Compare Stuff News » PubMed

The meteoric rise of mephedrone

Bob — Mon, 29 Mar 2010 12:09:51 +0000

Let’s start by comparing some old-school drugs, cannabis and cocaine, and their relative occurrence on web pages containing the numbers 2004 to 2010. In other words, let’s see how their web-popularity has changed in recent years:

They are almost identical. Let’s try another two, MDMA and heroin:

Again they are very similar, but with a slight tailing off for MDMA (harder to make in the UK since 2008).

Now let’s try mephedrone and cocaine. The difference is massive.

Meanwhile, it is almost completely absent from the medical literature.

Bugfixes

Bob — Sat, 20 Mar 2010 17:29:56 +0000

The Pubmed version had broken due to a change in the Pubmed results format. It was easy to fix and now you can query for all kinds of trends in the biomedical research literature. See the previous blog post for more examples.

I also removed the “powered by Yahoo” icon because Yahoo are not serving it any more. It seems we’re using a deprecated API now, but I guess it will work for a while to come.

And while we are here, I’ll post a graph for old times’ sake:

Happy comparing!

Trends in biomedical literature

Bob — Thu, 06 Sep 2007 08:49:28 +0000

A long, long time ago when Compare Stuff was young I made a version which made it easy to do quantitative biomedical literature search analysis. I have now blown away the cobwebs, improved it a little (now you can compare 8 things) and will briefly demonstrate its genius…

Note that it still uses the old vertical graph style like the legacy web version, but it’s still quite easy to figure out what’s going on.

With it you can… compare trends in research activity:

Sad, but true, obesity research is on the increase while cancer research is steady or perhaps on the decline.

You can also compare techniques and places:

Or you can compare places with respect to research activity:

Here is looks like Kobe and Osaka are better places to work on diabetes if you thrive on competition and discussion with colleagues.

And that’s only the beginning… you can look at trends across different diseases or types of cancer, different techniques, fields, authors, gene names… in fact anything you can think of (but remember that the numbers of “hits” should be large so that the stats are good).