绘图(图形)
荟萃分析
计算机科学
判别式
趋同(经济学)
索引(排版)
理论(学习稳定性)
数据科学
风险分析(工程)
运筹学
统计
人工智能
机器学习
数学
医学
经济
万维网
经济增长
内科学
作者
Jazeel Abdulmajeed,Luis Furuya‐Kanamori,Tawanda Chivese,Chang Xu,Lukman Thalib,Suhail A.R. Doi
标识
DOI:10.11124/jbies-24-00155
摘要
Introduction: In recent decades, clinical research has seen significant advancements, both in the generation and synthesis of evidence through meta-analyses. Despite these methodological advancements, there is a growing concern about the accumulation of repetitive and redundant literature, potentially contributing to research waste. This highlights the necessity for a mechanism to determine when a meta-analysis has conclusively addressed a research question, signaling no further need for additional studies—a concept we term an “exit” meta-analysis. Methods: We introduced a convergence index, the Doi-Abdulmajeed Trial Stability (DAts) index, and a convergence plot to determine the exit status of a meta-analysis. The performance of DAts was examined through simulation and applied to two real-world meta-analyses. Results: The DAts index and convergence plot demonstrate highly effective discriminative ability across varying study scenarios. This represents the first attempt to define an exit meta-analysis using a quantitative measurement of stability (as opposed to sufficiency) and its corresponding plot. The application to real-world scenarios further validated the utility of DAts and the convergence plot in identifying a conclusive (exit) meta-analyses. Conclusion: The new development of DAts and the convergence plot provide a promising tool for investigating the conclusiveness of meta-analyses. By identifying an exit status for meta-analysis, the scientific community may be equipped to make better-informed decisions on the continuation of research on a specific topic, thereby preventing research waste and focusing efforts on areas with unresolved questions.
科研通智能强力驱动
Strongly Powered by AbleSci AI