The content and value of letters of recommendation in the resident candidate evaluative process

John B Fortune
DOI: https://doi.org/10.1016/s0149-7944(01)00538-4
Abstract:Purpose: To determine the content and value of letters of recommendation (LOR) in the selection of residents for general surgery training. Methods: During the 1999 application process, we evaluated the source, content, and usefulness of 966 LOR for 288 viable applicants to our program. Each letter was reviewed to determine if the author had clinical contact with the student and whether the student's fund of knowledge, clinical achievements, psychomotor skills, and personal characteristics were described. The use and type of a "single word" summary was also noted. The program director then determined if the LOR was helpful in evaluating the student. Participants: All LOR in the application materials of candidates applying to the General Surgery Residency of the University of Arizona. Results: All of the candidates submitted 3 LOR and 34% submitted 4. The average length was 13.7 (range, 4 to 36) sentences. Department chairs wrote LOR for 66% of the students, and these comprised 20% of the total LOR. Program directors, division chiefs, and clerkship directors wrote 16%, whereas other general surgery faculty contributed 40%. In 17% of the LOR, no evidence of direct clinical supervision could be found. Personal characteristics were described in 89% of the letters, but reference to psychomotor skills occurred in only 27%. In 58% of the LOR, a single adjective was used to describe the students; of these, "outstanding" (or equivalent) was used in 37% and "excellent" (or equivalent) was used in 38%. Meaningful comparison to student colleagues appeared in 11%. In our program director's opinion, 24% of the LOR were helpful in the evaluation process. The department chair's letters were helpful only 19% of the time. Conclusions: Letters of recommendation are imprecise indicators of student performance for resident candidate evaluation. A "standardized LOR" may help provide a consistent and more objective evaluative tool, and the APDS should consider its development for future selection processes.
What problem does this paper attempt to address?