Abstract
Background:
Boundary issues, which regularly arise in therapy, can present dilemmas for most clinicians. There has been substantial literature on boundary excursions in clinician–patient relationships, however, very little empirical research exists and is documented. As mental health researchers, we need to investigate a wide range of sensitive topics to enhance our understanding of the many issues that arise in the psychotherapeutic frame.
Aims:
We set out to empirically explore perceptions of what may constitute a boundary violation among the mental health staff in the State of Qatar and their views on the subject.
Results:
A total of 50 participants (24 psychiatrists, 2 doctorate level psychologists, 24 psychiatric nurses) responded with a response rate of 80%. Participants rated each possible boundary violation according to its degree of harm and professional unacceptability. Three distinct groupings of boundary violations were obtained: (1) core, consisting of the most serious violations; (2) disclosure and greeting behaviour, involving disclosure of information about the therapist and greeting behaviour; and (3) separation of therapist and client lives, involving encounters between therapists and clients outside of therapy.
Conclusions:
It is important to ascertain these dilemmas so that these theoretical models can be integrated in clinical practice.
Introduction
Maintaining boundaries in human relationships is crucial in all aspects of life, more so in professional settings. However, the doctor–patient relationship is a rather complicated one especially in psychiatry and clinical psychology. Doctors are observed as being in a position of power over the patient particularly as they have legal obligations to deprive the patient of liberty if necessary, yet they must demonstrate a level of empathy and understanding. The risk of boundary violation is often higher in medical specialties involving physical contact with patients. In psychiatry, a doctor–patient relationship tends to be long term with deeper involvement in the patients’ personal life, exploring their personal feelings and experiences, which makes both psychiatrists and patients vulnerable to boundary excursions.
Gabbard and Nadelson (1995) defined professional boundaries as parameters that describe the limits of a fiduciary relationship in which one person (a patient) entrusts his or her welfare to another (a physician), to whom a fee is paid for the provision of a service. It is important to distinguish between boundary crossing and violation. Crossings are departures from usual practice that are not exploitative and may sometimes be helpful to the patient, while violations are harmful (Galletly, 2004). Repeated crossings may eventually lead to a violation. This area becomes blurred when considering that medical practice is shifting away from its paternalistic attitude to a more informal approach where the patient also has a stronger say in what is required.
In a psychotherapeutic environment, several factors can play a role. Establishing rapport with the patient is fundamental for the therapeutic process. The balance between personal and professional becomes delicate and may affect those who may already be emotionally vulnerable.
Boundary crossings and violations extend to many behavioural practices, ranging from touching a distressed patient to providing comfort, hugging at the end of an interview, to accepting of gifts, and the more serious physical and sexual exploitation. While there is a general agreement that pursuing a sexual relationship with a patient, or ex-patient, is a violation, other crossings or violations may be less well defined in certain cultures. Most notions of therapeutic boundaries have ignored cultural variations and this remains an under-researched area. A boundary breach in a certain culture may be acceptable to both patients and physicians in another (Miller, Commons, & Gutheil, 2006).
In the Middle East, anecdotal evidence suggests that some families may provide expensive gifts to their treating doctors and other staff. This practice has not only become accepted, but actually expected, particularly in surgery and obstetrics. The patient may offer an amount of money to the caring staff as a token of appreciation. Refusal to accept a gift may be taken as an insult. Furthermore, it is normal in this culture to call patients by their first name, yet a doctor is unlikely to accept being called by their first name by the patient.
Doctors may be more at risk of breach of boundaries especially when under stress, or when lacking personal or professional support when working in solitary specialties and isolated locations. Patients who have been subjected to emotional, physical or sexual abuse are also more vulnerable. Any breach of boundaries can seriously damage the doctor–patient relationship and have a negative impact on the patient’s trust of his/her doctor and the medical profession as a whole (Galletly, 2004; General Medical Council, 2006).
The objective of the current study is to explore how clinicians understand these dilemmas. We set out to integrate some of the theoretical constructs in the practical understanding of boundary issues in counselling and psychotherapy. An additional objective is to review what is seen as common boundary violations in the Gulf State of Qatar and make recommendations to deal with such boundary violations when these occur in therapy.
Method
The study was conducted in the Rumeilah Hospital at the Hamad Medical Corporation in Doha, State of Qatar. The survey was conducted from January 2008 to May 2008. A total of 50 participants responded: 24 psychiatrists, 2 doctorate level psychologists, 24 psychiatric nurses. The questionnaire was distributed to staff after a regular weekly staff meeting. Participation was voluntary and anonymous. All individuals were asked to fill out the questionnaire and return it anonymously in the blank envelope provided, within a period of four weeks. The response rate was 80%. There were no differences between the respondents and non-respondents. The sample was representative of the mental health professionals in the hospital. The study was exempted from review by the Institutional Review Board at the Hamad Medical Corporation in Doha as no patients were involved.
Instrument
Miller et al.’s (2006) questionnaire was amended following consultations with the author. A few additional questions assessed certain cultural professional experiences seen in psychiatry in Qatar. The specific boundary excursions asked about are shown in the Appendix. Participants were asked to rate (on a six-point scale) the percentage of cases in which it would be harmful to the patient if a colleague behaved in the manner described, and concurrently, to rate in what percentage of cases this same behaviour was professionally unacceptable. The technique of asking about colleagues’ behaviour, rather than the clinician’s own, was used to minimize defensive reactions that might otherwise occur. The rating scale asked participants to circle one of the following percentages: 0%, 2%, 16%, 50%, 84%, 98% and 100%. This scale reflects a linear z-score scale that was converted into percentages, using the cumulative normal distribution function, with a probit transformation. This scale was used because: (1) it represents the postulated distribution for most psychological characteristics; and (2) the differences at the ends of the scale are more important than those in the middle (Galletly, 2004). All the questions were in English as all participants had a good command of the language.
Student t-test, non-parametric Mann-Whitney, χ 2 and Fisher’s exact test (two-tailed) were performed appropriately. Principal component analysis (PCA) with varimax rotation was used to ascertain the factor structure of the clinicians’ behaviour in Qatar. Cronbach’s α reliability coefficients were also calculated to assess the internal consistency of the clinicians’ behaviour scale scores. The level p < .05 was considered as the cut-off value for significance.
Results
There were 50 respondents; the majority were physicians and there were more males than females in this group. Table 1 shows the socio-demographic characteristics of physicians and nurses by gender.
Socio-demographic characteristics of physicians and nurses by gender.
Cronbach’s α for the questionnaire about the professional unacceptability was .936 and for professional unacceptability and acceptability combined was .968. These findings are illustrated in Table 2.
Description of some clinicians’ behaviours.
PCA with varimax rotation was carried out by eliminating coefficients with an absolute value less than .5 in order to reduce the number of items giving weight to major contributors. Out of the total 176 questions to four distinct factors (Table 3), 117 items remained. To determine appropriate labelling for each factor, the most commonly occurring item in each was considered as a representative for the whole factor. Table 3 shows the four factors as follows: (1) building personal relationships and business excursion – 66 items (55.6%); (2) sexual violations – 28 items (23.9%); (3) touching or greeting – 12 items (10.3%); and (4) mixing personal and business in the profession – 12 items (10.3%).
PCA in clinical practice in psychotherapy.
A Rasch (Dattilio, Commons, Adams, Gutheil, & Sadoff, 2006; Rasch, 1980) analysis was carried out using WINSTEPS (Linacre, 2003) to determine whether or not there were gaps between the seriousness of various items. If not, there would not be a bright line separating boundary excursions and more serious boundary violations. The Rasch analysis would also help define rough equally spaced regions of seriousness. Figure 1 demonstrates boundary crossing and violations of 50 raters (N = 176). The Rasch analysis also showed that the scale values vary from -3 logits to +1 logits, with -3 indicating the most serious violations. Hitting the patient (-2.56) was seen as the most extreme. Only slightly less extreme was kissing the patient on the lips (-2.01). The items having sexual intercourse with the patient and paying the patient to do any of the above both had Rasch scores of -1.29, which for some reason was seen as less serious. At the other end, things such as greeting clients with only a handshake (.9) were seen as not at all serious.
Boundary crossings and violations of 50 raters (N = 176).
A value of 2 logits roughly describes the distance between a category where the boundary excursions are perceived as not at all serious and a category where they are viewed as extremely serious. The category with items perceived as the most acceptable contained items such as shaking hands with same gender patient and patting patient on the back (.72). One logit more serious is a category that would have items most likely characterized as being moderate boundary excursions, such as phoning patient about treatment after office hours and evaluatively commenting on patient’s partner (as a critical observation) (.14). Slightly more serious, but in the same range, is socializing with a patient outside the clinical setting (.12). In the extremely serious range is yelling at patients (-.76) and making sexist remarks (-.84). However, these divisions remain arbitrary.
Discussion
The findings in this study highlight some of the issues clinicians face on a daily basis. It is important both from an ethical point of view but also from a professional view to ensure that patient–clinician boundaries are not muddled and are clear. While interpreting these data it is important to bear in mind that these observations are from one group of clinicians in one hospital. These estimated values of perceived seriousness are based on a small sample size and generalizability across other cultures may be problematic.
In Qatar, the medical system is state sponsored and all nationals are provided free medical health care. This may make patients feel indebted to their treating physicians, which may further encourage patients to frequently offer gifts to express gratitude. In the Arab culture it is generally considered rude to refuse gifts. In our questionnaire, 50% of the therapists expressed they take gifts to some degree, whereas 24% stated that they were not sure. It is important to explore the meaning of the gift, with special attention to feelings of rejection if the gift is inappropriate.
The participants responded clearly and there appears to be significant agreement on a number of parameters. In this sample over 94% of the items fall on a single dimension of perceived seriousness of a boundary issue. This result may be contrasted with our previous factor analysis which suggested three clear factors (Blatt, 2001). But the single scale makes more sense in describing the severity of all boundary excursions. Less personally invasive excursions (e.g. going to a client’s funeral, visiting a client at home in pursuit of medical activity, patting a client on the back or displaying diplomas within the office) were also seen as comparatively less serious. More personally invasive excursions (e.g. ‘necking’ with a patient or borrowing money from a patient) were rated higher in severity, thereby indicating that there is a clearly understood boundary.
The scale is linear and smoothly continuous. This linear scaling provides support for the clinical and forensic observation of progressive boundary excursions (Blatt, 2001; Kroll, 2001; Strasburger, Jorgenson, & Sutherland, 1992), but this deserves to be explored in qualitative explorations as scales may not only be more convenient but they may also allow participants to answer in limited ways. As noted earlier, at times there is no clear line dividing the boundary crossings from the boundary violations. This further underscores the critical role of context in the analysis of boundary issues (Gutheil & Gabbard, 1998; Linacre, 2003). For the less serious boundary excursions, context matters greatly. For items with a value of -.83, behaviour begins to occur that almost guarantees a high likelihood of ethics complaint or litigation. Almost all the boundary issues with values more negative than -.83 become suitable legal issues, but does this mean that a cut-off point on the scale will allow such an approach? Context matters below this point because of the considerable ambiguity and the nature of the excursions. But after -.83, there is no escaping that these issues constitute boundary violations. These findings need to be duplicated in other clinical settings.
Despite the diversity in the definition of boundaries between cultures, and even between clinicians within the same culture, there is general agreement on the parameters that guide delivery of psychotherapy (Gabbard, 2001). All trainees in psychiatry and psychology are expected to acquire therapeutic skills and boundaries related to physical contact with patients, the extent of self-disclosure and maintaining confidentiality. However, recent discussions identifying the cultural relativity of boundaries show that acquiring this skill is more complex than previously assumed (Gabbard, 2001; Kroll, 2001). This may be because psychotherapy occurs within a framework created by managing parameters such as role, space, time, self-disclosure, fees, gifts and confidentiality. These are the same components that form a therapeutic boundary. In psychiatric practice too, similar factors play a role in identifying therapeutic boundaries. Clinicians should recognize the benefits of self-disclosure as well as its dangers (Psychopathology Committee of the Group, 2001). The choice of whether to self-disclose should always be based on the patient’s best interests. Guidelines, examples of good clinical practice and supervision are all necessary to make the best choices about self-disclosure. Although it is a component of many harmful boundary violations, it does not inevitably lead to them (Psychopathology Committee of the Group, 2001).
In Qatar most mental health workers acknowledge that they occupy a desk and thus sit across from the patient. This not only imposes an artificial distance between therapist and doctor but also reconfirms the doctor’s status of power in this generally paternalistic culture. Furthermore, when asked if they would refer their patients to same-gender therapists, 6% of therapists stated that they would most of the time, 16% stated some of the time, 10% stated little of the time and 18% stated they were not sure. Issues of gender match between the professional and the patient may be more relevant in Arab cultures and need to be explored further. The challenge is to see if gender matching between therapist and patient produces better outcome or simply better initial engagement.
Thus, diversity of the culture-bound understanding of boundaries, and what could be considered as acceptable in a certain cultural settings, raises the necessity of developing ethical guidelines that are culture specific (Godlaski & Clark, in press). A clear and reasonably specific set of principles or ethical standards is crucial. This should avoid the risk of having overly rigid rules that hinder meaningful practice (Gutheil & Gabbard, 1993). The guidelines should be disseminated to all health care providers and regularly monitored to ensure application. Regular training should include supervision and also teaching culture-specific guidelines on boundaries and how to enforce them (Hess, 1998).
Conclusions
The seriousness of a boundary crossing or violation depends less on the clinician’s belief on what is right or what is wrong than on the effect it may have on the patient. Although certain crossings may have some therapeutic role, what might be helpful for one client can prove harmful for another. Self-disclosure is an example of such crossing that may have different interpretations according to the doctor–patient situation.
Clinicians and managers must take the lead on developing and sticking to culture-specific guidelines and evaluate their use in appropriate settings so that patients can continue to feel and remain safe.
Footnotes
Acknowledgements
This work was supported by the Hamad Medical Corporation HMC, Doha, State of Qatar.