整合分析之探討與應用－文獻回顧

Exploration and Application of the Meta-Analysis-A Review

蔡鎰隆(Yi-Lung Tsai) ； 劉胤廷(Yin-Ting Liu) ； 張宏博(Hong-Po Chang)

《中華民國齒顎矯正學雜誌》 16卷1期 (2004/03) Pp. 36-40

https://doi.org/10.30036/TJO.200403.0005

摘要

最近十年在健康科學的領域中使用整合分析（meta-analysis）的文章明顯增加，其克服了傳統狹義文獻回顧低效力的缺點，是一項相當有效力的工具，特別在臨床研究與基礎醫學方面。但整合分析並非沒有爭議，本篇文章的目的是呈現整合分析的主要爭議，並探討進行整合分析與報告整合分析的結果時最重要的事項。整合分析受到的爭議包括研究者結合不同研究的能力，例如：研究的種族，實驗的設計，品質控制與選擇偏差等，如果進行整合分析時能嚴格遵守基本原則，就能避免產生錯誤的結果，因此要注意(a)搜尋文章的方法(b)減少偏差的測量(c)引用與排除文章的原則(d)如何從文章中攫取資料(e)資料如何進行統計分析。

關鍵字

整合分析 ； 選擇偏差

並列摘要

In the last decade, health sciences literature has seen a significant increase in papers that report findings of meta-analysis. Meta-analysis overcomes the drawbacks of conventional narrative literature review. It is a great tool, especially in clinical research and evidence-based medicine. But it is not without controversy. The purpose of this paper is to present the major controversies of meta-analysis and explore the most important issues of conduct and reporting of results derived from those individual studies.
The disputes include the ability of the researchers to combine studies with different aspects, such as study population, experimental designs, quality controls, and selection bias. We can avoid misleading results if some principles are followed:(a) the methods of literature search, (b) the measures taken to reduce bias, (c) inclusion and exclusion criteria, (d) how to extract data from the literature, and (e) how to analyze data statistically.

並列關鍵字

Meta-analysis ； selection bias