The Hidden Curriculum in Ethics and its Relationship to Professional Identity Formation: A Qualitative Study of Two Canadian Psychiatry Residency Programs

Les années de résidence comprennent la dernière période de formation officielle d’un médecin. C’est à ce stade que les stagiaires consolident les compétences cliniques requises pour une pratique indépendante et atteignent un niveau de développement éthique essentiel à leur travail de médecin, un processus appelé formation d’identité professionnelle (FIP). On pense que l’éducation à l’éthique contribue au développement de l’éthique et à cette fin, le Collège royal des médecins et chirurgiens du Canada (CRMCC) exige que l’éducation formelle en éthique soit intégrée dans tous les programmes de formation spécialisée postdoctorale. Cependant, un programme d’éthique formel peut fonctionner en parallèle avec des programmes d’éthique informels et cachés, ces derniers étant plus subtils, omniprésents et influents pour façonner les attitudes et le comportement des apprenants. Cet article fait état d’une étude des programmes d’éthique formels, informels et cachés de deux programmes de psychiatrie postdoctorale au Canada. Sur la base de l’analyse des sources de données, nous relions les divergences entre les programmes d’éthique formels, informels et cachés à deux aspects de la formation de l’identité professionnelle (FIP) pendant la formation en résidence en psychiatrie. Le

The residency years comprise the last period of a physician's formal training. It is at this stage that trainees consolidate the clinical skills required for independent practice and achieve a level of ethical development essential to their work as physicians, a process known as professional identity formation (PIF). Ethics education is thought to contribute to ethical development and to that end the Royal College of Physicians and Surgeons of Canada (RCPSC) requires that formal ethics education be integrated within all postgraduate specialty training programs. However, a formal ethics curriculum can operate in parallel with informal and hidden ethics curricula, the latter being more subtle, pervasive, and influential in shaping learner attitudes and behavior. This paper reports on a study of the formal, informal, and hidden ethics curricula at two postgraduate psychiatry programs in Canada. Based on the analysis of data sources, we relate the divergences between the formal, informal, and hidden ethics curricula to two aspects of professional identity formation (PIF) during psychiatry residency training. The first is the idea of group membership. Adherence to the hidden curriculum in certain circumstances determines whether residents become part of an ingroup or demonstrate a sense of belonging to that group. The second aspect of PIF we explore is the ambiguous role of the resident as a student and a practitioner. In ethically challenging situations, adherence to the messages of the hidden curriculum is influenced by and influences whether residents act as students, practitioners, or both. This paper describes the hidden curriculum in action and in interaction with PIF. Our analysis offers a complementary, empirical perspective to the theoretical literature concerning PIF in medical education. This literature tends to position sound ethical decisionmaking as the end result of PIF. Our analysis points out that the mechanism works in both directions: how residents respond to hidden curriculum in ethics can be a driver of professional identity formation.

Introduction
The residency years comprise the last period of a physician's formal training. It is at this stage that trainees consolidate the clinical skills required for independent practice and achieve a level of ethical development essential to their work as physicians. This is a process known as professional identity formation (PIF) (1). Ethics education is thought to support and facilitate this ethical development during residency. To that end, the Royal College of Physicians and Surgeons of Canada (RCPSC) 1 requires that ethics education be integrated within all of its accredited postgraduate specialty training programs (3). The RCPSC has made substantial efforts toward developing a formal ethics curriculum (FC) for residency training although neither the specific content of the curriculum nor its format are mandatory (4). Each program determines how it will fulfill its obligation to offer formal ethics education to its residents. For the RCPSC, formal ethics education tends to focus on clinical ethics but can also extend to research and organizational ethics depending on the context and learning needs of trainees within a given program and location.
Page 81 A formal ethics curriculum includes explicit, official teaching such as seminars, rounds, journal clubs, and case-based discussion sessions. Several authors have pointed out that education of this type operates in parallel with informal and hidden ethics curricula (5,6). The informal curriculum (IC) consists of unscripted, interpersonal forms of teaching and learning that take place between faculty and students, such as bedside teaching. The hidden curriculum (HC) refers to influences at organizational and cultural levels that inform the learning process (7). The HC is said to be more subtle, pervasive, and influential than FC and IC in shaping student attitudes and behavior (5,6). Indeed, previous analyses of the impact of the hidden curriculum have suggested that if its negative elements are left unaddressed during training, they can lead to ethical erosion instead of ethical development (8)(9)(10)(11). Thus, the HC has the potential to disrupt PIF of residents. As a result, the Future of Medical Education Postgraduate Project report (12), funded by Health Canada, recommends that postgraduate education organizations and faculties of medicine attempt to address the counterproductive elements of the HC.
Addressing the impact of the HC in postgraduate education requires knowledge of how it operates in that particular context. However, our understanding of it is based primarily on studies in undergraduate medical programs (13)(14)(15)(16). VanDeven and colleagues (17) investigated the HC in five postgraduate radiology programs. They found that the HC contained a central message specific to that specialty, namely that professional isolation is a norm. Our involvement in postgraduate education in psychiatry, and in ethics education specifically, led us to explore the specificity of the HC in psychiatry. We investigated the formal, informal, and hidden ethics curricula at two postgraduate training programs in psychiatry (hereafter P1 and P2). Our study began by identifying areas of divergence between the FC, IC and HC on specific ethics topics where they existed. We then sought to understand how these divergences arise, what sustains them, and how residents responded to them. Our analysis yielded two emergent themes, each of which was relevant to certain ethics topics where divergences existed. In an earlier paper, we discussed how 'the minimum standard of ethics' generates and sustains divergences between the FC, IC, and HC. We observed this 'minimum standard of ethics' applied to the topics of interactions with the pharmaceutical industry, dealing with colleagues' ethical lapses, and residents' performance evaluations (26). In this paper, we report on the second theme, which relates to the professional identity formation (PIF) of residents and how it relates to four specific ethics topics.

Study Design and Setting
We used qualitative case study as our overall research design because it is appropriate when 1) the phenomenon of interest is ongoing rather than historical and cannot be separated from the context in which it occurs and 2) researchers want to answer "how" and "why" questions in addition to determining what is happening in the situation under study (18,19). Specifically, we conducted an instrumental case study because we were using the cases (the two programs) to gain insights into the more general phenomenon of the hidden ethics curriculum rather than evaluating the programs themselves.
We also decided to conduct a multiple-rather than single-case study because we wanted to gain in-depth understanding of the issue by comparing data from two different contexts. We intentionally chose two programs of different sizes and from different regions of Canada to facilitate our comparisons. We assumed that certain differences between the two programs were likely to shape participants' experiences and views on ethics training in their programs (20)(21)(22), the most important being the program's focus, i.e., P1training clinician scientists and subspecialty clinicians versus P2training versatile general psychiatrists capable of making a long-term commitment to the province.

Sources and Data Collection
A hallmark of the qualitative case study design is collecting and comparing findings from multiple sources of data (triangulation) to gain a deeper and balanced understanding of the topic under study compared to what can be achieved from a single source (23,24). For each program we collected data from three sources to achieve triangulation: 1) documents from the program, hospital or faculty (e.g., ethics curriculum documents, ethics policy papers, and departmental mission statements); 2) key informants (KI, i.e., faculty and residents who had insider knowledge about important ethical issues because of their roles in their programs); and 3) resident volunteers. To preserve confidentiality, we do not cite the documents directly nor do we name the specific roles of key informants. Given the potentially sensitive subject matter of the interviews, the study team took particular care to recruit participants using confidential means (e.g., by individual mailings rather than group announcements) and to offer to conduct interviews away from their base hospital if participants wished. The study received formal research ethics board approval at both locations. Each participant gave informed consent before participating in the study.
We reviewed program documents relating to ethics in order to obtain contextual information about each program as well as any explicit statements or commitments regarding ethics education. The document review enabled us to describe the formal ethics curriculum for each program (what each program officially taught about ethics) and identify potential key informants. This was an information-gathering step to which we did not apply a specific analytic framework. We refined the interview guides through team discussion following a review of dominant themes in the literature on the HC in ethics education. Our document review enabled us to develop and add program-appropriate probe questions into the interview guide.
In the first phase of the project, we recruited 19 KIs from P1, (47% of whom were residents) and 5 from P2, (20% of whom were residents). One potential P1 KI declined to participate on the grounds that it was "too risky." Despite our enquiry, we were not able to obtain any further information about this decision. An interviewer at each site, unknown to the participants and working outside the program, conducted semi-structured interviews with KIs, following the interview guide. In the second phase, we invited participation from the entire resident body at each site. We recruited 16 resident volunteers in P1 (out of the approximately 100 remaining residents not identified as potential KIs) and 8 in P2 (out of approximately 14). These semistructured interviews were conducted by the project's resident co-investigators 2 , who were also unknown to the interviewees. The resident co-investigators acted as interviewers for the resident interviews at their counterpart program. The interview guides for KIs and residents differed somewhat. The KI guide included questions that probed insider knowledge of the hidden curriculum within each program (see Appendix A and B) while the resident guide focused directly on individual experience of the three curricula both in direct teaching and in the clinical training environment.
All interviewers participated in a formal training session with an experienced qualitative researcher. The training involved assigned readings and discussion of interviewing methods and skills, practice interviews using the draft interview guides, and feedback on each interviewer's interviewing technique. For logistical reasons the interviewer for the P2 KIs participated in training sessions via telephone. In providing this training, we aimed to achieve the greatest degree of consistency possible between interviewers.
Each interview was audiotaped and transcribed, except for one in which the participant declined to be taped but agreed to have statements documented in field notes taken by the interviewer. The interviewers reviewed the transcripts of their interviews to correct any transcription errors. The researchers sent the corrected transcript and a summary of key points to each participant for review, corrections, or withdrawals (there were no withdrawals).

Data Coding and Analysis
Our analysis followed methods described by various case study researchers whose guidelines are an accepted part of the qualitative research literature (22,23,27). Our data organization and analytic processes involved two or more analysts at all times. The analysts included the principal investigator and two research assistants, one person in phase one and a different person in phase two. The first research assistant was herself a qualitative researcher while the second was a learner who had been engaged in a studentship with the project from its outset and was thus familiar with the ongoing analytic discussions.
Our initial phase of data analysis was carried out in three steps. First, through re-reading and discussing the transcripts, the investigators identified in each program several ethics-related topics to which we attributed codes. We identified (if possible) whether these topics were addressed in the FC, IC, or HC of each program and what messages were transmitted about these topics. Second, we searched for convergences and divergences between the three curricula on the topics. Third, we looked for similarities and differences between the two programs in terms of the ethics topics of greatest importance to the participants. For those topics that were of concern to both programs, we contrasted the divergences between the respective curricula. We organized the interview data and codes using a qualitative data management software program (Nvivo7).
In phase one of the analysis, we also noted that there were certain broad concepts that participants raised repeatedly and deemed to be relevant to ethics but were not always in relation to a specific ethics topic. These included: loyalty to group and role of the resident (in the healthcare system/hospital/hierarchy). We asked ourselves: 1) How are these concepts related to ethics and ethics curricula? 2) How are these concepts related to each other (if they are)? and 3) How are these concepts related to the ethics topics for which there were divergences (if they are)?
To begin to address these questions, we hosted a focus group discussion at each site with resident volunteers (recruited via open invitation) to explore and refine these broad concepts. This exercise also enabled us to confirm or refute what we had learned about divergences between curricula for certain ethics topics. Because this focus group exercise was confirmatory only, we did not analyze them directly but instead, used the results to inform the second phase of our analysis of the individual interviews.
In phase two of our analytic process, each of the two analysts independently reviewed the transcripts and grouped passages according to the two broad concepts when applicable. In an iterative fashion, working between the analysts' groupings and team discussion, we clarified two high-level concepts that serve as a common thread to four specific ethics topics that emerged from phase one. This iterative process led us to rename the first concept 'Group Membership' because in our data, loyalty was only one aspect of group belonging. We renamed the second concept 'Role Ambiguity: Student versus Practitioner' because in their discussion of role, residents focused primarily on this tension. At this stage, we conceptualized the link between the ethics topics, the two higher order concepts, and the larger theme of 'Identity' presented in the Results and Discussion sections.

Results
In this section, we describe divergent messages between the formal, informal, and hidden curricula for four ethics topics of importance in one or both programs that relate specifically to PIF: 1) career planning/pathway, 2) 'locker room talk', 3) applying mental health legislation, and 4) prioritization of clinical service or training and education (see Table 1). We identified the Page 83 content of the formal curriculum through our document review as well as what was described to us by our participants. We identified the content of the informal and hidden curricula through our participant interviews. How residents reacted in the face of the divergences between curricula was mediated by two aspects of professional identity formation (PIF). One of these aspects is group membership (i.e., resident, doctor, psychiatrist). The other is residents' experience of their ambiguous identity as students, practitioners, or both. In what follows we will present group membership and ambiguous identity as they emerged in our data, followed by the ethics topics/divergences that illustrate their impact on residents' PIF. These findings are supported by direct quotations from 19 of our 48 participants (40%). Citations are indicated using the following system: Program 1 (P1) or Program 2 (P2), followed by KI (key informant) or R (resident), and then their project code number. Although several faculty and staff members participated as KIs in the study, all of the quoted participants were residents at the time of the study except P1-KI1, KI10, and P2-KI5 (indicated in the text following their citations).

Group membership
During postgraduate training, psychiatry residents' sense of belonging shifts between various groups. In our study, trainees described transiently belonging to various groups: 1) the resident group (vis à vis other learners such as medical students or fellows), 2) the doctor group (vis à vis other professionals), 3) the psychiatrist group (vis à vis other specialists), and 4) the mental health group (vis à vis other domains of practice). Participants affirmed that the resident group wasand ought to be their primary affiliation and afforded a certain degree of reassurance and/or safety: …the people that are having the most difficulty are the people that separate themselves from the other residents. And, then they find themselves on their own, and they perpetuate their own problems, and they don't have any resources to ask for help… (P2-R5) Recently there was an issue with the postgrad director wanting to come [to a closed door residents' meeting] and discuss really openly and get feedback from residents….I think that the meeting would not be a great place for him to do that because it's important that residents have their own place where they can talk about things openly. (P1-KI4) As they move between hospitals, rotations and training years, psychiatry residents leave certain groups and move into others. For example, while working in the emergency room, residents in P2 feel a sense of belonging to the emergency room group when interacting with their colleagues in psychiatry despite sharing a specialty. Residents may belong to multiple groups at once depending on what tasks they are engaged in causing group membership to shift and/overlap in relatively short periods of time.
Our interview data show that two factors are important in establishing group membership at a given time: loyalty and conformity. Loyalty is a feeling that emerges in safe circumstances which is often the case within one's primary group affiliation. Loyalty may have positive aspects as illustrated by the citation below: There's a sense that there's a certain respect for your colleague as part of this profession. This is a special profession. It's a privilege of being part of this profession. And our responsibility for us is to look after each other… (P1-KI10; faculty) In contrast, loyalty may also have negative aspects. In a lengthy segment that was difficult to reproduce for the purposes of this paper, a P1 participant described a supervisor who threatened to give residents poor evaluations when they had to be absent from their clinical rotations in order to attend obligatory, residency activities. The participant explained that although those responsible for postgraduate education were aware of the problem, the behaviour continued. The participant speculates, "…I think within some hospitals there's definitely a culture where colleagues kind of protect one another…." (P1-R8).
Residents belong to the resident group but aspire to become members of the psychiatrist group. The psychiatrist group with which residents have the most contact is the faculty group in their university departments 3 . Although residents are part of the university department, they are not members of this group. Conformity arises in circumstances where group membership has not yet been established (as is the case for residents with respect to the 'faculty' group) and one is still learning group norms. However, when membership in the faculty group is fully established, the divergent views held by certain members may be tolerated by the group.
I think there's maybe a little more scope for faculty to be mavericks. So I think if you're a big researcher, and you bring in lots of research dollars to your hospital and you're a maverick on an issue that will be tolerated to a greater extent… (P1-KI1; faculty) Departmental norms reflect those of the faculty group. Though residents participate in department life they are not department members. As a result, there are repercussions of resident non-conformity to these norms. In P1, the postgraduate training program had just been through the accreditation process required by the RCPSC. Faculty treated the accreditor's critical comments as evidence that residents violated the norms that faculty expected of them: There was something interesting that happened last year with the accreditation. What I heard had happened was there was awful lot of blaming within the department when that [critical] accreditation report came. That the residents were bitching too much. And that this poor accreditation was the residents' fault. (P1-KI5) Although officially, everyone's like oh no no, no we're not blaming the residents, we thank them. Unofficially, behind closed doors, two different staff psychiatrists from two different hospitals have told me, 'Ok off the record, we blame the residents.' (P1-KI9)

Career planning
There were two specific ethics topics that arose in both programs and which illustrate how the process of group membership influences, and is influenced by, the messages from the hidden curriculum. Both of these topics were emergent, that is, they were spontaneously raised by participants rather than probed by the interviewers. The first relates to career planning or choice of career path.
In both programs, there was no formal ethics curriculum concerning career pathways. The divergence existed between the informal and hidden curricula. The message from the IC is that there are multiple career options for psychiatrists upon completing their training, all of which are equally valuable. However, in P1, the message from the HC is that the best career path is to become a researcher. Some participants thought that career planning favouring membership in a particular group is an ethical choice rather than merely a personal choice.
In a series of reflective memos, the P1 resident co-investigator explained that choice of career path was related to ethics in P1 in the following manner. Being a researcher was more highly valued than being a clinician or teacher. Furthermore, only certain kinds of research were valued, namely randomized controlled trials often done with pharmaceutical industry support and neuroimaging or basic science research. The resident went on to explain that this kind of research located the problem at the level of the individual's biology rather than being person focused. It did not take into account the social, political and economic contexts in which people live. The inattention to such factors is an ethical choice that becomes systematized through an educational system designed to produce researchers who are individually rather than socially-oriented. In addition, caring for suffering individuals is an inferior task left to less talented psychiatrists.
There's always comments about, "in the community you never know what happens"…and the community seems like this cesspool of horribly incompetent people. And there's such an idealization of a research scientist in academic center… that I think it makes people feel inferior for not aspiring to do that…. I'm not sure how ethically sound that is given most people go to medical school to be healers or to help people or because they're interested in people and they want to work clinically. (P1-R1) In P2, there were two distinct psychiatrist groups: university-based and community-based; the term 'community-based' was a proxy for psychiatrists who were paid entirely on a fee-for service basis through the provincial health insurance scheme. The message from the HC is that regardless of which career path one chooses, one must be loyal to that group. Both examples illustrate that addressing the HC requires the residents to determine their identity, to which group they are 'aligned' which then helps to structure their future choices.

Locker room talk
A second ethical issue raised by participants in both programs was what one P1 participant called 'locker room talk' which means "…talking about patients in a very irreverent way." (P1-R4) Participant P1-R4 expressed shock about how prevalent locker room talk was and in the passage below explained that they received divergent messages about the ethical acceptability of this kind of conversation: Sometimes it's just the black humour that we need to get through the situations and I actually think it's quite acceptable, and sometimes it verges on a little cruel, the way we talk about different patients. And certainly in our formal teaching we're not talking about patients in that kind of disparaging way. (P1-R4) As other participants pointed out, engaging in locker room talk was something one did to belong to the group.
I think that trainees then understand…it's okay to mock patients with your other colleagues. So they learn ways of being a team, um, and that if you sort of look a little bit stiff about what they just said then you're kind of a stick in the mud in the team.…I think that when I've heard comments like that, I don't launch into a lecture. I think if somebody were to react in that way it certainly would be perceived as a stick in the mud. I never have, but there are times that I will be not as obviously laughing along or may even have a question about it. And there are times that I've caught things coming out my mouth that I've just been like 'oh my God', I'm totally taking on the culture of this place you know? (P1-R12) Being part of the team meant accepting locker room talk. Participating in or accepting locker room talk may reinforce group membership particularly when it is reserved for group members.
Behind closed doors amongst residents we definitely talk about the nursing staff that we have difficulty with, um, in a more derogatory way. The attendings are more careful but I have seen them talk about nursing staff amongst themselves. But we're less likely to talk to the attendings about the nursing staff or the other staff in a derogatory way. (P2-R2) The presence of outsiders was sometimes unavoidable in which case they must be educated about group norms as in the following example: Often what comes up is in terms of managing our countertransference with, typical groups of patients, so borderline patients…people especially when they're sort of venting and decompressing will be less than empathic and focus more on how pissed off we can get. And that's usually when residents or staff will make pretty negative comments about a patient but I think that if you take them into context it's not such a big deal. However there is a responsibility that if you've got more junior residents or more junior students, that they know to take it in the context and that the context is apparent to everybody because if it isn't then that's unethical because then that's what they're going to see is 'Wow! I can call her a bitch behind her back and that's fine' when it's not, it's more sort of processing of counter-transference, whereas you still have to treat people with respect. (P1-R16)

Role Ambiguity: Student versus Practitioner
A second aspect of PIF we wish to highlight is the constant negotiation residents must undertake because they straddle the roles of student and practitioner. As a matter of fact, residents are simultaneously students and functioning professionals. They have a restricted license to practice medicine as well as requirements to maintain their own professional malpractice insurance.
Program documents indicate that in P2, they are explicitly recognized as both students and practitioners, whereas in P1, their status is unclear. Residents are considered students for certain purposes (e.g., access to the library) and not others (e.g., access to the university complaints' processes). In practice, the difference between the student and practitioner identities of psychiatry residents is more than an administrative distinction. It has an impact on the ethical responsibilities that residents are expected to uphold. Those responsibilities are reflected in comments from P1-R16: To me to be ethical is to make the just decision. It's getting at what are the principles and ultimately acting justly in terms of acting professionally in terms of upholding the standards of your profession and then your personal, moral and ethical standards, you know, living up to that. So if you're in a position where you see that a patient is receiving sub-standard care because of 'x' reason, that you have the knowledge that number one that you feel something is wrong but also the willingness to engage with it to try to make some sort of a change and not just sort of sit back, medical student style and sort of just sit there and stew and feel uncomfortable but to actually go and do something about it.
However, even when acting as a professional, there are limits to what kinds of actions are acceptable.
I do sense this, maybe I'm wrong, but that one can't be, for lack of a better word, a 'shit disturber'. There's a place to be critical but you can't be too critical or perceived as too difficult or too dissident. In some ways to be professional and to be perceived as professional is to carry the party line to a certain degree. (P1-R4) Residents try to decide whether to behave as a student or whether to assume the role of a practitioner but their possibilities for action are also influenced by the norms of the workplace. In P1, the workplace culture is rules-driven, "You know, saying appropriate things, maybe not divulging too much about themselves, really keeping the boundaries between supervisor and resident pretty clear." (P1-KI7) By contrast, in P2 the workplace culture is relationship-driven: …for most of them there is a fairly good sense of collegiality between residents and teachers. That line between if you're a teacher your boundary is here and you're the student, it's not quite as fixed. And in my mind, I agree because how do you go from one day having all these boundaries to all of a sudden you get that certificate on the wall and now, we're instant colleagues. Relationships don't change like that… (P2-KI5; faculty) Indeed, collegiality was of central importance in P2 because there is very little out-migration, meaning that residents who train there tend to remain afterwards to work as psychiatrists. In this context, fostering sustainable relationships that can last for decades becomes a priority. As the second passage indicates, collegiality is also essential in order to deliver high quality care.
There is a push for us to, to that collegiality part, to be able to develop a sense of identity with the faculty knowing that many of us are going to be working with them as colleagues after our residency. So I know there's a number of psychiatrists that encourage us to act with them like we are colleagues already and not think of ourselves as residents, you know kind of under them. Whether it is going out for drinks with them, or they encourage us to confidently, when we're discussing cases, speak to them as a colleague, and not be afraid to speak our mind. (P2-KI4) Yes, it's [collegiality] very important actually. Even getting your things done or getting -you can pull some strings if you have a good collegiality relationship with your, actually your own department residents, with other department residents too. So we went through the PGY1, we rotated through all the rotations and I have good friends in almost every subspecialty, and it's always easier to get a hold of them when I need someone. It's easier for them to get a hold of me when they need someone or talking to my attending. That's very important, and again for your patient care it's very, very important. (P2-R3) In our data, the question of whether to adhere or not to the hidden curriculum in light of their ambiguous identity arose for residents in both programs. However, this question arose for different ethics topics in each program. Two examples are mentioned below.

Interpreting mental health legislation (P1)
P1-KI4 explains, "…residents are concerned about figuring out our own values… for example, learning how we're going to navigate the system around certification and legal issues." Residents are expected to exercise their own professional judgment regarding involuntary hospitalization. At the same time, they are being evaluated on their ability to do so by a supervisor or a medical hierarchy whose judgment may differ.
I've seen cases where staff will certify a patient and [the patient] is seen as able to contest so suddenly they're voluntary and I think if that was the case that you're going to make them voluntary anyways knowing that you're going to lose the Review Panel hearing, I find that unethical…I've done it. But it leaves me feeling after the experience thinking about it and why I did it. (P1-R11) Page 87 Professionals I was working with misused the mental health act. They misused the legislation, either because they didn't understand it or because they did it on purpose in order to intervene with a patient who was refusing an intervention, and to justify it under the basis of mental health legislation….Anyway, it was clear that there were some serious legal issues coming out of this case. My supervisor and I we both saw it, we talked about it, we tried to pursue it in the hospital, and the message that I got was -and I'm not paraphrasing now -"If I were you, I would not pursue this any further." It was said in a threatening manner, the intent was clear. (P1-KI1; faculty) 4 In both examples the participants were aware that they could and ought to exercise autonomous judgment on how to interpret mental health legislation. At the same time, they experienced and ultimately responded to pressure to carry out orders or respect hierarchy, which is part of being a student in the rules-based culture of the workplace in P1.

Prioritization of service or education (P2)
During their training, residents perform essential clinical tasks within the healthcare system. In P2, some participants questioned the ethics of faculty according greater weight to this service (particularly if it is personally beneficial to faculty) at the expense of resident education and training.
Another issue I think is residents feeling like we're more of a tool to help the faculty to make money sometimes. And not getting any education in return, that we're more of a service sometimes. There's a pressure for us to, when we're at clinics, to have a full clinic and see all of our patients very quickly even without consulting with our psychiatrists 'til the end of the day. So we would go make management plans and discharge patients without even talking to the psychiatrist we're working with. When things like that happen we feel as though, 'wow', 'cause the psychiatrist gets the bill for everybody we see, just like they saw them. They [the psychiatrists] make double the amount that day, and we don't necessarily feel as though we really got any education out of it at the time. I mean seeing patients is educational, and I learned a lot from just seeing patients, but I'm comfortable seeing patients on my own without having to necessarily discuss it with my psychiatrist. (P2 KI4) In this example, the resident feels pressure to take on the role of practitioner rather than student but is also ambivalent about the extent to which they need to be in a student role. Having determined that s/he is at ease acting as an independent practitioner, the participant can then face clinical practice.

Discussion
In our study of the HC in two Canadian postgraduate training programs in psychiatry we found that when there are divergences in the messages from the formal, informal, and hidden curricula, residents' attempts to resolve these conflicting messages are mediated by professional identify formation (PIF). Our analysis pointed to two sub-processes of PIF: group membership and role ambiguity. These sub-processes influenced how residents responded to conflicting messages between curricula. Further, residents' actions in responding to these conflicts themselves had an impact on PIF. Thus, we argue that residents' responses to the HC are in an iterative relationship with their evolving professional identities. In what follows, we discuss some of the key themes emerging from the literature concerning professional identity formation in medical education and postgraduate training more specifically. This literature develops the concept of PIF largely on theoretical grounds (1,27,40). We situate our empirical findings in relation to these theoretical discussions and identify those aspects of PIF which were supported by our findings.
PIF in medicine is characterized as the process of transformation of layperson to physician (1,27,28,(37)(38)(39)41,42). The expression is derived in part from the concept of 'formation' in clergy training, which is defined as a "process that prepares an individual to serve a spiritual calling; [including] engagement in service, reflection on experience, growth in self-knowledge and expertise, intense mentoring, and attention to one's inner life." PIF in medicine is considered to be a secular form of this process, "encompassing the personal and moral development of learners within the context of their evolving professional identities; its central task [involving] integration of personal values with the core values of medicine." (42). Medical education is thus framed as "a process of becoming a professional, which involves not only the achievement of competence in core domains but also the deepening of commitment to excellence, humanism, and altruism." (31). Part of PIF then, is the cultivation of ethical virtue and values during the training years. Phillips and Delgarno point out that this aspect of postgraduate training may be fraught with pitfalls. In their study of first year residents they found that, "as a learner progresses through medical training, he or she experiences a tension between professionalism ('being an ethical, compassionate and virtuous person and doing medicine in a moral and competent manner') and professionalization ('the process of entry into the profession of medicine, of identity formation via socialization and absorbing values that may be, but are not necessarily in keeping with professionalism')." Their point resonates with what we have noted, that divergences between formal and hidden curricula create this tension (43).
PIF has been conceptualized from a developmental standpoint in which sound ethical decision making (among other capacities) is the last and most desirable stage of development. For example, Kegan's stages of psychological development have been evoked to equate stage four as the autonomous professional self who is "capable of authoring his or her own principles and values within the professional practice context," having internalized and personalized the values and expectations of the profession (35,37). Socialization is the process through which a trainee moves through these various stages (27,30,32,33,35,36,42,(44)(45)(46)(47). Holden and colleagues point out that through socially-situated learning, which is typical of medical education, individuals come to experience themselves as accepted and competent in the 'community of practice,' and can then assume their professional identities (37). Environmental and contextual factors, the presence of role models, and inherent stressors of medical education are all thought to influence the developmental trajectory (37). Few authors refer explicitly to the impact of the HC on PIF, although it is alluded to by Coulehan and Williams (48) as well as Kelly and Mullan (49).
Our data provide empirical support to the theorized relationship between PIF and ethical decision-making, through the discrepancies between the FC, IC and HC. The residents in our study were confronted with ethical problems both in their clinical tasks and in their work relationships. The formal, informal and hidden curricula sometimes diverged on how such problems ought to be handled. These residents also responded to such ethical problems by appealing to their identities (as members of certain groups; as students or practitioners) and the choices they made further reinforced those identities. Thus, the present analysis also provides support to the arguments made by some authors, on conceptual grounds, that PIF is an iterative process rather than only an evolutive one (41). In other words, the capacity to respond to ethical problems is not only the end result of PIF but itself contributes towards PIF. As Burford states, professional identity "may be viewed in terms of selfcategorization rather than simply attainment…" (50).
We view group membership and the tension between the student and practitioner roles as aspects of the socialization process in postgraduate medical education. Burford argues that social identity theory can be applied to medical training (professional identity being an instance of social identity) and can further explain potential group membership dynamics at play (for example, how group norms can maintain observable 'unprofessional behaviours') (50). Social identity, or "an individual's self-concept in relation to his or her membership of social groups" is an active process of self-categorization containing a motivational component of a "positive distinctiveness on a valued dimension for one's own group (the in-group), compared with other groups (out-groups)". Moreover, multiple identities are accessible to individuals given that there are multiple groups in any given situation. Our analysis supports this argument and extends it by showing that social identity (as exemplified by the group membership sub-process) affects and is affected by the way in which residents' respond to the hidden curriculum. For example, locker room talk is something that one does to show one's belonging to a group (not being 'a stick in the mud' in the team), but also reinforces one's belonging when this behaviour is not known or understood by members of an out-group (such as junior residents, medical and nursing students).
Finally, our data support the argument made by Hoop who explores psychiatry residents' conflicting identities, or dual roles, between physician and learner, physician and supervisee, and physician and employee of a training institution (51). Hoop's point is that ambiguous identity has ethical implications. Our participants' experiences illustrate how options for action in ethically challenging situations can be influenced by their dual or ambiguous identities. For example, raising concerns about misuse of the mental health act permitted the resident to act as a practitioner in defence of certain ethical principle and legal requirements. Dropping the issue when instructed to do so forced the resident to show deference as a student.

Conclusions
This paper makes two important contributions to the literature concerning professional identity formation and ethics education offering, as it does, one of the few empirical descriptions of the hidden curriculum in ethics in postgraduate medicine. First, we described the hidden ethics curriculum in action and in interaction with the complex process of professional identity formation. We provided specific examples of divergences between the formal, informal, and hidden curricula which, although obtained in the context of psychiatry training, may be common to other specialties. Second, our results showed how residents' decisions about whether or not to adhere to the hidden curriculum were mediated by two PIF subprocesses: group membership and role ambiguity, i.e., student versus practitioner. This evidence helps to substantiate the theoretical links made between PIF and individual ethical development/decision-making in the literature as well as describe the mechanisms by which PIF and ethical development/decision-making are related. We conclude that PIF and ethical development/decision-making are in an iterative relationship during residency training rather than an evolutive one.
Postgraduate educators have been called upon to address counterproductive messages from the hidden curriculum. Because the content of the hidden curriculum is taken up and integrated into the process of professional identity formation, our analysis suggests that educators will not be able to counteract the hidden curriculum merely by adding more formal content. They must also be attentive to the processes, some of which we describe, that facilitate its uptake.
Reviewer evaluations are given serious consideration by the editors and authors in the preparation of manuscripts for publication. Nonetheless, being named as a reviewer does not necessarily denote approval of a manuscript; the editors of Canadian Journal of Bioethics take full responsibility for final acceptance and publication of an article. The editors follow the recommendations and procedures outlined in the COPE Code of Conduct and Best Practice Guidelines for Journal Editors. Specifically, the editors will work to ensure the highest ethical standards of publication, including: the identification and management of conflicts of interest (for editors and for authors), the fair evaluation of manuscripts, and the publication of manuscripts that meet the journal's standards of excellence.