Speaker 1: So you've got to do a literature review, but you're not quite sure where to start. Don't panic. Help is at hand. Now the biggest problem that people have when they do a literature review is they do a search in a search engine like PubMed and they actually find too many papers. And the trick is to narrow that search down so that you land up with a manageable number of papers without losing important publications. And I'm going to show you exactly how to do that using Boolean operators. Those are ands, ors, and nots and mesh terms in PubMed. So stay tuned. It's worth mentioning that a good search is only part of doing a good literature review. And if you want to learn more about all the other aspects of a literature review, then go to learnmore365.com and I've got a course on how to do a literature review right there. Okay, let's get stuck in. All right. So this is PubMed. PubMed is a database and a search engine that you can use to search biomedical literature. It's got over 28 million citations. If you're doing a search that's outside of the medical field, not to worry. Most of what we're going to talk about applies to any search engine. The temptation now is to jump right in and do a search in the search box. I'm going to say no. The first thing you want to do is you want to sign in and if you don't have an account, create one. It's free and you're going to find it extremely useful. Now we've signed in. And again, the temptation is we see a search box, let's stick in some search terms and start searching. Again, the answer is no. I'm going to suggest we go straight to the advanced section and you're going to see why in just a minute this is a much more powerful way of doing your search. So this is the advanced search builder section and it lets you build up a search query that will hopefully provide you with the results that you want. Let's just quickly have a look at what we've got here. At the top, we've got a section. This is where your search is going to land up being. This is down here is the builder. Now this is where all the action is going to take place and you're going to find this very exciting. And down below, we see a history and at the moment, the history is empty, but you're going to find this extremely useful as well and we're going to look at how to use that. Now remember, we're coming to the search engine having already done some thinking, right? So we've identified a purpose for our literature review and for example, if we were doing a literature review on the impact of poverty on HIV treatment outcomes, we've got three conceptual buckets, right? We've got poverty and there's search terms and keywords around poverty that we're interested in. Then there's HIV, same thing applies and then of course, there's treatment outcomes. So if we take our first conceptual bucket, we want to build up a search term that finds all of the papers that have anything to do with the idea of poverty. So we type in poverty and let's say for example, there's another word, let's say deprivation that might be of interest. Now there'll be many, many more, but we're just going to use two for the sake of this example. Now remember, we said we want to use or and not and because we want to include all the papers that meet either of these criteria, not the papers that intersect with both of them. Now instead of pushing search, if we push search here, it would do that search and tell us about all of and show us the papers that it found. We're not interested in seeing those papers at this point, we're simply interested in knowing how many papers it identified. So instead, we click on just add to history, right? So now in our history, we've got the search number, in this case number 18, that just means that I've done 18 searches today, don't worry about that. We've got the search that we did poverty or deprivation, so it will include any paper that fills either of those criteria and it has found 12,000 or sorry, 125,955 items. If we clicked on that, it would actually show us those items. We're not going to want to look at that number of papers, so there's no point in doing that at this point. Now we might look at that and we say, look, that might not be all the papers that are relevant to this particular conceptual bucket. We think to ourselves, well, perhaps we needed to include the idea of inequality, not a problem. We go down to our builder here, we click on add, it pops that search back up there, we add in another search term, let's say inequality. Again, remember to say or and not and at this point because we're building up one conceptual bucket. Now you'll notice that I've said all fields here. You could be searching by author, by title, by abstract. I usually just leave it at all because we want to be quite inclusive if we're doing a literature review. Then add to history. Now we can see search number 19, it is still this conceptual bucket that we're trying to build up and it's identified 512,000 papers. We would continue doing this, we would continue adding search terms and ideas, separating them by an or until we believe that we've identified within that bucket all the papers that might be relevant, but for the sake of this example, I'm just going to stop there. Now we move on to our next conceptual bucket, HIV. We want to identify all the papers in PubMed that may have anything to do with HIV, so we type in HIV and we include other search terms that we think might be important, let's say for example AIDS. We would continue adding search terms and keywords. We might include human immunodeficiency virus, et cetera, et cetera. In this case, we'll just stop with two for the sake of simplicity. Remember to separate them by all. In this case, I'm going to use all fields to be inclusive and we hit add to history and so search number 20 on HIV identified 432,000 papers. And then of course, we would do the exact same thing for treatment outcomes. We would develop a search term that identified all the papers that we thought were relevant to treatment outcomes. For the sake of simplicity, I'm not going to do that now, but you understand exactly what I'm getting at. So we've identified papers that have got something to do with poverty and we've identified papers that have got something to do with HIV. How do we identify the papers that exist in both of these searches? The way we do that is we click on add and add and you'll see these searches pop up into our builder here. In this case, we do want them separated by and because we want to find papers that fulfill this search query and fulfill this search query. Hit add to history and that has found 18,121 papers. All of these papers will be found in the results of both of these search queries. Now if we click on this 18,000 papers, it does the search, it pulls out the actual papers from PubMed and we can see them here. PubMed gives you the option of filtering this so we could say we just want to look at review papers and you can filter by whether or not the full text of the paper is free and available. And now of course you can click on any one of these papers and have a look at it. Now let's talk about MeSH terms. Before we carry on, I just wanted to say a big thank you to Biomed Central or BMC for sponsoring this video. Biomed Central is a publishing company that publishes open access journals and that means that the full text of any paper that they publish is available for free to anybody in the world. I have really been extremely impressed with Biomed Central as a company. They have integrity and they are truly making the world a better place. So check them out at BiomedCentral.com, I'll put a link in the description below. Now let's talk about MeSH terms. MeSH terms are fantastic and the reason is while we're doing everything electronically here, when it comes to MeSH terms, there have actually been people, indexes, human beings at the National Library of Medicine who have gone through the papers on Medline and identified what those papers are about and indexed them. And they've indexed them using these MeSH terms. So if you do a search by a MeSH term, PubMed will give you a result of all the papers that have been identified by these indexes that have got anything to do with that term. But in order to do a search by a MeSH term, we need to know what the MeSH terms are. And so we can do a search in PubMed of the MeSH term database. So we go up here, click on MeSH and type in a term. Treatment for example. Now PubMed is not producing a result of a list of papers, but instead it's giving a result of a list of MeSH terms, right, because we did a search for MeSH terms. And if we scroll down, we can see that there is a MeSH term for treatment outcomes. So here we have a description of the MeSH term. We've also got subheadings, so you might want to search within a particular subheading of that MeSH term. And we've got all of the entry terms that are used to develop that MeSH term. MeSH terms exist within a hierarchy, so it's like a tree, and you might decide that you want to go more general or more specific, but usually just keep into the MeSH term that you found is about right. To use the MeSH term, simply click on Add to Search Builder, click on Search PubMed and voila, we've got 23,000 papers that have got something to do with treatment outcomes. But just be careful, because we've still got these filters on, full free text and review, and if we take them off, we can see that the MeSH term actually contains 857,000 papers. Now if we go back to Advanced, we can see search number 30 is our MeSH terms for treatment outcomes, and there's the 857,000 papers. And now we might want to say, look, let's take our search that identified everything about HIV and everything about poverty, and let's add to that the treatment outcomes and look for the overlap between them, add that to history, and we've got 464 papers. And that's how you do a structured search within a search engine and database. If you want to learn more about doing a good literature review, then go to learnmore365.com and I've got a course on how to do a literature review right there.