Using the End of Curriculum Exam to Evaluate Student Readiness: Balancing the Need for Remediation Against Real-World...

This study describes the rationale, implementation, and outcomes of defining a program-specific end of curriculum (EOC) exam passing score to support fairness and transparency in student graduation and remediation decisions.

Using 6 years of data from Stanford University's Master of Science in Physician Assistant Studies program (graduating cohorts 2020-2025), predictive models of Physician Assistant National Certifying Examination (PANCE) performance were examined to explore the impact of iterative adjustments to EOC cut scores. Data from the first 4 cohorts (n = 110) were used to train predictive models, while 2 subsequent cohorts (n = 53) served as validation data. Generalized linear models using EOC scores alone (model I) and EOC plus selected end of rotation exam scores (model II) predicted PANCE performance.

Applying model I to test cohorts established empirical support for setting and adjusting EOC passing thresholds over time. Model II offered only marginal improvement over model I, suggesting that EOC performance alone sufficiently predicts PANCE outcomes.

While the program's mandated EOC remediation plan in 2024 and 2025 resulted in extended matriculation and postponed PANCE dates for 5 students, all students passed their EOC retake and subsequent PANCE. Results support using EOC scores as a reliable predictor of PANCE performance, provided appropriate remediation can be provided. Importantly, by viewing program extension decisions like mandated remediation in a consequential validity framework, the study underscores the need for PA programs to develop local, evidence-based, and equitable policies for decision-making.