Psychotherapy training has long been a defining feature and core value of psychiatric education. In recent decades, the advent of manualized psychotherapies, increased emphasis on evidence-based practice, advances in neurobiology, the increasing availability of relatively safe and effective pharmacologic agents, managed care, and split treatment have forged a challenging landscape for training programs (1–5). Some educators have expressed concerns about the future utilization of psychotherapy by the next generation of psychiatrists (3, 6). Yet, applicants to psychiatric residencies express strong interest in and desire for psychotherapy training (7, 8).
Spurred by the diminishing role of psychodynamic psychotherapy in residency training, the Association for Academic Psychiatry (AAP) and the American Association of Directors of Psychiatric Residency Training (AADPRT) jointly considered the importance of psychodynamic psychotherapy in psychiatric education and developed a model curriculum (9). The Residency Review Committee for Psychiatry subsequently mandated the assessment of competence in five psychotherapy modalities: brief, cognitive-behavioral, combined psychotherapy and psychopharmacology, psychodynamic, and supportive (10). This mandate ignited considerable controversy (11–13). In 2007 the requirements were integrated into the overall “Patient Care” core competence and reduced to three competencies (10).
The challenges for educators in implementing the requirements are significant (14). Residents’ attitudes are likely to influence the success of strategies for teaching and evaluation (15–17), yet minimal research has focused on residents’ attitudes toward psychotherapy in general, their views of the quality of their teaching and supervision, or their self-perceived competence in various modalities (17–20).
We examined residents’ views of the quality and quantity of psychotherapy training, perceptions of programmatic support for training, and perceptions of barriers to training. Residents’ self-perceived competence in the five (pre-2007) mandated forms of psychotherapy was assessed, and associations between self-perceived competence and views of the training environment were evaluated. We predicted that self-perceived competence would be positively associated with increasing years of training and with more favorable views of the quality of and support for psychotherapy training.
+
Development of Survey Instrument and Recruitment of Participating Programs
An initial survey draft was sent to faculty at the University of California, San Diego (UCSD)—with expertise in psychotherapy training, educational research, and survey design—for comment on item content, wording, and scope. The questionnaire was revised several times. To solicit interest from potential collaborators at other residency programs, a description of the study and an invitation to participate was posted by the UCSD residency training director (SZ) on the AADPRT electronic mail listing. Residents from 21 programs responded to the initial solicitation, and 15 programs ultimately participated: UCSD (coordinating site); Case Western Reserve; Emory University; Mayo Clinic; Michigan State University, Kalamazoo; Maricopa Health Systems, Phoenix, Ariz.; St. Elizabeth’s Hospital, Washington, DC; SUNY at Buffalo; SUNY at Syracuse; University of California, Los Angeles; University of California, San Francisco; University of Kentucky; University of Oklahoma; University of Texas-Southwestern; and University of Wisconsin. A penultimate draft of the survey was sent to the collaborating sites, and final changes in response to their suggestions were made. (The final survey is available from the first author upon request.)
In addition to demographic questions, the majority of the survey was comprised of 36 Likert-scaled items (1=strongly disagree, 5=strongly agree). These items explored attitudes toward the following aspects of psychotherapy training: overall quality, resources, teaching and supervision, perceptions of support for psychotherapy training, self-perceived competence, identity as a psychotherapist, and future plans to study and practice psychotherapy. The survey also included four yes/no items about personal psychotherapy experience and access to personal psychotherapy at their program, and three questions asked about patient and supervisory contact hours. It concluded with an open-ended question asking residents to provide general comments about psychotherapy training at their institution. In this article, we present data from the sets of items related to perceptions of quality of training and supervision in psychotherapy, level of support for psychotherapy training, and self-rated competence in different modes of psychotherapy (the other domains will be discussed in additional articles).
An e-mail was sent to all psychiatric residents and fellows at each participating program, inviting them to complete an Internet-based survey. The e-mail addressed informed consent and confidentiality and contained a link to the survey. The survey was reviewed and approved (or exempted from review) by each program’s institutional review board. Up to three follow-up e-mails were sent by the local investigator to encourage responses. Data from all sites were collected and analyzed using SPSS Version 12.0.1 (2003) by the coordinating site (UCSD).
Several repeated measures multivariate analyses of variance (MANOVA) of the responses to 5-point rating scale items were conducted, with the consistent structure of item (within subjects repeated measures) × training level (postgraduate year [PGY] 1 versus PGY-2 versus PGY-3 versus PGYs 4 to 6, between subjects) × gender (between subjects). Each analysis addressed a conceptually related set of items as the repeated measures factor; items were reverse scaled as appropriate for consistent direction in analysis. Cohen’s d is reported as a standardized measure of effect size. Correlations among responses to attitude items and residents’ self-perceived competencies in psychotherapy were examined. By prior agreement, program-specific analyses were not conducted to avoid singling out programs.
Surveys were completed by 249 of 567 residents (43.9%). Table 1 shows participant characteristics. The mean age of respondents was 32 years old (range=25–48).
+
Views of Resources, Teaching, and Supervision of Psychotherapy Training
Only a small percentage of respondents (6.8%, n=17) believed that there was too much psychotherapy training in their program, while just over one-quarter (27.7%, n=69) believed that their program did not allocate sufficient time or resources for psychotherapy training. Almost one-quarter (23.3%, n=58) believed that it was difficult to find good psychotherapy patients. There were no significant differences by gender or training level in responses to items concerning resources.
A majority of respondents (59.6%, n=149) agreed that their program’s didactic curriculum helped them to become a better psychotherapist. In response to the statement “My residency training program provides high-quality training in psychotherapy,” just over one-half (56%, n=140) agreed, 28.8% (n=72) were neutral, and 15.2% (n=38) disagreed. The majority (69%, n=172) agreed that their programs teach evidence for and against effectiveness of psychotherapy, pharmacotherapy, and combined treatment. Almost 10% of respondents (9.3%, n=23) agreed that their program’s faculty were not good at teaching psychotherapy, while the majority (72.4%, n=181) agreed that the faculty in their program were well-trained psychotherapists. No significant gender or training level differences were detected in responses to the psychotherapy teaching items.
The majority of respondents (64.2%, n=158) agreed that they received excellent supervision for their psychotherapy cases, and only a small minority (2.4%, n=6) agreed that their supervisors had negatively influenced their interest in psychotherapy. Residents in PGY-2 or higher agreed more than PGY-1 residents that they received excellent supervision for psychotherapy and agreed less that their interest in psychotherapy had been negatively influenced by supervisors (training level main effect F=16.18, df=3, 235, p<0.0001, maximum Cohen’s d=1.12). Men agreed more that they received excellent supervision than did women (item × gender interaction F=18.65, df=1, 235, p<0.0001, d=0.34).
+
Perceptions of Support for Psychotherapy Training
Only a small minority of respondents (3.5%, n=9) agreed that there was a general negative attitude toward psychotherapy in their training program. Similarly, few residents (4.7%, n=12) agreed that their training director did not actively support residents in psychotherapy training. In distinction, nearly two-thirds (65.7%, n=167) agreed that their departmental leadership actively supported residents’ psychotherapy training, while 15.8% disagreed (n=26). A majority (62.7%, n=158) agreed that their fellow residents actively support one another in becoming psychotherapists. Responses were largely neutral (48.2%, n=122) regarding fellow trainees in other disciplines, with the remainder evenly distributed between agreement (26.1%, n=66) and disagreement (25.7%, n=65). There were no significant differences by training level or gender in responses to the items concerning support for psychotherapy training.
+
Self-Perceived Psychotherapy Competence
Residents indicated their agreement (1=strongly disagree, 3=neutral, 5=strongly agree) that they would rate their abilities in five forms of psychotherapy as competent, given their current level of training (Table 2). Residents were neutral to somewhat positive about their level of competence in cognitive behavior therapy, brief psychotherapy, and psychodynamic psychotherapy (means=3.30 to 3.44) and more positive about their abilities in combined psychopharmacology and psychotherapy and supportive psychotherapy (means=3.75 to 3.90; item main effect F=37.98, df=4, 233, p<0.0001, maximum Cohen’s d=0.67). PGY-3 or later residents reported greater overall self-perceived competence than did PGY-1 and PGY-2 residents (means=3.81 to 3.82 versus 3.17 to 3.40, maximum Cohen’s d=0.98; training level main effect F=14.00, df=3, 236, p<0.0001). Residents’ self-perceived competence in supportive psychotherapy increased most with level of training (means from 3.36 to 4.27, d=1.01), and self-perceived brief psychotherapy competence increased least (means from 3.11 to 3.51, d=0.45; item × training level interaction F=1.92, df=12, 617, p<0.03). Residents overall disagreed with the statement “I am less skilled in psychopharmacology than I am in psychotherapy” (mean 2.07, SD=0.87). Men and women did not differ in self-perceived abilities.
+
Relationship Between Psychotherapy Competencies and Training
Greater self-perceived competence in each of the five psychotherapy modalities was modestly associated with less agreement that the program does not allocate enough time and resources for psychotherapy training (r=−0.11 to −0.28, p<0.05 in four of five cases) and that faculty are not good at teaching psychotherapy (r=−0.16 to −0.25, p<0.05 in all cases). Similarly, greater self-perceived competence in all five modalities was associated with greater agreement that the didactic psychotherapy curriculum is helpful (r=0.17 to 0.25, p<0.05 in all cases), that the program provides high-quality training (r=0.17 to 0.29, p<0.05 in all cases), and that faculty are well-trained psychotherapists (r=0.11 to 0.20, p<0.05 in four of five cases). For all five modalities, greater self-perceived competence was also associated with stronger endorsement of receiving excellent supervision (r=0.13 to 0.25, p<0.05 in all cases). For four of five modalities, greater self-perceived competence was associated with less agreement that supervisors have negatively influenced respondents’ interest in psychotherapy (r=−0.11 to −0.27, p<0.05 in three of four cases). For all five modalities, greater self-perceived competence was associated with less agreement that the general attitude toward psychotherapy in the training program is negative (r=−0.12 to −0.25, p<0.05 in four of five cases), that the residency training director does not actively support residents to be outstanding psychotherapists (r=−0.12 to −0.27, p<0.05 in three of five cases), and that the respondent’s interest in psychotherapy has decreased since the beginning of training (r=−0.11 to −0.18 for four modalities, p<0.05 in three of four cases).
Some worry that psychotherapy training in residency is in jeopardy (21) at a time when psychiatrists are providing less psychotherapy (22). Yet data are sparse regarding psychiatric residents’ views of the quality of their psychotherapy training and how these views relate to residents’ self-perceived psychotherapy competence. Our results highlight the complexity of answering these important questions.
Survey respondents overall viewed their psychotherapy training positively. They nevertheless cited areas for improvement and barriers to excellent training. Although only a minority felt that insufficient time and resources were allocated to training, including appropriate psychotherapy patients, their open-ended responses seem particularly compelling:
“My program provides excellent psychotherapy training and resources, but because of our heavy service requirements in psychopharm [sic] clinics, we spend less time proportionally on psychotherapy than I would like.”
“Hope to have greater emphasis [on psychotherapy]. Too many medication clinics and too few therapy-based clinics.”
Others indicated that logistical issues affected their ability to conduct psychotherapy during training:
“There does not tend to be enough flexible clinic time to get and keep therapy patients.”
“Time for psychotherapy is mostly given lip service. If we want to pursue more psychotherapy supervision as part of our curriculum, there are significant bureaucratic barriers that effectively prevent this. Conversely, if one wishes to minimize or omit psychotherapy from their training, this is relatively easy to accomplish.”
The implication that some residents may graduate without adequate exposure to psychotherapy may alarm some educators. Moreover, beliefs that psychotherapy training is undermined by logistical hurdles underscore the importance of identifying such barriers.
Faculty members were viewed as good psychotherapists and clinical supervisors, but whether they were considered good psychotherapy teachers was less clear. This finding seems consistent with a survey of chief residents who were asked if faculty responsible for teaching and assessing competencies in psychotherapy were qualified to do so; a substantial minority (26%) responded “somewhat” or “not at all” (17). These findings may reflect the complexity of teaching multiple modalities of psychotherapy, regardless of the teacher’s competence as a therapist or supervisor.
Although only a minority of residents appeared not to feel supported by their departmental leadership, this finding suggests that there is room for improvement if the Residency Review Committee mandate of having all graduating residents acquire competence in psychotherapy is to be fulfilled. Perhaps some residents perceive a mixed message: on one hand, they must become competent multimodality psychotherapists; on the other hand, they do not feel fully supported by their department’s leadership (examples provided in the survey were chairperson, division chief, and service chief). In addition, the different views of support, more favorable for training directors and less so for departmental leadership, suggest a misalignment of priorities that may confuse residents and undermine psychotherapy training.
Could this misalignment represent a “hidden curriculum” in psychiatric training (23, 24)? The hidden curriculum consists of the “covert, inferred tasks, and the means to their mastery . . . rooted in the professors’ assumptions and values, the students’ expectations, and the social context” (25). For instance, tables of contents of prestigious psychiatric journals hint at the relative value placed on different types of psychiatric research. Although intensely discussed in medical education (23, 26), the “hidden psychiatric curriculum” has received little research attention for its possible effects on psychotherapy training or the emerging professional identities of trainees.
Across years of training and modes of therapy, residents perceived their own competence in neutral to slightly positive terms, and this self-perceived competence increased as they progressed through training. Whether this increase reflects a general increased sense of self-efficacy as a psychiatrist or greater confidence in specific skills cannot be determined. Self-perceived brief psychotherapy competence did not increase across training levels as much as self-perceived competence in other psychotherapies; this could reflect less emphasis, a more diffuse focus on brief forms of psychotherapy (e.g., different types of brief psychotherapy being taught at different programs), or a lack of clarity to residents about the meaning of “brief psychotherapy” in the context of the survey. We speculate that the relatively stronger self-perceived competence in cognitive behavior versus other forms of psychotherapy in the latter years of training may stem from greater availability of trained supervisors and/or institutional support. Another possibility is that residents perceive greater mastery of cognitive behavior therapy, which is more manualized, sooner than the less operationalized skills of empathy, dealing with countertransference, and so on—skills that require an attitude of ongoing inquiry and self-examination.
Higher self-perceived competence for all five modalities was associated with more favorable views of the quality of training, helpfulness of didactics, faculty’s teaching skills, and quality of supervision. Because these data were cross-sectional, it is unclear whether quality of training leads to higher self-perceived competence or whether those who feel more competent feel more favorable toward their training. The latter possibility could reflect some unmeasured variable(s) such as overall disposition or self-efficacy. Greater self-perceived competence could also be based on an assessment that one’s time investment must have produced positive results.
Men agreed more that they received excellent supervision than did women, which corroborates a previous study of rated videotaped supervision sessions (27). A replicated finding in the patient-physician communication literature is that women physicians’ communication style is more patient-centered than that of their male colleagues (28, 29). Though highly speculative, it is possible that the women who responded had different expectations of supervisors than did the men, which could account in part for women’s lower ratings. This would be consistent with Shanfield and colleagues’ (27) finding that women more readily acknowledged problems in their supervision, particularly regarding what they perceived as empathic failures.
Our study has several important limitations, foremost of which are the possibilities of sampling and response biases, although the overall 43% response rate is comparable with response rates in other physician surveys (30). Residents who responded possibly had more positive views on psychotherapy, and those with more negative views on or who were disinterested in psychotherapy may have elected not to participate. This possibility is illustrated by the fact that women were overrepresented among respondents, compared with U.S. psychiatric residents as a whole (31). Collectively, women residents may have a stronger interest in psychotherapy, although a gender difference in willingness to complete the survey could also explain the sample’s gender distribution. Participating programs also may have had residents more interested in psychotherapy training. However, responses spanned the full range of possible answers, suggesting that the respondents represented diverse viewpoints. We also lacked corroborating data on performance-based competence and therefore cannot comment on the relationship between self-perceptions and actual skills. Future studies should include multidimensional assessments of psychotherapy competency, including supervisor and patient evaluations, patient outcome data, qualitative measures, and quantifiable examinations (32).
The Residency Review Committee psychotherapy competencies mandate and the ensuing debate have exposed the complexities of teaching psychotherapy as well as the diverse views among psychiatry faculty. Our findings add fuel to the debate. It is striking that most residents believe their faculty are well trained in psychotherapy, slightly more than half believe that their program provides high quality training, and almost three-quarters believe that their leaders support psychotherapy training, given the substantial challenges in residency that make psychotherapy training difficult. However, our findings also point toward areas that require ongoing attention, such as facilitating residents’ finding and/or recognizing “good psychotherapy patients,” providing better supervision, and allocating more resources to psychotherapy training. Clarification of the role of the hidden curriculum on psychotherapy training, including possible mixed messages from departmental leaders regarding the value of psychotherapy training, should be pursued. Finally, more comprehensive research including a more representative sample of residents and residency programs seems warranted.
The authors thank Katherine Green Hammond, Ph.D., for assistance with the statistical analysis. They also wish to thank the participating sites and residents: Irina Korobkova, M.D., Case Western Reserve School of Medicine; Deeba Ashraf, M.D., and Meg Weigel, M.D., Emory Univeristy; Ranji Varghese, M.D., Mayo Clinic; Preeti Kalani, M.D., Michigan State University, Kalamazoo; Shabnam Sood, M.D., Maricopa Health Systems, Phoenix, AZ; Ashraf Fanous, M.D., Diana Martin, M.D., Salman Wahid, M.D., and Snejana Sonje, M.D., St. Elizabeth’s Hospital, Washington, DC; Jennifer Haak, M.D., and Jesus Ligot, Jr., M.D., State University of New York at Buffalo; Georgian Mustada, M.D., State University of New York at Syracuse; Nicole Lanouette, M.D., University of California at Los Angeles; Melanie Deluna, M.D., and Robin Bitner, M.D., University of California at San Francisco; Jeffrey Tuttle, M.D., University of Kentucky; Nicole Washington, D.O., University of Oklahoma; Andrea Kim, M.D., University of Texas, Southwestern Medical Center, Dallas, TX; and Cynthia Singley, M.D., University of Wisconsin. The authors also thank the anonymous reviewers of this manuscript for their valuable feedback and observations.
At the time of submission, Drs. Calabrese, Bitner, and Tuttle reported no competing interests. Dr. Sciolla is a member of the Speaker’s Bureau for AstraZeneca. Disclosures of editors are published in each January issue.