Approximately 4.5% of medical students enter psychiatry annually; therefore, recruiting the best and brightest candidates is one of the most challenging aspects of a residency director's job (1, 2). Most residencies spend countless hours recruiting residents. Each program has its strengths and weaknesses. Features such as geographic location and university affiliation can be hard to change and yet are potential factors in an applicant's choice. The literature shows that the impression a student has at the conclusion of an interview is a strong predictor of how the student will rank a program (3). This is good news, since the recruitment process itself, which teaches the applicant about a program, is usually under the control of the residency director. Unfortunately, the literature is mostly on the recruitment of medical students into psychiatry as a profession and not on recruitment strategies into specific programs (4—8).

The main aim of this article is to review the processes and strategies we used as a model of how to change the recruitment process at your own institution. This was prompted by our own program having to participate in the "scramble" five times, from 1999 to 2004.

Information-gathering about the recruitment process commenced in 2004, 6 months after a new residency director started, with discussions with other program directors primarily from medicine and psychiatry residency programs in the same geographic region. Questions asked during the discussions included the following:

Feedback from medical students identified weaknesses in the psychiatry core rotation. Students felt the rotation was adequate, but unexciting. Their feedback about supervision, didactics, and clinical rotations led to the following changes: the addition of three more medical student electives, problem-based lectures, and the option of 1-day experiences in psychiatric subspecialties during their 4 week core rotation.

The residents were also interviewed. They suggested a website update, the introduction of a residency newsletter, and a thank-you gift bag. They undertook the task of developing a survey of their fellow residents. The survey asked for input on the "positives" about working and living in our community. This newsletter was included in the departing thank-you gift bag, with a snack for the trip home.

A post-match follow-up survey had been developed in 2002, and this process was continued, since we felt it was a good way to judge what was working and what needed improvement (Table 1). The survey is sent with a postage-paid envelope to the upper one-half of those ranked.

Examination of the post-match surveys and faculty/resident/medical student feedback between 1999 and 2004 identified the following weaknesses:

Recruitment Modifications

On the basis of the data collected, the recruitment process was restructured:

The development of the structured "Recruitment Day" allowed for up to 10 candidates to be interviewed individually by the Residency Director, core faculty, and one or two residents. The idea of a "Recruitment Day" decreased interviewing fatigue and created excitement and momentum for those involved.

A group lunch with the residents alone was provided to give candidates an opportunity to ask questions freely and give the residents an opportunity to speak with the candidates in a more relaxed setting. Residents were educated before the Recruitment Days about the department's mission and their important role in building the program and its future. Through this, residents felt a sense of ownership and pride in the program and the residents they recruited into it.

The development of the hospital's Simulation Center has been a great addition to the institution. The psychiatry residency program has used the center for PGY 1, 2, and 3 Objective Standardized Clinical Exams. Having the applicants tour the center with the residents allows for a first-hand experience with this innovative educational activity.

From a budgetary point of view, our new recruitment process does not provide a "dinner out" for applicants but we do provide hotel accommodations for out-of-state candidates. Also, we do post-interview follow-ups with applicants via e-mail and phone, specifically reminding those applicants whom we intend to rank highly that a "second look" interview is welcomed. Candidates who came for a "second look" were able to tour other treatment settings and, often, meet with the Chairman one-on-one.

Another strategy is recruitment events. These events usually include an interactive educational component, such as having the applicant participate in a lecture, attend a resident movie night, or attend clinical rounds on the inpatient unit. This gives the applicant an experience with how he or she will learn and allows the program to assess the candidate's abilities in a variety of psychiatric venues.

The revised recruitment strategy and plan described above was implemented for the 2004—2005 recruitment season. In the past five recruitment periods between 2004 and 2009, the program filled all six positions through a combination of the AOA and NRMP matches. (The program is dually accredited by both the ACGME and the AOA, which allows this program to participate in both matches.) The Wilcoxon 2-group rank-sum test revealed that average median USMLE 1 and 2 scores of incoming residents did not significantly change, but, average median COMLEX (Osteopathic Equivalent Exam to USMLE) 1 and 2 scores increased from 35.0 to 77.5. (p<0.012; see Table 2).

The Wilcoxon-2 statistic lumps together the 5 years before the change and the 5 years after into two comparison groups, which might explain the lack of difference in USMLE scores. However, if we simply look at mean score differences between the entering classes in 1999 and 2009, the 2009 entering class mean score on the USMLE 1 is 92, and USMLE 2 is 94. This is an 18% improvement since 1999, when the mean USMLEs were 77 and 78, respectively. The COMLEX 1 mean score improved by 86%, from 11 to 77, over the same period. The AMG/IMG ratio for the program changed from 7/16 in 1999—2000, to 19/5 in 2008—2009. Although the AMG/IMG ratio is not an absolute indicator of improvement, it suggests that this program was recruiting more applicants with knowledge of the American health system, since their medical school training had been in the United States. Also, four of six entering PGY-1 years in 2009 had done a medical school rotation with us. More residents are passing the ABPN Part 1 since these recruitment changes, as well, although we still do not have 100% of graduates taking the exam immediately after graduation.

Since we initiated the post-match survey, in 2002, the response rate has varied from 30% to 50%. Geographic location has been identified as an important factor in students' ranking of residency programs (3). Our program's post-match survey confirms this. The most common first reason students did not pick our program is its inner-city location. Since this cannot be altered, the program continues to emphasize the factors we can control, which include: the diversity of the clinical experience, the strong clinical teaching, and the quality of the educational programs.

Having an effective recruitment strategy involves many aspects of a program from the medical students, residents, faculty, and institution. Recruitment requires knowing the program's strengths and being able to articulate them. The Residency Director affects the outcome by constructing the process and selecting the faculty and resident interviewers.

It appears that the changes we made in the interview day were important in recruiting a full class, with better academic skills. It was organized, smooth, and well run. Poor organization and wasted time on interview day can be viewed as a reflection of how the program is run every day (8). Providing a residents-only lunch gave an opportunity for the resident interviewers to talk with all applicants, including those they may not formally interview, and the cost savings could be invested in "second looks" by applicants.

This program has had longstanding relationships with multiple local medical schools and has about eight students per month rotating in psychiatry. Most authors believe that the psychiatry clerkship is the most important influence on a students' choosing a career in psychiatry (4—6). Factors that improve recruitment include enthusiastic supervision (4), active student participation in patient care (5), well-defined student roles (4), and students seeing psychiatric treatment as effective (5). A review of our current medical student rotation allowed us to identify and correct weaknesses. Fourth-year student psychiatry electives also correlate with higher recruitment rates (7). Three new electives were implemented as part of our "recruitment strategy."

Although our program was dually accredited by the AOA and the ACGME before and after the changes in the recruitment process, we believe our "dual" status is beneficial to recruitment. This dual status expands the applicant pool to excellent osteopathic trainees.

Review of NRMP match statistics from 2000 to 2009 shows that the number of positions offered in the match is fairly static, with an increase of 119, from 944 to 1,063, and positions filled by American graduates remained virtually unchanged, 641 (in 2004) to 643 (in 2009) (4, 9). These data suggest that the change in our program's ability to attract American graduates from 2005 to 2009 is not because of increased American graduate interest but because of the implemented recruitment strategies.

The major limitations of this study are the fact that that these data only relate to one program over a 10-year cycle, 5 years using the old strategy and 5 years using the new strategy. Nevertheless, our data show that a focused recruitment process that emphasizes the strengths of the program and involves enthusiastic resident and faculty input and participation can greatly improve a program's success in the match process.

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