Amara

[more free HINARI training, including Universal Subtitles at hinaritraining.org/​]

Transcript of video:

Advanced PubMed: Literature searches for complex questions.

While a simple search in PubMed will often find relevant articles, you may find that sometimes you either need to be more thorough or that the simple search did not retrieve what you hoped to find. When this is the case, you may want to approach your PubMed search differently. This video will walk you through the steps of conducting an advanced search in PubMed.

To do an advanced search, it is helpful to understand what is in PubMed. Advanced searches also require that you learn to identify the key concepts in your topic, identify subject headings relevant to your topic, and use some PubMed search tools, including Search Details, the MeSH database, and Advanced Search.

First, what is in PubMed? We know that it is a database of citations to journal articles in medicine and the related sciences. But is the full text of the article also available within PubMed itself? Unfortunately not. PubMed contains only the citations to articles, and LINKS to the full text options.

This means that you are only searching the citation information. This includes the article title, journal title, authors, page number, volume and issue, and the abstract if it’s available. You are also searching something called the Medical Subject Headings, or MeSH terms assigned to the article by indexers at the National Library of Medicine. These MeSH terms are chosen from a list to describe what the article is about.

So to get articles on your topic, it is important to choose search terms that will best match the MeSH terms or medical subject headings.

In PubMed, your search terms are automatically mapped to MeSH terms, so many times this is done completely behind the scenes for you. However, it is a good idea to check the Search Details to check to make sure that the mapping has found a good MeSH term for your concept. If it hasn’t, then you can look for a better subject heading in the MeSH database.

Before starting your search, it is important to think carefully about your topic and what information you would like to find. So as you think about your topic, take a moment to identify the key concepts in your search. In our example of attitudes toward the HPV vaccine in developing countries, our main concepts are likely the HPV vaccine, attitudes and behaviors, and developing countries.

To be facilitate better mapping, it is a good idea to search one concept at a time. Let’s start with our first concept, HPV vaccine.

After logging into HINARI and clicking on the PubMed link, we are at the PubMed home page. We can type our first concept into the search box. [type HPV vaccine]

Note that as you type, suggestions may appear. These are not MeSH terms, but are rather terms that other people have used. Generally, it is safe to simply ignore these terms.

After clicking search, the results appear. In order to determine whether we were mapped to a relevant MeSH term, scroll down on the results page to view the Search Details box. To make this area larger, you can click on See more.

On this page, note that our term, hpv vaccine, was mapped to the term papillomavirus vaccines. The system also looked for the terms papillomavirus vaccines in all other fields of the citations.

Since our terms were hpv and vaccine, it also looked for those in all fields. So, we have been mapped to a MeSH term that is relevant to our topic.

We can now search for our second term, attitudes. Directly from this search details page, we can type our term into the search box at the top. [type in attitudes, and click search]

Note that we are not searching within our set of citations about the HPV vaccine, but are rather creating a second set of articles on our second concept. We will combine these at the end.

As we did with our first concept, we will now check the Search Details box to see what kind of MeSH term we were mapped to for attitudes [scroll down to search details].

In this case, we were mapped to a MeSH term for attitudes. This might be a good term, or it might be too broad. If we want to determine whether there could be a better MeSH term for our topic, we can scroll back to the top of the page and change the search option to MeSH. This will let us look directly in a database of MeSH terms for one that is the best for our concept. [change search dropdown to MeSH, click Search].

Our results are now MeSH terms that have something to do with attitudes. Note that these are not citations, as we are now in the MeSH database NoT the PubMed database.

From this list of terms, it appears that there might be more relevant subject headings for our topic. To see the definition of a term, click on it. [click on attitude to health] This term is specifically about attitudes to health, disease and the medical care system, and is more specific to our topic than simply the term attitudes.

Scrolling down on a MeSH term page will tell us more about the term. From the outline structure on the page, we see how it fits in with other MeSH subject headings. This is a broader term than health knowledge, attitudes, and practice, and includes such other terms as patient acceptance of health care and treatment refusal. These would be good terms to include in our search.

In PubMed, when you search with a MeSH term, the system will automatically include citations that are indexed to terms that are narrower than your term, unless you tell it not to. So if we search using this term, we will automatically also find citations that are about these narrower topics.

Once you have identified a MeSH term you would like to include in your search, you can send it directly from the MeSH database. [scroll back up] Simply click on the Add to Search builder button, then click on Search PubMed. This takes you back to the results page in PubMed for your selected term.

We are now ready to search for our last term, developing countries. [type in developing countries, click search].

Again, it’s important to check the Search Details to determine if this was mapped to a relevant MeSH term, and we see that it was. If it had not been, we could return to the MeSH database and find the appropriate term. In this case, we can now move on to combine our sets together.

To combine sets, click on Advanced Search. This page is where you can come to find your search history for the session. To combine sets from the search history, simply type in the set numbers associated with the best sets together with the best connecting words – And or OR.

AND is used for finding a set of articles that has both terms, and OR is used for finding a set of articles that has either term.

In this case, we will use AND, because we want our set of citations to include all of our concepts – hpv vaccine, attitude to health AND developing countries. [type in #1 AND #3 AND #4]. Note that you must use the # sign for the system to recognize these as set numbers.

When you have combined the terms, click Search.

As this is a fairly narrow topic, there are only a few results. In Video 3, we used the Limits page. Here, using Limits is probably not necessary as there are not many results to look through. Instead, we may want to look at the filters for Free Full Text and HINARI to see what can be accessed in full text.

In summary, to conduct an advanced search, you should first identify your main concepts. Search one at a time. Check Details to see if you were automatically mapped to a good MeSH term, or medical subject heading.

If you weren’t, go to the MeSH database to find a better term and send it to the search box.

Combine your best sets in Advanced search using ANDs and Ors.

Apply limits if necessary. And look for HINARI and free full text buttons to connect to the
full text of articles.

This concludes Video 4 – Advanced PubMed.