Well thanks thanks for giving me this opportunity. It's been a nice session this morning. Um so my topic is how to search the literature effectively. Um i think it's a little difficult topic. Because there are no set guidelines or principles that's governing. You know like we have a lot of uh big time researchers here they have all they all have their own way of doing things you know like so. There is more than one way of skinning. The cat like not many people will agree on. Okay this is the best way but i'll try to kind of put it show you tell you guys my way so this this is kind of one of the generally accepted way of searching the literature. First you find out what your research question is. And then you conceptualize it and then you decide where you are going to search for the answer and then you decide how your search strategy is going to be and then you decide how good the research results you got it and select which one you're going to say use and which one you're not going to use so to tell you the truth. No i'm not big on pickle like uh because when i was uh trained this speak out was not a big thing you know so i kind of still use. Uh what is my research question and then i start using the keywords and i start doing it but i think pico is one thing. I think there's a talk after this one once you decide what's your research question. You have a lot of options. You have a lot of options you can go to the pubmed you can go to the uh ov sp we can like uh uh proquest we have embase we. Have you know you can go to the md console science direct. You can cochrane. So how are you going to decide where you are going to search you know like everybody like those days when we are getting trained. Whoever can articulate that uh thought or whoever can talk loudly he's right but now there is data for everything you know so yes there is data for where to where to search. Actually if you look at uh the study by slover gn you know. Like if you use the pubmed and the embeds together uh you know you can get at least 91 percent of the primary studies.

And if you add the cochrane which is basically a review database of reviews you can get almost 97 of the series that's pretty decent and then you have to do some unconventional searching after that. Uh now google scholar is big in u.s. Uh and google scholar actually is uh sure has been shown to uh give like almost twice as many relevant articles as pubmed but still i have to tell you the truth like uh most researchers especially in big universities nih funded researchers they use the pubmed and uh when i when i started i used to go to the individual websites like the villain science the scopus this thing science direct all those things and then search independently but then now i realize that it's not yielding any significant uh new articles so i now stick to the pubmed maximum i may go to the google scholar. I have to tell you the truth that i don't use the m base because access is limited and ovied. I don't have access in india so and if you are using a google scholar like uh you better turn off the ip address sensor because it will only give you the what is pertinent for the indian subcontinent. That will actually list it. Uh that way so you have to kind of turn it off. Once you decide like i decided which database you are going to use. Now i use the pubmed the i go to the pubmed and then i use the search term like the keyword or there are two ways you can do it one. You find the article that very closely matches your uh research question like uh for example you want to talk. Uh do a research on external trochanteric osteotomy internal hip replace revision hip replacement. You just put in like if it is a non-clinical guy. Then he may use like a replacement and all those things so there are some specified mesh terms for those kind of guys but for surgeons who know the exact term so you can go and type find the article that closely matches your article and then you try to go to the related article.

That's that's what i usually do but the recommended method is first you you kind of identify the mesh terms that will closely match your research question and then once you have some uh kind of a results you go into the mid-stream of each result and then you try to kind of combine them with the booleans. The pubmed boolean so so these are the two ways as i said like either you can use the mesh terms or you can click on the relevant article. So if you don't use the uh mesh term your results will be too many and you'll be relevant so if you use the mess from it the accuracy improves and this is the uh first page like the home page of the pubmed. You see all those things the clinical queries the single citation matcher. You know you can put the clinical query there and get the result and take it from there so also you can do an advanced search in the problem go into the individual journals and actually do it but uh that's not the effective way of doing it so as i said like if you put an extended token request to item you'll have a bunch of things and if you see the rounded area that contains all the mesh terms and the boolean things so you can take these mesh terms and then add it with your other terms. You are thinking about maybe revision. Maybe you know like cables or something like that and then you can get generate more results once you have your results and you know what happens is it keeps giving you the same result again and again so i usually try to create my own collections. I have a login. You can use your google to login you can get into the pubmed and save all your searches there and uh once uh you save all your searches. So that's all the login you know so it will list your thing so the next time when you go for because the search sometimes happen over a period of time so later on when you go search you find more articles you can keep adding on to that and once you've done all your once you found all the relevant necessary article then you try to export it to the pubmed uh like the some of the citation managers.

That's the difficult process. You have to manually enter that that. I don't think there is any way you can. Effectively transfer the pubmed articles to one of the referencing softwares. And uh once you get go through all the abstracts what you have to do is try to find the full text of uh the reliable pertinent articles. I've had one resident uh who i asked him to write an article and he went through all the abstracts and wrote a manuscript. I said that's unacceptable. That's unethical you the relevant article you have to beg borrow or steal and get the full text and you have to then look at the back references and you have to snowball the references and find more articles. So um that's that's very important. You know like uh your uh literacy search will decide that you don't want your review or quote an article and saying like you have missed this article it's like it is it is a big uh and then i use an endnote efficiency is about doing it. Like effectively means you have to do it in a quick and reproducible manner uh less input and more output so i use endnote and i create a library there and i try to export all the citations there and uh there in in the endnote you have a option to search the pubmed so you can go through that or you can manually enter it so and then with some of my publications so apart from the index journals you have to also go through the gray literature. Grade literature is something that's uh published in non-indexed areas in the pubmed in the internet and also like all the conferences. Uh all those things so it is better to include if you want to have a comprehensive study and once you have all the articles you need to find choose because now there's so much literature you can find evidence to support anything you need to find not only about the studies but also the quality of the studies. I always look for where the study is from. Which by which country is a country with a very high human development index very regulated country or is it published in one of the top journals and you know see what kind of studies it is.

The case series is rct and the number power is very important. The power of the study is very important and then you include those studies you know with maximum citations so uh in natural you know like a. It's it's a very very difficult process when you're starting it. You may find it very intimidating so you need to have a thorough subject knowledge to find out what's relevant and what's not relevant you take the help from your guide and uh also you know with the experience. You'll know how to do it some references for that and uh my postgraduate moonish helped me with this presentation. I would like to thank him for that thank you.