Google Or PubMed: When To Use What

Locating an association’s website, looking up a recipe, checking the lyrics to a song - these are all search tasks perfectly suited for using Google. Searching for medical literature is not one of them. Why? Because there are better resources available, such as PubMed, that are nearly as easy to use and exist for the sole purpose of putting the searcher in touch with biomedical literature. In order to make an informed decision as to the appropriateness of using one resource over another, it is important to know a little about how they work and what they are searching. This article will provide of brief introduction to Google, Google Scholar, and PubMed.

Contrary to popular belief, Google does not search the entire Web. Google, like all search engines, searches an indexed file of its own database. While this database contains billions of documents, it does have limitations and cannot, for example, search inside the content of secure resources like CINAHL or UpToDate.

Chances are you will get very different results if you search for the exact same topic in Google and another search engine like Yahoo. Not only do these search engines search different indexes, they rank results differently. In addition to page content, Google uses their patented PageRank™ technology to determine which results are most important to your request. Google’s technology takes into account hundreds of different variables, one of them being the perceived importance of webpages that link to the webpage you requested. How this “importance” is calculated is not exactly clear. Google vaguely describes the process as using the “collective intelligence of the web” (“Corporate Information,” 2009) to determine a webpage’s importance. In other words, Google is the ultimate popularity contest.

Google Scholar [http://scholar.google.com] is a Google product that was created for searching scholarly literature. Google Scholar searches journal articles, abstracts, books, theses, and papers from a number of institutional repositories, educational websites, publisher websites, and universities. Scholar is also able to search the entire content of a document for a search term, not just the title, abstract, or keywords, as is the case with a database like PubMed. Add the simple, familiar Google search box and Google Scholar almost sounds too good to be true. Well...it is. There a few important things to be aware of before relying exclusively on Google Scholar for all your scholarly searching needs. First of all, Google tells us Scholar searches educational and publisher websites, yet they do not disclose which ones. Therefore, while a search may return some results, it’s always important to consider what else may be missing. Another drawback is not knowing when or how often the Google Scholar database is updated. We do know, for example, that PubMed is updated daily Tuesday through Saturday.

Like Google, Google Scholar has a built-in relevancy ranking system. Scholar considers a number of factors in its ranking including how many times search terms show up in the document text, prominence of the author, publication importance, and how many times the article has been cited in the scholarly literature. Because Scholar considers “cited by” in its ranking, it is less likely to locate brand new articles on a topic because no other sources have cited them yet. Ranking by citation analysis may be helpful in locating classic articles on a topic, but the most recent articles will be missed.

The PubMed database was created by the National Center for Biotechnology Information at the National Library of Medicine. PubMed provides free access to the Medline database. Medline currently indexes content from over 5300 biomedical journals published worldwide dating back to 1950. A complete list of the journals indexed in Medline is available online.

Like Google and Google Scholar, PubMed offers the convenience of a simple search box where search terms can be entered. One of the real values of PubMed is its ability to map search terms to Medical Subject Headings behind the scenes. Medical Subject Headings (MeSH) are applied by human expert indexers to every article in Medline. The use of subject headings makes it possible to retrieve citations on a topic even if your search terms do not exactly match those used in the article. For example, if you looked up “heart attack” in PubMed, MeSH will also provide you with articles that use the term “myocardial infarction.” You can see what MeSH terms your search is mapped to by clicking on the Details tab on the results page.

PubMed tops both Google and Google Scholar in the ability to limit a search. PubMed searches can not only be limited by variables such as date, patient age group, or publication type (systematic review, clinical trial, etc), its Clinical Queries feature makes it possible to limit to articles on etiology, diagnosis, therapy, prognosis, or clinical protection guidelines. You can access the complete list of PubMed limits by clicking on the Limits tab found at the top of each page or by clicking on Advanced Search located to the right of the search box. Google Scholar does offer a few limit options from its Advanced Search page such as date range and author name, but they are not nearly as specialized as the limits offered by PubMed.

There is no denying Google’s popularity. It’s accessible, easy to use, and 9 times out of 10 you will find some results on the first page that sort-of answer your query. But it is also important to know the limitations of Google, or any resource you are working with, and not rely on it as your single source of information. The next time you use Google or Google Scholar to search for medical information, consider popping over to PubMed and run a quick search there to see how the results differ. You will never regret being too thorough, especially when dealing with health.

References:

Freeman, M.K., Lauderdale, S.A., Kendrach, M.G., & Woolley, T.W. (2009). Google Scholar versus PubMed in locating primary literature to answer drug-related questions. The Annals of Pharmacotherapy, 43 , 478-484.

Google. (2009). Corporate information: Technology overview. Retrieved June 30, 2009, from http://www.google.com/corporate/tech.html.

Google searching vs. PubMed searching. (2007). Retrieved June 24, 2009, from
http://healthlinks.washington.edu/howto/googlechart.html.

Rethlefsen, M.L., Rothman, D.L., Mojon, D.S. (2009). Internet cool tools for physicians. Berlin: Springer-Verlag.

Schultz, M. (2007). Comparing test searches in PubMed and Google Scholar. Journal of the Medical Library Association, 95(4), 442-445.