中国临床药学杂志

目的 依据《儿童药品临床综合评价方法学指南》开展儿童川崎病药品临床综合评价试点，探索评价机构间治疗方案的一致性、合理性方法，验证基于多中心真实世界数据开展儿童药品临床综合评价自动分析的可行性。方法 以国内儿童专科医院抽取的脱敏数据为样本，计算治疗药物的治疗时间贡献比，采用ABC分析方法获得的A类的药物及ATC治疗学分类即为各医院主要治疗药物、主要的药物治疗学类别，以文本余弦相似度算法计算院间治疗用药相似度。描述性统计药品首次使用与诊断明确的时间差。结果4家样本医院主要治疗药物品种共计43种，无4家共用药物品种，3家共用的药物有4种。医院间主要治疗药物品种相似度平均为0.21。主要的药物治疗学类别共计19类，4家共用的有1类，3家共用的有3类。医院间主要的药物治疗学类别相似度平均为0.40。医院间主要治疗药物的首次使用时间差中位数、四分位距离基本相同。这与指南的推荐是一致的。结论 医院间用药品种存在习惯差异，主要治疗药物品种重叠度低，但主要的药物治疗学类别有一定集中度，符合临床实际和指南推荐。基于真实世界数据药物治疗时间贡献比，以ABC分析方法确定主要治疗药物、药物治疗学类别是可行的。

AIM To assess the correspondence and rationality of pharmacotherapy strategies across different locations, and evaluate the feasibility of automatically analysis of children's drugs based on multicenter real-world data according to the “Methodological Guidelines for Clinical Comprehensive Evaluation of Children's Drug”, conduct a pilot clinical evaluation of Kawasaki disease drugs. METHODS Based on the desensitization data collected from children's specialized hospitals in China, the time contribution ratio of therapeutic drugs was calculated. The class A drugs and ATC therapeutic classification obtained by ABC analysis method were the main therapeutic drugs and main pharmacotherapeutic categories in each hospital. The text cosine similarity algorithm was used to calculate the similarity of therapeutic drugs between hospitals. Describe the time between first use of a drug and definitive diagnosis. RESULTS Totally 4 children hospitals had used 43 mainly therapy drugs in total. There were no drugs used by all 4 hospitals, while three of the hospitals used 4 drugs. The average similarity of mainly therapy drugs between hospitals was 0.21. The four hospitals had used 19 mainly ATC classification groups in total. All of the hospitals used 1 group, while three of the hospitals used 3 groups. The average similarity of mainly ATC classification groups between hospitals was 0.40. The medians and quartiles of the time difference between mainly therapy drugs start and diagnosis made were similar across hospitals, and which were consistent with guidelines recommendations. CONCLUSION The mainly therapy drugs used in hospital practices are diverse. The similarity of mainly therapy drugs is low across locations, while the similarity of mainly ATC classification groups among hospitals is higher. It is feasibility to calculate the ratio of treatment length of a drug or a ATC classification group, and select the mainly therapy drugs or mainly ATC classification groups according to ABC classification on multicenter real-world data.