CHAPTER 2
Preliminary evidence gathering: snowballing and reverse snowballing
Previously1 you have identified that currently no reviews exist, are on the way to completion, or are not of a high enough quality to be considered reliable. Therefore, you need to start some preliminary evidence gathering.
Typically you know of one or two articles that currently exist in or around your topic of interest. Therefore, the first thing to do is to find out which relevant articles those papers cite, and then retrieve them. You should then check if those articles cite any other relevant articles, retrieve those articles, and continue that process until you cannot find any more relevant articles. This process is known as snowballing, and you will notice that you will only ever find articles published before the first article you located or originally knew about.
However, this process can also be implemented in reverse and used to find more contemporary articles, hence reverse snowballing. Large online databases such as ISI Web of Knowledge and Google Scholar offer an advanced facility known as citation tracking. By locating a known article within the database it will display which articles are referenced within the article, and importantly, also indicate more recent articles that cite your article of interest. You can then choose the articles which appear to be of interest and continue the process until you find no more relevant articles.
From this process you should have a number of articles relating to your topic, and the next step is to then identify the words in the title and abstract of the articles and Medical Subject Headings (MeSH) which define the articles that you have collected. To understand more about MeSH visit the homepage.2
Next chapter: reference management and identifying keywords.
- © British Journal of General Practice, 2007.