Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
First, library shelves groaned under the weight of volume after volume of Index Medicus. Then came Medline, a journal abstracting and search service initially available on CD (at a price) and, for the UK higher education community, by telnet and then on the web . . . and then there was PubMed.
It is a mark of its success that it already seems to have been with us forever . . . but it is only 5 years since former American Vice President Al Gore, in a breathtaking act of generosity, announced free internet access to PubMed. In a world where knowledge is increasingly viewed as a commodity (viz the human genome project), PubMed demonstrates the global power of the internet to transform access to information.
So what can it do? PubMed now contains more than 12 million citations (about a million more than Medline), and search fields include Medical Subject Heading (MeSH) keyword and title or abstract word or phrase. The search returns a full citation including an abstract of the article and, increasingly, a hyper-text link to where the article may be obtained online. In a recent PubMed search I had immediate full text access for 16 of 20 “hits”. The department photocopier will soon be obsolete, and in many respects the library already is.
There are some drawbacks. Firstly, citation searching is not available (but see www.mimas.ac.uk), so looking forward from important papers is not possible. Secondly, abstracts on your computer may seem as reliable as full text articles in a library across town, encouraging a superficial engagement with the literature.
Finally, PubMed can be a (dubious) source of diagnostic assistance: searching for “case report”, “deafness”, “retinitis pigmentosa”, and “myopathy”, for instance, returns only 24 hits, most of which describe a mitochondrial disease. Comparison of this approach with a more traditional line of diagnostic attack has yet to be subjected to a randomised controlled trial.