Psychotherapy skills are at the core of clinical psychiatry. Psychiatrists must have the ability to use clinical skills learned in psychotherapy training and to use psychotherapy as a primary treatment. Psychotherapy is currently less frequently provided by psychiatrists in a typical office visit, however, presumably because of advances in psychopharmacology and the impact of managed care (1). For a time, some psychiatrists could complete training without basic competence in any form of psychotherapy. In response, in 2001, the Accreditation Council for Graduate Medical Education (ACGME) mandated training to competency in five models of psychotherapy (2). The ACGME narrowed the requirement to psychodynamic, cognitive-behavioral, and supportive treatments in 2007 (3), with the goal of graduating psychiatrists with a broad range of psychotherapy skills. Psychodynamic psychotherapy was the dominant model of psychotherapy taught to residents before this mandate, but the evolution and evidence-base for cognitive-behavioral therapy (CBT) made it a logical choice as an equally important and distinctly different approach to require. Supportive therapy is a model that utilizes elements of psychodynamic psychotherapy and CBT, but has a separate set of theoretical principles, areas of emphasis, formulation, technique, and applications (4).

Although a substantial literature exists about learning psychotherapy, pedagogy concerning sequencing, essential didactic content, and the minimum amount of clinical experience and supervision is lacking. Many experts have written encouraging the further delineation of guidelines for competency-assessment and evidence-based practices in psychotherapy training (5). Residents often must learn multiple models of treatment simultaneously and frequently desire to implement integrated approaches (6). Several authors have described educational approaches that use the common elements of CBT and psychodynamic psychotherapy as a foundation for learning core skills of psychotherapy before learning specific approaches (7, 8).

In an attempt to determine the current state of training in psychiatric residency programs, the authors surveyed training directors about the amount of education and measures of competence used in teaching several models of psychotherapy and asked them to describe perceived obstacles to training.

Using the American Association of Directors of Psychiatric Residency Training (AADPRT) training directors’ mailing list, in September 2009, we designed and e-mailed a 26-item survey to adult residency training directors. The survey was modeled after a national survey of psychotherapy training (9), but was specifically tailored to psychiatric education. It inquired about hours of didactic training, supervision, and clinical experience in a variety of psychotherapeutic methods. Questions included how competencies were assessed within the program and what obstacles existed to psychotherapy training. (Copies are available on request.) A second e-mail with added instructions to not send in a duplicate response was sent out 1 month later. Survey data were returned by e-mail and inspected to ensure that each program was included only once. Responses were collated by Survey Monkey.

Data from a previous survey conducted by one of us (DMS) about cognitive-behavioral therapy training just before the ACGME requirements in 2001 (10) were compared with the amount of current training in CBT.

Data analyses were done with SPSS Version 15. For comparisons of the three modalities (CBT, psychodynamic, and supportive) in the dataset on numeric measures, a Friedman rank ANOVA was used because the data analyzed were based on a 0–5 scale (0: none through 5: 70+). Significant Friedman ANOVAs were followed up with Bonferroni-corrected Wilcoxon tests, both suitable for scale data. The minimum number of patients was analyzed similarly because data were highly positively skewed.

Comparisons between the two CBT data sets were performed with the Mann-Whitney U test, suitable for between-group comparisons.

We received 82 valid returned surveys, for a total response rate of 45%. Not all respondents answered all questions, but almost all (76/82) answered the questions about required seminar and supervised clinical hours and minimum number of patients. The number of hours of required didactics, supervised clinical experience, and patient numbers are for the entire 4 years of training.

Post-hoc analysis of Friedman ANOVAs indicates that psychodynamic therapy has a significantly higher number of hours and that CBT and supportive therapy do not differ.

The largest amount of didactic training occurs in psychodynamic psychotherapy. Approximately 70% of programs provide over 30 hours, with nearly 25% reporting 70 hours or more. Almost one-third offer less than 30 hours over the course of residency training. Less time is devoted to CBT didactics, with two-thirds of programs offering less than 30 hours; one-third, more than 30 hours; and almost none, greater than 70 hours (see Table 1). However, this is a significant increase from 2001, when only 13.1% offered more than 30 hours. Supportive-therapy didactics show 85.3% reporting less than 30 hours, but it is not possible to know whether this is included as part of psychodynamic or CBT training.

Supervised clinical experiences show similar time differences, with psychodynamic psychotherapy at 73.3% offering more than 30 hours, nearly half of which require greater than 70 hours, and one-quarter of programs with less than 30 hours total. CBT supervised clinical experiences have dramatically increased over this decade, to 46.8% with over 30 hours and only 6.5% with no hours of required supervised clinical work reported (as compared with almost 50% in 2001). Approximately half of the programs report more than 30 hours of supervised clinical work in supportive therapy, but it is not possible to know whether this is part of case-supervision (including psychopharmacology) or as separate supportive psychotherapy supervision.

The minimum number of patients treated show supportive therapy as the highest, with a mean of almost eight patients. This is not surprising, since the majority of outpatients seen during residency require medications and supportive clinical management. There is a modest difference between the other two modalities, with psychodynamic psychotherapy close to a mean of four total patients and CBT at three patients. It is of note that the question of duration of these treatments and the number of cases taken through a planned termination were not addressed by this survey.

The results of this survey strongly suggest that residency programs are responding to the ACGME mandate for psychotherapy training. Many programs, however, report a number of hours that would not be adequate to produce resident competence in any modality. Psychodynamic psychotherapy is taught with the widest variability; CBT training has increased substantially since 2010; and supportive-therapy education is the least consistent. Psychodynamic psychotherapy training has the largest number of didactic hours and required supervised clinical experience, with 25%–38% of programs teaching this model for more than 1 year. The hours of training reported in psychodynamic psychotherapy speak to the complexity of the technique and its history as the primary psychotherapy modality practiced in American psychiatry.

In 1990, a group of prominent educators from AADPRT and the Association of Academic Psychiatrists (AAP) coauthored a landmark paper on psychodynamic psychotherapy training and its future, including a model curriculum (11). The amount of time specified per week for training was: PGY1: 2 hours, PGY2: 6 hours, PGY3: 8–11 hours, and PGY4: “slight drop” from PGY3. Our survey, 20 years later, and following a specific RRC mandate for training to competence in psychodynamic psychotherapy, reveals that most programs are far below this baseline for minimal training. There were no contemporary survey data to compare the amount of training time designated for this model before this survey. Psychodynamic psychotherapy training would be much enhanced if, similar to CBT, core didactic content for residents was described along with recommendations for the amount of supervision and patient care experience. Current psychodynamic psychotherapy texts (12–14) can serve as a template for further curriculum development. Consensus is needed concerning content and sequencing of training (15), requisite clinical experience, amount of supervision, and evaluation of clinical competency (16). Evaluation methods need to be developed, validated, and disseminated. The Columbia Psychotherapy Exam (17) can measure knowledge, but observed therapeutic work rated with standardized evaluation instruments is the benchmark of therapist competence (9).

CBT training has greatly increased in scope since 2001, but the amount of required supervised clinical experience remains significantly less than that for psychodynamic psychotherapy. CBT has a long history of specifying effective methods of training, supervision, and evaluation, which results in improved therapist competence and patient outcomes (18–22). Effective training programs that improve therapist ratings from "advanced beginner" to "competent" (18, 23, 24) consist of a minimum of 1 year of didactic training and 50–70 hours of supervised clinical experience. Therapy sessions are evaluated with validated instruments that measure competence and adherence (e.g., the Cognitive Therapy Scale [CTS] Young JE, Beck AT: Cognitive Therapy Scale Rating Manual. Philadelphia, PA, University of Pennsylvania Center for Psychotherapy Research, 1980; unpublished). Such instruments measure the skill of the therapist and ensure that therapy is faithfully and competently delivered. One study of assessment of competence in CBT indicated that reliable assessment of competence must include multiple work-samples evaluated by several instruments (e.g., tape recorded session material and written case conceptualizations) (25).

Educators in CBT responded to the 2001 mandate with recommendations for efficient methods for training and guidelines for didactics, supervision, and evaluation methods (26). Although the number of hours varies, one-third of responding programs report using live patient supervision, video, or audiotape review of CBT case-material to assess resident performance. This is a tremendous step toward increasing the quality of CBT education and an indicator that programs are adopting recommended guidelines for CBT training. Textbooks designed for resident education, some with DVD demonstrations (27–29), now exist to facilitate learning key principles and therapy techniques. Seminars and extramural workshops for residency training directors and psychotherapy educators were held at AADPRT, the APA, and the AAP, and in prominent CBT training centers (e.g., The Beck Institute for Cognitive Therapy and Research). A parallel process in psychodynamic and supportive therapy would greatly enhance psychotherapy training.

A problem common to both CBT and psychodynamic psychotherapy indicated by the survey is the wide variability in hours of required supervised patient care. The majority of programs require 1 hour or less of supervised patient care per week for 1 year of training. A related area is the minimum number of patients required to treat. Both forms of treatment are practiced with variable numbers of patients; 72.7% of programs required three-or-fewer patients treated in psychodynamic therapy; 92.9% of programs required three-or-fewer patients treated with CBT; 95.5% of the respondents cited difficulty finding patients who have the motivation and capacity for insight-oriented treatment as an obstacle. Effective psychotherapy training requires suitable patients. We did not assess how long patients are treated, but residents often provide therapy in settings where treatment can be brief, inconsistent, or incomplete. Furthermore, there are no clear guidelines as to what the optimal didactic and supervised clinical experiences are to produce "competent" residents.

The data on training in supportive therapy are revealing. Supportive psychotherapy, by far the most widely practiced form of therapy, has the least-defined pedagogical base and the most limited focused training. Controversies exist about its definition, its links to psychodynamic psychotherapy, CBT and other specific modalities, its evidence-base, and even whether or not it should be taught as a distinct discipline (4). Our data indicate a wide divergence of patient hours and supervision and substantially less didactic training. The lack of clarity about the model of supportive treatment, the lack of specialized training centers, and the absence of a clearly-defined curriculum or training protocols makes training in this modality a challenge. Supportive therapy needs its own defined pedagogy and guidelines for effective supervision.

Several weaknesses exist in this study. First, only 45% of the cohort responded to the questionnaire. Our sample may not be adequate to fully comment on the state of psychotherapy training. However, our speculation is that respondents could be in those training programs most committed to psychotherapy education. Furthermore, although we were blinded to the answers given by each respondent, we are aware of the identities of the programs that responded and the fact that we had a wide range of program sizes, geographical distribution, and types of sponsoring institutions. Only one-third of our respondents answered the questions about obstacles to training, and our findings are not statistically significant. However, our findings suggest that obstacles exist (i.e., lack of faculty availability was cited by 60% and 77% regarding CBT didactics and supervision, respectively; 55% and 58%, respectively, for psychodynamic psychotherapy didactics and supervision). Residency programs need trained faculty to effectively teach and supervise. The experience in CBT over the past 10 years demonstrates that it is possible to efficiently train faculty when the curriculum is well defined and resources are deployed. Programs for distance learning (30) and supervision exist in CBT, with participants attending several weekend seminars and then receiving weekly supervision by telephone. CAPA (China-American Psychoanalytic Alliance) is a recent example of a national effort to provide psychodynamic supervision and psychotherapy (via SKYPE) to students in China (31). Similar strategies could be utilized in the United States.

There has been considerable progress over the past decade in psychotherapy education, but inconsistencies remain in the amount and quality of training experiences offered to residents. CBT education has become much more consistently developed in residency programs. Psychodynamic therapy is widely taught, but establishing consensus regarding its content and more rigorous evaluation strategies would improve resident competence. Supportive therapy, the most widely practiced form of treatment, needs better definition and pedagogy development. Core content, sequencing of training, length of training, patient availability, and faculty education must receive ongoing attention and development to ensure more effective psychotherapy training for psychiatrists.

The authors thank Edward Gracely, Ph.D., and Philip Rodgers, Ph.D., for their assistance in the preparation of this manuscript