Boundaries, Professionalism, and Malpractice in Psychiatry

Although the concept of boundaries in psychiatry has undergone considerable development over the past 40 years, the concept first appeared in medical literature under the heading of the patient-physician relationship. The origin of boundary maintenance is traced to the Hippocratic Oath. Along with establishing the early principles of beneficence and nonmaleficence, the oath described the professional obligation of maintaining confidentiality and provided guidance on the physician’s relationship with patients in a manner that has come to be known as professional boundaries: “In every house where I come I will enter only for the good of my patients, keeping myself far from all intentional ill doing and all seduction, and especially from the pleasures of love with women or with men.”

Numerous sources and contributors have helped shape the understanding of boundaries in the patient-physician relationship. Both professional codes of ethics and the psychiatric literature have contributed to this understanding. A review of developments in the American Medical Association (AMA) Code of Ethics and in the American Psychiatric Association (APA) annotations for psychiatrists is particularly informative in defining normative standards for managing boundaries. In addition, numerous authors have offered concepts that can be helpful in defining professional boundary obligations and managing boundary dilemmas.

The AMA’s Code of Ethics has evolved since the first version in 1847. The concept of boundary management, although not explicit in the 1957 revision, was implied under several principles and in the preamble to the code (6). As the Code of Ethics moved toward its current structure, the AMA made it clear that the practice of medicine is a profession that “must recognize responsibility to patients first and foremost” (7) The principles following the preamble provide detailed standards of conduct that stress the importance of virtues such as honesty, compassion and respect for patients, and adherence to the obligation to maintain patient confidences and make explicit that the responsibility to the patient is primary, requiring that physicians never place self-interest over care of the patient.

In the Code of Ethics’ current structure, the opinions fall under three headings: Patient–Physician Interactions, Treatment and Use of Technologies, and Professional Self-Regulation. Most information involving boundary issues falls under the first and third headings. In Opinion 1.1.1, the AMA makes it clear that the practice of medicine and the encounter between a patient and a physician is “fundamentally a moral activity that arises from the imperative to care for patients and to alleviate suffering” (8). In the special issues section under Patient-Physician Interactions, one can find several opinions pertaining to boundary issues, including guidance on self-treatment and treatment of family members, accepting gifts from patients, and ethical issues in telemedicine. In the Professional Relationship section are opinions on intimate and romantic relationships with patients and key third parties, as well as policy on sexual harassment (9). Opinion 9.1.1 is titled “Romantic or Sexual Relationships With Patients.” This opinion states that clinicians should terminate the patient-physician relationship before initiating a personal intimate relationship, and it also cautions that a relationship with a former patient “may be unduly influenced by the previous physician-patient relationship . . . if the physician uses or exploits the trust, knowledge, emotions, or influence derived from the previous professional relationship” (10).

Recognizing that the practice of psychiatry involves unique and complex dynamics in the patient-doctor relationship, the APA in 1973 began to publish the Principles of Medical Ethics: With Annotations Especially Applicable to Psychiatry to offer more nuanced and specific guidance for psychiatrists. In the 2013 edition, the annotations address boundaries. In Section 1.1, which considers the physician’s obligation to provide competent care with compassion and respect for human dignity and rights, the annotated section reads,

Section 2 addresses standards of professionalism. Annotation 2.1 provides additional commentary on boundaries:

In their early writings, Gutheil and Gabbard offered a theoretical risk management approach to boundary crossings and boundary violations that has guided a generation of psychiatrists (12). They responded to the growing number of sexual misconduct claims in civil litigation, ethics committee hearings, and complaints filed with medical licensing boards. With the view that minor boundary crossings sometimes lead to significant boundary violations and in the hope that managing minor crossings would reduce the risk of more severe violations, their model offered a typology of behaviors in the therapeutic relationship of which practitioners should be aware. Often described as a “slippery slope” management strategy, this model was supported by some evidence that those practitioners who eventually engaged in sexual violations with patients began with more minor intrusions into the patient’s space (13). Gutheil and Gabbard offered a list of areas in which psychiatrists should consider boundary dynamics, including role; time; place and space; money, gifts, and services; clothing; language; self-disclosure and related matters; and physical contact.

Gutheil and Gabbard offered a model that continues to be useful but predates the variety of psychiatric practices of today because the dominance of psychoanalytic and psychodynamic approaches has receded (12). Most psychiatrists are not psychoanalysts, and many use a variety of theoretical approaches and techniques, including the prescribing of psychotropic medication, the intervention that defines much of psychiatric practice. In such a psychiatric practice, terms such as transference, countertransference, and the categories of interaction offered by earlier authors may be less useful in managing boundaries. In addition, the treatment of patients from a variety of ethnic and cultural backgrounds complicates boundary considerations.

In response, authors have questioned the negative impact of overly rigid adherence to boundaries. After all, not all boundary crossings lead to boundary violations, and some boundary flexibility allows for individual response to individual patient needs. Martinez offered an alternative to the slippery slope model of boundary analysis, labeled the “graded-risk model” (14). In this model, in an attempt to move away from proscriptions against certain categorical boundary situations, all boundary dilemmas should be individually processed. Each boundary dilemma requires an analysis of six factors: the potential harm to the patient and the patient-physician relationship; the potential benefit to the patient and the patient-physician relationship; the presence or absence of coercive and exploitative elements in the boundary crossing; the professional’s motives and intentions; the professional’s aspiration to professional ideals of care; and the context, including the cultural considerations, of the boundary crossing. Through such an analysis, clinicians can consider whether a boundary crossing would be encouraged, permitted, or prohibited, depending on the totality of these factors.

For example, in deciding to have an intimate relationship with a patient, the potential for harm and the unlikely benefit to the patient and patient-physician relationship are obvious. The potential for coercion and misuse of personal information obtained in the patient-physician relationship would be high. The physician’s motivation and intention are unlikely to represent altruistic values and are certainly counter to placing the patient’s interests above those of the physician. It is doubtful that the physician’s aspiration to some ideal care could be the justification for such behavior. This analysis would lead to the conclusion that such a boundary crossing is clearly a boundary violation and should be prohibited. At the other end of the spectrum, simple boundary crossings such as offering a patient food or refreshment during treatment, or even interacting with a patient in a socially appropriate setting, might very well be permitted and even encouraged depending on context. Still, the inevitable social interaction may raise issues for the patient and the therapist and create gratifications and curiosity; this may be grist for the mill in treatment.

Prohibitions against physician-patient sexual relationships date back to the Hippocratic Oath (15). One of the first malpractice claims involving sexual boundary violations in the patient-psychiatrist relationship involved a psychiatrist who engaged in a sexual relationship with a patient under the claim that the sexual relationship was part of the patient’s therapy. In Roy v. Hartogs, Ms. Julie Roy sued Dr. Hartogs for emotional and mental injuries (16). Dr. Hartogs was found liable; the court awarded Ms. Roy compensatory and punitive damages. Because sexual misconduct is not considered part of professional services, many malpractice carriers, while providing coverage for the defense in a claim, may exclude coverage for losses.

Guidelines dating from the 1990s include prohibition of sexual relationships (or caution about them) with former patients (15). A more recent addition is prohibition of certain sexual relationships with relatives of patients. Gladieux v. Ohio State Medical Board (17) was a 1997 case in which it was alleged that Dr. Gladieux, a pediatrician, had sexual relationships with at least seven patients’ mothers, including the mother of a child patient who eventually died from her bone cancer (15). In this case, the medical board imposed a two-year suspension, calling the behavior reprehensible. The pediatrician appealed this punishment in court and lost at the trial and appellate court levels. In a subsequent report, the AMA discussed “key third parties,” which included spouses, parents, and various others. Next, Long v. Ostroff (18) focused on Dr. Ostroff, an internist having an affair with the wife of his patient, Mr. Long. During an evaluation for chest and back pain, Mr. Long had told Dr. Ostroff that he was worried about his marriage falling apart. Dr. Ostroff had not mentioned the affair, nor had he recommended that Mr. Long find another doctor (15). In this case, the court found that there was no therapist-patient relationship.

Psychiatrist-patient sexual relations are not only considered boundary issues and malpractice but are also illegal in many states. Therapists may face criminal charges for engaging in sexual acts with patients. For example, even after resigning her state therapist license, one therapist involved with her patient was charged with two counts of sexual exploitation by a therapist (a felony) and was jailed (19).

A range of state laws criminalize sexual acts between therapists and patients (20). For example, the Ohio Revised Code defines the offense of sexual imposition as follows:

No person shall have sexual contact with another, not the spouse of the offender . . . [including when] the offender is a mental health professional, the other person or one of the other persons is a mental health client or patient of the offender, and the offender induces the other person who is the client or patient to submit by falsely representing to the other person who is the client or patient that the sexual contact is necessary for mental health treatment purposes. (21)

Of note, this would not then appear to include all cases of psychiatrist-patient sexual acts, as long as the psychiatrist did not tell the patient that the sex was part of the treatment.

In other states, the fact that a patient consents or assents to the sexual acts with the physician is not exculpatory. New York State Penal Law Article 130 indicates that a person is unable to give consent to sexual acts if they are younger than age 17 years or mentally disabled or incapacitated, or if they are

a client or patient and the actor is a health care provider or mental health care provider charged with rape in the third degree . . . criminal sexual act in the third degree . . . aggravated sexual abuse in the fourth degree . . . or sexual abuse in the third degree . . . and the act of sexual conduct occurs during a treatment session, consultation, interview, or examination. (22)

Again, there appear to be loopholes—in New York, although multiple charges may be brought against the psychiatrist, these charges appear to relate only to sex in the consultation room.

In the United Kingdom, the General Medical Council (GMC) investigates complaints against physicians alleged to have had an inappropriate relationship with a patient (or to have made an inappropriate advance); 91% of the 90 physicians investigated in 2013–2014 were male (4). The GMC states, “Patients should be able to trust that their doctor will behave professionally towards them during consultations and not see them as a potential sexual partner” (4). Such relationships not only negatively affect the individual patient but also erode public trust in physicians.

In malpractice cases, forensic experts must establish whether the behavior was a boundary crossing or a boundary violation (23). Malpractice insurers usually do not cover sexual boundary violations, bringing into question whether the most common of online complaints would be covered, being that they may fit into the boundary violation domain. Claims involving boundary issues are rare in psychiatric malpractice. This may be because sexual activity with patients is now illegal as well as grounds for a malpractice claim.

The digital revolution has allowed for various advancements in the practice of medicine. However, conundrums exist regarding social media, as well as e-mail—both solicited and unsolicited. In fact, in boundary violations, both email and social media are frequently the vehicle (personal communication, Cash D, 2018).

The AMA has published recommendations regarding the caution physicians should use in social media. The AMA Code of Medical Ethics requires physicians to uphold professional standards, to respect the rights of others, and safeguard patient privacy. The AMA’s Council on Ethical and Judicial Affairs (CEJA) has made recommendations on guidelines for online professionalism based on AMA policy (24). CEJA noted that “the internet fosters disinhibition and feelings of anonymity and invisibility, which can promote either bad behavior or behavior that an individual would not engage in offline” (24). Behaviors and interactions should be the same online as they would be in the clinic. CEJA also noted that physicians who are aware of unprofessional online conduct by colleagues are obligated to address it. This is because physicians must protect patients and public welfare, as well as trust in their profession, and part of physicians’ obligation is to report impaired or incompetent or unprofessional colleagues (24). Therefore, CEJA’s recommendations include maintaining patient confidentiality and privacy in all settings; using privacy settings to safeguard personal details and monitoring one’s Internet presence; maintaining appropriate boundaries with patients on the Internet just as in person; understanding the responsibility to bring unprofessional behavior by a colleague to that colleague’s attention and to the appropriate authorities when needed; and recognizing that online actions may affect one’s reputation and career (24). Finally, for example, medical school talent shows, which may use dark humor and song to deal with stress, may not be appropriate for public consumption as an online video (24).

The American College of Physicians (ACP) and the Federation of State Medical Boards (FSMB) also authored a position paper on online medical professionalism (25). They noted that “standards for professional interactions should be consistent across all forms of communication between the patient and physician, whether in person or online,” reflecting the importance of maintaining public trust in physician-patient relationships (25).

When a physician uses social media and patients seek out the physician online, there is a risk of blurred boundaries between the personal and the professional. These blurred boundaries can affect how psychiatry is viewed by the community at large (25). The creation of distinct personal versus professional social media personas for online social media use, as well as carefully reviewing what personal information is in the public domain on social media, can help prevent a blurring of personal and professional boundaries (25). Venting or complaining about patients in online forums or on social media is unprofessional. One should also be aware that online postings are permanent. Moreover, the ACP-FSMB discourages interactions with patients via Facebook or other social media outlets, and strict privacy settings are recommended (25). When using social media or dating Web sites, psychiatrists may inadvertently come in contact with patients and engage in undue familiarity (personal communication, Cash D, 2018 ).

Using social media or search engines to research patients may be helpful in understanding the patient’s behaviors and online presence (25). However, doing so may decrease the patient’s trust in the physician-patient relationship (25). Moreover, the psychiatrist may unwittingly be exposing the patient to the physician’s own social media account, as a Facebook friend suggestion, for example. The rationale behind doing so should be considered, including whether there is a benefit to the physician-patient relationship versus whether the search is because of voyeuristic interest (25).

In forensic psychiatric evaluations, there is no traditional doctor-patient relationship, and data about the evaluee may be gathered from the Internet (26). This online research can sometimes provide valuable insight into the evaluee’s mental status and belief systems or symptoms; however, what the evaluee posts online may constitute impression management rather than the objective truth. It is important, though, even in forensic evaluations, that psychiatrists be respectful of the evaluee in their approach to the gathering of information.

Box 1 lists risky or problematic online behaviors. A 2012 survey of directors of American medical and osteopathic boards asked about violations of online professionalism and the boards’ actions (27). Respondents represented those responsible for medical licensure and disciplinary actions for 88% of physicians in the United States. The most commonly reported violation was inappropriate Internet communication with patients (i.e., sexual communications), followed by Internet prescribing without a relationship, misrepresenting credentials online, violations of patient confidentiality online, not revealing one’s conflict of interest online, derogatory remarks about patients online, photos taken while intoxicated, and discrimination (27). Disciplinary proceedings were the most common outcome after online professionalism violations occurred, and most were followed by informal warnings. Greyson and colleagues suggested that “these violations also may be important online manifestations of serious and common violations offline, including substance abuse, sexual misconduct, and abuse of prescription privileges” (27). Other online ethical violations include a failure to report other clinicians’ violation of patient privacy on social media, bullying colleagues, and negative communications about employers (28).

E-mail or text communication with patients allows the physician to be more accessible and to answer simple questions without waiting for an appointment. However, risks include the replacement of face-to-face appointments, misunderstandings of communications (resulting from the lack of tone of voice or body language and nonverbal cues), and the potential for problems with confidentiality (5, 25). The ACP-FSMB position paper recommended that physicians establish guidelines about what issues may be communicated digitally and only use e-mails or texts with patients who are attending follow-up appointments (25). Patient consent should be obtained, and communications should be copied into the patient’s medical record (25). Another risk is that a psychiatrist may unintentionally misdirect an e-mail, leading to a breach of confidentiality (personal communication, Cash D, 2018). Box 2 lists some questions for ethical self-reflection.

Rather than initially occurring over e-mail, the boundary violation has often already occurred, and the patient and psychiatrist are merely communicating about the violation electronically. For example, if a patient and psychiatrist are romantically involved, their e-mail communications may demonstrate this. These e-mails may come to the attention of the patient’s partner, for example, and expose the inappropriate relationship. In e-mail interactions, self-disclosure by the therapist, inappropriate seductive language, breaches of confidentiality, countertransference issues (e.g., about the time spent on this often unbilled work), and lack of boundary with the end of session (and into the e-mail) are all potential problems (5). Of note, “some malpractice cases based on boundary issues turn on emails that are seen as markedly seductive, erotic, or revealing undue familiarity; inappropriate pictures also may be included” (5).

Unsolicited e-mails pose a greater ethical conundrum, and the clinician may incur liability if care is not taken in deciding whether to respond (29). It is currently rather simple for a physician’s e-mail address to be found through an Internet search. If a nonpatient emails a physician about a medical or psychiatric problem, what is the most appropriate response? There may seem to be no harm in replying. However, physicians may respond about specific medical issues and, in so doing, create a doctor-patient relationship. However, not replying to a person who is writing to a physician about medical concerns may seem callous.

Unsolicited e-mails from unknown individuals are not necessarily reliable or objective sources of information (29), such as, for example, if a plaintiff’s attorney e-mails a defense expert witness about a clinical issue, pretending to be a potential patient. There is also a risk of creating a duty of care if one replies, despite the lack of a patient in front of the physician to examine. Also, licensing and malpractice insurance may be at risk if the clinician replies to someone e-mailing from a state in which the clinician is not licensed nor covered by malpractice insurance (29). Potential solutions for physicians receiving the unsolicited e-mail include not opening unsolicited emails from unknown senders or sending a standardized autoreply. Clinicians should use caution when sending a personal reply to an unsolicited e-mail (29). Recupero has recommended that an appropriate reply may be a generic statement reminding the person of the lack of a doctor-patient relationship and encouraging the person to seek help (30). Replying in a specific way and giving specific advice to the person emailing about a situation might appear to establish such a doctor-patient relationship, even if one is merely trying to be helpful. Also, the e-mailer may misinterpret the return e-mail as a source of medical information and not seek actual medical help in person (29). If a clinician does reply, it is prudent to either politely state that the clinician is unable to help or to give a general reply clarifying the lack of doctor-patient relationship and lack of medical advice, as well as encouraging help seeking locally (29).

The concept of boundary continues to be useful and instructive in processing decisions involving the relationship between the patient and psychiatrist. Although the concept was initially created to manage the risk for sexual misconduct and harm to patients, the concept has utility in understanding many of the interpersonal domains that occur in the patient-psychiatrist relationship. From managing social media with patients, to accepting gifts or attending social functions, to making decisions regarding prescribing for family members and other dual relationship considerations, the concept of boundary crossing and boundary violation can be helpful to consider before making decisions and assist in retrospective understanding of boundary challenges.

The authors appreciate the input of David Cash, J.D., of Professional Risk Management Services, and of colleagues in the Group for the Advancement of Psychiatry’s Committee on Law and Psychiatry, in the preparation of this article.