BMC Medical Education最新文献

Background: Despite the objective structured clinical examinations (OSCE) being widely used in the assessment of clinical competency, the optimal scoring systems remains debatable with limited data from sub-Saharan Africa. This study compared the performance, reliability, and agreement between global rating scales (GRS) and checklist scores at a comprehensive exit examination for undergraduate medical students in Rwanda.

Methods: This cross-sectional descriptive study was conducted during the final 'Exit Exams' of undergraduate medical students at the University of Global Health Equity, Rwanda. The OSCE included 15 stations spread across major clinical specialties and subspecialties. Each station had a checklist with a total score of 20 marks and a three-level GRS (failed, borderline, or passed).

Results: A total of 36 students took part in the OSCE examinations. The mean (standard deviation) checklist score was 84.3 (5.3) %, which was lower than the mean GRS score, 94.3 (6.9) %, p < 0.001. All students achieved overall scores above the standard-setting pass mark of 64.4% [set from modified Angoff method], on both scoring systems. While no student failed any station using checklist scores, the GRS identified failures in stations 5 and 6 (one student each), and stations 7 and 13 (two students each). Overall internal consistency (Cronbach's Alpha) across all the stations was 0.760 [ranged from - 0.216 (station 15) to 0.746 (station 9)]. Pearson correlation demonstrated a very strong positive correlation between the checklist and GRS (r = 0.924, p < 0.001). Bland-Altman plot showed a mean (standard deviation) of difference of 10.01 (2.8) in favor of GRS, with a lower and upper limit of agreement of 4.44 to 15.58 respectively.

Conclusion: Checklist and GRS scores in the OSCEs demonstrated strong positive correlation, but GRS showed a higher discriminatory ability, identifying performance differences that checklists did not capture. Incorporating GRS alongside checklists may enhance the robustness of high-stakes clinical assessment.

Objective: This study was aimed to compare the accuracy of four state-of-the-art large language models (LLMs)(ChatGPT-5, ChatGPT-4o, DeepSeek, and Gemini 2.5 Pro) on Dentistry Specialization Exam (DUS) questions in Prosthodontics, and to assess the effect of querying method on performance.

Materials and methods: A total of 128 multiple-choice DUS questions (106 knowledge-based; 22 case-based) from 13 exams (2012-2021) were administered. Each model was tested within a single 24-hour window (September 9, 2025). Two protocols were used: (1) Independent Query (each item in a fresh chat) and (2) Sequential Query (exam-like blocks of 10 items). First responses only were recorded and scored against official answer keys. Statistical analyses were performed using the chi-square (χ²) test to compare categorical outcomes (correct/incorrect responses) across different large language models and query strategies. Inter-rater agreement for question classification was assessed using Cohen's kappa coefficient, and all analyses were conducted at a significance level of α = 0.05.

Results: With Independent Query, the accuracy for knowledge questions was found to be 91% (GPT-5), 86% (GPT-4o), 71% (DeepSeek), and 88% (Gemini 2.5 Pro) while total accuracy was detected as 86%, 83%, 66%, and 85% respectively. In this context, statistically significant differences were reported across models, also due to the low scores of DeepSeek. Case-based accuracies (64%, 68%, 45%, and 77%) did not differ significantly. With Sequential Query, knowledge accuracies were 75%, 73%, 63%, and 82% and case accuracies were 77%, 68%, 64%, and 91%, respectively. Total accuracy still differed across models (75%, 72%, 63%, and 84%). Within-model comparisons showed significant drops for knowledge items from Independent to Sequential querying for GPT-5 (91%→75%) and GPT-4o (86%→73%). DeepSeek and Gemini 2.5 Pro showed no significant changes. Notably, Gemini 2.5 Pro yielded the highest performance in case-based questions (91%) with sequential query.

Conclusions: While the present study findings highlight current limitations in clinical reasoning, they support the conclusion that LLMs can be used as supplementary educational tools for DUS-style knowledge assessment but should not replace expert judgment or patient-specific decision-making.

Introduction: Academic performance in medical students is influenced by a wide range of psychosocial variables, including emotional distress, motivation, emotional intelligence, and social support. Understanding how these factors relate to academic outcomes is essential for designing effective educational and mental health interventions.

Objective: To systematically evaluate the association between psychosocial factors and academic performance among medical students and to identify which determinants show the most consistent evidence across studies.

Methods: A systematic review was conducted following PRISMA 2020 guidelines. Searches were performed in PubMed/MEDLINE, ScienceDirect, Scopus, and Web of Science for studies published between 2014 and 2024. Observational studies examining psychosocial factors in relation to academic performance were included. Quality assessment was conducted using the Newcastle-Ottawa Scale.

Results and conclusions: Forty-five studies met the inclusion criteria, most of which were cross-sectional. Academic performance was evaluated through GPA/CGPA, examinations, and licensing or OSCE scores. Psychosocial variables assessed included stress, anxiety, depression, motivation, emotional intelligence, resilience, burnout, and social support. Across studies, stress, anxiety, and depressive symptoms showed predominantly negative associations with academic performance (reported in 67-80% of articles addressing these constructs), while intrinsic motivation, resilience, emotional intelligence, and perceived social support demonstrated mostly positive associations (60-85% of relevant studies). Some studies reported null findings, reflecting heterogeneity in measures, populations, and outcomes, or a true null effect. Evidence supports consistent associations between psychosocial factors and academic performance among medical students. Stress and negative emotional states are generally linked to poorer outcomes, whereas motivation, social support, emotional intelligence, and resilience tend to predict better performance. The predominance of cross-sectional designs and varied measurement tools precludes causal inference. Strengthening psychosocial resources may help improve both academic success and well-being among medical students.