药物流行病学中处方序列对称分析研究的系统综述

Prescription sequence symmetry analysis in pharmacoepidemiology:a systematic review

目的系统评价和分析国内外处方序列对称分析(PSSA)的原始研究,总结研究进展和方法学细节,为我国今后开展同类研究提供参考。方法以PSSA相关的中英文关键词,分别检索中国知网、万方数据知识服务平台、维普数据库及PubMed、Embase、Cochrane,筛选截至2020年6月30日发表的PSSA原始研究。采用Excel软件构建信息提取表,采用Stata软件进行统计学分析。结果共纳入45篇文献,从2013年开始文献发表数量显著上升;研究地区主要分布在日本(11篇,24.44%)、中国(10篇,22.22%)、丹麦(9篇,20.00%)、澳大利亚(8篇,17.78%),大部分研究使用的数据库为医疗保险理赔数据库。纳入的PSSA研究共涉及16类药物,其中精神类药物和他汀类药物的研究数量最多(8篇,17.78%),人体各系统的不良反应均有涉及。方法学上,明确给出了洗脱期和间隔期的研究数量分别为35篇(77.78%)和43篇(95.56%),其中相应有14篇(31.11%)和9篇(20.00%)给出了确定的方法或理由。另外分别有16篇(35.56%)和18篇(40.00%)文献报告了敏感性分析和亚组分析。结论PSSA作为有效的安全性信号挖掘方法之一发展迅速,但方法学处理和结果报告还需提高。我国PSSA研究仍处于起步阶段,需注重研究质量,推进方法学探索。

Objective To systematically evaluate and analyze the original research of prescription sequence symmetry analysis(PSSA),summarize its research progress and methodological details,and provide a reference for the future use of this method.Methods The keywords related to PSSA were used for literature retrieval from Chinese databases(CNKI,Wanfang,and VIP)and English databases(PubMed,Embase,and Cochrane).Original articles that were related to PSSA and published before June 30,2020 were selected.Information form was developed by Excel.Stata was used for the statistics analysis.Results There were 45 eligible articles included in the research.Since 2013,the number of studies using PSSA has increased rapidly.These studies were mainly conducted in Japan(n=11,24.44%),China(n=10,22.22%),Denmark(n=9,20.00%),and Australia(n=8,17.78%).Medical claim database was used most commonly when PSSA was implemented.The included studies involved 16 types of drugs,of which the number of studies of psychotropic drugs and statins was highest(n=8,17.78%),and adverse reactions of almost all human systems were involved.In terms of methodology,35(77.78%)and 43(95.56%)studies clearly reported the run-in period and interval period,of which 14(31.11%)and 9(20.00%)respectively gave the method or reason for determining the duration.In addition,16 articles(35.56%)and 18 articles(40.00%)reported sensitivity analysis and subgroup analysis results,respectively.Conclusions PSSA,one of the effective methods for safety signal detection in healthcare databases,has developed rapidly,but the methodological details and result reporting need to be improved.In China,PSSA research is still in its infancy,and it is necessary to pay attention to the quality of research and promote methodological exploration.