Meta – Analysis Evidence Based Research

Meta-Analysis:

Meta-analysis is the combination of conceptually similar studies using statistical principles to attain a higher degree of accuracy towards the truth. It is extensively used in healthcare, especially in clinical trials where it is applied to find the common effects of same treatment, where the effect of the treatment remains consistent through multiple cases of study. Meta-analysis is at the pinnacle of levels of evidence in the pyramid of evidence-based healthcare. The pyramid specifically helps in understanding various levels of evidence available and thei level of bias and reliability.

As one proceeds to the top from every level of evidence in the evidence pyramid the evidence becomes less biased as compared to previous level. Thus, Meta-analysis  can be regarded as the apex of healthcare evidence
 At the bottom of the pyramid of evidence are studies with sufficient bias and as we move to the top the studies become more and more reliable with stark reduction in bias as the level of scrutiny and methods are highly reliable and scientific. Meta-analysis is at the apex of it, as it is combines various scientific studies to provide an estimate of the true value (Haidich, 2010).

“Meta-analysis refers to the analysis of analyses" (Gene v. Glass, 1976)

Origins of meta-analysis can be traced to 17^th century when the probability theory was applied to astronomy. Meta-analysis as we know it today emerged in the 20^th century. The technique and naming is credited to the American scientist Gene V. Glass, who in 1976 address provided several examples of meta-analytic work which included, among other examples, a combination of findings of psychotherapy on himself and his colleagues. Since then, they have become a common way for synthesizing evidence and summarizing the results of individual studies(Chalmers et al., 1977)
 

The Need Of Performing Meta-Analysis:

Result of a single study is not sufficient to validate a hypothesis. Results vary for each study by a slight or heavy margin depending on the various conditions, study subjects and samples, and other confounding factors. Thus raises the need for synthesizing data across multiple studies.  Meta-analysis uses statistical principles for this end.

The combination of different studies provides more data which is advantageous in increasing precision and accuracy of the statistical estimate. Even individual studies that are underpowered can be improved by combining with a meta-analysis of similar studies. Thus meta-analysis ensures improving statistical power in detecting a particular effect

Steps To Perform A Meta-Analysis.

Following are the basic steps involved in carrying out basic Meta-analysis (Field & Gillett, 2010)

1. Identification and formulation of the problem
Define the question which needs to be answered.

2. Carry out a literature review.

Study and search for various databases that include different reviewed articles. Some commonly used platforms this purpose includes PubMed, Embase, Scopus, etc.

3. Define inclusion and selection criteria.
This process helps in identifying and selecting relevant evidence from studies for the problem statement.

4. Extracting Data.
Extract data from studies which are relevant for your purpose for combining in the final set of analysis.

5. Performing Meta-analysis.

Use one of the many software tools available to carry out meta-analysis. Software like Review manager and comprehensive meta-analysis provide a good analysis.

Learn More : Meta analysis Evidence based Research