Background

The idea that sexual relationships between patients and doctors should be proscribed was first stated in the Hippocratic writings (c. 4 bc). The Hippocratic oath quite clearly states that ‘Into whatever house I enter, I will go into them for the benefit of the sick and will abstain from mischief and corruption and from the seduction of females or males, of freemen or slaves’ (Reference JonesJones 1923). The implication here is that doctors are in a position of influence that could be used for non-therapeutic purposes.

The ethical implications of sexual boundary violations in therapeutic relationships and the harm they cause are well documented (Reference Bouhoutsos, Holroyd and LermanBouhoutsos 1983; Reference Pope and BouhoutsosPope 1986). Textbooks have been published (e.g. Reference Gabbard, Pope and GabbardGabbard 1989; Reference JehuJehu 1994) and scholarly articles in professional journals appear with regular frequency. All point to the harm done to patients and the wrong done to professional integrity when doctors have sexual relationships with their current patients.

However, there has been less discussion of the ethical implications of relationships between physicians and former patients. Although ethicists such as Reference Spiegel, Colella and LuptonSpiegel (2005) reluctantly agree that there is ‘potential’ for harm, in fact there is no empirical or documented evidence of any grade that sexual relationships with former patients are harmful.

However, absence of evidence is not evidence of absence. For current patients, there is good empirical evidence that a sexual relationship with a patient is almost always harmful. There is some evidence, however, to say that clinicians themselves view sexual relationships with former patients as less ethically problematic than sex with current patients. Although the percentage of clinicians who view sex with current patients as wrong hovers around 95%, the range is wider regarding former patients. It may be as little as 50% for psychologists practising as psychotherapists (Reference Pope, Tabachnick and Keith-SpiegelPope 1987) to 70% for psychiatrists (Reference Herman, Gartrell and OlarteHerman 1987).

Ethics and sexual boundary violations

The main arguments against a physician–patient sexual relationship rest on three principles of ethics: trust that the physician has the patient's best interests at heart; respect for the power imbalance between doctors and their patients; and respect for patient autonomy by obtaining consent (Reference ZelasZelas 1997). The deontological approach to bioethics establishes whether a proposed course of action is consistent with the accepted principles. Where there is an ongoing therapeutic relationship, it is clear that sexual boundary violations constitute a breach of trust (the physician is considering their own interests), an exploitation of the patient's vulnerability and dependence, and a failure to obtain consent for a change of role.

How might this argument apply once treatment is terminated? If consent has been given to terminate treatment, then the possibility of other types of relationship arises, including friendship, business and sexual relationships. The power discrepancy is no longer the issue, nor is there an issue of trust. There are documented situations when treatment is terminated in order to facilitate a sexual relationship; but in these cases, it is often unclear whether there has been ‘true’ consent on the patient's part and the breach of trust issue remains. It is also likely that the clinician has not considered the potential harm of prematurely terminating treatment.

But what about relationships that begin after treatment has been appropriately terminated with proper consent and closure, and when a ‘reasonable’ (however defined) period of time has elapsed? The power imbalance arguments do not hold good anymore – why should a former patient always be personally less powerful in any future relationships, let alone with a professional? Some ex-patients may be more powerful than the professional, for example in status or personality. It is generally assumed that ‘patienthood’ is a role occupied only when one is actually ill or receiving treatment; it would be odd to say that someone who had their leg fracture pinned 2 years ago is still an orthopaedic ‘patient’. If psychiatric and psychotherapy patients are always ‘patients’ in this sense, it suggests that they are being treated differently from medical patients in a way which sits oddly with anti-discriminatory practice.

The trend in modern healthcare is to treat patients as partners, equals and consumers. Consumers, many healthcare providers of today will argue, are more powerful than the most powerful of manufacturers. Respect for autonomy, always a key principle in bioethics (Reference GillonGillon 2003), is now understood as the patient's right to be the final arbiter of their medical treatment including a right to end and refuse treatment. As deference to the medical profession decreases, so the power imbalance between doctors and patients diminishes. Psychiatric and psychotherapeutic treatment in particular aims to enhance patient autonomy and make it possible for the patient to see the therapist as a useful ally instead of an authority figure. Reference Appelbaum and JorgensonAppelbaum & Jorgenson (1991) argue that any power held by the therapist over the patient substantially decreases over time once the treatment is terminated. A decision by an ex-patient to pursue a relationship with their former therapist may be unwise: but it is still their decision, and competent adults are allowed to make unwise decisions. If this respect for autonomy is not to be allowed for former psychiatric patients, there needs to be a cogent argument given for this position.

It might be argued that all intimate relationships have aspects of power discrepancies. More radical feminist discourses have argued that all heterosexual relationships involve a disparity of power between men and women, in which women are subordinate. Such a feminist approach has informed the literature on sexual boundary violations by doctors because of the evidence that the majority of such relationships are between male doctors and their female patients. It is also undeniably true that all patients are dependent on their doctors to some degree and that for many patients ‘illness’ is synonymous with ‘dependence’. The argument then would be that this power discrepancy never ceases, even if the clinical need is over, and there may be some reason for thinking that this is so, as I will discuss in more detail below. But if the power differential has been reduced through treatment, then a subsequent relationship might be acceptable. There is also the question of who decides the nature and size of the power discrepancy: who does the maths and is there a sliding scale of power?

It might be said that the mere fact that there are power discrepancies between intimate partners does not make them ipso facto ethically unjustifiable. In all relationships there will be important differences between partners: it is, in my view, how those differences are approached by each person in the relationship that is crucial. In relationships that are working well, partners may take it in turns to be more dependent on the other, especially at times of distress or crisis. The question is whether either party is attempting to maintain or exploit the former power differential (‘Well, honey. You know that you were pretty borderline a few years ago when you were in treatment with me, so you really aren't fit to make this decision about the children’ or ‘I could make your life quite difficult by telling your colleagues that I used to be your patient’).

Consent

That leaves us, for the moment, with the issue of consent. It is usually argued that there can never be ‘true’ consent to a sexual relationship in any power-disparate relationship such as a therapeutic one. This feels counterintuitive in the light of the preceding discussion, especially when we consider that we can accept as ‘truly’ consenting those decisions that place patients in danger such as the refusal of life-saving treatments. It is accepted in both ethics and the law that competent adults can make unwise decisions which may threaten their health – outcomes that are no less dangerous than an unsuccessful relationship. What makes these decisions valid in terms of consent is the informed choice made by the patient in discussion with their doctor.

There is no particular reason to assume that an ex-psychiatric or psychotherapy patient is less competent to take decisions than other patients. It might be said that a decision to have an intimate relationship with someone is highly emotionally charged and that this will be ‘too much’ for an ex-patient to bear. In counterargument, it might be said that the decision to embark on any emotional relationship is always highly charged, and in some sense no one is perfectly competent to make such decisions. It is pertinent at this point to note that most legislation only requires individuals who wish to marry to be of legal age, not that they be emotionally competent.

The chief concern about former patients is that they are more vulnerable than other potentially ‘dateable’ people and that exploitation of the vulnerable is wrong. The problem here is how to define vulnerability in ways that do not result in overly paternalistic practice – an image modern medicine is striving to shed. It is also demeaning, as much vulnerability goes hand in hand with experiences of trauma and suffering, and like these experiences can be mastered or negotiated.

There have been attempts to define vulnerability for the purposes of deciding whether there has been exploitation. For example, Minnesota law lists vulnerability factors such as ‘when the former patient is emotionally dependent’ on the therapist (Minnesota Office of the Revisor of Statutes 2008: Section 609.341). Deliberate exploitation of a patient's (or ex-patient's) vulnerability for nefarious purposes will almost always be seen as wrong. Even the law courts have ruled that, after treatment has ended, engaging with the patient in the very acts for which they sought help (such as drug use or seductive behaviour) can be clearly negligent ( Noto v. St. Vincent's Hospital and Medical Center of New York (1988)). The barn door cases are not going to be the problem; the thousand shades of grey will be.

In the wake of much publicised cases of egregious sexual boundary violations by physicians (not limited to psychiatrists), the government urged the UK's medical regulatory bodies to tighten up guidance on relationships with former patients. The GMC gives the following renewed guidance (General Medical Council 2006a) and incorporates it into Good Medical Practice (General Medical Council 2006b), a de facto code of practice against which the doctor is judged in all cases of conduct or capability:

Similarly, the current position (taken by the GMC and most medical boards in the USA) is that doctors should not treat immediate family or close friends. The rationale for this seems to be that the intimacy, closeness and non-clinical nature of the relationship may impair the doctor's judgement in making important decisions. If this is so for family and friends, then perhaps there is a case for saying that doctors may not be able to make good judgements about former patients. Equally, it might be argued that if an ex-patient is not explicitly told about the potential pitfalls of embarking on a relationship with a previous therapist, or not told about the prevailing professional view that it is wrong to engage in such a relationship, the validity of any consent is reduced.

Transference and relationships with former patients

In addition to the three ethical principles described earlier as a rationale for proscribing relationships with ex-patients (Box 1), psychotherapists (and increasingly the law courts) have added a fourth argument: namely, the relevance of transference in professional relationships. Transference is not a principle but a phenomenon. Classic analytic theory and recent research based on attachment theory make it clear that patterns of relationships with childhood caregivers are frequently replicated in later relationships with adults, especially where there is an imbalance of power and one party is dependent on the other. The majority of psychotherapists would argue that sexual boundary violations take place most commonly in clinical relationships where transference is most likely to be activated; for example, therapeutic relationships involving long-term dependence with opportunities for mutual intimacy, such as general practice and psychiatry. There is no good reason for thinking that the transferential aspect of physician–patient relationships ceases just because the clinical relationship has ended; therefore, ex-patients will always relate to their therapist as a transferential figure and any relationship will not really be between equals. Doctors, even ex-doctors, will always be like mothers or fathers, or some other authority figure, to the patient.

Although classic analysts will argue that transference is omnipresent and highly influential in therapeutic relationships, other psychiatrists do not accept its power, even if they are prepared to accept its existence. Outside psychiatry and psychotherapy, transference is hardly recognised. Therapeutic relationships between surgeons and patients, for example, are assumed to not involve transference; only psychotherapeutic relationships claim that it is an important aspect of therapy which needs to be worked on (and possibly worked through). Transference is arguably not an issue where the patient and the doctor have little personal contact and where therapy is of brief duration. The relationship between a patient who needs emergency surgery and the surgeon who operates on them without getting to know them either before or after the operation, cannot be said to involve transference. One assumes that it would therefore be OK for the surgeon to date the patient once they have recovered. However, it is plainly wrong to assume that transference is not present in this case. The surgeon may feature as an important person in the patient's mind, even if the patient has not met the surgeon when conscious. The patient could very well have ‘ready-formed’ (and powerful) transferential fantasies about the surgeon because the latter's involvement with the patient involves intimate access to their body (including cutting into it), power of life or death over them and so on.

So should sexual relationships between former patients and their doctors be proscribed only when the transference is more clearly an issue, i.e. in psychoanalysis or, slightly more broadly, psychiatry? The position taken by Reference GabbardGabbard (1991) is that ‘all therapeutic situations are coloured to some degree by transference’, i.e. attitudes and feelings that reflect past relationships with carers. Since many, if not most, patients in therapy will have had abusive or exploitative relationships with their parents that are likely to be enacted with therapists, any intimate relationships, especially sexual ones, are proscribed. There is some evidence that suggests that patients who have a history of incest are more likely to be involved in boundary-violating relationships with doctors (the sitting-duck theory; Reference Kluft and KluftKluft 1990). Reference GabbardGabbard's (1991) view extends to include former patients, since he argues that therapy patients (especially those who do have a history of incest) will never be free of unconsciously child-like attitudes towards their doctors.

There are two problems with this argument. First, one might argue that successful psychotherapeutic relationships aim to work through the transference, and one good outcome of therapy is that the patient no longer relates to the therapist (or indeed anyone) as a parental figure. If this is so, then it should in theory be possible for patients who have successfully completed therapy to relate to their ex-therapists as potential partners. The issue then would be whether it is really possible to show that transference no longer exists.

The other problem is that it is clear that all intimate relationships are influenced by past experiences with parental figures. There is good evidence that early patterns of attachment are replayed in romantic relationships with adult partners (Reference SimpsonSimpson 1990), and often highly insecure people, who have had difficult and failed relationships with parents, choose partners to whom they will relate in similar ways. These relationships in ordinary life are not proscribed: so why should they be for ex-patients and ex-therapists? If, as Gabbard suggests, the nature of the unconscious mind is timeless and intrapsychic attachments to parental or authority figures are persistent and highly influential in the real world, then all patients relate to all doctors as parental authority figures, which may explain the traditional deference to doctors and the idealisation of the medical profession.

The answer to at least the second argument does rest in transference and the re-enactment of early attachment patterns. The key issue here is that transference works both ways: the therapist also forms a transferential relationship with the patient. Insecure patients will form insecure relationships with carers (Reference Dozier, Tyrrell, Simpson and RholesDozier 1998); but insecure therapists, with unresolved childhood distress, may relate to their patients in highly insecure ways. Arguably, the wish to have a relationship with someone who used to be a patient reflects an insecure relationship between therapist and patient; what the attachment literature would call an ‘enmeshed’ or ‘preoccupied’ relating style.

However, the argument remains: insecure relationships in other domains of life are not proscribed. If two people who have previously been in a therapeutically intimate relationship can come together and mutually agree that they want to start a different relationship, then the possibility of unconscious transferential attitudes on both sides does not make the relationship ethically unjustifiable. It may make it more likely to be unsuccessful, but with a divorce rate in the Western world approaching 50%, this does not seem to make physician–patient relationships stand out. The practical question raised by Reference Appelbaum and JorgensonAppelbaum & Jorgenson (1991) is whether it is possible to set some time limit for the change in nature of the relationship. In other situations where there are concerns about enmeshed or unresolved attachments, such as divorce or bereavement, an interval of 12 months is recommended before a new relationship is attempted.

There is an interesting difference in relation to the time question in the opinions held by the American Psychiatric Association (1989) and the American Psychological Association (1992). The former holds that it is almost always unethical for a therapist to engage in a sexual relationship with a former patient (irrespective of time after termination); the latter, that at least two years should elapse before a relationship can take place, and only if the therapist can demonstrate that there has been no exploitation in light of all the relevant factors (Box 2).

Psychological processes and consequences

It might be argued that even if the situation is ethically justifiable and there is no exploitation in physician–patient relationships, there is still a concern that it will affect how doctors, especially psychiatrists, are seen. Reference Gabbard, Pope and GabbardGabbard & Pope (1989) argue that if the taboo of sex with patients is not extended to sex with former patients, then the process of therapy would break down. Patients might see therapy sessions as a dating service (either consciously or unconsciously) and predatory therapists could justify terminating therapy in order to exploit their next victim. For some patients, there could even be a disincentive to get better.

It might be argued that when an ex-patient consents to engage in a relationship with their former psychiatrist or therapist, they are basing this decision on trust of the medical institution as a whole. It is not the individual person who is the object of trust but ‘doctors’ as a genus. Such relationships generally impair trust in the profession, even if there are individual cases that work well, and relationships with ex-patients diminish trust in the profession as a whole. This is an argument from consequence not principle: society has an investment in maintaining trust in the profession so that patients in future will not be deterred from trusting doctors with their intimate secrets.

We know (and have come to accept as a matter of law) that confidentiality does not die with the patient. Similarly, Reference Gabbard, Pope and GabbardGabbard & Pope (1989) argue that transference, at least in its most benign paternalistic form of continuing professional responsibilities, persists even after treatment is completed. Just as there is a compelling state interest in preserving secrets shared between a patient and a therapist to enable future patients to trust their doctors, so there is a public interest in keeping the personal and professional worlds of doctors highly separate, which means no intimate relationships with patients, past or present.

In my view, the transference issue is crucial. Even if transference does influence all emotional relationships between adults, psychotherapeutic and psychiatric relationships are different because it is past attachments and transferences that are the clinical issue, or symptom, if you will. The danger that the therapist will make any presenting problem worse outweighs, in my opinion, the danger of unfair discrimination. The analogy again is with incest: even when a child grows up and becomes an adult, there can be no lawful consent for parent–offspring sex.

Sir Graeme Catto, President of the UK's GMC, has stated that ‘a relationship between a doctor and a patient is never really equal’ (General Medical Council 2006a). I agree with him, much as I agree with Reference SheatherSheather (2006), of the BMA's Ethics Committee, that this is dangerous territory, ‘complex [in] ways to the ebb and flow of power’. Political correctness aside, once the recent trend of treating the patient as a consumer passes, we will be left with the ancient situation of distressed and diseased individuals coming for help to the very people society says are capable of easing their suffering. The currency of the physician–patient relationship is trust, and this trust comes at a price. Some would call this a priori reason. Some would call it inviolable principle. One might give it a utilitarian spin of maintaining the reputation of the profession. Whatever one calls it, the comment of one of America's most famous psychoanalysts, Frieda Fromm-Reichman, is surely relevant and sobering. She is reported to have said (perhaps half-jokingly), ‘Don't have sex with your patients. You will only disappoint them’. With this, you might well include former patients.

MCQs

1. 1 The GMC's recent guidance on relationships with patients:

   1. a makes a distinction between former and current patients

   2. b advises a total ban on relationships

   3. c makes it a criminal offence

   4. d can be extrapolated to other professionals

   5. e has been welcomed by all.

2. 2 A physician–patient relationship:

   1. a is sometimes acceptable

   2. b is universally condemned

   3. c is thought to be equal in power

   4. d puts too much unnecessary stress on transference

   5. e is viewed differently in different jurisdictions.

3. 3 Sexual relationships with former patients:

   1. a are being increasingly seen as acceptable

   2. b treat the patient as an equal

   3. c may be acceptable in some circumstances

   4. d are not acceptable even after a time gap

   5. e are permitted by the American Psychiatric Association and the American Psychological Association.

4. 4 Ethics of professional boundary violations:

   1. a are necessarily consequence based

   2. b are mainly deontological (principle based)

   3. c mean that patients are not treated as individuals

   4. d are at odds with the feminist literature

   5. e de-humanise professional relationships.

5. 5 The government's action plan on professional boundary violation is:

   1. a an unwelcome intrusion on professional practice

   2. b fairly recent and exaggerated

   3. c not driven by recent cases of serious sexual exploitation

   4. d spearheaded by the GMC

   5. e summarised in the Department of Health's funded ‘Clear Boundaries’ project.

MCQ answers

1		2		3		4		5
a	t	a	f	a	f	a	t	a	f
b	f	b	f	b	f	b	f	b	f
c	f	c	f	c	t	c	f	c	f
d	f	d	f	d	f	d	f	d	f
e	f	e	t	e	f	e	f	e	t