Scholarlink

Background/Objectives <p>Standardized exams such as the NBME Comprehensive Basic Science Examination (CBSE) are essential in medical education and licensure worldwide. However, data on predictors of performance in Latin American settings remain limited. This study aimed to identify academic and demographic factors associated with NBME CBSE performance among medical students at Universidad San Francisco de Quito (USFQ) to support early identification of at-risk students and guide academic interventions.</p> Methods <p>We conducted a retrospective cohort study of medical students at USFQ in Ecuador. Pre-admission data, academic performance in basic science and clinical courses, and demographic characteristics were analyzed. Spearman rank correlations and Wilcoxon rank-sum tests were used to assess associations with NBME CBSE scores. A multivariable linear regression model was developed using variables available by the end of the third year and internally evaluated using 10-fold cross-validation.</p> Results <p>GPA showed a strong positive correlation with NBME scores (rho = 0.585), whereas PAA had a weak correlation (rho = 0.245). Several foundational courses including Genetics, Microbiology, Biochemistry, and Foundations of Medicine 1 were moderately associated with exam performance. Although Pediatrics demonstrated a strong correlation with NBME CBSE scores in exploratory analyses, it was excluded from the primary prediction model because it is completed after the intended early-intervention period. Age was negatively associated with scores, largely due to course retakes. Geographic background was not significantly associated with examination performance. The final end-of-third-year model incorporated GPA, course-retake history, and grades in Foundations of Medicine 1, Biochemistry, Genetics, Pharmacology, and Microbiology. The model explained 42.9% of the variance in NBME CBSE scores and had an in-sample mean absolute error of 5.82 points.</p> Conclusions <p>Academic performance metrics available by the end of the third year may help identify students at risk of lower NBME CBSE performance. GPA was the most informative predictor in the multivariable model. This internally cross-validated early-warning model may support timely academic interventions, although external validation is needed before implementation in other settings.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Relationship Between Academic Performance and NBME CBSE Scores: A Retrospective Cohort Study and Predictive Modeling for Early Identification of At-Risk Students in a Latin American Medical School

  • 期刊论文

  • 发表日期：

    2026年7月20日

查看全文
  • Camila D. Lopez,
  • Cesar Erazo,
  • Marisol Bahamonde,
  • Ahmed Eissa,
  • Gabriel Cedeno,
  • Carlos Ruiz,
  • Henry Vasconez

摘要

Background/Objectives

Standardized exams such as the NBME Comprehensive Basic Science Examination (CBSE) are essential in medical education and licensure worldwide. However, data on predictors of performance in Latin American settings remain limited. This study aimed to identify academic and demographic factors associated with NBME CBSE performance among medical students at Universidad San Francisco de Quito (USFQ) to support early identification of at-risk students and guide academic interventions.

Methods

We conducted a retrospective cohort study of medical students at USFQ in Ecuador. Pre-admission data, academic performance in basic science and clinical courses, and demographic characteristics were analyzed. Spearman rank correlations and Wilcoxon rank-sum tests were used to assess associations with NBME CBSE scores. A multivariable linear regression model was developed using variables available by the end of the third year and internally evaluated using 10-fold cross-validation.

Results

GPA showed a strong positive correlation with NBME scores (rho = 0.585), whereas PAA had a weak correlation (rho = 0.245). Several foundational courses including Genetics, Microbiology, Biochemistry, and Foundations of Medicine 1 were moderately associated with exam performance. Although Pediatrics demonstrated a strong correlation with NBME CBSE scores in exploratory analyses, it was excluded from the primary prediction model because it is completed after the intended early-intervention period. Age was negatively associated with scores, largely due to course retakes. Geographic background was not significantly associated with examination performance. The final end-of-third-year model incorporated GPA, course-retake history, and grades in Foundations of Medicine 1, Biochemistry, Genetics, Pharmacology, and Microbiology. The model explained 42.9% of the variance in NBME CBSE scores and had an in-sample mean absolute error of 5.82 points.

Conclusions

Academic performance metrics available by the end of the third year may help identify students at risk of lower NBME CBSE performance. GPA was the most informative predictor in the multivariable model. This internally cross-validated early-warning model may support timely academic interventions, although external validation is needed before implementation in other settings.