Purpose Despite the importance of early intervention and remediation, the relatively short duration of physician assistant education programs necessitates the importance of early identification of at-risk learners. This study sought to ascertain whether machine learning was more effective than logistic regression in predicting remediation status among students, using the limited set of data available before or immediately following the first semester of study as predictor variables and academic remediation as an outcome variable. Methods The analysis included one institution and student data from 177 graduates between 2017 and 2019. We employed one data mining model, random forest trees, and compared it to a traditional predictive analysis method, logistic regression. Due to the small sample size, we employed leave-one-out cross-validation and bootstrap aggregation. Results Data provided evidence that the random forest algorithm correctly identified individuals who would later experience academic intervention with a 63.3% positive predictive value, whereas logistic regression exhibited a positive predictive value of 16.6%. Conclusions This single-institution study indicates that predictive modeling, employing machine learning, may be a more effective means than traditional statistical methods of identifying and providing assistance to learners who may experience academic challenges.

Purpose Physician Assistant Education Association (PAEA) End of Rotation™ exams are used by programs across the country. However, little information exists on the predictive ability of the exams' scale scores and Physician Assistant National Certifying Exam (PANCE) performance. The purpose of this study was to evaluate End of Rotation exam scores and their relationship with poor PANCE performance (PPP). Methods In an IRB-approved, multi-center, multi-year study, associations between PAEA End of Rotation exam scale scores and PANCE scores were explored. A taxonomy of nested linear regression models with random intercepts was fit at the program level. Fully adjusted models controlled for year, timing of the exam, student age, and gender. Results Fully adjusted linear models found that 10-point increases in End of Rotation exam scores were associated with a 16.8-point (95% confidence interval [CI]: 14.1-19.6) to 23.5-point (95% CI: 20.6-26.5) increase in PANCE score for Women's Health and Emergency Medicine, respectively. Associations between exams did not significantly vary (P = .768). Logistic models found End of Rotation exam scores were strongly and consistently associated with lower odds of PPP, with higher exam scores (10-point increase) associated with decrements in odds of PPP, ranging between 37% and 48% across exams. The effect estimate for the Emergency Medicine exam was consistently stronger in all models. Conclusions PAEA End of Rotation exam scores were consistently predictive of PPP. While each End of Rotation exam measures a specialty content area, the association with the overall PANCE score varied only by a change in odds of low performance or failure by a small percentage. Low End of Rotation exam scores appear to be consistent predictors of PPP in our multi-center cohort of physician assistant students.