A Data-driven Approach to Selecting Pulmonary and Critical Care Fellows for Interviews

Abstract

Background: Training programs around the country receive many applications every year with a limited time window to send out invitations for interviews. This poses major barriers to conducting holistic application reviews. Objective: To create and implement a data-driven selection process that promotes holistic reviews within a tight timeline to select applicants for invitation to interview. Methods: In 2022, we conducted a survey of clinical faculty and fellows to ascertain the experiences, attributes, metrics, and characteristics deemed important for success in our training environment. We formed a selection committee and used the survey results to construct an automated screening tool and a faculty-completed application review form with resultant data summarized to aid in data-supported selection decisions. Results: Among the 60 survey respondents, 42 (71%) were faculty members and 17 (29%) were current fellows. The six most important items for trainee success fell under the domain of leadership attributes. Survey results were used to create a weighted screening score that was used for initial triaging of applications and a weighted faculty-reviewed application score standardized to each faculty reviewer and used to select applicants for interviews. These sequential scores allowed a holistic review of 306 applications by 20 faculty in a time-sensitive manner. Conclusion: Survey methods can be used to generate weighted and standardized application assessment tools that allow data-supported fellow selection decisions and facilitate holistic application reviews.

Keywords: data science; data-driven decision making; postdoctoral admissions; postdoctoral fellows training; program improvement.

Figure 1.

Timeline and diagram of admissions workflow, combined pulmonary and critical care medicine, 2022. ERAS = Electronic Residency Application Service.