Search OVID EMBASE and Get MEDLINE for Free…. without knowing it

19 10 2010

I have the impression that OVIDSP listens more to librarians than the NLM, who considers the end users of databases like PubMed more important, mainly because there are more of them. On the other hand NLM communicates PubMed’s changes better (NLM Technical Bulletin) and has easier to find tutorials & FAQs, namely at the PubMed homepage.

I gather that the new changes to the OVIDSP interface are the reason why two older OVID posts are the recent number 2 and 3 hits on my blog. My guess is that people are looking for some specific information on OVID’s interface changes that they can’t easily access otherwise.

But this post won’t address the technical changes. I will write about this later.

I just want to mention a few changes to the OVIDSP databases MEDLINE and EMBASE, some of them temporary, that could have been easily missed.

[1] First, somewhere in August, OVID MEDLINE contained only indexed PubMed articles. I know that OVID MEDLINE misses some papers PubMed already has -namely the “as supplied by publisher” subset-, but this time the difference was dramatic: “in data review” and “in process” papers weren’t found as well. I almost panicked, because if I missed that much in OVID MEDLINE, I would have to search PubMed as well, and adapt the search strategy…. and, since I already lost hours because of OVID’s extreme slowness at that time, I wasn’t looking forward to this.

According to an OVID-representative this change was not new, but was already there since (many) months. Had I been blind? I checked the printed search results of a search I performed in June. It was clear that the newer update found less records, meaning that some records were missed in the current (August) update. Furthermore the old Reference Manager database contained non-indexed records. So no problems then.

But to make a long story short. Don’t worry: this change disappeared as quickly as it came.
I would have doubted my own eyes, if my colleague hadn’t seen it too.

If you have done a MEDLINE OVID search in the second half of August you might like to check the results.

[2] Simultaneously there was another change. A change that is still there.

Did you know that OVID EMBASE contains MEDLINE records as well? I knew that you could search EMBASE.com for MEDLINE and EMBASE records using the “highly praised EMTREE“, but not that OVID EMBASE recently added these records too.

They are automatic found by the text-word searches and by the EMTREE already includes all of MeSH.

Should I be happy that I get these records for free?

No, I am not.

I always start with a MEDLINE search, which is optimized for MEDLINE (with regard to the MeSH).

Since indexing by  EMTREE is deep, I usually have (much) more noise (irrelevant hits) in EMBASE.

I do not want to have an extra number of MEDLINE-records in an uncontrolled way.

I can imagine though, that it would be worthwhile in case of a quick search in EMBASE alone: that could save time.
In my case, doing extensive searches for systematic reviews I want to be in control. I also want to show the number of articles from MEDLINE and the number of extra hits from EMBASE.

(Later I realized that a figure shown by the OVID representative wasn’t fair: they showed the hits obtained when searching EMBASE, MEDLINE and other databases in Venn diagrams: MEDLINE offered little extra beyond EMBASE, which is self-evident, considering that EMBASE includes almost all MEDLINE records.- But I only learned this later.)

It is no problem if you want to include these MEDLINE records, but it is easy to exclude them.

You can limit for MEDLINE or EMBASE records.

Suppose your last search set is 26.

Click Limits > Additional Limits > EMBASE (or MEDLINE)

Alternatively type: limit 26 to embase (resp limit 26 to medline) Added together they make 100%

If only they would have told us….


3. EMBASE OVID now also adds conference abstracts.

A good thing if you do an exhaustive search and want to include unpublished material as well (50% of the conference abstracts don’t get published).

You can still exclude them if you like  (see publication types to the right)

Here is what is written at EMBASE.com

Embase now contains almost 800 conferences and more than 260,000 conference abstracts, primarily from journals and journal supplements published in 2009 and 2010. Currently, conference abstracts are being added to Embase at the rate of 1,000 records per working day, each indexed with Emtree.
Conference information is not available from PubMed, and is significantly greater than BIOSIS conference coverage. (…)

4. And did you know that OVID has eliminated StopWords from MEDLINE and EMBASE? Since  a few years you can now search for words or phrases like is there hope.tw. Which is a very good thing, because it broadens the possibility to search for certain word strings. However, it isn’t generally known.

OVID changed it after complaints by many, including me and a few Cochrane colleagues. I thought I had written a post on it before, but I apparently I haven’t ;).

Credits

Thanks to Joost Daams who always has the latest news on OVID.

Related Articles

Advertisements




Medical Information Matters 2.8 is up!

15 10 2010

The new edition of Medical Information Matters (formerly Medlibs round) is up at Danielhooker.com.

The main theme is “Programs in libraries or medical education”.
Besides two posts from this blog (A Filter for Finding Animal Studies in PubMed” and more on the topic: An Educator by Chance) the following topics are included: a new MeSH (inclusion under mild librarian pressure), PubMed in your pocket, embedding Google Gadgets in normal webpages and experiences with introducing social bookmarking to medical students.
If you find this description to cryptic (and I bet you do), then I invite you to read the entire post here. I found it a very pleasant read.

Since we are already midway October, I would like to invite you to start submitting here (blog carnival submission form).

Our next host is Dean Giustini of the The Search Principle blog. The deadline is in about 3 weeks ( November 6th).

Related Articles





The New PubMed: Trick or Treat?

31 10 2009

31-10-2009 8-53-21 the new pubmed entry

The New PubMed: Trick or Treat?

After a long days work, when looking at the screen,

there it was: PubMed’s new interface, so it seemed,

But one blink – and it had gone

To come back the following dawn.

The change itself was long announced,

we could play with the new “Advanced”.

Still I postponed and procrastinated,

Pointless: the new PubMed couldn’t wait

any longer, but this Redesign isn’t it for me….

Sure, the front page looks web-2-ish, minimalistic & clean,

which is perfect for the Google-Generation,

the hurry-don’t think-just-slash-i-got-one-publication-

PhD’s, for whom all alterations have been made. 2989360212_882aff28d8 trick or treat

Some people think you just have to wait

& see and get used to it.

but I’m already fed up with it.

I know you all think it is just a Librarian-rant.

Librarians they can stick with the new “Advanced”,

“Advanced” however, is just Limits & Index…

But boy did they make this page look complex!

Sure, the basic researchers seem to be quite pleased.

Busy physicians too, they think it is more easy.

They tell me librarian not to wine:

Go MEDLINE OVID! we stay with this design.”

This is no new idea, didn’t you know:

I long seek refuge in OVID MEDLINE, although1810987271_9044fb5ca0 candy

only for exhaustive searches, that much is true.

So why -having this alternative- am I still feeling blue?

Well, I’m not complaining for myself, but for you.

I don’t speak as a searcher, but as a teacher too.

It is so frustrating that I have to explain to you

that each step you take is now multiplied by two.

NLM says all functionalities are still there.

The problem is you have to find where

I don’t mind the present front page,

but the so called “Advanced” gives no advantage,

at least not for doctors searching evidence.

I teach them “Googling doesn’t make sense“.

Just choose the most important concepts,

work from the History and search words separately.

Begin to find the MesH-terms, and although it is complex

add textwords too, to find papers not yet indexed.

Combine synonyms with “OR” and concepts with “AND”,

Go to the Clinical Queries and use the appropriate command”..

But now it takes so many steps. It is a BIG FAIL

sometimes. You start at the front page, look at the Details,

mapping is wrong, go to Advanced, scroll, scroll, scroll..

to Mesh, “send to Pubmed”, where am I? out of control,

again on the Start page? Go to Advanced again.

Away with Limit and other boxes! – I don’t need them!

The Index yields a MeSH that doesn’t exist?!

Darn, via automatic mapping the multi-term-word is split

in 3 separate words, complete out of context,

as I see In the Details -so I have to re-enter them,

And where have the Clinical Queries gone?

Right, have to scroll the entire “Advanced” page… Yawn…

While it is true that I’m a “bit” exagerating,

my point is that the new PubMed creation

could have been so much better:

not only the functionality, the route also matters.

The redesign is a missed opportunity,

to build an entire new PubMed you see.

The interface is still quite orthodox.2946761628_2eb3e8b009 bittersweet

I want clickable and movable boxes

with MESH in clouds thru which you can “walk”

and Clinical Queries that you can drag and drop

with a mapping tool-you can adjust,***

and savings of your settings, that is  a must.

“But the new PubMed”, you ask me

“what is it: a-trick-or-a-treat?”….

“It looks like a nicely wrapped candy,

but tasting a bit bittersweet?!”

Notes

* These links come from Eagle Dawg-blog: Pubmed: All in the attitude

** doesn’t apply to quick and dirty searches on the front page

*** i.e. allow to split or not

Photo Credits:

Reblog this post [with Zemanta]