试论中药人用经验资料收集、数据质量与证据形成

Data collection, quality and evidence formation for human use experience of traditional Chinese medicine

收集和总结人用经验资料,形成高质量数据和可用于评价的证据是中药人用经验研究关键环节。该文论述人用经验资料收集、整理和总结,指出人用经验收集应围绕中药新药处方来源展开,基于辩证思维、用药经验、组方特点和临床应用进行归纳;收集内容包括处方、工艺、临床定位与适用人群、疗效数据和安全性数据等;收集方法包括访谈、临床资料总结和数据挖掘等。基于人用经验资料形成的人用经验数据作为药品注册资料时,要保证数据来源的合法、合规,知识产权归属清晰;数据来源须符合医学伦理学要求,为避免利益冲突宜委托独立的第三方进行数据分析;制定人用经验数据质量控制措施,以保证数据可溯源性、完整性、一致性和准确性,避免数据偏倚;人用经验数据应包括准确的临床定位与适用人群、公认的临床疗效和安全性等关键数据。形成人用经验数据后,要制定人用经验数据的统计分析计划,通过严格的数据处理、统计分析和临床解读,从而产生可用于评价的人用经验证据。

Collecting and summarizing human use experience(HUE) data, forming high-quality data and evidences that can be used for evaluation are the key links of HUE research on traditional Chinese medicine(TCM). The collection, collation and summary of human experience data were discussed in this paper. It is pointed out that the collection of HUE should be focus on the source of prescription of new traditional Chinese medicines, and be summarized based on dialectical thinking, experience in medication, characte-ristics of prescription and clinical application. The collected contents include prescription, process, clinical location and applicable population, efficacy data and safety data, etc. The methods include interview, clinical data summary and data mining. When the data formed based on HUE information is used as drug registration information, it is necessary to ensure that the data source is legal and compliant, and the ownership of intellectual property is clear.Data sources should meet the requirements of medical ethics. To avoid conflict of interest, data analysis should be conducted by an independent third party. It is necessary to develop the quality control measures of HUE data to ensure the data traceability, integrity, consistency and accuracy, and avoid data bias.The data of HUE should include the key data such as accurate clinical location and applicable population, recognized clinical efficacy and safety.After the formation of HUE, the statistical analysis plan of empirical data of human use should be formulated. Through strict data processing, statistical analysis and clinical interpretation, HUE can be produced for evaluation.