Using Random Survival Forest Modeling to Predict Research and Career Outcomes for the National Institutes of Health's Medical Scientist Training Program Trainees

• Activity Number: 181 - Contributed Poster Presentations: Government Statistics Section
• Type: Contributed
• Date/Time: Monday, July 31, 2017 : 10:30 AM to 12:20 PM
• Sponsor: Government Statistics Section
• Author(s): Katie Patel* and Deepshikha Roychowdhury and Katrina Pearson
• Companies: National Institutes of Health, Office of Extramural Research and National Institute of Health, Office of Extramural Research and National Institutes of Health, Office of Extramural Research
• Keywords: National Institutes of Health ; Research Outcomes ; Career Outcomes ; Random Survival Forest ; Trainees

Abstract

The Medical Scientist Training Program (MSTP) is a training program supported by the National Institute of General Medical Sciences (NIGMS) within National Institutes of Health (NIH) to train students with outstanding credentials and potential to pursue careers in biomedical research and academic medicine. This study primarily uses NIH's administrative data to analyze research and career outcomes of MSTP trainees. The outcomes include: applying for, or receiving, a NIH Research Project Grant (RPG), or securing a faculty appointment at a university or medical college after completing the training program. Random survival forest modeling was used to determine the main predictors from a list of predictors, including trainee age, gender, matriculation year, years since training program started, and some organizational characteristics, associated with each outcome. The same method was applied for a cohort of NIH T32 trainees with MD-PhD degrees to compare outcomes, and predictors associated with those outcomes for those trainees vis-à-vis MSTP trainees.