A variety of factors seem to affect medical students' attitudes toward psychiatry, and these can be divided in two main categories (1). The first category includes factors that influence students before their entrance into medical school, such as demographic and personality characteristics (1—3), cultural factors (4), and the historicosocial climate of a particular period of time (5). The second category includes factors that influence students during the medical school years, such as the medical school's attitude toward psychiatry, the view of nonpsychiatric faculty about psychiatry and psychiatrists, and the undergraduate psychiatric education itself (1,6—8). During the last 15 years, there has been a growing interest in the international literature on the impact of psychiatric education programs on medical students' attitudes toward psychiatry. Most of the studies on the impact of psychiatric education programs on medical students' attitudes toward psychiatry have been stimulated because there has been a steady decline—mainly in the United States and Canada—in the percentage of students choosing to specialize in psychiatry (6,9—11). Undergraduate psychiatric educational programs have been criticized as being unresponsive to the needs of the discipline (9) and as being largely responsible for creating negative attitudes toward psychiatry (12). However, several studies have shown that students' attitudes toward psychiatrists, psychiatric treatment, psychiatric patients, and psychiatry in general change significantly in a positive direction after their psychiatric training (7,11,13—16).

A variety of scales and questionnaires have been used by different investigators to measure student attitudes; however, little work has focused on students' attitudes in terms of underlying conceptual or ideological models (17—19). Augoustinos et al. (19) reported a significant conceptual/ideological shift in attitude toward psychiatry, with the development of a more positive view of the discipline, in medical students after a 6-week psychiatry course. In two prior studies (20,21), we found similar results. In addition, we found that this shift seems to be sustained even 2 years after completion of training (21). Our original study (20) was conducted in 1985. Since then, the basic psychiatric training at the institution under study has been moved from the fifth to the fourth year, with additional optional training in the sixth year. Furthermore, the duration, the structure, and the content of the training have also been modified. The aim of the present study is 1) to see if the previous results can be replicated using a sample of fourth-year medical students in 1993, 2) to compare these students' attitudes toward psychiatry, before and after their training in psychiatry, with the corresponding attitudes of the students in 1985, and 3) to compare both groups of students' attitudes at the end of their education program with their teachers' attitudes.

The sample consisted of 155 fourth-year medical students receiving psychiatric training during 1993 at the B' University Department of Psychiatry of the Aristotelian University of Thessaloniki in Greece. The students attended 4 hours of theoretical lectures per week, for 16 weeks, focusing on the history of psychiatry, phenomenology, diagnostic classification, etiology, and treatment of psychiatric disorders. In addition, the students were divided into groups of 15—20 and spent another day per week rotating in an acute or chronic ward of a state hospital and a mental health center. The content of this second educational component was mostly practical, including case presentations, the study of video recordings of interviewing techniques, and diagnostic procedures and treatment interventions, and it required active student participation. Nine staff psychiatrists were involved in the training program.

The Libertarian Mental Health Ideology Scale (LMHIS) by Nevid and Morrison (22) was used to assess attitudes. This scale measures attitudes toward two popular and sometimes competing conceptual frameworks in psychiatry, the medical model and the radical psychosocial or "Szaszian-libertarian" position (22). This latter position supports the idea that mental illness is a myth or a metaphor that has outlived its usefulness (23). The LMHIS, which is one of the most suitable scales for assessing attitudes of mental health professionals and students (22) has been translated and adapted in Greek with unanimous consensus by three psychiatrists and one clinical psychologist. Two of these translators were bilingual, and the other two were fluent in English. The scale is composed of four factor subscales: Mental Illness Mythology (MIM), Anti-Medical Model (AMM), Social Deviance Control (SDC), and Anti-Coercive Treatment (ACT). The MIM and AMM subscales attempt to distinguish between a radical psychosocial and a medical position in the conceptualization of the nature of mental illness and its etiology. The psychosocial end of the MIM suggests that mental disorders should not be classified under the category of "illness" but instead should relate to personal and social issues rather than medical ones. The medical end of the MIM supports the belief that mental disorders are illnesses like any others and their treatment has to follow the procedures applied to somatic illnesses. The medical end of the AMM subscale stresses the position that mental/behavioral problems are organic in nature. The radical psychosocial end of the AMM denies the contribution of any organic factor to the development of mental/behavioral problems. The SDC and ACT subscales deal with civil, ethical, and legal issues in psychiatry. The libertarian end of the subscales expounds the belief that compulsory hospitalization and treatment are completely unjustifiable in all cases and that mental health professionals become agents of the state to control any social deviance. The medical/authoritarian end of the subscales accepts the need for involuntary hospitalization and coercive treatment when indicated and denies the aforementioned view about the role of mental health professionals.

The questionnaire includes 39 Likert scale items, with lower scores indicating a more medical/authoritarian orientation and higher scores indicating a more libertarian orientation. Scores range from 39 to 195. There are also subscores of each of the four subscales. The LMHIS has high split-half reliability and highly significant intercorrelations among factor and total scores (22).

The questionnaire was administered to the students at the beginning and again at the end of their psychiatric training. To allow for subsequent follow up, the students reported their names. Therefore, the students were reassured that their answers would not in any way contribute to their final grade in psychiatry, since the questionnaire items elicit matters of opinion and do not have right or wrong answers. Of the 1993 fourth-year medical students, 90.3% (140 of 155) completed the questionnaire. All teaching staff members (N=9) also completed the questionnaire. The results of the 1993 student group were compared with the results of a prior study conducted in 1985 (20). The sample of this 1985 study consisted of 51 fifth-year medical students receiving psychiatric training in our department. Their educational program was mainly theoretical and shorter in duration (i.e., 12 three-hour lectures/week). Following the same procedures as the 1993 group, these students had completed the LMHIS, which had also been completed by all the teaching staff members of that period (N=12). Eight of the aforementioned teaching staff members also participated in the 1993 program.

T1 presents the LMHIS scores (total and subscales) of the 1993 and 1985 students before and after their psychiatric training. As shown, significant differences exist in both groups, between the pre- and posttraining scores, indicating a considerable shift in the students' attitudes toward a more medical orientation at the end of the training. Furthermore, the 1985 students have significantly higher pretraining means on all but one subscale (ACT) and on the total LMHIS score, compared with the 1993 students, a finding suggesting a more libertarian and less medical orientation. The greatest differences appear on AMM and SDC. This less medical orientation of the 1985 students, compared with the 1993 students, continues to exist even after their courses in psychiatry. The 1985 students continue to manifest significant after-course differences, except on AMM, compared with their teachers, with the students showing a less medical orientation (T2). On the other hand, the 1993 student group posttraining and the teaching staff show no significant differences on any of the LMHIS scores (T2). It is worthwhile to mention that both student groups had significantly higher pretraining subscales and total scores compared with the teaching staff, with the exception of the 1993 group on scale AMM.

The present study shows that a significant change in students' attitudes toward psychiatry and mental illness occurs after their exposure to a psychiatry teaching program. This study confirms the results of our earlier studies (20,21) and supports the findings of other investigators in different countries, including the United States (6,12), Canada (7,11), Great Britain (14,24), Saudi Arabia (15), Korea (16), and Chile (25), using different questionnaires. Augoustinos et al. (19) in Australia, also using the LMHIS scale, got similar results and suggested that this change reflected a global conceptual/ideological shift in the way these attitudes are organized and arranged and was not merely an isolated attitude change. In addition, the present study found significant differences in the students' pretraining opinions toward psychiatry between the 1993 and the 1985 students. The 1985 students were more libertarian and less medically oriented. Medical students' attitudes are likely based on opinions they form earlier in their lives (26), which are closely related to social, political, and cultural factors (4,5). Although speculative, differences in the overall political climate in Greece from the mid-1970s to the 1990s may have contributed to the higher libertarian scores of the students in 1985. Changes in the provision of psychiatric services over this period may also have contributed to the different scores in 1985 and 1993. In 1985, psychiatric units in general hospitals in Greece did not exist. These units play an important role in the attitudes about psychiatry of both psychiatric and nonpsychiatric medical professionals, whose views strongly affect the students who train under them (6). By 1993, psychiatric units of general hospitals were well established and offered educational benefits and consultation-liaison services to nonpsychiatric medical professionals. This factor may have contributed to the pretraining less libertarian and more medical orientation of the 1993 students compared with the 1985 students.

The finding that the 1985 students continued at the end of their training to have a more libertarian orientation than the 1993 students is probably due to some extent to the fact that these students had a more libertarian orientation before the beginning of the course. However, we believe that another reason for that continuing difference is that the educational program was reformed; therefore, the educators in 1993 had the opportunity to be more effective in clarifying and correcting students' false beliefs than they were in 1985. This view is supported by the finding that the pre—post total score changes in 1985 were 11.7 points, whereas in 1993 the pre—post total score changes were 15 points. Furthermore, although the 1993 students' postcourse attitudes were very similar to those of their teachers, the 1985 students' postcourse attitudes were not. The 1993 students attended a longer, more practically oriented course, with more case presentations, live or videotaped, and their participation was more active. There are reports that students are favorably impressed when they observe live or videotaped interactions between psychiatrists and their patients (6) and that more active student participation contributes to the development of a more positive attitude toward psychiatry (27,28).

The results of the present study suggest that 1) if these medical student groups represent the larger Greek society, then significant changes may be occurring in the Greek society about attitutudes toward psychiatry and 2) psychiatric education in medical school is a significant determinant in molding students' attitudes toward psychiatry. Students' attitudes influence not only psychiatric recruitment but also the ability of future nonpsychiatric physicians to identify and appropriately refer patients with psychiatric disorders (6,11). This ability is essential, especially for physicians who work in primary care settings. Furthermore, the students of today will be nonpsychiatric physicians tomorrow and will influence the attitudes of their own students (6). One should also bear in mind that nothing in the literature suggests that Greek students differ significantly from students in other countries, including the United States, to make these implications irrelevant for psychiatric educators outside of Greece. Our findings indicate that psychiatrists who participate in undergraduate psychiatric education have a good reason to try to improve their educational programs (10,29), since the quality of psychiatric education is an important factor, perhaps the only one within the educators' control, in molding their students' attitudes toward psychiatry (30).