Abstract: Objective To summarize the types and new applications of mainstream Common Data Models (CDMs) in the medical field, analyze their applicability in public health and epidemiological surveillance, drug research and clinical trials, and quality and performance management of health care. Methods The leading characteristics, technical architecture, and applications of ten mainstream CDMs were introduced. In combination with typical cases of research at home and abroad, the current applications of CDMs in the medical field, main challenges and developments were analyzed. Results CDMs could effectively support semantic interoperability of multi-source heterogeneous medical data by unifying the data structure and terminology system, so they were highly applicable to cross-border and real-world research, epidemiological surveillance, and response to public health emergencies. The distributed analysis framework based on CDMs could ensure large-scale and collaborative use of data while protecting patients' privacy, which significantly reduced the threshold for multi-center research. However, such issues as the strongly unstructured clinical data in Chinese, low efficiency of distributed analysis, heterogeneity of international data standards, and the imperfect domestic integration system still restrict their large-scale implementation in China. Conclusion CDMs can break down medical data silos and provide core technical support for generating high-quality real-world evidence. It is recommended that such technologies as large language models and lightweight federated learning be combined to enhance automated mapping and distributed analysis, promote multi-standard interoperability and construction of the industrial ecosystem, and accelerate their extensive applications in medical research and management.

Key words: Common Data Models (CDMs), Public Health, Drug Development, Medical Big Data, Data Standardization, Real-World Evidence, Multi-Center Study