Objective To conduct a systematic analysis of a clinical surgery theoretical examination based on knowledge points, thereby providing a reference for establishing a scientific and comprehensive test paper analysis system.

Methods Using the difficulty coefficient (P-value) and discrimination index (D-value) as indicators, a detailed evaluation was performed on 6 question types and 40 knowledge points from the Clinical Comprehensive Course: Surgery examination for the 2023 clinical medicine pilot class at Peking Union Medical College. Data were visualized using frequency distribution charts, radar charts, and scatter plots.

Results A total of 23 examination papers were collected. The students' average score was 78.63±7.58, with a pass rate of 100%. The overall examination difficulty was relatively low (P=0.79), and the overall discrimination was poor (D=0.15). Among the 40 knowledge points, 2 (5.0%) were difficult, 9 (22.5%) were moderately difficult, and 29 (72.5%) were relatively easy. Regarding discrimination, 3 (7.5%) knowledge points showed excellent discrimination, 8 (20.0%) demonstrated good or acceptable discrimination, and 29 (72.5%) had relatively poor discrimination. Some knowledge points e.g., "Benign Prostatic Hyperplasia" (P=0.26, D=-0.17) presented issues of high difficulty coupled with poor discrimination. A considerable number of knowledge points require corresponding adjustments.

Conclusion Knowledge-point-based test paper analysis can provide references for optimizing question type design, thereby enhancing the scientific nature of assessments and offering a basis for teaching improvements.