Abstract
BACKGROUND: Progress testing is a longitudinal assessment method used to monitor the acquisition and retention of knowledge throughout medical training. While progress tests (PTs) have been widely adopted internationally through collaborative networks of medical schools, in Peru, their implementation has been primarily institutional. This study aimed to evaluate longitudinal trends in PT scores at a Peruvian medical school. METHODS: We conducted a longitudinal analysis using data from PTs administered annually between 2017 and 2024. The PT assessed students' knowledge based on the subjects completed at the time of testing. Scores ranged from 0 to 250 and were converted to a 20-point scale. Independent variables included number of PTs taken (1-7), year of entry into medical school (entry cohort; 2017-2024), year of test administration (2017-2024), and sex. Generalized estimating equations (GEE) were used to assess score trends over time, applying an identity link function with a Gaussian distribution and robust standard errors clustered by student ID. RESULTS: We included 1,899 test scores from 669 medical students. The mean score across all tests was 9.19 (standard deviation = 2.34). No consistent upward trend in PT scores was observed over the study period; scores decreased by 0.088 points per additional year (CI95% CI: -0.147 to -0.029, p = 0.003). Students who completed five PTs scored significantly higher than those who took four (β = 1.40; 95% CI: 0.79 to 2.01). When stratified by entry cohort, no sustained improvement in scores was observed within cohorts over time. CONCLUSION: Over an eight-year period of administering a progress test at a Peruvian medical school, student performance remained stable, with an average of approximately 50% of questions answered correctly per test. Longitudinal analysis did not reveal a sustained increase in scores as students advanced through the curriculum. This pattern may be explained by the PT design, which assesses only the content covered by students at the time of each administration, in contrast to other PTs that measure end-of-curriculum knowledge across all cohorts. Nevertheless, an increase in median scores was observed during the transition from basic science to clinical subjects.