Abstract: Objective: To implement intelligent full-process management of nursing records through the application of DeepSeek Large Language Model and evaluate its application effectiveness, aiming to enhance the quality and efficiency of nursing record management. Methods: In February 2025, the nursing department and information management department in a tertiary grade A hospital in Qingdao completed the application of DeepSeek. Intelligent full-process management of nursing records was implemented, utilizing an intelligent interactive interface to assist in nursing decision-making, documentation, quality control, and other workflow stages. We compared the time required to complete nursing records, error rates, quality and system usability of nursing records before and after the DeepSeek model's deployment. Results: After the deployment of DeepSeek, the time required to complete a nursing record was reduced from (3.96±1.56) hours to (2.47±0.53) hours, and the error rate of nursing records decreased from 6.33% to 1.56%. Meanwhile, nurses' ratings of the medical record quality and system's usability were significantly improved. All the differences were statistically significant (P<0.05). Conclusion: With the support of DeepSeek Large Language Model, in-depth analysis and mining of nursing record data have been conducted, improving the accuracy of record quality assessment, promoting the standardization of nursing practices, and ensuring the quality and safety of patient care.

Key words: DeepSeek; Large Language Model; Artificial Intelligence; nursing record; quality control; health education