In the last tutorial, I showed you how to formulate a focused, clinical question using the PICO model. Well, the next step is to run individual searches in OvidSP for each of the concepts, or key terms, identified by the PICO model. The way OvidSP works, is that you do a separate search for each of the concepts, or key terms, and then combine and limit those searches in various ways to achieve the best and most relevant result set. As I've said in the last tutorial, I'm going to be using this question in my example: "In elderly patients with high blood pressure, are ACE inhibitors more effective than beta-blockers in controlling blood pressure and minimizing adverse effects?" From the PICO model, we know that the first concept, or key term, we want to search in OvidSP is "high blood pressure". So, again, this is the main search page of OvidSP's MEDLINE database. You can see, by default, I am under "Advanced Search". Now there is a basic search, and I could simply type in "high blood pressure" in this search box under "Basic Search", and OvidSP's natural language processing algorithm will take that term, and bring back a list of results. But to explain the search process more clearly, and to have more control over the precision of the search, I'm going to be using "Advanced Search" here, and run the search step by step. This also helps you better understand how OvidSP works, and the structure of MEDLINE data. So here I'm going to leave everything as default, including this "Map Term to Subject Heading" box, We will explain what that does later. And I'm going to type in "high blood pressure". And click on the "Search" button. Now what I can see here is the mapping display screen. What this does, is that it takes my query, which is "high blood pressure", and tries to automatically map it to standard medical subject headings. You can see that "Hypertension" is what OvidSP thinks to be the standard term that should be used in my search Now why do I want to do that? Why can't I just use "high blood pressure"? Don't they mean the same thing? Well the first problem, is that people often use different words to describe the same concept. For example, some authors may use the term "high blood pressure" but others may use "hypertension" to refer to the same condition. Running only a keyword search, for "high blood pressure" without checking the map term to subject heading box, will only give me those references that specifically have the phrase "high blood pressure" in them. And I will not get all those references that only have the term "hypertension" in them. The vast number of synonyms in human languages creates problems for running thorough searches in databases. We have to find all synonyms for all concepts in our question. Additionally, the terms we enter in the search box could be ambiguous themselves. Our term "high blood pressure" here, of course, is not that ambiguous. But suppose we enter the term "fall" in the search box. And that could mean many different things in the English language. And therefore we need to restrict the meaning of the word to exactly what we mean by it in our context. And, believe it or not, it is a very hard thing to do. So we have two problems here: the synonym problem, and the ambiguity problem. Now MeSH, or "Medical Subject Heading", is designed to solve these problems. MeSH is a standardized list of terms, and their definitions, used to describe the content of MEDLINE articles. Subject specialists at the National Library of Medicine read every single article to be indexed by MEDLINE, and assign MeSH terms that best describe its content. All articles about the same concept are assigned the same MeSH term. For example, articles on both "high blood pressure" and "hypertension" are assigned the term "hypertension", which is the standard MeSH term for the condition. So if I use the MeSH term, "hypertension" in my search, I will not only get references that use the term "hypertension", but also those that use "high blood pressure". I don't have to think of all the synonyms for the concept in my question. MeSH headings also solve the ambiguity problem. each MeSH heading is clearly defined and means very specific things, and nothing else. So, obviously, the most efficient way to search MEDLINE, in OvidSP is to use the MeSH terms. But how do we know if the search terms we put in are standardized MeSH terms? Well, as we have seen here, the beauty of OvidSP is that it can automatically map the search terms you enter to relevant MeSH terms. So it is important to remember to leave the "Map Term to Subject Heading" box checked in the previous screen. I highly recommend that you always run your search like this. As I've mentioned before, each MeSH term has very specific meaning. If you're not completely sure what exactly a MeSH term means, you can click on the blue information icon to read the "Scope Note" of the term. So in this case, we can see a definition of the term "hypertension" and note on how the term is used in indexing. Down here, we can see some MeSH terms related to hypertension, and some common terms that people often use to mean hypertension. And we can see that our term, "high blood pressure", is here. Let me just click on this "Previous Page" link up on the top to go back to the previous page. If I click on the term, in the mapping display page, I can view the position of the term in the MeSH tree structure. By the way, "MeSH" is not a simple list of terms. It also describes the relationship between terms, so it is a hierarchical tree structure, outlining the medical or related concepts from the broadest to most specific For example, I can see that the term "Hypertension" is under "Vascular Diseases", and has a number of more specific terms listed under it, such as "Renal Hypertension", et cetera. Another important thing to know about MeSH and MEDLINE is the rule of specific entry and indexing practice. which means an article is assigned MeSH terms at, and only at, the most specific level possible. For example, an article on hypertension will be assigned the MeSH term "Hypertension", and not its more general term "Vascular Diseases" at the same time. Likewise, articles that are assigned more specific terms under hypertension, such as "Renal Hypertension", are not assigned the more general term "Hypertension" at the same time. Therefore, if we select only "Hypertension" we will miss all those articles that are assigned terms at more specific levels. Fortunately, OvidSP allows us to explode our search to include those articles that are assigned more specific MeSH terms. For example, exploding our term "Hypertension" will automatically include articles that are assigned more specific terms indented under it, such as "Renal Hypertension". I don't have to go to this tree structure page to explode a term. Let me just go back to the mapping display page, and you can see that OvidSP explodes MeSH terms in searches by default. This is known as "Auto Explode". Another thing to note here is this "Focus" box. MEDLINE articles are typically assigned multiple MeSH terms. The indexers at the National Library of Medicine are required to identify, among these terms, three or four major ones for each article. These major terms reflect the primary topics of the article. So checking this box here will allow us to focus our search to only those articles that have the term "Hypertension" as a major term, and will filter out those articles that only have "Hypertension" as a minor topic. However, just to be on the safe side, I suggest that you begin with a broader search by not restricting the focus. If you get too many references, you can always do the search again, and use focus to narrow the search to more relevant results. So, here I'm just going to leave it as it is, and click on the "Continue" button. Okay, the next screen is "Subheading Display". Most MeSH terms have subdivisions called "Subheadings". Subheadings help further refine your search by narrowing it to specific facets, or aspects, of a MeSH term. In this example here, I am interested in the drug therapy aspect of hypertension, so I can check the "Drug Therapy" box to choose that subheading. I can select multiple subheadings, if that's what I want. If I'm not sure which subheading I really want, I can always check the "Include All Subheadings" box. By the way, I can also click on the blue information icons to see the definitions of the subheadings. I'm just going to check the "Drug Therapy" box and click on the "Continue" button. By now we have finished the search for our first concept, "High Blood Pressure", and the search strategy, and the number of records retrieved, are posted to the search history table of the main search page. Finally, I'm just going to mention that a lot of the OvidSP features we just talked about, such as the MeSH term mapping, tree structure view, subheading view, and the Scope Note, the "explode" function... these are also available under "Search Tools" You can see that the default selection here is "Map Term" This will be the same as the "Map Term to Subject Heading" box under "Advanced Search". So if I enter a term here, such as "high blood pressure", it should give me "Hypertension", and I can continue to do my search from here. This "Tree" tool is for directly viewing the position of the MeSH term in the tree structure. It'll take a MeSH term, such as "hypertension" So it should give me the position of the term in the tree structure. We've seen this before. The "Permuted Index" is like a simple keyword search for MeSH terms. This would be especially useful if I can't remember exactly what a MeSH term is, and just want to verify the term. So, let's say, if I type in the word "compliance", I should get all MeSH terms that have the word "compliance" in them. Going back... The "Scope Note" tool, will take a MeSH term, and give you the scope note of that term. The "Explode" tool will take a MeSH term and explode that term for a search. And you probably know by this time, that the "Subheading" tool will take a MeSH term, and let you select subheadings for that term. So anyway, you can either use the advanced search to run a step-by-step structured search like we did previously, or you can use any of these tools to fine tune your search terms, and start your searches from there. So in this video, we learned how to use OvidSP to run searches for a specific concept, or key term, identified in our PICO model. In the next tutorial, we will run similar searches with other concepts in our PICO model. And I'm going to show you how to combine and limit those searches in various ways to further fine-tune the search results to the most relevant articles. I'll see you next time.