Showing posts with label pubmed. Show all posts
Showing posts with label pubmed. Show all posts

Sunday, December 13, 2015

Advanced PubMed Searching




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 vaccine, which is relevant to our topic. Note that the system also looked for the terms papillomavirus vaccine and hpv vaccine in other fields of the citations to be sure to pick up newer 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 box 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 the Advanced search link [click Advanced]. This page is where you can come to find your search history for the session. To combine sets from the search history, simply click ADD next to the best set from your searches. We will choose the sets associated with hpv vaccine, attitude to health, and developing countries. This will place the sets into the search builder.
Remember that AND is used to find a set of articles that has all of your terms, and OR is used for finding a set of articles that has any of your terms.
In this case, we will use the defaule AND, because we want our set of citations to include all of our concepts. We will now click Search [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.

How to Search PubMed in HINARI




Transcript of video:
Using PubMed through HINARI can help you find articles on a topic and connect to HINARI or other free full text versions of them quickly.
PubMed is the US National Library of Medicine’s search engine for the MEDLINE database. It searches over 20 million citations to journal articles in medicine, basic sciences, nursing, dentistry, and related biomedical fields.
HINARI links to a special version of PubMed that makes it easier to find full text articles in HINARI and articles that are available for free on the Internet. Let’s click on this link to open PubMed.
The example topic is malaria.
Type a term into the search box. As you type, terms other people have searched for appear below. These are not recommended or preferred terms, so it is best to ignore these. Click search.
This search was very broad and retrieved many results. PubMed has many citations, so it helps to be specific. For example, if I am treating pregnant patients with malaria, then I could improve my search by adding AND pregnancy after the initial search term. This refines our search toward our topic.
Note that using AND between search terms narrows your search to only those citations that have both concepts. Using OR between search terms broadens your search to retrieve citations with either concept.
Once you have a set of articles on your topic, use filters to refine your results more. Filters appear to the left of results.
I can narrow my search by date (such as the last 5 years), publication type (such as clinical trials), and even patient characteristics. More filters are available by clicking show additional filters, for example age. If we were interested in articles on effects of maternal malaria on infants, we would choose the infant filter. If we want articles on malaria in either mothers or infants, we could check limits for adults and infants.
Be careful not to choose full text or free full text yet. Choosing these may remove some of the articles that you could get through HINARI.
I now have results that are clinical trials from the last ten years about malaria and pregnancy in adults and newborns.
The default view for results is called the Summary view, and lists only the basic information about the article. If you would like to read more, click on the article title. This will take you to the abstract, if it is available.
This view also shows you access options. The first button is the publisher button. The second is the HINARI button. To get to full text, click on the HINARI button.

Look for the PDF option for the easiest saving and printing. The PDF looks like the paper version, has full-sized graphics, and page numbers. Other journals will look different, as they may be published by different companies or societies.
To get back to the results, close this journal window. Then click back on the browser.
Some articles will be available for free – see the Free in PMC note below the citation. This article is available for free from the journal, through HINARI, and for free in PubMed Central, the open access repository tied to PubMed. Click on any of these buttons to get to the full text.
Other articles may not be available through HINARI or for free. If you see NO buttons on an abstract, then it is not available.
If you see only a publisher button (NOT HINARI and NOT free full text), then you will likely be confronted with the option to buy the article, rather than the full text itself.
If you want to browse only those citations that you know are available via HINARI or for free, try the filters to the right of the results.
• The HINARI filter limits you to citations available in HINARI journals
• The Free full text filter limits you to citations available for free.
Browse both lists, as there may be overlap and unique citations between the lists.

This concludes the video on How to Search PubMed in HINARI. For more advanced searching techniques, see the video on Advanced PubMed Searching.