Mentorship is an important part of a successful career in medicine but is difficult to define. A mentor is not simply a role model. The mentor/mentee relationship has distinct characteristics: active interest and engagement of both parties, a commitment to mutually agreed upon objectives, the devotion of necessary time and energy, and the expectation that the mentee will become increasingly independent (1, 2). Although there is a general paucity of research on mentorship, an exploration of the recent scientific literature (3–16) reflects an increased interest in this area. Clinicians, clinical fellows, and academic faculty with mentors have all reported higher levels of career satisfaction compared with their colleagues lacking mentors (12, 17–20).

Mentored residents are nearly twice as likely to describe themselves as having excellent career preparation (12). In a study of U.S. maternal-fetal medicine fellows (21), those who were satisfied with their training were significantly more likely to have a mentor (88.2% versus 55.8%, p<0.001) and/or a faculty adviser (87.3% versus 64.9%, p=0.002). In addition, those fellows in the study with a mentor or faculty adviser were more likely than those without mentors to recommend their fellowship. In a survey (22) of 3,031 faculty from 24 medical schools, those faculty who clearly identified mentors had significantly higher career satisfaction scores than those without mentors. Physicians with mentors publish more frequently in peer-reviewed journals and receive more research grants (18, 23, 24). Levinson et al. (18, 25) surveyed 558 women ages 50 years or younger listed in the faculty roster of the American Association of Medical Colleges (AAMC). Those who had a mentor averaged 13.1 publications, versus 10.3 for the nonmentored group. In the study of maternal-fetal medicine fellows (21), the presence of an identified mentor on the faculty increased the likelihood of research completion from 52.3% to 83.5% (p<0.001).

Despite the important role that mentors play in the development of an academic career, surveys of internists show that only 39% to 61% feel adequately mentored (12, 14, 18, 25, 26). Ramanan et al. (26) surveyed 8,397 internists with Harvard Medical School appointments and showed that only 39% had an identified mentor, despite the fact that 85% of respondents stated they were committed to a career in academic medicine. In addition, only about half of respondents with identified mentors reported satisfaction with their current mentoring (26). Sambunjak et al. (14) reviewed 43 articles on mentorship in medicine and found that fewer than 50% of medical students and fewer than 20% of faculty reported having a mentor.

A review of the research corpus leads to the conclusion that having a mentor has positive effects on career development. However, most of the studies have had small sample sizes, were not blinded, and have been performed in specialties that may not generalize to psychiatry.

In the absence of research on mentoring psychiatry residents, APA Committee of Residents and Fellows took advantage of an opportunity to conduct a cross-sectional survey of psychiatry chief residents attending the National Chief Residents Leadership Program conferences held in 2004 and 2005. The goal of the survey was to answer the following questions: What proportion of chief residents report having a clearly defined career development mentor? What is the association between having a clearly defined mentor and feeling adequately mentored? Which residents and residency program characteristics are associated with better mentoring outcomes? Is having a mentor or feeling adequately mentored related to a resident’s perception of feeling prepared to practice psychiatry independently at the end of residency?

We surveyed 229 chief residents from Accreditation Council for Graduate Medical Education (ACGME) accredited psychiatric residency training programs who attended the APA National Chief Residents Leadership Program in 2004 and 2005. Each residency training director nominated one or two chief residents to attend the conference. Travel, lodging, and incidental expenses were paid for by an unrestricted educational grant funded by Eli Lilly and Company.

The survey assessed domains such as demographics, home and family life, recruitment into psychiatry, quality of didactics received, professional interests, mentorship, and transition to postresidency. The survey also asked five questions about mentoring and self-perceptions regarding preparedness to practice psychiatry upon graduation: Do you have a clearly defined career development mentor? Who provides mentorship in your program? Do you feel you are adequately mentored through your residency program? Given your schedule, what do you consider is an ideal frequency to meet with your primary mentor? How well prepared do you believe you will feel to practice psychiatry independently when you graduate your residency?

The survey was piloted by its author on members of the APA Committee of Residents and Fellows to assess question content, relevance, length, and coherence. Minor changes to questions were made after the pilot phase, primarily to clarify wording on several questions. Surveys were administered prior to the lunch break on the first day of the APA National Chief Residents Leadership conference each year. In 2004, 173 chief residents attended the program, and 124 returned the survey, for a response rate of approximately 70%. In 2005, 195 chief residents attended the program, and 105 returned the survey, for a response rate of approximately 53%. In total, 229 surveys were completed by separate individuals over 2 years.

Demographic data are summarized in Table 1. Approximately equal numbers of men and women completed the survey, and diverse ethnic backgrounds were represented. Sixty-nine percent of the chief residents who completed the survey were from academic training programs, and 31% were from community-based training programs. More than one-third of respondents (37%) were international medical graduates (IMGs). Nearly half (49%) of residents reported that they did not have a clearly defined mentor, while 39% reported that they did not feel adequately mentored. Most residents (46%) stated that the ideal frequency of meeting with their mentor was weekly.

We assessed for associations among resident characteristics, residency program descriptors, and responses related to resident mentoring experiences. We found that residents’ reports of having a clearly defined career development mentor and feeling adequately mentored were not associated at a significant level with gender, race, marital status, having children, being a U.S. medical school graduate versus an IMG, or attending residency at an academic versus community-based training program. Chief residents who reported writing a peer-reviewed publication were significantly more likely to report having a clearly defined mentor.

Our a priori hypothesis was that adequate mentoring would be associated with a self-report of greater preparedness to practice psychiatry after residency. We investigated the effects of the dichotomous variables, “having a clearly defined mentor who helps you in your career development” (no/yes) and “feeling adequately mentored through your residency program” (no/yes), on the outcome variable, “how well-prepared do you believe you will feel to practice psychiatry independently when you graduate residency?” The two mentoring variables (“having a clearly defined mentor” and “feeling adequately mentored”) were found to be highly associated with each other (χ²=42.98, df=1, p<0.0001), such that residents with a clearly defined mentor are twice as likely to feel adequately mentored than those without a clearly defined mentor [relative risk=2.10, 95% CI (1.63, 2.70); OR=7.03, 95% CI (3.82, 12.96)]. Therefore, we separately estimated the effect of each of the two mentoring variables on preparedness for practice. We employed cumulative logistic regression modeling because “preparedness” was defined in the survey as an ordinal variable with seven levels, ranging from 0=poorly prepared to 6=extremely well prepared. We adjusted for baseline characteristics, including gender, race (Caucasian versus non-Caucasian), academic versus community-based training program, and U.S. medical school graduate versus IMG. The decision to adjust for these characteristics was based on prior studies (4, 5, 8, 12, 14, 15, 17, 26, 27) that indicated these variables could affect resident perceptions of preparedness to practice independently.

The results for the two cumulative logistic regression models are shown in Table 2. Residents who identified themselves as having a clearly defined mentor showed two times higher odds of perceiving themselves as better prepared to practice psychiatry independently at graduation than those without a clearly defined mentor (OR=1.99, 95% CI=1.18–3.36, p=0.01). Residents who reported feeling adequately mentored also had higher odds of feeling prepared to practice (OR=1.65, 95% CI=0.97–2.79, p=0.06), although this association only trended toward significance. In both models, chief residents who attended a U.S.-based medical school reported more than twice the odds of perceiving themselves as prepared to practice psychiatry independently at graduation compared with IMGs (p<0.01 in each model).

Our findings that less than half of surveyed psychiatry chief residents report having a clearly defined mentor during residency replicate similar studies for internists (1, 13, 15, 19, 26). After adjusting for other variables, our study indicates that psychiatric chief residents with mentors report feeling better prepared to practice after residency compared with those lacking mentors. For cost and logistical reasons, there was no attempt to incorporate standardized, objective measures of residency performance, such as the Psychiatry Resident-In-Training Examination (PRITE), or to longitudinally track the career progression of respondents through follow-up surveys. Therefore, our study does not address whether the perception of feeling better prepared to practice translates into genuinely being better prepared to practice. Given that all responses are based on self-report, we do not know exactly if self-perception correlates with measured outcomes of preparedness. Clearly it would be informative to test the reliability of the results presented here with objective proxies of preparedness to practice (e.g., PRITE scores, National Board of Medical Examiners pass rates, psychodynamic psychotherapy competency exam scores) or longitudinal follow-up to assess objective level of preparedness to practice independently after graduation.

Our study is limited by the inability to survey all chief residents nationwide. We achieved a 62% response rate among those chief residents who registered to attend the conference. Yet if we estimate there to be two chief residents per program, our sample represents approximately one-third of U.S. chief residents. The lower response rate in 2005 as compared with 2004 was likely due to fewer staff members being available to distribute and collect the survey in the second year. Even though our sample was nearly half non-Caucasian and half women, the percent of African American and Hispanic respondents was low and may reflect an underrepresentation of these groups in medicine in general.

Another limitation is that the psychometric properties of the different questions have not been studied. Rigorous qualitative and quantitative assessment of the construct validity of each question was beyond the scope of the survey designer, and the survey represents a de novo attempt to capture qualities of psychiatric residency experiences. Furthermore, the chief residents surveyed may not be representative of psychiatric residents at large. Chief residents are often at the top of their peer group and must interact closely with faculty by virtue of their position and responsibilities. Therefore, they may have greater opportunities for mentoring by faculty than other residents within their program. For this reason, we posit that psychiatric residents as a whole receive less mentoring, and that this study represents an overestimation of mentoring for these residents. Thus, one must use caution in extending the findings of this study to psychiatry trainees who are not chief residents.

Our study did not address the responsibilities of the mentee. What is the role of the mentee? Is the mentee’s responsibility simply to meet with the mentor, or is it broader? Common sense suggests that productive mentorship requires a good fit between mentor and mentee. Lee et al. (10) published recommendations for both mentors and mentees. They suggest that the mentee should create an annual professional development plan, meet regularly with his or her mentor, ask for support and guidance, and provide an annual review of the mentor to the program director or department chair.

Finally, we were also limited by the small number of questions devoted to mentorship in the larger survey from which our data emanate. More probing questions about different aspects of preparedness, such as specific skills, attitudes, and knowledge, could reveal greater information about the role of mentorship in residency. Such questions could highlight important differences in the characteristics of residents with and without mentors. Furthermore, it is unclear if formal mentorship programs would lead to the establishment of more mentor-mentee relationships and if mentoring relationships from a formalized program established within a residency would successfully benefit the residents’ development. Some positive outcomes have been shown from formal mentorship programs in both objective measures (publications, funded proposals) and qualitative assessments of the mentees (28–31). However, these outcomes are not universal and are not attributable solely to these programs.

Our results show that although only half of psychiatry chief residents in our survey report having a clearly defined mentor, those with mentorship perceive themselves to be better prepared to practice psychiatry upon graduation. Structural changes in psychiatric residency training that enhance the provision of skilled mentorship for more residents could contribute to substantial improvement in residents’ perceived preparedness to practice and could enhance the educational experience of residents. The roles played by mentors and mentees must be further clarified and studied in order to institute effective mentorship programs that best address the needs of residents.

At the time of submission, the authors declared no competing interests.