Department of Clinical Medicine (Pharmaceutical Medicine), Graduate School of Pharmaceutical Sciences, Kitasato University, Shirokane 5-9-1, Minato-ku, Tokyo, 108-8641, Japan.
Pharmaceut Med. 2021 Nov;35(6):327-338. doi: 10.1007/s40290-021-00406-8. Epub 2021 Nov 1.
In recent years, real-world data (RWD) have been actively used in the field of pharmaceutical research. Database (DB) study, one of the observational studies using RWD, is a comprehensive, continuous, and rapid research method that plays an important role in the postmarketing stage of drugs, although the interpretation of the results may be limited. DB studies are often focused on drug safety, and previous research reviewing DB studies on drug effectiveness across different disease areas have been limited.
The objective of this review was to reveal the current status of DB studies on drug effectiveness in various therapeutic areas and to provide information that allows researchers to consider conducting appropriate DB studies on drug effectiveness in the postmarketing stage.
For this systematic review, we searched Embase and MEDLINE for DB studies on drug effectiveness published between 1 January 2018 and 31 December 2019. We reviewed the title, abstract, and methods to identify studies on drug effectiveness using medical information DBs, and excluded non-medical studies, studies on non-drug, and studies on drug safety, actual use, or cost outcomes that did not include any effectiveness outcomes. The name and type of the DB (administrative claims DB, clinical DB, pharmacy DB, and DB linkage), study design, comparison group, type of outcome, and presence or absence of reference to the outcome definition were extracted and summarized according to disease areas.
We obtained 225 articles on DB studies on drug effectiveness using DBs that integrate large-scale medical data for secondary use across different disease areas. Among the DB classifications, administrative claims DBs (70%, 158/225) were most commonly used, while pharmacy DBs were used in only three studies. The largest number of reported studies were associated with cardiovascular, respiratory, and infectious diseases. Outcomes were often inpatient diagnosis, and some ideas included defining effectiveness based on drug use. While various outcomes were uniformly used in studies for the treatment of infectious diseases and respiratory organs, death (overall survival [OS]) and drug continuation (progression-free survival [PFS]) in patients with cancer, laboratory values in the endocrine system (mainly diabetes) were used as the main outcomes. Outcome validation within the article was limited. New user design (32%, 73/225), propensity score analysis (58%, 131/225), and sensitivity analysis (40%, 90/225) were used as measures to reduce bias in these studies. Sixty-eight studies (30%, 68/225) were supported by pharmaceutical companies.
This systematic review summarized the status of cross-disease research articles on DB studies on drug effectiveness. While considering the strengths and limitations of DB studies, we hope that our comprehensive results would help to promote appropriate DB studies on drug effectiveness in the postmarketing stage.
近年来，真实世界数据（RWD）已在药物研究领域得到积极应用。数据库（DB）研究是使用 RWD 的观察性研究之一，是一种全面、连续、快速的研究方法，在药物的上市后阶段发挥着重要作用，尽管结果的解释可能存在一定的局限性。DB 研究通常侧重于药物安全性，之前对不同疾病领域药物有效性的 DB 研究的综述较为有限。
本综述旨在揭示不同治疗领域 DB 研究药物有效性的现状，并提供信息，使研究人员能够考虑在药物上市后阶段进行适当的 DB 研究药物有效性。
本系统综述通过 Embase 和 MEDLINE 检索了 2018 年 1 月 1 日至 2019 年 12 月 31 日期间发表的关于药物有效性的 DB 研究。我们对标题、摘要和方法进行了审查，以确定使用医疗信息 DB 进行药物有效性研究的研究，并排除了非医疗研究、非药物研究以及不包括任何有效性结果的药物安全性、实际使用或成本结果研究。根据疾病领域，提取并总结了 DB（行政索赔 DB、临床 DB、药房 DB 和 DB 链接）的名称和类型、研究设计、对照组、结果类型以及是否参考了结果定义。
我们获得了 225 篇使用整合了不同疾病领域大规模医疗数据进行二次利用的 DB 进行药物有效性研究的文章。在 DB 分类中，行政索赔 DB（70%，158/225）最常用，而仅有三项研究使用了药房 DB。报告的研究数量最多的是心血管、呼吸和传染病相关研究。结果通常是住院诊断，有些研究的想法包括根据药物使用情况定义有效性。虽然在传染病和呼吸器官治疗研究中统一使用了各种结果，但癌症患者的死亡（总生存[OS]）和药物持续使用（无进展生存[PFS]）、内分泌系统（主要是糖尿病）的实验室值则是主要结果。文章中对结果验证的程度有限。新用户设计（32%，73/225）、倾向评分分析（58%，131/225）和敏感性分析（40%，90/225）被用作这些研究中减少偏倚的措施。68 项研究（30%，68/225）得到了制药公司的支持。
本系统综述总结了跨疾病 DB 研究药物有效性研究文章的现状。在考虑 DB 研究的优势和局限性的同时，我们希望我们全面的结果将有助于促进药物上市后阶段进行适当的 DB 研究药物有效性。
