A search of PubMed to identify published studies on mentorship in the orthopaedic literature was performed with use of the search terms "mentor," "mentorship," and "mentoring." Six articles were identified. These studies discussed the issue of mentorship and its challenges, yet none investigated the prevalence of mentoring among orthopaedic residents or the perceived value of mentoring from the residents' perspective4-6. The AAOS staff and the AAOS Washington Health Policy Fellows designed a census survey of residents with regard to their experience in and opinion of mentorship programs and the prevalence of such programs. A web-based survey was developed and sent to all orthopaedic residents within the AAOS database (see Appendix). The questions were formatted and the data collection tool was developed by the Department of Research and Scientific Affairs of the AAOS to ensure a proper survey format and to minimize response bias7-9. The final survey was reviewed by the AAOS Washington Health Policy Fellows to ensure the relevance and clarity of the questions. The survey utilized a real-time modification of the questions offered to the respondent depending on the respondents' answers. Thus, questions that were not pertinent because of initial responses were eliminated from the survey. Respondents were asked to answer most questions by rating their satisfaction with or the importance of various aspects of their mentoring environment or experience on a scale of 1 (not at all) to 5 (extremely). No formal definition was provided to the respondents regarding what constituted mentoring or a mentoring program. Respondents made these distinctions themselves.
In January 2008, residency program directors were contacted to apprise them of the survey and to ask for their assistance in securing responses from the residents. One week after the announcement to the program directors, the survey announcement and link was distributed to 2210 residents from postgraduate years (PGY) 1 through 5 by means of electronic mail and was mailed through the U.S. Postal Service to 1445 residents without e-mail addresses on file, for a total distribution of 3655 surveys. A reminder e-mail was sent two weeks later to the nonrespondents with e-mail addresses. No repeat postal mailing of the survey was performed. Data collection was concluded four weeks after the initiation of the survey. The respondents were required to provide their e-mail address as part of the survey, but they were assured of the confidentiality of their responses. This information was used to confirm that responders were invited to participate, as well as to ensure that an individual did not respond more than once to the survey. Identifiers were also used to compare the representativeness of the response group to the entire population of residents in terms of postgraduate year. At all times, the survey results were maintained with strict confidentiality. After verification, the identifiers were removed so that no responses could be tracked to the respondent. The data were coded, tabulated, and analyzed for accuracy by the Department of Research and Scientific Affairs of the AAOS. The data are summarized and presented as the percentage of responses to items and the mean ratings where applicable. When it was appropriate, statistical analysis was performed with the use of analysis of variance and independent group t tests.
The survey was distributed to a total of 3655 residents, with a response rate of 14% (506 residents). According to AAOS demographic information, the distribution of residents in each program year across the nation is 20% except for PGY-2 (21%) and PGY-5 (19%). The respondents were distributed through all years of training, with 11% in PGY-1, 17% in PGY-2, 22% in PGY-3, 23% in PGY-4, and 26% in PGY-5. Thus, the respondents represented a smaller proportion of PGY-1 residents (11% compared with 20%) and a greater proportion of PGY-5 residents (26% compared with 19%) than are found in the AAOS resident database.
Slightly over half (52%) of the 506 respondents had a mentorship system within their residency training program; 26% had a formal program, while 26% had an informal system. Fifty-one percent (258) of the 506 respondents had a mentor at the time of completing the survey. The respondents with mentors reported that 52% of those mentors had been assigned and 48% had been sought out by the resident. Of the residents with an assigned mentor, 70% were assigned to a mentor during PGY-1; 22%, during PGY-2; and the remaining 8%, during PGY-3 to PGY-5. Of the residents who obtained their own mentor, 19% obtained their mentor during PGY-1; 33%, during PGY-2; and 41%, during PGY-3. Only 7% sought out a mentor during PGY-4, and no respondent obtained a mentor during the PGY-5 year. Some residents noted that they were assigned a mentor as PGY-1 and PGY-2 residents and then selected their own mentor as PGY-3 or PGY-4 residents. Forty-one percent (106) of the 258 respondents with mentors met with their mentor every six months, 20% met monthly, while 15% met annually. Seven percent stated they met weekly.
Residents with formal mentoring programs and those with informal programs had significantly higher rankings for satisfaction with their mentoring environment than did those without a mentoring program (3.9 and 3.5, respectively, compared with 2.8; p < 0.01 for both) on a scale of 1 to 5. Only 17% (eighty-six) of all 506 respondents were extremely satisfied with their mentoring program or environment, 28% (142) were somewhat satisfied, 31% (157) were neither satisfied nor dissatisfied, and 24% (121) were somewhat or very dissatisfied. The satisfaction of the respondents with their mentoring environment also varied by whether a mentor was assigned or chosen. Respondents with a self-selected mentor had a higher ranking for satisfaction with their mentoring environment than did those with an assigned mentor (4.0 compared with 3.6; p < 0.01). The respondents with self-selected mentors rated them as more helpful than those with assigned mentors in aiding with career decisions (4.2 compared with 3.2), for supporting the educational experience (4.3 compared with 3.4), for professional development (4.3 compared with 3.2), and in providing networking opportunities (4.0 compared with 2.8) (p < 0.01 for all values) (Fig. 1). The majority (89%) of the residents with mentors, whether assigned or self-selected, felt that their mentor provided them with advice regarding career, employment, or difficult cases.
Nearly all respondents (95%) believed that mentoring should be a part of their residency program. Of that group, 20% thought that it should be required and monitored; 35%, that it should be required but not monitored; 40%, that it should be offered with resources available but not required; and 5%, that it should be a resident's decision to seek out a mentor if the resident desired. Thus, while 95% of the residents responding to the survey believed that mentoring should be part of their residency program, only 55% thought that mentoring should be required or mandated in some form. While 24% of residents felt that mentors were critical to their training, 72% believed that mentors were beneficial but not critical. Only 4% thought that mentors were neither beneficial nor critical to training. Thus, 96% of the residents who responded to the survey felt that mentors were either critical or beneficial to their training. Characteristics of mentors that were important to orthopaedic residents in the selection of mentors included approachability (78%), subspecialty interest (65%), similar research interests (46%), a good reputation as a mentor (32%), and an ideal practice environment (23%).
The level of importance placed on having a mentor varied by the presence or absence of a mentor. The respondents with a mentor, defined as those who had a current mentor, those who previously had a mentor, or those who anticipated having one, rated the importance of having a mentor higher than did those who did not have one (4.0, 4.2, 3.9, and 3.3, respectively; p < 0.01 for all values). However, the perceived importance of having a mentor was not found to vary by whether the residency program had a mentoring program in place (p = 0.377). In other words, the importance of having a mentor was rated similarly by residents with a mentor irrespective of whether they had a formal or informal mentoring program.
As in the days of Hippocrates, the art and science of surgery has been passed from generation to generation through a student-teacher apprenticeship model. The model of a handpicked apprentice being educated by an experienced surgeon was formalized in the United States in the late 1800s by William Halsted10. Although this tradition of training was effective in transferring knowledge from teacher to student, several authors have pointed out that it has distinct disadvantages, including being paternalistic, creating extended work hours, promoting intense competition among students, and having the possibility of indefinite tenure10,11. While Halsted's influence in the American model of surgical training continues to a degree, some of the strengths of this model have been eroded by the modern graduate medical system. For example, a resident goes through a matching process and is therefore not handpicked by a master surgeon to be a protégé. That is, the personal investment in the selected resident is less than if the surgeon had chosen the individual specifically. Moreover, even if a resident were handpicked in the current system, the individual would rotate with many staff during training for short periods of time, thus again not fully experiencing the benefits of the master surgeon as teacher and exemplar. One modern model of transferring knowledge from a master surgeon to an apprentice that can capture some of the strengths of the Halstedian model is mentorship. Wilson defined mentorship "as a relationship, whether formal and/or informal, between a novice and one or more senior persons in the field for the purposes of career and personal development and preparation for leadership."6 Hill stated that the goal of mentoring is "to provide a young aspiring professional with a tangible and immediate role model."12
We showed that only 26% of the responding residents reported having formal mentorship programs. The reasons for this low rate are legion, and several authors have attempted to identify the causes. One proposed reason has been that clinical revenue previously could subsidize the uncompensated physicians' costs of time and effort to mentor10,13,14. However, in the current financial climate of residency education, faculty members have less time available for teaching and education because of increased pressures to be financially solvent and academically productive10,15. Thus, while academic institutions stress the importance of teaching and education, faculty performance and qualification for academic promotion are measured predominantly by research or clinical productivity and minimally by educational productivity10,16.
Pellegrini noted that many common personality traits found in those pursuing surgical careers are antithetical to the traits of a good mentor5. Two of the traits he identified were competitiveness and self-reliance, while asking for help or guidance was considered to be a sign of weakness. Interestingly, in the present study, we demonstrated that approachability was the most common characteristic, listed by 78% of the respondents who selected their own mentor, as a reason the mentor was chosen. This suggests that the faculty members who are perceived by the residents as not having the traits identified by Pellegrini are more sought after as mentors.
Even though the modern mentorship model for orthopaedic education has a relatively low prevalence among the resident respondents, other investigators have shown that it is an effective and important paradigm in medical education. In a systematic review of the medical literature, Sambunjak et al. noted that mentorship had a substantial impact on personal development, career guidance, career choice, and research productivity3. In a study conducted at the University of California, San Francisco, medical students rated having a research mentor as the most important factor in influencing their specialty choice3,17. A survey among pediatric residents showed that nearly 80% of the respondents felt that having a mentor was very useful or critical to surviving residency2. Similar to those studies, we found that 96% of responding orthopaedic residents thought that mentoring and mentorship were important elements of their education and professional development. In fact, 95% of the respondents believed that mentoring should be part of their residency program.
The present study substantiates the hypothesis that most residents would want mentors and mentorship programs but that many did not have them. Other authors have made this assumption without data, and they have commented on ways to improve the mentoring environment in orthopaedic education5,6,12. Pellegrini identified two critical attributes to make an effective surgeon mentor: the ability to "revel in succession planning for the next generation" and the "skill of deriving personal satisfaction from the accomplishments of the protégé as a direct and unspoken extension of the mentor's own achievements."5 Wilson proposed guidelines for developing an effective mentoring program, including quarterly mentoring meetings, collecting subjective and objective data regarding the accomplishments of the mentoring relationship, having mentors participate in workshops to improve mentoring skills, ensuring that mentors have sufficient experience and perspective in their field, and ensuring that core values and career plans are established in the first few sessions6. Both Wilson6 and Hill and Boone12 believed that the mentor should be chosen or assigned on the basis of the matching of professional interests and the personal attributes of the mentee. We found that 65% of the respondents who selected their own mentors reported that an element in their selection of a mentor was finding one with similar interests. Importantly, those who selected their mentors were more satisfied with their mentoring experience, thus substantiating the suggestions of Wilson as well as Hill and Boone that mentors and mentees should be matched because of similar interests.
Only 44% of the respondents were either satisfied or extremely satisfied with their mentoring program. This begs the question as to why the respondents had differing levels of satisfaction. One factor was that only 52% had a mentorship program or a mentor. An additional important factor was the respondents' ability to choose their mentor as those who did so were significantly more satisfied. This seemed to be because the resident was able to choose a mentor who was approachable, who had similar interests, and who had good mentoring skills or a strong mentoring reputation. Moreover, those who had selected their mentor found those mentors to be significantly more helpful in aiding in career decisions, supporting their educational experiences, fostering professional development, and providing networking opportunities. Thus, having a mentor who actively participates in the training and development of the mentee was an important factor in the increased satisfaction among the respondents who had selected their own mentor.
This study had several weaknesses. The first is the low rate of response of 14% to the survey, although this rate is not uncommon among orthopaedic residents. In a recent survey of orthopaedic residents regarding work-hour limits, which had been distributed by mail, Kusuma et al. received a similar response rate of 13.2%18. Similar surveys of medical trainees and faculty across various medical disciplines regarding mentors have reported response rates as low as 5%3. Although we were able to identify significant differences in these data among those with and without mentors, or mentor programs, and those who selected their own mentors or were assigned mentors, a more aggressive data collection strategy could have secured a higher response rate and strengthened the results and the generalizability of the conclusions. Additionally, this survey was voluntary, and the data obtained might be unduly influenced by those who chose to participate in the survey because of a state of either considerable satisfaction or dissatisfaction regarding their mentoring environment. A follow-up with nonrespondents would provide further information on the perspective of this group.
Several aspects of this study warrant further exploration. An area of further study would be to determine what the actual effects of mentoring are on the residents. This could be investigated by determining a resident's academic and clinical performance, measured by such things as Orthopaedic In-Training Examination scores, board examination scores and pass rates, success in practice, and longevity of employment positions. Additionally, it would be interesting to determine whether mentors influence residents to pursue careers in academic or private practice. If mentors were shown to encourage the pursuit of academic careers, this would be a reason to establish successful mentoring programs. Moreover, because of the constraints of administering a concise survey to encourage its completion, many other interesting aspects of mentoring were not investigated. It would be important to investigate more deeply the qualities and attributes that make an effective mentor and educator and to study the reasons for the dissatisfaction expressed by many of the residents regarding their mentoring environment.
While making specific recommendations regarding the process of establishing a mentorship program is beyond the scope of this paper, we can report that residents with mentors find them to be a valuable resource. Nearly all respondents thought that mentoring should be part of their residency training program. The residents who selected their own mentors were more satisfied with their mentoring than were the residents with assigned mentors. Despite not having a mentor, the residents without one ranked the value of mentorship very highly. The residents with formal mentorship programs were the most satisfied. We believe that orthopaedic residency training programs should consider instituting formal mentoring programs, encourage active mentoring of residents, encourage residents to seek mentors early in residency, and allow residents to select their own mentors. Furthermore, we recognize that mentoring is a challenging task and recommend that faculty actively develop mentoring attributes and skill sets. Mentoring should be recognized by the educational institution as a professional activity valued similarly to other activities in academic medicine. Individuals desiring more information regarding mentoring and how to implement a mentoring program are referred to the works of Pellegrini5 and Wilson6 for further suggestions.
Note: The authors thank Heidi Schmalz of the AAOS Department of Research and Scientific Affairs for her help with the statistical analysis, survey format, and management of the collected data. They also thank Jeanie Kennedy of the AAOS Washington office for her assistance in designing the survey and in the preparation of the manuscript.
The questions used in this study are available with the electronic versions of this article, on our web site at (go to the article citation and click on "Supplementary Material") and on our quarterly CD/DVD (call our subscription department, at 781-449-9780, to order the CD or DVD).