to situations that arise after the termination of treatment and not during active treatment.

In summary, the use of the concept of transference to resolve legal issues arising after termination of treatment produces more problems than it resolves. The term arose from psychoanalysis and has not been consistently defined. Its definition has varied not only over time but also with different theoretical perspectives. There continue to be disagreements among psychoanalysts and psychodynamic therapists about its extent and nature. The extremes are represented by those who do not accept the concept of transference to the opposite of those who see transference operating in nontherapeutic settings in the ordinary course of life. The need for preciseness in court testimony as a basis for establishing liability stands in contrast to the vagueness and amorphousness present in clinical situations. These are major difficulties in using such a framework to assess boundary violations when legal issues arise after termination of treatment.

While the sardonic phrase—“Once a patient, always a patient”—may serve as a rhetorical device in an advocacy context, it does not do justice to the complexities of the problem nor to the tremendous adverse personal and social consequences that arise from adopting such a stance. Apart from the question of the scientific validity of the concept of transference per se and the degree of its subsequent influence after termination for diverse therapeutic models, there is the realization that not all clinicians rely upon nor accept the validity of a transference model. Whether they should be told that they have to conform to a transference model in their treatments is a separate and quite controversial position.

However, for those who use a transference model, a persistent problem is the need to assess what role transference may or may not play. Even for those adhering to a transference model, it does not mean that people once in therapy cannot thereafter make informed and reasoned choices in their lives. When opinions are expressed in expert testimony about a boundary violation after a proper termination, the most subjective and vulnerable parts of psychiatry in the courtroom are revealed. We have then extended ourselves into an area that has the danger of subjectivity without the counterweight of meeting legal standards for scientific objectivity.

Presented in part at the 153rd annual meeting of the American Psychiatric Association, Chicago, May 13–18, 2000. Received July 7, 2000; revision received Nov. 7, 2000; accepted Nov. 27, 2000. From the Department of Sociology, University of Minnesota; and The Cambridge Hospital/Harvard Medical School, Cambridge, Mass. Address reprint requests to Dr. Malmquist, Department of Sociology, University of Minnesota, 909 Social Sciences Building, 267 19th Ave. South, Minneapolis, MN 55455.