Clinical data is one of the key sources of real-world data, while the reliability of such data is often hindered by unscientific and irregular data management. Based on various problems found in the application of clinical data to scientific research, this study formulated a standard operation procedure(SOP) for real-world data management, and organized the data in terms of content and logic consistency, completeness and standardization. The clinical data of syphilis at a tertiary hospital were used as an example for verification. The clinical data of syphilis patients in the dermatology department from January 1, 2008 to January 1, 2018 were derived from the basic information database, diagnosis database, laboratory information database, and treatment information database of its hospital information system, and a total of 71 705 pieces of relevant information were extracted. SOP analysis showed that 6 816 articles were completely repeated. There were content or logical inconsistencies in name, medical identification number, resident ID number, age and gender, being 152, 360, 88, 107 and 457 respectively. There were many missing entries for marital status, ethnicity, gender, birthday, and ID number, being 1 711, 1 077, 457, 496 and 355 respectively. The standardization situation was generally good, and the number of irregularities in filling out occupational items was 1 884. The standardized and collated data could effectively count the proportion of patients in different stages of syphilis and the cost of diagnosis and treatment, indicating that the real-world data management and analysis SOP proposed in this study could be used to build a clinical-research oriented dual-functional database based on medical practice data.