Factors associated with the career choices of hematology and medical oncology fellows trained at academic institutions in the United States

Abstract

Purpose: Factors that influence hematology-oncology fellows' choice of academic medicine as a career are not well defined. We undertook a survey of hematology-oncology fellows training at cancer centers designated by the National Cancer Institute (NCI) and the National Comprehensive Cancer Network (NCCN) to understand the factors fellows consider when making career decisions.

Methods: Program directors at all NCI and NCCN cancer centers were invited to participate in the study. For the purpose of analysis, fellows were grouped into three groups on the basis of interest in an academic career. Demographic data were tested with the Kruskal-Wallis test and χ² test, and nondemographic data were tested by using the multiscale bootstrap method.

Results: Twenty-eight of 56 eligible fellowship programs participated, and 236 fellows at participating institutions responded (62% response rate). Approximately 60% of fellows graduating from academic programs in the last 5 years chose academic career paths. Forty-nine percent of current fellows ranked an academic career as extremely important. Fellows choosing an academic career were more likely to have presented and published their research. Additional factors associated with choosing an academic career included factors related to mentorship, intellect, and practice type. Fellows selecting nonacademic careers prioritized lifestyle in their career decision.

Conclusion: Recruitment into academic medicine is essential for continued progress in the field. Our data suggest that fewer than half the current fellows training at academic centers believe a career in academic medicine is important. Efforts to improve retention in academics should include focusing on mentorship, research, and career development during fellowship training and improving the image of academic physicians.

Fig 1.

(A, B) Relationship between career and survey questions. The relationship between the response to each survey question and to “How important is an academic career to you?” are summarized in the table (C). The percents in each row sum to 100%. For example, of the respondents who indicated that an academic career was not important, 74.3% also responded that basic science research was not influential to them, 22.9% responded it was influential, and 2.8% responded it was very influential. The top line of the table (C) is represented in the top line of the smaller graph (B). Looking down the columns of the large graph (A), we see the trend that the more important an academic career is to someone, the more likely that person is to respond that basic science research is very important. This trend is visible in the graphic as a stair-step pattern. Had the percentages in each column been the same, there would not be a relationship between the response to the survey question and the question “How important is an academic career to you.” The graphic for a table with no relationship would not have this stair-step pattern but would still indicate how important each survey question is, regardless of interest in an academic career. The large graph (A) represents 35 of these tables; (*) indicates survey questions that were statistically related (in a proportional odds model) to interest in an academic career after accounting for differences in age, sex, fellowship year, current educational debt, and whether the respondent had published research.