Medical humanities (MH) courses blend humanities with medical training to help students develop essential skills and empathy. Yet, how MH influences practical clinical performance during internships is not well established. This study investigated the effect of MH learning on internship outcomes. We analyzed the academic records of 1,364 medical students from eight admission cohorts. Student performance in basic sciences, clinical skills, and MH courses was evaluated, alongside internship grades. Machine learning models—including support vector machines, logistic regression, and random forests with ten-fold cross-validation—were applied to predict internship performance. Multiple regression was also used to determine MH’s independent impact. MH was a significant predictor of internship performance. Removing MH variables from models notably reduced predictive accuracy (e.g., logistic regression AUC dropped from 0.781 to 0.742). Regression analyses revealed that MH had the most potent independent effect on internship grades (OR: 1.29, P < 0.0001), accounting for nearly 30% of the variance. Within MH, courses in Medical Sociology and Cultural Studies, and Communication Skills and Interpersonal Relationships, were particularly influential (AUC 0.710 and 0.705). Performance in MH courses is strongly associated with clinical competence during internships, even more so than basic medical science scores. These findings suggest that integrating the humanities into medical curricula can better prepare skilled and compassionate physicians. Further studies should explore the long-term benefits of MH engagement.
