Problem-based learning (PBL) is an educational method that is growing in use in medical schools (1—3). PBL has been used and evaluated in medical schools to teach a broad variety of topics, including those related to psychiatry (4,5). There are far fewer reports of its use in residency education, although it could be argued that clinical case supervision is similar to PBL (6). A PBL curriculum replaces most or all lectures with group meetings in which a clinical problem is examined (7). Learning issues are identified, and the students seek information from published material and faculty. Each student then returns to the group and reports his/her findings as the group tries to understand and manage the clinical problem. The group leader's role is to facilitate the process rather than to give answers. PBL was developed to address the difficulty of ever-expanding medical information; PBL models a method of learning and investigation to be continued after graduation (8).
Most published reports on the effectiveness of PBL are favorable. Recent research has suggested that clinicians educated in PBL as medical students had more knowledge in the treatment of hypertension than their colleagues taught in a traditional format (9). PBL students may learn a different method of thinking, which could limit the accuracy of comparisons with students educated in traditional formats (10). Nonetheless, those comparisons have indicated that PBL students may score better on more clinically oriented tests, but lower on examinations that emphasize basic sciences (11). A meta-analysis of studies comparing PBL with traditional education supported this finding and revealed that PBL was rated more highly in students' evaluations and clinical performance, and rated about equal in clinical and factual knowledge (12). Although PBL appears to be able to cover necessary topics, students and faculty worry about missed material, perhaps because traditional lectures cover topics in a more organized format (13).
It is probable that residency education in forensic psychiatry is conducted primarily through lectures, except when a faculty member supervises a resident in the evaluation or treatment of a defendant or patient (14—17). This direct exposure to forensic cases varies among programs and individual resident experiences. There are efforts to standardize a didactic forensic psychiatry curriculum, but programs probably still vary in terms of the number of sessions, depth, and focus of the lectures given to residents. Although forensic psychiatry is a recognized subspecialty with a growing number of fellowship programs, educating general psychiatrists (and other physicians) about medicolegal issues is still an important task of the academic forensic psychiatrist (18—21). The authors noted that whereas residents' interest in actual cases was rather high, interest in the didactic issues varied.
To evaluate its effectiveness, PBL was used in a pilot project to teach forensic psychiatry to psychiatric residents. The residents were given written problem descriptions and read them as a group. They attempted to solve each of the problems by reasoning, investigation, and sharing knowledge, with some resource material provided by the facilitator. Our hypothesis was that the residents would report an increased comfort with forensic psychiatry issues at the conclusion of the PBL course.
The participants were 23 residents in psychiatry. The residents' mean age was 35.7 years (standard deviation [SD]: 4.63 years). The residency is an Accreditation Council for Graduate Medical Education-approved university program, with a mixed urban/rural multisite setting and a multicultural population. The PBL program was offered over 2 consecutive years. There were 11 trainees in the 1993 section and 12 in the 1994 section. There were no significant differences between the participants' precourse comfort with forensic issues or in their knowledge base between the two years' sections. Five residents were in their first year of residency, 13 in their second year, and 2 in their fourth year. Three individuals were visiting medical students or declined to identify the year of residency.
The PBL materials consisted of four written problems involving hypothetical patients. Each problem had multiple page breaks to serve as pauses for the generation of incipient hypotheses, learning issues, and discussion. The first problem involved a young woman who was vague and approximate in the telling of her history, then more clearly psychotic. Shortly after initial antipsychotic treatment, she was observed outside of the clinic appearing very well. The psychiatrist then was asked to testify regarding a previously unrevealed criminal matter. The issues of malingering, tardive dyskinesia, Ganser's syndrome, and testimony as a treating clinician were identified as potential learning issues. The second problem involved a manic patient on a surgical service who initially refused surgery. After an alliance was formed, he admitted to a prior violent felony and thoughts of hurting someone else. The issues in this case included confidentiality, competence in medical settings, reporting of prior felonies, and Tarasoff duties. The third case was of a young prisoner referred to a psychiatrist for violence. The man complained about prison rape. There were indications of planned violence on his part, but then a history of epilepsy was revealed. The case ended with a request for the psychiatrist to conduct a forensic psychiatry evaluation of the man's dangerousness. Identified learning issues included prison violence, prison culture and sexuality, directed violence vs. nondirected violence, the treatment of violence, and the prediction of dangerousness. The last case was a patient with a history of past lawsuits. She presented with depression and back pain. After other treatments were ruled out, the psychiatrist prescribed a monoamine oxidase inhibitor. The patient later fell, which resulted in a questionable, but possible, back injury. Her attorney alleged that she experienced hypotension and had not been warned about it. Alcohol may have been a major factor. The psychiatrist testified at her deposition and trial. At the latter, one expert witness for the plaintiff had a very different form of clinical practice, the other appeared to have a general stance against medications. Issues raised by this case included informed consent, litigiousness, malpractice, standard of care, and the ethics of testimony. There were three consecutive sessions in which these cases were analyzed. The residents attempted to solve each case's dilemmas and identify issues for further learning.
The evaluation measures consisted of four Likert-type scales, each measuring subjective individual comfort with the forensic issues of violence, testimony, liability, and competence on an ordinal scale, from a comfort level of 1="none" to 5="high." A similar 5-point Likert-type scale was used to subjectively measure forensic knowledge base. As part of the course evaluation, a similar postcourse 5-point Likert-type scale was used to register resident satisfaction with the educational program.
This study used a within-subjects repeated-measures design. Testing occurred in two sessions. The pretest was administered before beginning the PBL program in forensic issues. The posttest was administered at the end of the PBL course. Data were analyzed by using Statistical Program for Social Scientists.
A small sample size and the use of multiple measures were anticipated to present some confounds to the research design. To compensate for the likely inflation of significance level resulting from multiple measures, a Bonferroni adjustment was made, holding the experiment-wise significance level at alpha=0.05. It was also noted that this analysis would be able to detect effect size of 0.5 SD. Small-to-moderate changes in comfort level or knowledge base were less likely to be detected.
Increased knowledge was expected to result in increased comfort with forensic issues. Thus, it was anticipated that the residents would become more comfortable with the forensic issues on postcourse measures and that there would be a concomitant gain in knowledge base. Therefore, a one-tailed analysis was used for these measures.
A number of residents did not attend all of the sessions. There were 12 complete sets of data. The attendance compared favorably with didactic sessions given to the same groups in the same years. F1 shows the results of change in the measured domains of forensic psychiatry.
The residents' comfort level with the issues of testimony, liability, and competence increased significantly (one-tailed paired t-test: t=8.4, 6.5, and 3.3, respectively, df=22, P<0.01). The mean levels of comfort for these topics were testimony: precourse: 2.28 (SD=0.89), postcourse 3.33 (SD=0.97 ); liability: precourse 2.33 (SD=0.97), postcourse 3.28 (SD=0.83); and competence: precourse 2.78 (SD =0.73), postcourse 3.33 (SD=0.91). The residents also reported an increase in comfort with the issue of violence: precourse comfort 3.11 (SD=0.83), postcourse 3.33 (SD:=0.69). These values just missed the predetermined significance level for the analysis (t=1.81, df=22, P=0.048). The subjective ratings of the residents' knowledge base in forensic psychiatry also increased significantly at the end of the PBL course. F1 shows the increase from a mean subjective rating of 2.50 to 3.06 on the 5-point Likert-type scale (t=1.81, df=22, P<0.01).
The residents reported a moderately high level of satisfaction with the PBL course format. The mean level was 4.33 (SD=1.03, df=18).
There is much less literature on PBL for resident education, compared with medical student education. Evidence from medical student education indicates the PBL may activate prior learning and that the discussion process stimulates learning and interest (22). Some residents expressed discomfort in the transition from traditional learning to PBL; this finding is similar to that seen in medical students who found PBL interesting but questioned its applicability (23). The initial change to a PBL program may be the most difficult part (24). As reported in the literature on medical students, the residents in this study were worried that all topics were not covered. Nonetheless, the residents appeared to be engaged and interested, and they gave the program moderately high ratings. The fact that there was the least increase in comfort regarding violence may suggest that knowing more about violence may lead to more caution, and thus may mitigate "comfort," although the residents' reported level of comfort at the start of the course was higher for violence than for the other categories. This study addressed self-reported knowledge base, which rose, but it was not compared with an external measure of forensic knowledge.
This study did not compare PBL with traditional lectures. Data from the residency program indicated that the PBL course was rated highly and equivalent on average to a similar didactic course given by the same presenter at a different time. Subjectively, the residents appeared to be more enthused when the cases were discussed, either in the PBL format or as part of a lecture. It seems possible that forensic cases may be more interesting than didactic forensic issues. This program seemed to successfully link the two. However, this PBL program was a supplement to didactic lectures given at other times (not during the PBL course). Further research may need to be directed at a stand-alone program of PBL education for forensic psychiatry, with direct comparisons with traditional formats.