The story of how, over the past 35 years, psychiatrists held and then lost leadership positions in community mental health centers (CMHCs) is well known (
1,
2). During the past decade, however, leadership positions for psychiatrists at CMHCs, typically as medical directors, have been partly reestablished.
Methods
In February 1998 we mailed a questionnaire to all members of the AACP. The questionnaire asked respondents to indicate their specific job title and to classify their level of operation as agency medical director, program medical director, staff psychiatrist, or consultant. Medical directors were also asked whether their breadth of supervisory responsibility included medical staff only, medical and other clinical staff, or all staff— that is, medical, clinical, and administrative. The questionnaire established six job types for medical directors: a program medical director who supervises medical staff only, an agency medical director who supervises medical staff only, a program medical director who supervises medical and other clinical staff, an agency medical director who supervises medical and other clinical staff, a program medical director who supervises all types of staff, and an agency medical director who supervises all types of staff.
The questionnaire also asked respondents to indicate how often they perform a variety of tasks in three domains: direct service, including providing medication or psychotherapy, overseeing medical care, and negotiating care with other providers; clinical collaboration, including supervising medical and nonmedical staff, providing informal consultations, participating in team meetings, and conducting formal training; and administration, including developing policy, carrying out routine administration, providing quality assurance, negotiating contracts, and serving as a link to other agencies, regulatory bodies, and boards.
Of the 479 questionnaires mailed, a total of 286 individuals returned them, for a response rate of 61 percent (12 forms were undeliverable). Of the 286 forms, 21 were marked "not applicable," in most cases indicating that the addressee was in private practice or retired. A total of 265 completed questionnaires were received. This study analyzed data from the 168 respondents who classified themselves as medical directors.
Results
The mean±SD age of the 168 medical directors was 49±9 years. Most were male (135 respondents, or 80 percent), Caucasian (153 respondents, or 91 percent), and graduates of U.S. medical schools (142 respondents, or 85 percent), and most worked in urban areas (127 respondents, or 76 percent). Approximately half were employed by not-for-profit agencies (88 respondents, or 52 percent). Thirty-eight (23 percent) worked in state facilities, 29 (17 percent) in municipal facilities, ten (6 percent) in for-profit facilities, and two (1 percent) in federal agencies.
A total of 106 of the medical directors (63 percent) classified their level of operation as an agency medical director, and 62 (37 percent) classified it as a program medical director. As expected, the size of the respondent's work group was larger for agency medical directors than for program medical directors (means of 116±176 persons and 40±50 persons, respectively). A one-way analysis of variance indicated the difference was significant (F=10.46, df=1, 157, p>.001).
Agency medical directors also reported somewhat greater breadth of supervision than program medical directors. Among 106 agency medical directors, 37 (35 percent) supervised only medical staff, 35 (33 percent) supervised medical and other clinical staff, and 34 (32 percent) supervised all types of staff. Among the 62 program medical directors, 25 (40 percent) supervised only medical staff, 20 (32 percent) supervised medical and other clinical staff, and 17 (27 percent) supervised all types of staff.
Considerable congruence was found between job titles and job types with regard to breadth of supervision. Forty-nine of the 61 medical directors who supervised medical staff only (80 percent) listed their job title as medical director or chief medical officer. Thirty-four of the 55 medical directors who supervised both medical and clinical staff (62 percent) also listed their job title as medical director. However, a majority of those who listed their job title as clinical director or chief clinical officer— ten of 13 respondents (77 percent)— actually supervised medical and clinical staff. Finally, of the 50 respondents who supervised all types of staff, 30 (60 percent) listed their job title as director.
Level of operation
We used the Wilcoxon signed-rank test to compare respondents' involvement in the three task domains— direct service, clinical collaboration, and administration. Program medical directors performed administrative tasks significantly less often than clinical collaboration and direct service tasks (for clinical collaboration, Z=-6.63, p>.001; for direct service, Z=-4.94, p>.001). The frequency with which they performed clinical collaboration and direct service tasks did not differ significantly. Agency medical directors performed clinical collaboration tasks significantly more often than direct service and administration tasks (for direct service, Z=-3.80, p>.001; for administration, Z=-5.40, p>.001). The frequency with which they performed direct service and administrative tasks did not differ significantly.
Compared with program medical directors, agency medical directors performed fewer direct service tasks (F=8.40, df=1, 166, p>.01) and more administrative tasks (F=12.01, df=1, 164, p>.001). The two types of medical directors did not differ in their performance of clinical collaboration tasks.
Breadth of supervision
We used one-way analysis of variance to assess the relationship between breadth of supervision and task domains. Breadth of supervision was significantly related to performance of direct service tasks (F=14.47, df=2, 165, p>.001) and administrative tasks (F=22.34, df=2, 163, p>.001) but not to clinical collaboration tasks. Pairwise comparisons indicated that medical directors who supervised all staff performed significantly fewer direct service tasks than those who supervised medical staff only (Bonferroni mean difference=-1.96, SE=.38, p> .001) or medical and clinical staff (Bonferroni mean difference=-1.64, SE=.39, p>.001). However, those who supervised medical staff only and those who supervised both medical and clinical staff did not differ significantly in the performance of direct service tasks.
Medical directors who supervised all types of staff performed more administrative duties than those who supervised medical staff only (Bonferroni mean difference=1.73, SE=.26, p>.001) or those who supervised both medical and clinical staff (Bonferroni mean difference=.89, p>.01). Medical directors who supervised both medical and clinical staff reported performing more administrative tasks than those who supervised medical staff only (Bonferroni mean difference=.83, SE=.25, p>.01).
Level of operation and breadth of supervision
A two-way analysis of variance was used to assess the joint effects of level of operation and breadth of supervision and their interaction for the three task domains. Consistent with the results of the one-way analyses, this analysis found significant main effects for level of operation and breadth of supervision both for direct service (for level of operation, F= 7.72, df=1, 162, p>.01; for breadth of supervision, F=10.54, df=2, 162, p> .001) and for administration (for level of operation, F=12.20, df=1, 160, p>.001; for breadth of supervision, F=22.26, df=2, 160, p>.001). The interaction terms were not significant.
By contrast, main effects for level of operation and breadth of supervision were not significant for clinical collaboration; however, the interaction term was statistically significant (F=3.83, df=2, 160, p>.05). Clinical collaboration tasks were performed least frequently by medical directors with the least responsibility— program medical directors who supervised medical staff only— and those with the greatest responsibility— agency medical directors who supervised all types of staff.
Discussion
We identified six discrete types of medical director positions based on level of operation and breadth of supervisory responsibility. This classification is congruent with actual job titles, supporting the idea that it is a meaningful way to understand the wide variety of roles of medical directors that psychiatrists perform in community mental health organizations.
Our findings indicate that it is clarifying to label administrative positions according to whether they are at an agency or program level of operation. Breadth of supervision is best elucidated by reserving the term "medical director" for positions with line authority over medical and nursing staff only and using the term "clinical director" for positions with supervisory responsibility for medical and other clinical staff. The term "director" is generally used to indicate supervisory responsibility for all types of staff, both administrative and clinical staff.
This analysis has created a language to facilitate discussion of the role of community psychiatrist-administrators. In practical terms, it can guide agencies in the recruitment of psychiatrists as administrators. Moreover, it can help psychiatrists negotiate their own positions within agencies more effectively, allowing them to better match authority and responsibility. We believe that such an approach will improve the functioning of psychiatrist-administrators within community agencies.
The survey assessed both job types and job satisfaction, the latter of which will be reported separately. A copy of the survey form is available from the authors and online at http:// cpmcnet.columbia.edu/dept/pi/ppf/role.html.