165 Predicting Success: A Mixed Model of KL2 Trainee Profiles and Outcomes

Abstract

OBJECTIVES/GOALS: Penn State CTSI supports KL2 career development awards for faculty seeking to become leaders in clinical and translational research. CTSAs can benefit from a better understanding of KL2 applicant profiles and trainee outcomes. Predictive modeling of KL2 records provides insights into institutional processes and continuous improvement goals. METHODS/STUDY POPULATION: Collecting KL2 application records at Penn State CTSI from 2017 to 2023, comprising both accepted and not accepted candidate profiles, this study used a generalized logistic mixed model with binomial distribution to understand the factors predictive of KL2 trainee acceptance, (n=47). The following factors were modeled as potentially predictive of scholars’ acceptance: Institution-specific Processes—Campus; Terminal Degree Type; College of Residency, Applicant Demographics and Portfolio—Minoritized or Protected Groups; Mean Application Score; Rurality Focus; Gender, and Outcomes—Post-Program h-index. RESULTS/ANTICIPATED RESULTS: Only Campus and Degree were significant factors predictive of trainee acceptance (r<.0001), with a particular campus and the MD degree-designation both exerting selectional pressures on acceptance rates. Applicant demographics were not significant historical factors in selection despite the most recent trainee cohort comprised of all women. Similarly, while our CTSA focuses on rural inequality and accessibility, a research proposal focused on rurality was not a significant factor for acceptance. Notably, NIH-scaled application scores and post-program h-indices were not significant for accepted and non-accepted applicants. DISCUSSION/SIGNIFICANCE: The absence of applicant-focused selectional pressure is striking—Penn State CTSI does not significantly select for gender, URM, or URP status. Administration is now empowered to intentionally engage, recruit, and retain from our other affiliated campuses and colleges.