In mid November 2019, the National Library of Medicine, USA announced the launch of the new PubMed. They mentioned that the new one will become the “default” from Spring 2020 – which will be around the 19th of March, 2020. It is time for us to start getting used to the new PubMed – both the good and the not so good, about the new PubMed. Here are some points of learning:
The site address
The new PubMed is available at https://pubmed.ncbi.nlm.nih.gov/ . If you have bookmarked the old, one, you will get a message asking you to try the new PubMed. But as I mentioned, in mid march, the new will become the default (and the old may disappear then)
Truncation (Wild card searching)
The best thing that we have noticed is that it has a very powerful truncation feature. In the old one, if we typed – hot* – in the search box (hoping to search for hot, hotter, hottest and all variations), we would get a message that said:
Wildcard search for ‘hot*’ used only the first 600 variations. Lengthen the root word to search for all endings.
IN the new PubMed – we do not get any message and it does search for all variations.
Results – Sort order
The default sort order is “Best match”. In the old PubMed, it was “Most recent”. In the old, both sort orders gave different results. In the new, both give the same results
New filter – Associated Data
In the new PubMed, we have a new filter – Associated Data. If you run a search (eg malaria/drug therapy[mesh]) and choose the filter Associated Data, then we get a bunch of results where the link to data associated with the study, is available in a clinical trials registry
Saving large volumes of data
In the old PubMed, you could save an unlimited number of results as a file (various formats). In the new, it permits only 10,000 results to be downloaded as a file. (I wonder if that number has been chosen based on some standard!)
There are more differences that we are tabulating. Most of them will make sense only for those who have done our courses. If you have done our course or the online workshop, and if you have any specific queries, we can share answers from our compilation