Meta-analysis example

What is the advantage of a meta-analysis?

The traditional practice of writing reviews has one significant flaw - excessive subjectivity. There are no systematic researches in descriptive papers. If the original articles are built according to the scientific methodology, descriptive reviews may contain inadequate assessments, since the data obtained from different sources are generalized without a quantitative analysis of the results of the studies. As a result, a review reflects a predominantly subjective opinion of an author, which contradicts the scientific principles.

To overcome this problem, a meta-analysis technique was developed. It can be defined as a "quantitative systematic review of the literature". The advantages of each example of a meta-analysis are most obvious when the results of individual researches contradict each other, the amount of information is small, and the organization of a large-scale study is too costly.

Examples of meta-analysis

The first relatively simple meta-analysis was carried out by Karl Pearson in 1904. He attempted to solve the problem of reducing the power of statistical criteria in researches with small samplings. Pearson analyzed the results of eleven studies of the effectiveness of typhus vaccine and combined them into one group to obtain a more accurate assessment of the effect.

Later, complicated statistical methods of meta-analysis used up to now were formed in pedagogical practice. For the first time, this approach was widely applied to the analysis of scientific data in sociology and psychology. Subsequently, it was extended to clinical research. By the way, the term "meta-analysis" was proposed by the psychologist G. Glass in 1976.

A classic example of meta-analysis which proves its great importance is the publication of Meir J. Stampfer and co-authors in The New England Journal of Medicine in 1982. In their review, eight studies of the intravenous introduction of streptokinase in case of myocardial infarction were examined. The publications reported about mortality during the first 45 days after the development of myocardial infarction and receiving the drug.

Three studies showed a statistically significant reduction in mortality. In two studies, mortality reduction did not reach a significant level. And three other publications reported an excess of mortality, although the veracity of the differences was not proven statistically.

Obviously, using the traditional descriptive approach, one could only make a conclusion about the need for further study of the problem with more patients, since the mentioned publications contradict each other. Nevertheless, this meta-analysis research example showed that the studied preparation reduced mortality by 20%.

Reasons for conducting a meta-analysis

This practice is advisable if the researcher sets the following goals:

• obtaining a more stable assessment of the effect of certain factors;
• revealing the differences between studies and generalizing their results;
• analyzing the effect of some phenomenon on individual subgroups;
• planning of future research.

Features of meta-analysis writing

Compilation of each qualitative example of meta-analysis research paper was carried out in several stages. The work begins with a formulation of tasks and the search for necessary literature. At this stage, it is necessary to clearly define the criteria for choosing publications. Statistical analysis is then carried out, including an assessment of the uniformity of the results and the reasons for their variability. In addition, the calculation of summary indicators and stability analysis of the findings are performed.

It should be noted that a probability of making mistakes exists already at the stage of searching for publications. Omissions can negatively affect the quality of the whole research. The widespread cause of the appearance of defects in the search for literature is the authors' inclination to publish positive results and hush up the negative ones. In addition, the work that showed the greatest effectiveness of some method is more often quoted and published in international periodicals. Consequently, it is possible to incorrectly reflect the state of things in the considered area.

Unambiguous criteria for selecting publications for review are needed. As a rule, an author of a well-written meta-analysis article example takes into account the structure of the study, the size of the sampling, the nature of the method, the publication of the research in the press, and so on. It should be remembered that all studies should coincide in the main parameters.

After the data collection is completed, a turn of statistical analysis comes, which should answer three basic questions:

1. Are the results of studies comparable? Otherwise, they cannot be combined.
2. Is there a correlation between the differences in the results of individual studies?
3. What is the best estimate of the considered method?

The analysis can be based on two models: with fixed and random effects. The first model is used when the analyst has access to the results of all studies on a given problem, the second one - if there is only a sample of the whole scope of work.

At present, scientists tend to pay increased attention not to the statistical significance of generalized results but to the scale of the effect from the application of a certain method. Its influence can be measured in qualitative or quantitative terms. In the first case, the evaluation is carried out using the odds ratio. In the second case, the standardized mean difference or correlation coefficient is applied.

In conclusion, it should be said that only the correct adherence to the methodology of meta-analysis will enable the researcher to obtain information that cannot be collected by performing a traditional descriptive literature review. Despite the huge number of computer programs created to automate the meta-analysis, the reliability of the results is completely dependent on the analyst.

So, prepare to the work thoroughly. If you received such a task for the first time and do not know how to conduct a meta-analysis, an example written by an experienced scientist will help you see many nuances that are not disclosed in the textbooks. Gain the reading experience, and soon you will be able to come to a brilliant discovery. Good luck!