Problem-based learning (PBL) has a long and rich tradition, based on 30 years of educational research. In the 1970s, Howard Barrows and his colleagues at McMaster University decided to modify medical education to help students cope with the explosion of medical knowledge. They also wanted to integrate the basic sciences more fully into patient care and promote self-directed learning to prepare medical students for a lifetime of learning. They understood that learning takes place best in a context, and so they created PBL.

PBL fosters good cognitive activities, such as elaboration; where new information is added onto existing knowledge. Elaboration is promoted through case discussion, note-taking, or answering questions about the case. PBL also promotes collaboration with peers and the ability to reason out and discuss various problem-solving approaches. Self-directed learning includes setting goals, selecting strategies for achieving goals, and monitoring learning; it also involves cognitive and motivational self-regulation. Tutors act as facilitators of the small groups of about eight trainees, probing the trainees' knowledge, ensuring that all trainees are involved in the process, monitoring the educational progress of each trainee, and modulating the challenge of the problem. The role of the tutor is to scaffold student learning, stimulating elaboration, integration of knowledge, and interaction between trainees by asking questions, seeking clarifications, and application of knowledge.

However, PBL has not been shown to be more effective than traditional learning (1), although studies show benefit in specific aspects of PBL, such as the ability to transfer concepts to new problems (2), improved collaboration, problem-solving skills, skills relevant to running meetings, and working independently (3). Further refinements of PBL have occurred over the years as teachers have responded to specific criticisms of PBL, such as its high faculty-to-student ratio and inefficiencies of process.

One exciting iteration of the PBL approach is the clinical-presentation curriculum model (CP), developed at the University of Calgary Medical School (UCMS) (4). This model focuses on teaching medical students the common presentations of illnesses (e.g., headache, chest pain, confusion). Educators teach the students cognitive schemas that illustrate the reasoning strategies used by experts, to think through a clinical problem and reach the correct diagnosis. These schemas were developed by asking experts to talk out loud while they solved problems, and their thinking strategies were mapped out onto clinical algorithms. This clinical-reasoning strategy is called scheme-inductive reasoning, and it has fivefold greater odds of diagnostic success than using traditional hypothetico-deductive reasoning (4). At UCMS, 120 cases are elaborated in total, each case with 6 possible underlying causes, for a total of 720 disease categories. These cases are roughly grouped into courses such as OB/GYN, endocrine, neurological, etc. Medical students review the schema for each presenting problem, see a demonstration of how a schema is thought through, attend lectures covering the related basic science, and complete that course by tackling PBL cases in small discussion groups. The schemas for all 120 clinical presentations are compiled into one book, the University of Calgary "Black Book," which is given to each medical student upon entering medical school.

Teaching an integrated biopsychosocial (BPS) approach to patient care is a clear priority in psychiatry. Psychiatric illness is frequently taught as two paradigms: 1) the biological; and 2) the psychological/social perspectives of illness, with not enough attempt to integrate these two dimensions into a cohesive understanding of the patient. Residents express difficulty with this training (5) and have expressed a desire for more learning in developing BPS formulations and treatment plans (6).

The integration of psychotherapy teaching with the rest of psychiatry and medicine is an important training focus (7). Glen Gabbard states that "a contemporary psychotherapy teacher needs to be conveying to psychiatric residents that psychotherapeutic principles are applied in all settings where psychiatric treatment is delivered (p 333)." This goal is reached by integrating the teaching of psychotherapy through PBL or case-based learning.

Patients also benefit from a case-based BPS approach. Patients can get a very different view of their problem, depending on the training background of their psychiatrist. A psychiatrist who is able to develop an integrated BPS formulation and plan with the patient and family will have much more to offer them. Through role-playing, residents gain an appreciation of the perspectives of their peers and how this affects patient care, and, through role-playing, residents also have the opportunity to try out different formulations and plans.

PBL and role-playing can be incorporated into psychiatry in several ways. On a small scale, we developed a PBL case-based course to teach psychosomatic medicine to PGY2 psychiatric residents. Psychosomatic medicine lends itself well to PBL because of the large number of differential diagnoses in patients with medical and psychiatric illnesses and the difficulties in presenting a complex case-formulation to the patient and family. We added role-playing to give residents practice in presenting complex material to the patients and families and the medical team. The curriculum consisted of three cases discussed over eight sessions of PBL.

Residents are excellent at generating differential diagnoses but are less skilled in presenting an integrated BPS formulation to the patient and family. Residents may have a strong belief about what type of treatment to deliver but often do not consider the benefit of discussing all treatment options with the patient and family in a collaborative process of decision-making. The addition of role-playing exercises, where different versions of the formulation can be tried out with the patient and family, provides residents with an opportunity to practice working in a collaborative way with the patient and family. In summary, we use a developmental teaching model to meet the residents where their knowledge and skill end, and help them develop new skills in communication and collaboration. For example, in one case, some residents favored a biological treatment approach, whereas others favored a psychotherapeutic approach. We encouraged the residents to determine what the benefits of each approach might be and how to combine modalities in practice and then how to discuss these options with the patient and the family. In this way, residents can meet many of the proposed Academy of Psychosomatic Medicine core competencies, for example, interpersonal and communication skills, systems-based practice, and practice-based learning and improvement.

In developing our course, we chose three cases: two cases from the outpatient psychosomatic clinic in a general-hospital setting and one case from the inpatient medical setting. In this case-based model of training, residents are given a short write-up of the case with a presenting complaint. At Week 1, after discussion about possible differential diagnoses, the residents develop a set of learning objectives to be reviewed and presented at the next class. At Week 2, the residents present their literature review and discuss how that knowledge can be applied to the case at hand. The residents discuss a case-formulation and review treatment options. A comprehensive biopsychosocial treatment plan is generated and presented to the patient and the family. Role-playing follows. Residents play the patient, and family members and the remaining residents take turns in being the psychiatrist. Role playing helps the residents develop clinical skills in presenting a complex case-formulation to the patient and his or her family. An example of one case is illustrated, with descriptions of the classroom activity over 3 weeks (see: Case of Jane Strong).

The Case of Jane Strong

Week 1

An 18-year-old woman presents to clinic with her parents with complaints of asthma being out of control, anxiety, and alcohol abuse. Past history is significant for adolescent difficulties resulting in parents "kidnapping" the child and sending her to an Outward-Bound program. At interview, the patient is anxious and tearful, and there is significant family conflict.

The group discusses the case and formulates provisional differential diagnoses. They develop learning objectives as homework assignments.

Examples of Learning Objectives:

Week 2

Residents briefly discuss key issues learned from homework literature review and apply them to the patient and family. They generate a BPS formulation and treatment plan. Residents engage in role-play, presenting diagnoses, BPS formulation, and plan to "attending team" and "patient and family." Roles are assigned by resident choice. Residents switch off playing the physician managing the patient and family/attending team.

The residents are given Part Two, which states that "The patient is resistive in the interview, and her parents, especially mother, want clear direction about how to manage the situation at home. Father is passive." The residents are asked to discuss how they will manage the family situation.

Examples of further learning objectives:

Week 3

The residents briefly review their homework, reporting to the group on the articles they consider pertinent, and they refine their BPS formulation and treatment plan. Residents role-play working with the family on initial treatment efforts. The session ends with a discussion of difficulties and uncertainties raised by this case.

A more radical shift in educational perspective would be to introduce full case-based teaching to the residency program, in a similar way to the UCMS curriculum. In this scenario, common psychiatric presentations, such as depressed mood, anxiety, and hallucinations, are taught, using cognitive schemas developed by expert clinicians. These common presenting problems are coupled with didactic lectures reflecting the basic science related to the disorder. PBL sessions follow, with residents learning to apply the cognitive schemas to a clinical case. For anxiety disorders, experts generate schemas illustrating how to reach a differential diagnosis. Didactics focus on the pathophysiology and neuroscience of anxiety, and the course ends with PBL cases. The addition of role-playing to traditional PBL allows the residents an opportunity to practice presenting their case-formulation and treatment plan to the patient and family, thus developing expertise in formulating cases at the end of each patient/family interaction.

Nine PGY2 residents took part in the first year of implementing this program. Residents and faculty stated that they enjoyed the experience and found role-playing helpful. Criticisms were that the residents wanted more cases presented, wanted to move through the cases faster, wanted more information presented, and wanted the learning objectives given to them at the start of class.

The residents and faculty, in the first year of introducing this course, had some difficulty shifting to a PBL role-playing approach. The comments made by the residents are typical of learners in a PBL setting. However, as the trainees get used to and learn to trust the process, they understand that learning is generated by their needs, and not by the supervisor's needs. The faculty also needs to introduce the concept of PBL in a more complete way and help individual residents meet their learning goals. It is interesting that the residents generated the almost identical list of learning objectives that the faculty did before the class. This shows that the cases were constructed with the relevant case material. Despite these criticisms, the residents and faculty enjoyed the role-playing and felt that they understood and could better manage the questions of the patient, family, and referring physician. Our next challenge is to develop a structured and effective tool for assessing the effects of PBL role-playing to determine whether the PBL role-playing design meets the core competencies as effectively as traditional teaching.

Measurement of the effectiveness of PBL is a significant challenge that has not been well met over the years since PBL was introduced. Multiple-choice exams, such as the PRITE and the American ABPN written board examination, do not measure clinical skills or assess how to think about and manage patients; that is, development of clinical reasoning skills.

Psychiatry, at this time in its life course, has a great deal to gain in adopting new training strategies that bring cohesion to the field. A problem-based, case-based, or clinical-presentation model of training resolves the question of how to teach psychiatry as an integrated biopsychosocial approach to patient care. A key component of any new teaching method is, of course, the ability to assess its effectiveness. Computerized case-simulation (CCS) offers an interactive assessment tool that is suited to the assessment of clinical skills in psychiatry (8). This type of method can be duplicated across many training sites and will, hopefully, allow resident assessment to become more reflective of resident practice.

We thank and gratefully acknowledge the contribution of Frederick Wamboldt, M.D., Professor of Psychiatry, who co-teaches this course.

At the time of submission, the author reported no competing interests.