Predicting and Explaining Professionalism Issues in Medical Students with Time-to-Event Models: An Exploratory Longitudinal Learning Analytics Study

Medical professionalism links physicians’ duties with societal expectations and is fundamental to competent healthcare. Early unprofessional behavior in medical school has been linked to future professionalism issues in clinical practice. However, the longitudinal dynamics and potential predictors of professionalism issues among medical students remain insufficiently explored, with prior research largely focusing on group-level analyses. To address this gap, we developed a time-to-event analytical framework to examine the timing of professionalism issues and factors associated with their occurrence. We analyzed a longitudinal dataset comprising sociodemographic, pre-admission, and in-school performance variables from a single medical school. Using survival analysis, the results suggest that several factors might be associated with the risk of professionalism issues within this institutional context. Specifically, higher pre-admission writing quality scores and academic examination scores were associated with lower risk. In contrast, higher scores in certain pre-admission Multiple Mini Interview (MMI) stations (e.g., Professional role awareness and Professional integrity) were linked to increased risk. Overall, this study proposes a preliminary analytical framework for examining professionalism issues using longitudinal educational data. The framework may support future research on professionalism development in medical training and may provide a methodological approach that could be adapted for use in other educational contexts.