Therapist-client sex in psychotherapy : attitudes of professionals and students towards ethical arguments

INTRODUCTION: Data suggest that a substantial proportion of psychotherapists engage in therapist-client sex (TCS), violating national and international ethical guidelines. The objective of our study was to find a new and effective starting point for preventive interventions. METHOD: Using an online questionnaire, this study explored professionals’ attitudes toward aspects of a TCScase example influencing the tendency to pursue colleagues’ TCS, including self-interest and responsibility ascribed to clients. RESULTS: A total of 421 participants expressed preferences for courses of action and rated given information in a questionnaire. Results indicate that TCS is most often condemned for its inherent carelessness towards clients, its exploitative nature, the abuse of dependency and for counteracting the inherent intention of psychotherapy. Partial responsibility for TCS was attributed to clients by 41.3% of the respondents. Although self-interest related information was rated as an acceptable reason against pursuing TCS, a strong tendency exists to confront an abusive colleague, even at the risk of own disadvantages. DISCUSSION: In the detailed discussion ethical arguments against TCS (other than the certainly inflicted, but hardly measurable harm) are elaborated. In particular the incompatibility of TCS with a psychotherapeutic relationship, the responsibility for TCS in the asymmetrical clienttherapist relationship and the legitimacy of self-protection are discussed. CONCLUSION: Reasoning against TCS can and should be based on explicit, ethical requirements for psychotherapists. Furthermore, integrating the topic in psychotherapists’ training is encouraged and a discrete procedure to report a colleague’s TCS is requested.


Introduction
For many decades, sexual assaults of clients by psychotherapists occurred beyond public perception.In the 1970s, the first few case descriptions emerged, suggesting a considerable number of unreported cases [1,2].Even though there is an ongoing debate about the current prevalence and extend of therapist-client sex (TCS) [3,4], there is little doubt that a substantial proportion of psychotherapists engage in sexual interactions with their clients [2,[5][6][7].Such behaviour contrasts significantly with the unanimous disapproval of TCS expressed in the literature [8][9][10] as well as in professional ethical guidelines, codes and laws, which in general prohibit TCS [11][12][13][14].Investigating incidence and presumable severe consequences might seem to be the most logical approach in fighting TCS [6,[15][16][17].But since it is nearly impossible to gain reliable and objective data, especially concerning the inflicted harm by accomplished and attempted TCS, one major objective of this study is to depict a different approach: we decided to explore the preferences of arguments concerning TCS of young physicians and psychologists, as well as students of medicine and psychology, and their adherence to professional and ethical guidelines.Openness to other ethical reasons against TCS than the inflicted harm could set a different direction for the argumentation against TCS in education and literature.This approach allows access to potential deficits in the occupational role perception of prospective and young practitioners.In particular, we ascertained whether clients are blamed for TCS occurring.As a third objective, we explored the influence of self-interest on willingness to pursue and to prescribe the pursuit of a colleague's TCS.The legitimacy of self-interest and its consideration in the pursuit of TCS are discussed.We hope that our findings provide a new and more effective starting point for preventive interventions and hopefully give reason to adjust education and, hence, indirectly protect patients.

Study design
Participants were recruited in Germany during December 2008 via e-mail by using public mailing lists for physicians and psychologists, students of medicine and psychology.The e-mail included a description of the study and the re-quest to participate, as well as a request to click on a linked counter if the recipient was not willing to participate.

Instrument
Data were collected using an online questionnaire constructed with the online survey software "unipark", the academic program of QuestBack.The questionnaire started with an introductory page that informed about the research goal, the processing time, and about voluntariness and anonymity, among other things.On the second page, the demographic data of the participants were collected.This was followed by two parts of the questionnaire, in which six case studies (dilemmas) were presented, inter alia a case example in which a psychotherapy apprentice becomes aware of her supervisor having had a sexual relationship with a client (see table 4 for complete case study and questions).In the first part, the participants were asked to take the perspective of the psychotherapy apprentice and to choose their preferred course of action on an eight-point scale.A medium response option was not offered, because in reality a decision, at least between action and inaction, always has to be made.Since for methodological reasons the decision for the preferred course of action had to be measured on one dimension, only two options for action where given.The phrasing and the verbal anchors of the scale are shown in table 1.
The questionnaire included five additional clinical case studies that served as distractors for the TCS case study, as well as for research concerning general aspects of moral decisions like potential trans-situational factors.Subsequently, the case study about TCS was presented once more, but this time the participants were asked to decide from a third-party perspective.This prescription of action was measured on a nine-point scale (with a medium response option), since the moral assessment of two options can be equal.Afterwards, the information given in the case study was presented again as separate items (called information items in the following) with a seven-point rating scale recommending one or the other of the two offered courses of action.The phrasing of the instruction and three representative items can be seen in table 2. To calculate the significance of the difference between the decision about an individual's own assumed course of action and the prescribed course of action, the variable measured on the nine-point scale was transformed (x2 = 1 + (x1 -1) * 7 / 8), to prevent artificial significance of the t-test.The final question asked for which course of action the person in the case study should decide.The answer to this item is categorised as a prescriptive statement, since the participants prescribe an action to the psychotherapy apprentice.The phrasing and the verbal anchors of the scale are shown in table 3.

Sample characteristics
A total of 1,012 persons responded to the e-mail and 421 respondents completed the questionnaire; 250 responses were incomplete, 20 persons indicated that they were not a physician/psychologist (or student), and 321 indicated their unwillingness to complete the questionnaire (minimal response rate 0.424).Only the 421 complete datasets were used for further analysis.The study population consisted of 87 physicians, 37 psychologists, 81 psychology students and 216 medical students from middle and north-western Germany.Of these participants, 318 were female, 103 male.The sex ratio approximately corresponds to the ratio in the population of the addressed professions and programmes of study in Germany (as well as in Switzerland).Female participants were slightly overrepresented compared with the basic population.The mean age was 24.7 (±5.1) years.Statistical analyses were done with SPSS 19.Descriptive statistics (mean, standard deviation, range and median) were used for characterising the sample.

Research findings
According to the study participants, most information supported the option of confronting the psychotherapist, with mean values between -2.54 and -0.02.The fact that the instructor had contributed to the client's psychological distress without remorse was weighted most strongly (table 4).According to the study participants, only a few points spoke against a confrontation.The item deliberating a possible negative impact on the grading of the psychotherapist-in-training owing to the confrontation of the instructor in the portrayed hierarchical relationship (mean value -1.1) was foremost rated as an argument against the confrontation.Overall, 41.3% of the participants judged the information about the client initiating the TCS as a valid reason not to pursue the colleague's TCS.Nevertheless, data suggest that 75.5% of the participants favoured pursuing TCS of colleagues even though the decision in the presented case example highly affects an individuals own interests.
The resulting mean for the transformed item relating to the prescribed course of action is M = -1.779, the standard deviation equals SD = 1.676.The mean of the ratings about one's own assumed course of action is M = -1.260and the standard deviation is SD = 1.942.The two-sided t-test for paired samples with 420 degrees of freedom between the decision about an individuals own assumed course of action and the prescribed course of action results in a test statistic of T = 7,440 and a significance of p <0.001.Even though t-tests are "known to be virtually immune to violations of the normality assumption" ( [18], p. 544) [19], the Wilcoxon Signed Ranks Test was calculated and confirmed the significance.
The items with the four highest absolute medial ratings (items 11, 10, 12 and 9) relate to ethical reasons to confront the therapist.These items address the therapist's carelessness towards the client, his lacking integrity, the exploitative character of the therapist's behaviour, the contravention of his client's expectations, and the violation of obligations mentioned by professional ethical guidelines.The following two items with the fifth and sixth highest rating concern the inflicted harm (item 5) and the prevention of further harm (item 8).The tendency to recommend the pursuit is slightly, but significantly, higher than to pursue a colleague's abuse of a client personally.This difference implies that self-interest presumably has a diminishing influence on the tendency to actively pursue a colleague's abuse of a client, since self-interest should have a stronger influence on one's own decisions than on the prescription of decisions of others.Nevertheless, the willingness to pursue a colleague's abuse of a client can still be seen as high.

Discussion
This study highlights the attitudes towards ethical reasoning of medical and psychological professionals and students about a topic of major concern: TCS remains a challenge to professionals and institutions whenever it is revealed.Thus, empirically based insight into the convictions, attitudes and concerns of the background disciplines of psychotherapists is paramount to develop strategies that are able to shape the ethical behaviour of psychotherapists and hence, protect patients.One major objective of this article was to explore new possibilities for the prevention of TCS.Using an online questionnaire including a TCS case example elicited information which the participants themselves suggested to be important for reaching a decision in a representative case of TCS.The results suggest that the integrity of the therapist, justice, the assumable expectations of clients and adherence with ethical directives are valued as primary crucial reasons to counteract TCS.However, in the literature arguing against TCS, the focus is primarily placed on the expected harm to clients [4,9,20].Not only accomplished, but also attempted TCS and less severe or less apparent boundary violations can be harmful; in these cases it might also be harder for the clients to recognise the abusive character of the therapist's behaviour.Accordingly, it could be helpful to support ethical reasons against TCS in the literature.From an ethical point of view, sexual abuse of clients does not become acceptable even under the hypothetical and counterfactual assumption that it would not harm clients.Accordingly, doubts concerning the validity of the findings about harm inflicted through TCS [3,4] do not challenge its ethical condemnation, but encourage misconducting professionals to deny any severe consequences of TCS for the client, for instance by ascribing the harm due to TCS to other influences such as pre-existing psychopathology of the client.
Based on the data of this study, we recommend that reasoning (and action) against TCS should not only rely on the argument of (hardly measurable) empirical harm resulting to the client.Although non-maleficence is a strong principle, both in medical tradition and in recent medical ethics, as well its codification [21], condemnation of TCS should not depend on providing empirical proof of harm beforehand.Rather, the nature of TCS in contradicting psychotherapeutic essentials such as trust, empathy, the therapist's neutrality and support for the client's autonomy calls for an explicit ethical stance that can be formulated not only on the grounds of principlistic approaches [22,23] that do not only rely on consequentialistic reasoning, but also on premises: it is a matter of consistency with psychothera-peutic values [23] and values of healthcare in general [22] to deny the legitimacy or ethicality of TCS.From the ethical values inherent in psychotherapy, the incompatibility of TCS with a psychotherapeutic relationship is a logical conclusion.As soon as a psychotherapist engages in inappropriate relations with the client (of which sexual contact is one extreme form), the therapeutic relation is terminated by definition.It is self-evident that a psychotherapist, who is not willing or able to keep up his professional therapeutic relation to his clients by frequently terminating it through misconduct or even TCS, not only contradicts his own professionalism, but forfeits his privilege to practice as a psychotherapist.
Abstinence from TCS should be based on explicit, general, undoubted and mandatory ethical requirements for psychotherapists, as they treat individuals who are in need of special protection.These ethical requirements should be an essential part of any training and evaluation of potential psychotherapists such as prospective and young physicians or psychologists.It may well be that 'abstinence' from TCS needs to be learned by apprentices in different ways in present times, as compared with the early history of (psychoanalytic) psychotherapy.Another objective of this article was to find out whether clients are blamed for TCS occurring.An important issue in many professional and ethical guidelines concerns the notion (widely unchallenged among experts) that the therapist bears sole responsibility for TCS [24], even in cases where the 'initiative' came from the client.Therefore, the median tendency to value the information about the client as the driving force behind TCS, against the investigation of a colleague's TCS, reveals an obvious discrepancy between the experts' opinions (as e.g., condensed in ethical guidelines) and the prospective psychotherapists' opinions.The consent to TCS, or even the pleading of the client in favour of it, does not authorise the therapist's misconduct [24,25]; that is, it fails to give an ethical justification of such behaviour.Since the client-therapist relationship is asymmetrical and a central matter of professional competence, the sole responsibility is with the latter.It thus seems necessary to confront the misguided, but presumably widespread, 'notion of victim responsibility' ( [26] p. 227).Therefore, we suggest that students of medicine and psychology, as well as apprentices of psychotherapy, should explicitly be educated about basic ethical obligations and boundaries for psychotherapists.This should aim at developing an ethical position and role-concept that helps psychotherapists to maintain their integrity regarding possible TCS (and other boundary violations).
The third objective of this article concerns the influence of self-interest on the willingness to pursue and to prescribe the pursuit of a colleague's TCS.In the attempt to reduce the prevalence of TCS, psychotherapists who come to know about the TCS of a colleague could become active and helpful by confronting the colleague or informing the relevant professional associations or authorities.However, the tendency to recommend the pursuit is higher than to pursue a colleague's abuse of a client personally, anticipated disadvantages for the individual seem to diminish the willingness to take action.Furthermore, as the high median rating of item 14 indicates, self-interest is not only seen as

Original article
Swiss Med Wkly.2015;145:w14099 Swiss Medical Weekly • PDF of the online version • www.smw.ch an acceptable reason not to pursue a colleague's TCS, it is even considered as an acceptable reason for the recommendation not to pursue a colleague's TCS.It may be that respondents believe that a certain amount of self-protection is advisable or even necessary to become and work as a professional.This aspect is particularly interesting since the regulations, guidelines, professional codes of ethics and other normative texts of the relevant professional associations differ in their prescriptions regarding the handling of colleagues' mistakes that one might notice.For example, the Metacode of Ethics of the European Federation of Psychologists' Associations expresses the obligation first to contact the (suspected) misconducting colleague and, only if appropriate, report to professional associations and authorities.The explanation of Lindsay, Koene, Øvreeide, and Lang is: "Not giving feedback to a colleague, whose professional behavior is perceived not to be in accordance with ethical principles, is disloyal to this colleague."([23] p. 150).In contrast, most other guidelines, for instance the International Code of Medical Ethics of the World Medical Association or the Ethical Principles and Implementing Procedures of the International Psychoanalytical Association, do not recommend first contacting the colleague concerned.Contrary to Medau, Jox and Reiter-Theil [27], we recommend that the right to contact professional associations and authorities on potential TCS observations without informing the misconducting colleague beforehand should be granted to psychologists, psychiatrists and psychotherapists for several reasons.First, individual professionals lack resources, skills and justification to search for proof; second, a central arbitration body would be able to combine information from different sources and ensure the validity of a complaint about a particular therapist; third, important information may not reach the authorities, preventing them from taking appropriate further steps; and fourth, a certain amount of self-protection seems not only to be legitimate, but even advisable for psychotherapists.The risk of wrongfully denunciating a colleague may be reduced without endangering one's own interests by providing a confidential procedure to contact the colleague under suspicion of TCS anonymously.Even though data about false accusations of sexual offences are hardly available, the risk and especially the potentially devastating personal and occupational consequences of false accusations for professionals should not be underestimated.Nevertheless, this should not become a major point against effective prosecution.The risk of approving false allegations by mistake may be reduced further without endangering professionals or complainants by providing a standardised, confidential procedure of investigation.In addition, clients could and should be informed about the ethical obligations of their therapist and about the option to report unethical behaviour to specified bodies.
Regarding the high tendency among our participants to condemn TCS, this behaviour cannot be seen as the consequence of a more or less widely shared well deliberated opinion of psychotherapists, even if there is a minority of psychotherapists who, contrariwise, declare TCS as a 'cure' [28][29][30][31].Rather, TCS seems to be the result of the therapist's inability to care for one's own needs in an ap-propriate manner [32].According to the literature [33], a chance of prevention by sensitization, transparency and education exists only for a small proportion of potentially misusing psychotherapists.Since abusive psychotherapists damage the professional standing of all psychotherapists, it is arguable that abusive psychotherapists should not be allowed to continue their therapeutic work -at the possible expense of another potential victim.In utilitarian terms, the cost-benefit ratio of every new chance drops with each further victim.Therefore, we advocate an occupational ban or at least explicit and supervised restrictions for abusive psychotherapists (see also [8]) especially for repeat offenders.
One limitation of the study is that the sample consists only of German-speaking subjects and, accordingly, the original questionnaire was in German.To generalise the findings, conducting the test with an English-speaking sample should be considered.Since we intended to examine a target population of potential medical and psychological psychotherapists who are still developing their occupational identity, the mean age is quite low, as most of the participants were still students.Therefore, no conclusions concerning the full population of psychotherapists of all ages are possible.Nonetheless, the results suggest the need to include this topic early in the training of medical, psychological and psychotherapy (and other disciplines) students.
An additional limitation of our study is that our interpretation of the items and our strategy to increase the validity of the data through the presentation of a case example instead of approaching attitudes directly cannot be evaluated.Taking account of the low average professional experience of the participants, their assumption how they would react in a given scenario may be wrong.Furthermore, the presented case example implies an unrepresentative asymmetrical relationship between perpetrator and observer, since we attempted to enhance the urgency of a self-interested decision.
Even though recent studies [34][35][36] show that surveys with low and high response rates predominantly yield results that are statistically indistinguishable, the response rate of 42% might be a limitation of our study.Of 1,012 persons responding to the e-mail, 671 (66%) took up its invitation to participate in the survey.Three reasons for not responding could be the unpopular topic, an assumed lack of time of the target population and the absence of extrinsic rewards.Of these 671 participants taking the survey, 421 (63%) completed the survey.Since the whole questionnaire contained more than 120 (partly quite serious) questions, it is impressive that nearly two thirds of the participants answered them all.Nevertheless, the representativeness of the findings could be diminished.The strengths of this study are that, through an innovative approach to this unpopular topic, new data are generated, which allow a new and hopefully fruitful perspective.This approach to data collection contributes to the validity of the data, since the realistic case example enables the test persons to put themselves in the position of the practitioner.Thus, the study helps to fill a gap in research and to find a way to respond to occurring TCS.Moreover, it facilitates objectifying the frequently heated discussion about TCS.

Conclusions
Despite the limitations of our study the results indicate a strong tendency of the participants to pursue TCS.But still uncertainties concerning the role-concept of psychotherapists, especially their responsibility or the question of appropriate self-protection remain.In conclusion, we suggest that the sensitisation to boundary violations such as TCS should be integrated into the training of students of medicine and psychology, together with other ethics education, in order to help establish principled ethical reasoning.Furthermore, we endorse the development of a procedure and guidelines for dealing with colleagues' TCS-related misconduct in order to protect the interests of both clients and informants.We also support the harmonisation of existing legal regulations, ethical guidelines and professional codes towards a unanimous position on reporting a colleague's TCS.Professional associations and societies should provide hotlines and readily available help for victims of TCS.In addition, guidelines and procedures for reliably informing patients about the ethical obligations of therapists and possible help in the case of therapeutic misconduct should be provided before starting psychotherapy.

Table 1 :
Instruction and verbal anchors to the own decision.Please decide on the basis of the given information with which certainty you would prefer which course of action.

Table 2 :
Instruction and verbal anchors to three examples of information items.

Table 3 :
Instruction and verbal anchors to the prescriptive statement.
The psychotherapy apprentice definitely should not question her instructor concerning this incident 4Original articleSwiss Med Wkly.2015;145:w14099 Swiss Medical Weekly • PDF of the online version • www.smw.ch

Table 4 :
Descriptive parameters.Asked by the psychotherapist-in-training, whether she might confront her instructor concerning this incident, her friend states that she does not feel comfortable about it, but would accept it, if her friend thinks it to be best.She knows that the behaviour of her instructor has violated the obligations in the ethical directives of the German guidelines (DGPs, BDP) as well as in comparable Austrian or Swiss guidelines.Please decide on the basis of the given information which course of action the psychotherapy apprentice should choose and with which certainty.
14.In the further course of her training, the psychotherapist-in-training will have to take examinations administered by this instructor, and confronting the instructor about the incident might have a negative impact on her grading.The instructions for the items and the verbal anchors to the corresponding scales are shown in tables 1-3.SD = standard deviation Original article Swiss Med Wkly.2015;145:w14099 Swiss Medical Weekly • PDF of the online version • www.smw.ch It can be supposed that our data originate from persons of the target population whose interest in the topic is above average.If we assume further that interested persons give more elaborated answers, the performance of the target population concerning the examined ethical questions should be lower, which would emphasise the necessity for more exhaustive ethical training of the target population.Although we intended to present an ethical dilemma with no socially preferred course of action, nonetheless social desirability could have biased the data.Especially, it could Swiss Med Wkly.2015;145:w14099 Swiss Medical Weekly • PDF of the online version • www.smw.chhave diminished the influence of self-interest on our findings.