Advanced Search

Regular Article | December 01, 2000

The Impact of Impaired Supervisors on Residents

Karine J. Igartua, M.D., C.M.

Academic Psychiatry 2000;24:188-194. 10.1176/appi.ap.24.4.188

View Article Information

Impaired SupervisorsAlcoholism

Dr. Igartua is a psychiatrist at the McGill University Health Centre, Department of Psychiatry. Address reprint requests to Dr. Igartua, McGill University Health Centre, Montreal General Hospital Site, Department of Psychiatry, 1650 Cedar, Montreal, Québec, H3G 1A4, Canada. e-mail: Kigart@po- box.mcgill.ca

Article

Tables

References

text A A A

Abstract

The author estimated the proportion of Canadian psychiatry residents who had ever worked with impaired supervisors and explored the residents' reactions. A 22-item questionnaire was distributed to all 600 Canadian psychiatry residents; a total of 229 completed surveys were returned. Seven percent of surveyed residents reported working with an impaired supervisor. Female and senior residents were more likely to report such an experience. The most common reactions to working with supervisors perceived to be impaired were alterations in their work patterns (e.g., taking on more responsibilities, working without supervision), anger, conflict, and loss. Forty percent of residents in this situation recounted symptoms of depression and anxiety. Supervisors who are perceived as impaired seem to have a large impact, both professionally and personally, on the residents they supervise. Training programs need to actively address impairment in psychiatrists affiliated with their institution in order to aid both their colleague and the residents that they supervise.

Figures in this Article

The American Medical Council on Mental Health has defined physician impairment as "the inability to practice medicine adequately by reason of physical or mental illness, including alcohol or drug dependence." This has been distinguished from physician incompetence, which refers to "the inability to provide sound medical care secondary to deficient knowledge, poor judgment, or substandard clinical skills" (1). Impairment is usually examined in the context of the doctor—patient relationship, yet many physicians are also teachers. Like the doctor—patient relationship, the supervisor—resident relationship is a fiduciary relationship in which the physician, because of his medical knowledge and clinical judgment, is invested with trust. In psychiatry, because of the nature of the work, the supervisory relationship may be even more meaningful than in other specialities.

Conflicting data have emerged regarding the prevalence of impairment among physicians (2). A conservative estimate would be that they are at least as likely as the general population to be impaired by alcoholism, and more likely to abuse prescription drugs (2). There is some evidence that physicians may be at increased risk of depression and suicide (3); psychiatry may be among the specialities at highest risk of mental impairment (2,4,5). What impact an impaired psychiatrist has on the residents he or she supervises and whether there are consequences for the resident's professional and personal development is still unknown.

A wide literature search revealed a handful of articles on the impaired resident and only one addressing the resident's responses to impaired physicians (1). In this study, internal medicine house officers were given a case scenario of an impaired physician and asked what actions they would take. Seventy-two percent indicated they would tell their chief resident. The authors acknowledge that the actions actually taken in real-life situations may differ because of "institutional or psychological barriers," which they do not explore. None of the articles dealt with the psychological or professional effects of having an impaired physician as a supervisor. This is unmistakably a delicate subject to explore because there are concerns for confidentiality and loyalty to teaching institutions and supervisors.

In order to conceptualize a model to examine the potential effects of an impaired supervisor on a resident, we sought literature on other fiduciary relationships (therapist—patient, teacher—student, boss— employee). None of these analogies is perfect; however each of them offers similarities to the supervisor—resident relationship. In the boss—employee relationship, the angles of performance and evaluation are most prominent. In the teacher—student relationship, it is learning that is in the forefront. In the therapist—patient relationship, nurturing and growth are the key features.

The literature on both the teacher—student and boss—employee dyads barely touched on impairment. Only four articles were found, none of which addresses the effect of impairment on the student or employee. The most fruitful area of literature dealt with therapist impairment. A dozen articles were identified; half of which addressed patients' reactions to therapists' medical or mental illnesses. Factors that affected patients' reactions included manifestations of illness (6), therapist's manner of handling the illness (6,7), therapist's vulnerability (6,7) and his or her absence and unplanned termination (6—9). Patients' reactions included denial, numbness, abandonment, distancing, anger, depression, hopelessness, panic, compassion, and concern (8,9).

Psychotherapy journals are peppered with articles debating the teach—treat controversy in psychotherapy supervision (10—12). The dilemma about how personal supervision can become illustrates that the supervisory relationship may at times best be seen as a therapist—patient relationship, whereas, at other times, it best fits a teacher—student model. This seems quite intuitive when examining the process of supervision of a psychotherapy trainee, but the same dual relationship exists in the supervision of graduate students. For instance, Hockey (13) states that the two principal roles of the Ph.D. supervisor are to provide intellectual expertise and to counsel and bolster students' confidence and morale.

It is common knowledge that the therapist—patient relationship often replays aspects of a parent— child dyad. Although perhaps less obvious, the analogy between the supervisor—trainee dyad and the parent—child dyad also exists in the literature. Plaut likens the role of teacher to that of parent or therapist because of the obligation to foster independence and autonomy while providing guidance and nurturance (14).

There has been considerable literature on the impact of parental mental illness on children and adolescents. In broad terms, parental mental illness has been associated with emotional distress (e.g., isolation, conflicts over loyalties, anger, guilt; 15), behavioral disturbances (e.g., rebellion, withdrawal), psychopathology (e.g., depression), disturbances in school functioning (16), or, at the opposite end of the spectrum, hyperfunctioning and hypermaturity (17).

Using the data on the impaired therapist and the impaired parent, one can begin to formulate hypotheses as to the effect of supervisor impairment on residents. For instance, emotional reactions may take the form of denial, anger, guilt, concern, or even depressive and anxious symptoms. Residents' "disturbances in school functioning" may manifest as deteriorating work performance or taking on fewer patients or academic activities. On the other hand, a resident may become "hyperfunctionning" by taking on more patients or academic activities, or become "hypermature" by working independently without supervision, taking over medical-student education, and feeling more confident in their clinical abilities. Impairment in a non-supervising psychiatrist may also have an impact on residents, be it by identification or empathy. Resident reactions, though, are likely to be less intense because of the lack of a fiduciary relationship.

The objectives of this study are to estimate the proportion of Canadian residents who worked with impaired supervisors and to determine whether these contacts varied with age, gender, or level of training; and to explore the residents' actual reactions to these supervisors and whether these reactions were modulated by age, gender, or level of training. In order to examine the interaction between impairment and supervision, the effect of perceived impairment in a supervisor was compared to the effect of perceived impairment in a non-supervising psychiatrist.

METHODS

Measures

A two-page survey was constructed for this study. So as to increase participation, it was specifically designed to take very little time to answer. Residents who had neither been supervised by nor had encountered an impaired psychiatrist had five demographic questions to answer. Those who had had such an experience were asked to rate 22 statements on a 3-point Likert Scale (0=not at all, 1=somewhat, 2=very much) in addition to the five demographics questions.

In order to construct the survey, semistructured interviews were conducted with four psychiatrists and two residents who had had an impaired supervisor during their training. Fourteen statements were formulated on the basis of the themes elicited. Because of concerns for confidentiality and risk of supervisor identification, no details were sought about the type of impairment or the type and duration of the supervisory relationship.

Supervisor impairment was based on residents' reports. We do not know how the residents determined the presence of impairment (e.g., their own clinical judgment or outside sources of information). The AMA definitions (1) were provided at the top of the survey so as to assist residents in making a judgment about the presence of impairment and avoid confusing impairment with incompetence.

A pilot version of the survey including 14 statements and an open-ended section was tested on 27 residents in one university. On the basis of their responses, eight more items were added before national distribution of the questionnaire.

Subjects

All residents enrolled in Canadian psychiatric programs in the 1996—97 academic year were targeted for this survey. In Canada, there are 16 programs, ranging in size from 11 to 120 residents, for a national total of 600 residents.

The survey was distributed to all Canadian psychiatry residents with the help of each program's COPE resident representative. (COPE is the national Committee On Postgraduate Education in psychiatry. It brings together the program directors of each Canadian program as well as one resident from each school.) The method of distribution varied from site to site: individually in smaller programs, during teaching seminars in mid-size programs, and via mailboxes in larger programs.

In total, 229 completed surveys were returned, yielding an overall response rate of 38%. Forty-three residents (19%) reported perceiving impairment. Seventeen residents (7%) had perceived impairment in a supervisor; 32 (14%) had perceived it in a psychiatrist who was not their supervisor. Of these 43 residents, 6 residents had had both experiences, and 4 residents had had the same experience twice, thus accounting for a total of 53 experiences with impairment.

Although responders were representative of target population for available demographic data (Table 1), residents who had perceived impairment in supervisors were more often female and most often in the fourth year of residency at the time of survey completion. However, they were in their first 3 years of residency when they worked with the impaired supervisor. They represented 9 of the 16 schools surveyed.

Analysis

Residents' reactions to perceived impairment in the context of supervision were tabulated and compared with residents' reactions in a non-supervisory context by means of chi-square analysis. In order to simplify presentation of the data in this article, residents are categorized as having endorsed a statement if they selected either 1 (somewhat) or 2 (very much).

Point prevalence of perceived impaired supervision was tallied. The overall risk of perceiving impaired supervision during residency was computed by Kaplan-Meier survival analysis. Age, gender, and level of training of those having worked with impaired supervisors were compared with those of residents without such experience by means of chi- square and t-tests.

The effect of gender, age, and year of residency in modulating the experience of impaired supervision was examined, although the total number of experiences was too small to glean significant patterns.

Correction for multiple comparisons (Bonferroni correction) set the significance at P<0.0023 in order to maintain an overall Type 1 error at P<0.05. Similarly, trends toward significance were noted when P<0.02.

RESULTS

None of the residents were being supervised by impaired psychiatrists at the time the survey was filled out. However, the proportion of residents in the sample having encountered an impaired supervisor at some point in their training was 17/229, or 7.4%. The Kaplan-Meier statistic determined that the risk of having worked with an impaired supervisor was 11% by the end of 5 years of residency, with most of the risk occurring during the first 3 years. Table 2 highlights the reactions of residents to impairment both when perceived in supervisors and in non-supervising psychiatrists. We present percentages of respondents endorsing each statement with either "somewhat" or "very much."

The most common reactions to supervisors perceived to be impaired were the following: working independently, without supervision (94%); assuming primary responsibility for patients (88%); feeling that other staff were too passive in dealing with the impairment (88%); needing to share feelings (87%); feeling conflicted about reporting (81%); and turning to others for supervision (81%).

Comparing residents' reactions to impaired supervisors vs. impaired non-supervising psychiatrists, the two groups differed in terms of reported behaviors directly relating to the lack of appropriate supervision (working independently [94% vs. 43%]), seeking alternate supervision [81% vs. 39%], and taking on more responsibilities [71% vs. 24%]), but also on some statements reflecting more personal effects (developing symptoms of anxiety or depression [38% vs. 7%], re-evaluating one's own lifestyle [56% vs. 31%], and loss of role-model/mentor [75% vs. 29%]). Overall, impairment had a greater impact when it occurred in the context of a supervisory relationship; the other statements did not reach a difference that was statistically significant.

Impairment in a non-supervisory relationship had much less of an impact, but it did give rise to some similar feelings. Both groups of residents reported feeling angry (65% with supervisors and 55% for non-supervisors) at the impaired psychiatrist and conflicted about reporting him or her (81% and 57%, respectively). Some residents did report the psychiatrist, and this was more likely to occur if the psychiatrist was supervising the resident (69% vs. 24%). Both groups admitted needing to share their feelings (87% and 63%) and feeling that other staff were too passive in dealing with the impairment (88% and 73%).

By and large, residents did not endorse potential positive effects of having an impaired supervisor. However, residents with impaired supervisors were more likely to report increased confidence in their clinical abilities (71% vs. 29%).

DISCUSSION

The results of this survey suggest that working with an impaired supervisor is not a rare occurrence. In our sample, the risk of encountering a supervisor perceived as impaired during residency training was 11%. The low response rate (38%) makes it difficult to draw conclusions. However, if we presuppose that none of the nonresponders had worked with impaired supervisors, the risk of working with an impaired supervisor during residency training would be 4%. A reasonable estimate is that the incidence is at least 4% and probably higher.

This study also highlights the importance of supervision beyond the mere conveyance of knowledge. As residents go through their training, they look to their supervisors for guidance and nurturing. Younger residents, still in the process of forming their adult identity, may look to their supervisors for role- modeling for both personal and professional development. Junior residents may be quite dependent on their supervisors early on. As they gain experience and autonomy, the dyad moves from a fiduciary relationship to a peer relationship. The resident's professional identity is thereby formed.

Working with a supervisor perceived to be impaired is therefore not a benign experience, yet residents for the most part were able to cope with the situation by becoming prematurely independent: by working without supervision (94%), taking on more patients or academic activities (71%), and assuming primary responsibility for the patients on the service (88%). Although personal and professional growth may arise through adversity, the risk is that these residents may not have poor clinical practices corrected, may be delayed in achieving competence, or even suffer from psychological distress. In fact, nearly 40% of our sample reported developing significant symptoms of anxiety or depression.

In this survey, proportionately more women and junior residents worked with impaired supervisors. Junior residents may be at a greater risk than senior residents because they may have less knowledge about the different supervisors and less choice of whom they work with. By the fourth and fifth year, residents usually are doing electives and therefore, to a larger extent, are choosing their supervisors (a choice not usually made blindly).

Usually, the resident does not have any information about the existence of illness or substance abuse. What the resident encounters, then, are odd or inappropriate behaviors. The notion of impairment is therefore inferential and subject to the resident's judgment. Until one has had several good supervisors, one does not have a frame of reference with which to contrast the impaired supervisor. Diagnostic skills and experience with supervision are necessary to make sense of the behaviors observed.

It is interesting to note that a disproportionate number of senior residents reported working with impaired supervisors, even though all experiences occurred in the first 3 years of training. Residents may not perceive impairment as they are working with their supervisor but come to this conclusion only in retrospect. In this study, residents' perception of impairment, rather than impairment per se, was actually measured. Although there is likely a correlation between residents' perception and reality, the results would have been strengthen had an independent assessment of impairment been possible.

Gender also seems to influence one's risk. Although the gender ratio of the entire sample was 57%/43% female/male, 69% of the residents who reported having worked with an impaired supervisor were female. Future studies could explore such hypotheses as: Are women more likely to be assigned to, to discern, or to choose, impaired supervisors? Are there differences in the supervisory relationship with women that make them more aware of impairment? Are women who work with impaired supervisors more likely to report it?

Despite our designing the study to examine the effects of gender, age, and level of training on the reactions of residents to impaired supervisors, the sample size was too small to glean significant trends. Other factors not addressed that would be expected to modulate the experiences were the quality, intensity, and duration of the supervisory relationship, as well as the type and severity of impairment.

CONCLUSIONS AND RECOMMENDATIONS

Supervision is a professionally and personally meaningful process. It goes beyond the transmission of clinical skills and knowledge. Working with an impaired supervisor is not an innocuous experience. Unfortunately, it is also not rare.

In Canada, conducting this study with the assistance of COPE and presenting preliminary results back to this national committee raised awareness of the problem and prompted universities to adopt guidelines such as the ones below:

A rapid intervention will perhaps limit professional and personal damage for the impaired psychiatrist by prompting him or her to seek treatment and/or arrange for an alternative type of practice. Not only will it precipitate alternate supervisory arrangements for the residents, but it will also validate the residents' perceptions (and clinical observations) and provide positive role-modeling. When faculty actively address impairment in their colleagues, they teach residents that psychiatrists do take care of their own. They also show the public that we are a profession capable of self-regulation.

ACKNOWLEDGMENTS

The author acknowledges the 1996—97 COPE representatives for their help in distributing and collecting the survey. She also thanks her supervisors, Drs. Richard Montoro, Viviane Zicherman, and Pascale DesRosiers, without whom this project would not have been possible.

TABLE 1. Demographics of target population and sample

View Large | Download Slide(.ppt)
Add to My POL

TABLE 2. Percentage of residents endorsing reactions to impairment

View Large | Download Slide(.ppt)
Add to My POL

References

Reuben DB, Noble S: House officer responses to impaired physicians. JAMA 1990; 263:958-960[PubMed][CrossRef]

Centrella M: Physician addiction and impairment—current thinking: a review. J Addict Dis 1994; 13:91-105[PubMed][CrossRef]

Serry N, Ball JRB, Bloch S: Substance abuse among medical practitioners. Drug and Alcohol Review 1991; 10:331-338[PubMed][CrossRef]

Doyle BB: The impaired psychiatrist. Psychiatr Ann 1987; 17:760-763

Arboleda-Florez J: The mentally ill physician. Can J Psychiatry 1984; 29:55-59[PubMed]

Osterheld J, Buckman D: Four aspects of therapists' acute illness and injury that trigger transference reactions. Psychotherapy in Private Practice 1989; 7:41-53

Counselman E, Alonso A: The ill therapist: therapist reactions to personal illness and its impact on psychotherapy. Am J Psychotherapy 1993; 47:591-602

Simon J: A patient-therapist's reaction to her therapist's serious illness. Am J Psychotherapy 1990; 44:590-597

Reynolds J, Jennings G, Branson M: Patients' reactions to suicide of a psychotherapist. Suicide Life Threat Behav 1997; 27:176-181[PubMed]

Laverman L: The multi-level supervision model and the interplay between clinical supervision and psychotherapy. Clinical Supervisor 1994; 12:75-91

Aponte HJ: How personal can training get? J Marital Fam Ther 1994; 20:3-15[CrossRef]

Sarnat J: Supervision in relationship: resolving the teach- treat controversy in psychoanalytic supervision. Psychoanalytic Psychology 1992; 9:387-403[CrossRef]

Hockey J: Establishing boundaries: problems and solutions in managing the Ph.D. supervisor's role. Cambridge Journal of Education 1994: 24:293-305

Plaut SM: Boundary issues in teacher-student relationships. J Sex Marital Ther 1993; 19:210-219[PubMed][CrossRef]

Dunn B: Growing up with a psychotic mother: a retrospective study. Am J Orthopsychiatry 1993; 63:177-189[PubMed][CrossRef]

Beardslee WR, Bemporad J, Keller MB, et al: Children of parents with a major affective disorder: a review. Am J Psychiatry 1983; 140:825-832[PubMed]

Kaufman C, Grunebaum H, Cohler B, et al: Superkids: competent children of psychotic mothers. Am J Psychiatry 1979; 136:1398-1402[PubMed]