E-mail communication between doctor and patient is an expanding feature of modern medical practice.1–7 The ability to pose thought-out questions to one's treaters, to avoid phone tag, or to report on side effects or other urgent matters certainly aids the practice of medicine. Even online psychotherapy has found a niche.7 Despite these clear advantages, which will not be considered here, the use of e-mails in psychiatric communication may also pose potential problems with therapeutic boundaries. This last point is the subject of this article.
In an earlier artcile,8 we described how the interval at the end of a therapy session “between the chair and the door,” when both parties rise at the close of the session, revealed two aspects. First, as an undefined period in time and space, which appeared in an illusory fashion to exist outside the session itself, the interval represented a common occasion for boundary problems arising from both therapist and patient. Such problems may constitute the beginning of the “slippery slope,” a term used to indicate that most boundary problems begin with small deviations and become progressively larger as one slides down the “slope.” Second, for that very reason, behavior during the interval could serve as an early warning sign of boundary issues requiring exploration in the subsequent session.8
The core dynamic here is the feeling that the walk between chair and door is extra-therapeutic, and material that arises there is somehow “off the record.” Failure to explore in the next session whatever of significance occurred between chair and door constitutes a technical error that could lead to further difficulties.
Example 1. While moving toward the door, a female patient, who had been preoccupied during the session with concerns about her husband's fidelity, impulsively asked the male therapist: “You look like you've lost weight; have you been working out?” The therapist, caught off guard, made some noncommittal response but failed to discuss it the next session. The patient took this silence as permission to ask ever more personal questions.
In the modern era, communications using the internet may evoke similar “off the record” perceptions and fantasies, especially in relation to e-mails. In the managed care era, however, characterized by ever briefer encounters with patients and by greater focus on medication issues, the boundary dimension may be unexplored. Indeed, brief encounters, often months apart, leave little time or continuity for such exploration. Instead, e-mails between physician and patient are common, containing questions about dosage, reports of side effects, physician advisories, and the like.
In contrast to office settings, where confidentiality may be effected easily, the internet in general is a far less secure context. Wireless use, misdirected e-mails, e-mails sent accidentally to a list instead of a person, and similar problems represent potential lapses of confidentiality.
Besides representing a nonconfidential form of communication (itself a boundary issue), e-mails are recognized as possessing some characteristics different from those of either face to face conversation or telephone calls. In addition, they have more immediacy and faster effects than exchanged letters. Face to face conversation allows “broadband” communication by face, voice, and body language, as well as the therapist's empathic responses. The telephone permits tone of voice and various vocal inflections to add nuance to the communication, in the absence of sight of the person. Letters are the most uninflected form of communication, but, of course, writing style, stationery characteristics, and word choice have always conveyed a vast amount of information about the writer. The latter point might be made for e-mails as well, although the delay intrinsic to postal delivery allows whatever heat exists in the relationship to cool down somewhat between exchanges of letters.
In contrast, e-mails have the stripped-down cool neutrality of letters combined with the hotter immediacy of an ongoing dialogue, at times in real time with the use of instant messaging. Even more than letters, e-mails are faceless and de-personalized at the time of writing or sending; the physical and vocal absence of the sender, providing a relative poverty of sensory data, allows fantasies to flourish relatively unchecked. This point was captured in a recent cartoon showing a dog seated before a computer saying to another dog, “On the internet, no one knows you are a dog.”
We suggest that the relative anonymity of e-mail fosters the development of more intense transference. Little is available in the literature on this point, and future research would be valuable. In an analogous connection, Martin Orne (personal communication, 1974) described an experiment in which therapists and patients were in separate rooms, invisible to each other, communicating via microphones; the therapists were allowed only to grunt without words as a response (“Um-hmm,” for example). Orne reported that in this format, instantaneous transferences were generated.
In an online publication, Stofle7 offers several practical considerations for e-mail therapy. Although e-mail therapy is not our topic, the description of the qualities of e-mail appears to have direct relevance here:
- Lack of nonverbal cues.
- Starkness and potential coldness of text-based communications.
- Potential for misunderstanding.
- Lack of control when the client is not in the room with you.
- Concern about the ability to establish a therapeutic relationship without seeing the client.
- Concern about the client's identity.
- The therapist's ability to communicate and get around online.
E-mails may serve as transitional objects9 for patients dependent on those items. Like other objects — letters, bills, stationery, prescriptions — given to the patient or stolen by the patient from the therapist's office, an e-mail from the therapist may represent a symbolic part of the therapist that can kept as a tangible reminder of the relationship. An analogy might be drawn to those dependent patients who call the therapist's answering machine when the therapist is known to be away for the specific purpose of hearing the therapist's voice as a form of support or contact.
In the office setting, the patient is also an observer in the dyad, a fact that may aid in constraining inappropriate behavior by the therapist. In sending e-mails, patient and therapist usually are isolated and thus free from the inhibiting scrutiny of observers, including each other. In a similar vein, just as the space “between the chair and the door” may foster greater informality because the formal session is ostensibly “over,” an informal tone may enter into e-mail communications, different from the more careful formality of the session. Patients eager to “get more” from the therapist may attempt to shift the majority of the therapeutic interchange into the e-mail realm for this reason. Finally, e-mails may serve as a form of durable documentation of either good therapy or bad therapy, or a mixture of the two.
E-mails and Boundaries
Several of the classic boundary transgressions10 may appear with e-mails. These may include inappropriate self-disclosure, inappropriate language (such as seductive content), breaches of confidentiality of the recipient patient or other patients, and billing issues (if e-mail contact is not billed, is it a social letter?). Concerns about billing or not billing may also introduce countertransference dynamics into the therapy; if the therapist decides not to bill for e-mails, does spending large amounts of time dealing with e-mails promote countertransference anger? Again, discussion of this point is lacking in the literature.
E-mails also easily can constitute the boundary crossing of inappropriately extending the session in a manner similar to the patient who is allowed by the therapist to stay in the session beyond the agreed-on end point. Again, the wish to prolong the contact may be intensified because of managed care constraints on adequate time with the clinician. On the other hand, patients phobic about some aspect of face to face contact or traveling to the office may be able to use e-mails to bypass these anxieties constructively but temporarily. Other problem boundary areas such as physical contact may not apply.
National medical organizations such as the American Psychiatric Association (APA) and the American Medical Association (AMA) have attempted to address the psychiatric and general medical aspects of the internet in their policies. The main focus of those policies appears to be cautions about inadvertent creation of a doctor–patient relationship over the internet and cautions about providing medical advice or treatment recommendations online.
The APA website,11 in the section on frequently asked questions, mentions in only one line the boundary-related caution that overuse of e-mails may be a therapeutic issue. The point is not further explained. The site also states, “The prudent physician will want to make a record of e-mail communications” and “[Compared with a progress note,] e-mail is more like a transcript and therefore is more revealing.”
The AMA guidelines5 also illustrate the professional concerns noted above but do not specifically address boundary issues except in one reference: “When using e-mail communication, physicians hold to the same ethical responsibilities to their patients as they do during other encounters.”
The Risk Management Perspective
Apparently based on little actual insurance loss experience, Professional Risk Management Services1 has proffered suggestions for dealing with e-mails in the clinical setting. Although their focus is more on avoiding liability risk and ensuring caution when establishing a doctor–patient relationship, some of their points may be relevant to the questions of boundaries. The 2004 online edition of the publication Risk Management Resources for Psychiatric Practice12 notes:
“At this point in time, there is little specific, detailed risk management advice to give regarding the use of e-mail with patients ... and there is no currently established e-mail liability law to study. ... [P]sychiatrists must always insure that legal and ethical standards regarding undue familiarity are met.”
The risk management advice also includes cautions about the inability to recognize someone in a chat room who may be a present or former patient, and they stress that the potential for sexual misunderstandings is increased in online communication. The document specifically urges that all e-mail messages be printed out and made a part of the record. We agree.
Example 2. A male therapist was treating a personality-disordered female patient. At one point the patient wished to show the therapist something she had written, and asked for his e-mail address. Although the therapist had never given out his e-mail address, with this patient he felt moved to make an exception. The patient sent him some writings that were not actively explored in the session. The patient continued to send more and more autobiographical and journal material without its being explored. The relationship deepened “outside the office,” with an eventual sexual relationship resulting. In ensuing litigation, the therapist lost his license.
Note here that the valuable boundary issue “red flag” — “I don't usually do this with my patients, but in this case I ...” — was ignored or not appreciated.9 As earlier noted,8 the same perceived suspension of the “usual rules” that occurs between the chair and the door may be used by the therapist to track and monitor the boundary state of the relationship. Similarly, the examination of e-mails as communications also may alert the therapist to the state of boundaries.
Example 3. Therapist “Jane Jones” noted that a patient who called her “Dr. Jones” in the office began e-mails to her with the greeting “Dear Jane.” When this was explored in session, the patient revealed wishes for closeness that had not been addressed previously.
Example 4. A psychotherapist in a clinic setting treated a female patient weekly for depression from relationship loss and borderline personality traits. The therapist maintained sound treatment boundaries initially. The therapist and patient began communicating by e-mail regarding questions and answers about her treatment. The e-mails increased in frequency, gradually becoming more social and eventually more intimate, with sexual content. A sexual relationship began outside the clinic setting. The therapist and patient agreed to terminate psychotherapy, but they continued to see each other socially. The therapist's wife discovered the e-mails and confronted her husband, threatening him with divorce and financial ruin. He immediately terminated the relationship with the patient. The patient, emotionally devastated, became suicidal and required hospitalization.
Six months later, the former patient filed a lawsuit against therapist and clinic claiming sexual exploitation and abandonment. Numerous erotic e-mails sent by the therapist were placed into evidence. No evidence of boundary violations was found in the patient's official record during the treatment, but neither were any e-mails found included as part of the patient's record. After the lawsuit was filed, the clinic established a policy requiring that all e-mails sent and received be made part of the patient's record. The case settled for six figures before trial.
Example 5. An inpatient with borderline personality disorder rapidly developed highly erotic transferences to almost all her caretakers, and e-mails flew back and forth. They were not made part of the chart. In a later action, the e-mails, both in ambiguous content and explicit frequency, were used to bolster the patient's (probably false) claim of sexual misconduct against her treaters (Strasburger L, personal communication, 2005).
Example 6. “After careful consideration and deliberation the Committee on Ethics ... voted unanimously to reprimand Dr. _____ for his conduct and particularly with regard to his e-mail and instant messenger communications with a patient. ...” (written communication, June 2005).
Example 7: A consultee reported that he maintained strict boundaries in the session but in subsequent e-mails he disclosed a great deal of highly personal information. He reported that it did not feel at the time like a boundary issue, since he was not in the room with the patient. The problem was caught in time and explored.
Example 8. A patient sent a huge number of highly erotic e-mails to the therapist. Although the therapist did not respond, questions were later raised at an administrative proceeding as to why the number had been allowed to get so high.
The risk-management issue here appears to lie between the lack of control that therapists have over patient's activity beyond limit setting and the dilemma of when to consider a patient as untreatable and to stop treatment. Consultation would doubtless be helpful, as well as printing out all the e-mails and including them in the chart. Progress notes should document the discussions about the content.
The course of psychotherapy is occasionally marked by acting out and acting in.13 The point has been made9 that one fact that may distinguish a boundary crossing from a boundary violation14–17 is whether it is discussable in the therapy and whether it is acknowledged (with the reasoning involved) in the record. Failure to “bring it in” when the patient attempts an extra-therapeutic communication with the therapist (be it phone call, letter, e-mail, encounter outside the office, or other form) avoids this discussion and any possible therapeutic mileage that can be gleaned from it. The therapist's silence may be seen as both an attempt at concealment in later litigation and as a permission to continue to act outside the dyad (or, for that matter, the therapy group).
Some malpractice cases based on boundary issues turn on e-mails that are seen as markedly seductive, erotic, or revealing undue familiarity; inappropriate pictures also may be included. Clinicians should be aware that computer hard drives may be subpoenaed and used as evidence, even after attempted deletions (which may not actually remove the materials). Beyond malpractice litigation, other consequences may include complaints to boards of registration and licensing and to ethics committee of professional organizations; the latter actions usually are not covered by malpractice insurance.
Obviously, it is a matter of clinical judgment whether large amounts of therapeutic attention should be paid to the rare neutral e-mail (“Confirming move to new time Friday”). Huge numbers of “neutral” e-mails might merit therapeutic exploration and possible limit setting, however.
Senior supervisors should recall that trainees today may be far more familiar with use of the internet in general and e-mails in particular but not accustomed to saving them for the medical chart. This point should be screened for and called to the trainees' attention. However, care also should be taken in how such extra-therapeutic communications are addressed during sessions. A trainee therapist, nonplussed at having received a written note from the patient between sessions, could err in starting the next session by demanding, “Now see here, what's the meaning of this?!”
Rather than proposing this beginner's error, we suggest a nonconfrontative exploration of all extra-therapeutic communications. The general principle is to try to bring extra-therapeutic material or contact into the therapy for processing. Examples might include, “Thank you for this thoughtful card. What were your thoughts when you decided to get it? How did you choose this one?” Or, “Thank you for your letter/e-mail. I have it here so we can talk about it. May I ask you now to read it aloud and call our attention to what you find yourself feeling?”
The written materials and the resulting discussion should be dated and placed in the chart. A strong and general argument could be made for charting any and all extra-therapeutic contacts.
In addition, clinicians should develop the habit of asking themselves why they are using e-mails in this instance. A useful touchstone is, “Would I say this same thing to the patient in the room, and if not, why not?” It also is important to maintain in the e-mail the same professional tone and language that one uses in the office. This includes avoiding inappropriate jokes, unprofessional social or self-disclosing remarks, and similar expressions.
Whether or not useful clinical material emerges from the exploration, the therapist's attention to the communication and insistence that it become part of the in-office therapy conveys an important message: the session is the place where the important communications take place, so that both parties can experience and strive to understand them.
- Borowitz SM, Wyatt JC. The origin, content, and workload of e-mail consultations. JAMA. 1998;280(15):1321–1324. doi:10.1001/jama.280.15.1321 [CrossRef]9794310
- Eng TR. The eHealth Landscape: A Terrain Map of Emerging Information and Communication Technologies in Health and Health Care. Princeton NJ: Robert Wood Johnson Foundation; 2001.
- Sands DZ. Guidelines for the Use of Patient-Centered E-mail. Massachusetts Health Data Consortium. 1999. Available at: http://www.mahealthdata.org/data/library/guidefiles/1999_PatientCtrdEmailGdlns-Sands.pdf. Accessed November 22, 2005.
- Kane B, Sands DZ. Guidelines for the clinical use of electronic mail with patients. J Am Med Inform Assoc. 1998;5(1):104–111. Available at: http://www.amia.org/mbrcenter/pubs/email_guidelines.asp. Accessed November 22, 2005. doi:10.1136/jamia.1998.0050104 [CrossRef]9452989
- Guidelines for Physician–Patient Electronic Communications. American Medical Association. 2004. Available at: http://www.ama-assn.org/ama/pub/category/2386.html. Accessed November 22, 2005.
- Levenson D. Online counseling: opportunity and risk. NASW News. 1997Sept:3.
- Stofle GS. Thoughts about Online Psychotherapy: Ethical and Practical Considerations. 1997. Available at: http://members.aol.com/stofle/onlinepsych.htm. Accessed November 22, 2005.
- Gutheil TG, Simon RI. Between the chair and the door: boundary issues in the therapeutic “transition zone.”Harv Rev Psychiatry. 1995;2(6):336–340. doi:10.3109/10673229509017154 [CrossRef]9384919
- Winnicott DW. Transitional object and transitional phenomena: a study of the first not-me possession. Int J Psychoanal. 1953;34(2):89–97.13061115
- Gutheil TG, Gabbard GO. The concept of boundaries in clinical practice: theoretical and risk-management dimensions. Am J Psychiatry. 1993;150(2):188–196. doi:10.1176/ajp.150.2.188 [CrossRef]8422069
- Frequently Asked Questions: E-mail in Psychiatry. American Psychiatric Association Council on Psychiatry and Law. Available at: http://www.psych.org/psych_pract/clin_issues/cybermedfaq.cfm. Accessed November 22, 2005.
- Professional Risk Management Services. The Risk Management Resource for Psychiatric Practice. American Psychiatric Association. 2004.
- Macdonald JM. Acting out. Arch Gen Psychiatry. 1965;13(5):439–443. doi:10.1001/archpsyc.1965.01730050053009 [CrossRef]5846705
- Gutheil TG, Gabbard GO. Misuses and misunderstandings of boundary theory in clinical and regulatory settings. Am J Psychiatry. 1998;155(3):409–414. doi:10.1176/ajp.155.3.409 [CrossRef]9501754
- Simon RI. Sexual exploitation of patients: how it begins before it happens. Psychiatr Ann. 1989;19:104–112. doi:10.3928/0048-5713-19890201-13 [CrossRef]
- Simon RI. Treatment boundary violations: clinical, ethical, and legal considerations. Bull Am Acad Psychiatry Law. 1992;20(3):269–288.1421558
- Miller PM, Gutheil TG. Rasch scale of perceived seriousness of boundary crossings and violations. Presented at: Association for Behavior Analysis 31st Annual Convention. ; May 27–31, 2005. ; Chicago IL. .