16 Summary and Quiz

EMBASE summary:

• As you have seen, EMBASE indexes each drug that is mentioned in an article.  This is, in part because EMBASE is, initially, machine-indexed and it’s easiest to ‘teach’ the computer to index each mention of a drug name.
• As you’ve seen, EMBASE also tries to index any administration routes that are mentioned by the authors.
• EMBASE keyword searches can be made more specific by using proximity operators (NEAR/x and NEXT/x) to indicate the number of words your are willing to see between the terms for your search concepts and by using field tags to limit your search to the fields you select.

• Proximity searches are extremely useful in drug related searches.   Think briefly about each search topic below, then click on the topic to see a way proximity operators can be used to make the search easier.

• EMBASE field tags can’t be applied to the individual terms inside a proximity search, but tags can be applied to the whole proximity search by adding an additional set of parentheses and placing the field tags immediately after the final parenthesis.

Versus PubMed:

• PubMed only indexes an individual drug if that drug receives more than a passing mention.
• But if you think searching PubMed is the answer when you want to avoid single mentions of a drug, think again.  PubMed will usually index the drug class rather than the individual drugs when several members of a drug class are mentioned in an article even if the individual drugs receive a lot of attention.
• Pubmed indexing does not provide route subheadings, but rather uses route headings.  As a consequence, you can’t associate a route with a specific drug.   A search for — “Topical administration”[mesh] AND “Celecoxib”[mesh] –( without additional concepts) will retrieve a lot of records about topical administration of other drugs (e.g. corticosteroids) when celecoxib is given by the oral route.
• PubMed does allow use of field tags (like the [ti] tag for title and the [tiab] tag for title/abstract).
• PubMed  field tags have to be placed after each keyword.
• PubMed introduced a proximity searching feature in 2022.  However, it is so under-powered that anyone with another proximity searching option is unlikely to use it.

PubMed proximity searching (for those who are curious)

As already mentioned,  PubMed introduced an underpowered proximity searching feature in 2022.  This next section is provided for those who are curious about PubMed proximity searching.  If you don’t care to learn how PubMed proximity searching works, feel free to skip this section and proceed to the next tutorial chapter now. 

Characteristics of PubMed proximity searching:

• only allows searching for two keywords in proximity (not sets of keywords in proximity)
• doesn’t allow use of the asterisk wildcard.
• uses the format:

“keyword1 keyword2″[tiab:~x]

“keyword1 keyword2″[ti:~x]

 You replace the x with the integer of your choice (number of words you’re willling to see between keyword 1 and keyword 2)

example:

The following is as close as one can get to representing our EMBASE proximity search in PubMed.

(“celecoxib topical”[tiab:~4] OR “celecoxib transdermal”[tiab:~4] OR “celecoxib ‘trans dermal’“[tiab:~4] OR “celecoxib cream”[tiab:~4] OR “celecoxib creme”[tiab:~4] OR “celecoxib salve”[tiab:~4] OR “celecoxib ointment” OR “Celebrex topical”[tiab:~4] OR “celebrex transdermal”[tiab:~4] OR “celebrex ‘trans dermal’“[tiab:~4] OR “celebrex cream”[tiab:~4] OR “celebrex creme”[tiab:~4] OR “celebrex salve”[tiab:~4] OR “celebrex ointment”)

The purple-highlighted terms contain “celecoxib”.  The yellow highlighted terms contain “celebrex”.