Computer generated transcript

Warning!
The following transcript was generated automatically from the content and has not been checked or corrected manually.

Thank you very much for that kind introduction. Let me just share my screen now. Um I hope you can see that. Yes. Yeah. Yep. Perfect. OK. So I'm gonna just move this cause it's a little distracting to see myself. Perfect. Um So thank you. Um I'm gonna talk a little bit about um literature reviews and I plan to share with you lots of the things that I wish somebody had shared with me at the beginning of my um somewhat accidental research um journey so far. Um I'm gonna give you some hopefully practical examples and talk you through some of the papers that I've written um in an attempt to kind of show how the principles that I'm describing can be applied to um the work that you might want to do um in your radiology um academic career. So I'm hoping that the slides have changed. OK. So um I've been kindly introduced already but just a very, very brief background. Um I did a degree in physiology before I went to medical school, um wrote a dissertation as part of that degree and that's probably my first true experience of research and that was a long time ago and a lot of things that I did during that degree I would never do. Now, um lots of things were done longhand um without the benefit of kind of having somebody to guide me in. This is an easier way to manage your references or this is a better way to write or structure a paper. So that was quite a lot of um teaching myself through that process. Then I went to medical school. Um and then I started my radiology training, got my FCR. And after I did all of that, I then started a phd in 2016 and I just um was awarded that um earlier this year. And I've got my graduation in a couple of weeks. So I now work as a um consultant G I radiologist at Saint Mark's um in Harrow, which is the National Bowel Hospital. And all of my research or most of my research has been into um CT colonography which we use to image the bowel and um colorectal cancer. Part of that work that I've done and part of my research and pictures shown on screen is one of the workshops that I did as part of my phd. My phd was called Perfect. And um part of that work has now informed the National CT Colonography Training and Accreditation program. So if you do um come into radiology or if you're already in radiology and you are a trainee, you will be doing some of the training that I have developed and built um which is delivered online. So that's a really nice combination to my research to kind of have this practical thing that I've been able to, to deliver to the radiology community. Um I've published um a few papers and um the ones that I'm most proud of are in um radiology um which was the main research paper from my um phd um my systematic review which um is published in the Lancet Gastro um Gastroenterology and Hepatology. Um and then a couple of other reviews um which I'm gonna talk to you a little about. So those are my qualifiers and that's why I guess I'm sitting here speaking to you. And so I want to share some of the things that I've kind of learned along the way, which will hopefully um be really useful for you. So if we just go back to first principles, I'm gonna try and answer these questions during the course of the presentation and we'll have time for some Q and A at the end as well. So firstly, what is a literature review so very broadly speaking, it's analysis of all of the existing literature on a specific topic. Actually, I say all, sometimes it's not all, but it's looking at a specific um area of interest and reading around it and then writing about that and it provides an overview of knowledge. So if you read lots and lots of papers. It's gonna be what have you understood from what you've read. How can you, um, kind of redeliver the existing knowledge? How can you explore what gaps there might be in the existing research? Can you identify any trends that can inform the research that you're gonna go on and do so? Why would you do a literature review? Well, primarily it helps you to gain a deep understanding of a particular area which is always helpful, but not only that, it helps you to pull together ideas, concepts and theories from lots of different places and then pull them together into kind of one cohesive insight that you have. And so that can be very valuable because it can um it's how we move research forward by picking up different um trends between different labs and different institutions and seeing. Ok, there's a gap here. No one's really thinking about this. How can we explore that area? How can we improve things in that way? So, um very valuable. And I think also if you end up going down even a um AAA small research journey as opposed to doing a phd or an MD, um A literature review is probably one of the first things that you'll have to write or that you'll have to do. And it's really good to have an understanding of how to do one and what the um kind of tips and tricks are for having a successful um literature review published. So we just broadly talk about the different types of literature review. These are the three main types, the narrative review, um scoping review and then a systematic review. And they basically in that order, get more detailed and more complicated to perform. So I'm gonna go through each of those in order um with examples from work that I have published. So a narrative review is basically you've sat down and you've decided I want to learn like I did everything about ct colonography and I'm gonna read lots of papers and I'm gonna just do like a summary and an interpretation of the existing current literature around that topic. And then what I'm gonna do is when I've read all of that stuff, I'm gonna provide insights and say, OK, well, we know quite a lot about um what the quality of ct colonography should be like, but we don't have a lot of information or there's not a lot of research on how cost effective it is compared to other ways of imaging the bowel. And in reading around the literature, you can expose those gaps, you can see where people have not really done research yet and that can potentially inform your own research question. Now, a narrative review doesn't follow necessarily a specific search methodology. Again, I'm gonna go on to discuss that a little bit um in a second and it doesn't um there's not strict criteria about the studies or the articles that you include in your narrative review. So when I'm doing my narrative review search, I'm looking for papers that are um in line with what I want to talk about rather than looking at all the papers about this particular subject. So this is an example of a narrative review that um I wrote um a couple of years ago now this was published in BJR. And this um the yellow arrows are basically highlighting the main um sections in the at the beginning of the paper. So obviously, you have the abstract in the gray box, you have an introduction. There's no method. After the introduction, it's a narrative review. We're not really interested in my methodology. We're just interested in the topics that I'm gonna talk about. So I go straight after the introduction, which includes some of the background context. We go into the logistics of image based screening and then I talk about CT colonography and the different aspects of that. Now to view is also a nice um it's a nice um paper to read, to get an understanding of a of a topic that you might not know much about. So if you were, I don't know, wanting to learn about rectal cancer, for example, you might not want to dive straight into the very detailed um research papers looking at the frontiers of imaging of rectal cancer. But what you would want is a nice narrative review to give you an overall understanding of the landscape. And so this was um what I wrote for CT colonography. So I looked at all of the different aspects of ct colonography, like the logistics cost effectiveness, et cetera, you can see in the title. OK. Um So if we go on to the scoping review, so this now is a little bit more detailed. Um It maps the literature and um around a particular topic and it provides an overview based on the evidence that you've got. And in this type of review, you want to have a more systematic approach to your research um to your sorry to finding your studies. So you might not do a full critical appraisal like you do in a systematic review, which I'm gonna mention. But you have a bit more structure to um how you do your searches and in your write up, you're going to talk about where you did your searches, where you got the articles from that have informed your scoping review. And again, you want to identify key concepts, you're gonna talk about the types of studies that you've included, the types of evidence. And again, you're gonna highlight um some research gaps. Now, a scoping review is a really good type of review to do before you do a full blown systematic review, which is a major major piece of work. So a scoping review is a little bit like dipping a toe in to kind of see what's going on and that'll give you a steer for how much literature there is out there on a particular subject. So for example, this is a scoping review um that I published last year. And so the yellow arrows now are highlighting again the introduction. But you can now see that there's a methods um section and in the yellow box, you can see um you might not be able to see, but I talk about what um search terms I use specifically and the database that I used to find the articles that I then go on to talk about in my results section. So in this type of review, you have an introduction method results ie the results of all of the studies that I have found what they show. And then um you go on to a discussion at the end. So generally, it's good practice, you probably won't get published if you don't have this. Um or certainly the reviewers will ask you to put it in if you don't have it. And that is um a study selection flow diagram. So this you normally would have in a scoping review and a systematic review. And it basically shows all of the steps that you went through to find the articles that you've then gone on to analyze and discuss. So the yellow arrow at the top shows that I search PUBMED for studies according to the search terms that I've mentioned in the main text and then you're gonna say how many studies um where, where the, what came out from that search. So in this particular instance, it was 987 studies. Often when you run a search through any of the big databases, you'll get duplicates. So you'll get articles that come up a couple of times. So you have to obviously remove them. Once you remove them, you will have a certain number depending on your topic of abstracts and um article titles that you can then review. So normally what you do is you would read through the titles and the abstracts and you'd be like, OK, this is about CT colonography. This one is about this one crept in by accident. It's actually about cardiac CT I'm not interested in. So that's not eligible. So I'm removing that and actually this isn't primary research. This is another review. So I'm not interested in that. So I'm gonna dump those ones to you. So you go through this exclusion process until you're left with um hopefully a much smaller number of articles for which you are now going to read the full text. So in this case, I had 27 articles that I was going to read the full text of and in reading those, you then realize actually a certain number of those don't um aren't um in line with my research topics, I'm going to dump them. So maybe it's the wrong study design. Maybe it involved animals, but you only want to focus on humans or maybe there was an extra element um that you think actually I don't want to go down that road. I want to focus just on this small area over here. So you're gonna exclude those. And in this scenario, I had 19 studies left that I then went on to analyze and do my um scoping with you about. So, um I only, and it's important to make the distinction that for the scoping review, I only use PUBMED. So I wasn't trying to find as my database to search. I wasn't trying to find every single article about a particular topic. Now this is different to if you do a systematic review. So a systematic review is hard work. So systematic views is different in the sense that it's much more rigorous and it's comprehensive and if it's done well, you should find all of the literature about your particular topic or question that is out there. Um It follows a predefined protocol. So you're gonna write how you're going to do the systematic review before you do it. You're going to write what your structured approach to your search is going to be and you're going to analyze all of the relevant studies and you're going to say what you intend to extract from each of the studies and you say that before you start and it's important that you do that before you start. So that you can mitigate against introducing any bias into either the papers that you've selected or how you interpret the papers that you've selected. So you're gonna establish a specific research question before you begin, then you're going to write the protocol. Hopefully, think about every different kind of permutation around your topic and then you are going to ideally publish that protocol before you've even done the systematic review. Now, if you want to do a systematic review, do it well and get it published in a high impact journal, then you have to make the protocol publicly available journals like the Lancet Radio, they won't even accept the submission if you haven't done that as a first step. And it's a good process to go through because it forces you to have a really robust um plan and to interrogate it um and then submit it for publication. So on screen, you can see um the sys the protocol for the systematic review that I went on to do was published in a journal called Systematic Reviews, which is where lots of people publish them. And the particular systematic review I um did was um trying to work out what the post ct colonography interval cancer rate is. So basically, if you've had a CT C and it was reported as normal, what is the rate of missed cancer? So what is the rate of patients that actually have a cancer that wasn't detected? And so in colonoscopy um there's been lots of research about post colonoscopy cancers. But in ct colonography, there was basically nothing has ever been published in a systematic review format. So that was the gap in the literature that um we wanted to fill. And so um we wrote the protocol and published the protocol. And then the other thing that a systematic review often includes, not always because it depends on the quality of the data that you get from the systematic review is a meta analysis. And this is a statistical technique that combines the data from multiple different studies and then summarizes them to give you an estimate. And that is what I wanted to get. I wanted to get an estimate for the interval cancer rate from looking at all of these different studies that had looked at missed cancers or interval cancers after a CT colonography. And so um I'm going to discuss that in a little bit more detail in a second. So one of the ways or two of the ways that we improve the quality of our systematic reviews and the credibility of your reviews is if you adhere to um the prisma guidelines, and if you register your systematic review on pros, which is like this international of all of the systematic reviews. And it's like um when you register a trial and you get a trial number, it shows that it's legitimately gone through a process and it's um people can read about the protocol, people can see what you're doing and there's this greater accountability and transparency. So pros is an online database where you can register your systematic review before you start. And also it allows you to see if anybody else out there is also doing a similar systematic review. So it avoids duplication and it also promotes collaboration between different centers or different labs by keeping this comprehensive record of ongoing or completed systematic reviews. Now, Prisma Prisma P um is a guideline that allows um it shows researchers how to report their systematic review. So basically, it's this long checklist of stuff that you have to include in your write up. Again, if you submit your systematic review to a um a high impact journal, they will want to see that you have a adhered to the Prisma P guidelines and literally you have to show for this, for this guideline. So this step on in the guideline, I have written the relevant thing on page four and on page six, I've written the thing that I needed to write for 1.3 and it's really, really specific and you have to show evidence that you've followed the guideline and then direct um the reviewers to the specific parts of your manuscript. If you do this, then you're much more likely to get your systematic review published in a high impact journal. So I'm just gonna talk you through um my systematic review, as I mentioned, this was published um in the Lancet um Gastro H back in 2018. And um, you can see, I've just circled, this is the first page um of the publication and I've just highlighted again. Usually you want to have systematic review and meta-analysis in the title of the publication. Um Not least because when people are doing their own searches on PUBMED, you want, you want to be able to be um located and identified. So as I mentioned, uh the question that I specifically wanted to answer was, what was the post ct C colorectal cancer rate? And so I did a systematic review to try and determine that, which involved doing a search of all of the databases that I could find um with really specific search terms which I'm gonna um share with you in a second. So this as you can see compared to the first two samples I showed you is much, much more detailed paper. Um for the lancet, they have an additional um insert that they like you to write. So not the abstract, not the main manuscript, but this thing that's in the pink box, which is the research in context. So what they want to um want you to do is basically summarize your systematic review into what was the evidence before you did this work? What value have you added through your systematic review? And what are the future implications of what you have found? Again? Really helpful exercise? For you to do to reflect on the work that you have done. And I think it's really actually helpful to see what the publications look like before you start. So that as you're thinking about research questions and as you're thinking about what I could do work on, you're already thinking about what this could look like when I need to write it up for the journal. And if you keep that at the back of your mind, I just think it's a really helpful um way to kind of keep you on track. OK. The second um arrow um highlights that we search, not just PUBMED but basically MEDLINE M base, the Cochrane um register of control trials and that we use specific um they're called um mesh terms. So um medical subject um headings, I'm gonna show you a sample um of some of those. And again, you want to be able to demonstrate your specific search terms, the date on which you did the search and the dates between which you ran the search. And this is also that if I gave my um search terms to you and you ran it at the same time, you should come up with the same articles. And it shows that we have done a thorough comprehensive um investigation of all of the literature that's out there. OK. So, um on the next page, the yellow box highlights here that we've adhered to um the Prisma um guidelines and that the protocol is published and is widely available. Um And then we have to do you have to write a whole section on how you analyzed your data. Um again, for um robustness of the, of the systematic review, you want to have more than one person doing the extraction of data from the article. So once you've got your 12 articles, for example, um I'm going to look at them and I'm going to extract all of the data that we said in our protocol prospectively that we're gonna extract. So for example, the author, the year of the publication, the sample size, the um the country where it was done if it was a multicenter or a single center trial, if it was an observational trial or a randomized control trial, all of these metrics um which we've agreed beforehand, I'm gonna extract and then someone else is gonna extract and then we're gonna compare and hopefully we've extracted the same data, we've inferred the same things and got made the same analysis from each of the studies. And if we haven't, then we need a third person to, to have a look and basically arbitrate and that just ensures that the data that you've extracted is um accurate and that if anybody else was to extract it, they would also come to the same conclusion. So it just increases the quality of your systematic review. The yellow arrow here highlights that um meta analysis, how it was done, the software that it was done. Who did the metro analysis? What the parameters were for? Um um confidence intervals, um statistical significance, et cetera. Um The, yeah, the first yellow box um highlights the pros registration number. Again. If you want to be published in a high impact journal, you have to have that and you have to include that in the manuscript. And then the bigger yellow box is showing um the study selection process um which is similar to what I mentioned for the um scoping review. It's just for the systematic review, you have to be much more specific about um why you've excluded a study and why it didn't meet the inclusion criteria and why it wasn't eligible. So you can't just say um three studies were um the ro were wrong. For example, you have to say why they were wrong. So I've put here inadequate follow up or inadequate ct colonography technique um or that the CT cally was performed at the wrong time for what we were trying to um investigate. So this is a just a diagram to show you how the study selection works. So you start with loads of studies at the beginning. So we searched those databases that I mentioned. We had almost 3000 3000 search results, then you're gonna take out all of the duplicates. So even after doing that, we still had over 2000 titles and abstracts that we had to, to screen. So I screen the list and then my colleague screened the list and then we put our list together and obviously some of them I would have dumped and he kept and vice versa. So then you put those together. So that was 63 papers that we agreed needed to be read to determine whether or not they were eligible. And then we discarded a whole bunch because they weren't eligible. They were either the review articles, they weren't primary research or they were editorials um or um et cetera. And that left us with 12 studies and those were the 12 studies that we then performed the systematic review. And then the results of that is then what we did the meta-analysis on. So hopefully, that's all clear. So now we're just gonna go through some steps on how to do a literature review and I'm gonna hopefully give you um a few examples and talk about what are good things to try and include. So the first thing that you want to do is define a topic. Um Obviously, this can be an area that you're personally interested in or something that's already happening in your department or if you're in an academic center, research that's already going on. If you do end up going on to do further research, then as I mentioned before, this process can probably be, will go towards being like the introduction for your thesis or the introduction to your next big piece of, of work. So you can start with an idea. Don't be afraid to adjust it if you find out that actually there's like a million reviews that have really been written on this. Um But then, um so don't be afraid to change your scope and then you want to break down your research question into like key words or concepts. And again, as I mentioned before, I think it's really helpful to think about either the target journal, um or the target journals that you might want to publish in and look at the kinds of things that they publish, um the length of um of their literature reviews just to kind of have that in the back of your mind. And also I think that it's good to kind of manifest your publications. So once you've got your topic, you're gonna do some searches now. So where are you gonna search? So, databases, which I've already mentioned, there's also great things like Research Gate and Google Scholar, um professional society website. So like RSN A buzz ESG IG I, so I only know all the G I ones, but there are, there are other ones that are not G I related, um which you can also search for. And so you want to search for primarily articles, but also you can get some great information from books from theiss. Most of the universities will have open access to their phd and MD graduate thesis. So when I finished my thesis and defended it, you have to submit a version online um which now is like publicly available. So you can get great information by reading other people's thesis. Um conference abstracts and posters. Often um people will submit um an abstract to a conference but may not then end up writing it up. But they have still got really interesting or valid results and also not to forget, clinical guidelines can also be really helpful. Now, one of the things that I found really useful was um going through reference lists. So if I find an article, I'm like, oh my goodness, this article is the business. It's talking about all the stuff that I wanna know, blah, blah, blah, blah, blah. I will spend time specifically after reading the paper going through the reference list um or when I'm reading through. And I'm like, oh, this is a really good point where have they look at who they referenced and where they got that information from and then go back to that paper. And it's just a really great way of kind of networking the papers together and following those trains of thought. So how are we going to do our searches? I've mentioned it a little bit already or alluded to it that you're gonna use um specific commands um or terms and so most of them are universal. Um But there'll be specific commands that are relevant to specific databases. So commands like and or, and not, you can use to refine your search if you have one term and then you put and, and then another term, then your search will combine um those two topics together. Um You can also truncate words by using um an asterix. So um in the box I've shown, so I wanted, this is the search, the search terms for the systematic review. So I wanted to, to search for any article step one that had either CT sometimes people spell it out. So commuted. So computed tomography and then I needed to add colonography, right? Um And so if I put an asterix towards the end of the word and truncate it, then you get all the different possible endings, right? And the idea is that you have as broad um a net as possible so that you can capture every possible article that relates to your topic. This is for systematic review specifically, then the dot A F is all filled. So wherever that's mentioned, if it's in the title or the abstract or if they've mentioned any of those terms, I want you to pull it. But then I also want you to pull because CT colonography is relatively new terminology and CTC used to be referred to as CT virtual colonoscopy and that's kind of died out of favor now, but there will be lots of publications that use that um as a descriptor. So actually I need to include that in my search, which is step two, which is virtual with a asterix to make sure I get all the different permutations and colo with an asterix et cetera. And then I want to combine one and two or one or two so that we don't um we can reduce a few of the duplicates, et cetera. So um dot sh is subject heading. So if um step four, you can see if those words are in the subject heading. I want those as well dot um PT is the um publication type. So for example, I want a specific journal article, I don't want a textbook, for example, for the systematic review. Um And so you can go through and that's why it's important to break down the key words that make up your topic so that they can inform your search strategy. Um er you can also do phrase searching by using um speech marks um around the phrase to make sure that the articles that you search for include that phrase. Now you don't have to do all of this on your own. Um Most hospitals, if they have an education center will have a librarian who can support you and at least give you some advice or direction. If you work at a hospital that's attached to university, then just go and spend an afternoon with the university library. They normally have specialists in search who can pretty often just help you write out this um the search terms list and also it helps um in your write up if you say that you um consulted um a search technician. And so basically, you're gonna get a combination of, of these um search terms to formulate um your list of relevant articles. Now you're gonna want to filter out your articles by relevant publication date language, for example. Um If you are um doing academic work often with other centers, then you might have to include like Dutch Spanish or French. Um in your search in your searches, um geographical location. If you really want to do UK based or America or Africa based papers, then you can filter by that. And um it's helpful to, to focus on more current um literature sources. Again, you will have to defend the decision of when you decided to have the cut off if you decide to have a cut off for me. CT colonography was first described in 1994. So I did all of my searches from 1994 onwards because before 1994 nobody was doing CTC. So that was my justification. The other tip I have is to save your searches. Sometimes you do a search and you're like, oh my gosh, these are really great papers and then you close the tab or you go away and come back and you haven't saved it and then you run the search again and you didn't do it exactly the same. So you don't pull the same things. It can be really frustrating. So it's better to save your searches as you go and then you can refine and dump them if you decide. Actually, that's not what, what I want. So then, um, I cannot stress this enough that you need to organize your references from the beginning of your life. And, um, you, I use N notes, but that's because my supervisor used end note and, and shared his um library with me. But you can, um which you have to pay for. Um, end note isn't free that I know of. Um, but you can use Mende and Zz Zorro I think is free. Zo is free emend, I think is free. But then you maybe have to pay to kind of get like the super version. And the idea is that um as you were doing your searches and finding lots of interesting papers, you put them into your reference manager and that acts as your database. You do not want to be doing searches and then putting everything in Excel and then trying to manage that. The other great thing about the reference managers is that you can integrate them with words so that when you're writing, you don't have to do all of your referencing by hand. And this is like the one thing that will save hours, days, weeks, months of your life if you do it from the beginning and you do it properly. So, um this is just a screenshot from my end note library in the yellow box. You can see that I've categorized my papers into subjects. So literally any time I find a paper that is interesting, even if I haven't read it, which is basically most of them, um I will download it and put the reference into my end note. So at least, um I, I've got it, I can come back and read it at another time. So I've categorized these into colorectal cancer papers, um colonoscopy papers, papers on CTC and then within CTC, I've got loads of different topics. Um Sometimes I'll create a specific folder for a paper. So for like the colorectal cancer screening paper, there's 204 references er papers in there because I've, those are the papers that I've read for that particular um article that I was writing. Um the good thing about end note and um the other reference managers that I mentioned is that you can get a plug in um to use with word. So the arrow is just basically showing you hear the tab that you get on word when you um install the plug in. So you can insert the citation as you're writing and then automatically end note will be populating a reference list at the end of your document and you can change the formatting according to the formatting of the journal. So if it's Vancouver, um or if it's um I lance it have a very specific one. So then you can create one so that when it spits out your reference list is already formatted according to the journal preferences. And so the pop up box which you can see here is just when I want to insert a reference, I would click insert citation, then I'm gonna type Fiddler. It was the, the author that I wanted to cite find it in the list and then it will um click on it and then it will embed it in your manuscript and spit it out at the end in your reference list. The fact that this can be done automatically when I have done this previously in my earlier life, long hand is incredible. Do not do this longhand. I've spoken to so many junior trainees who got given a something to write by their bosses and were never advised on how to do this prospectively. And then of course, when you send a draft to your boss, he wants to change everything and then you have to redo all the references from scratch. And honestly, life is too short for that. So take notes, you've got your articles, you've got your abstracts, you're skimming through, you're trying to extract all of the juice all of the relevant information. You want to er maybe use mind maps to link different ideas between different topics. You want to outline categories, things to really pay attention to is how the study was designed? Is it a small sample? Is it a massive sample? Is it multicenter, what was their methods? Is there any bias there that you can see? How have they compared um their two different parameters? Um Is it a fair comparison et cetera? So um this is um as you can see, this is one of my notes from 2016 when I was first starting my thesis at my research. And so these were the things that I wanted to capture for each of the articles that I was reading. So the categorization of the paper, the journal where it was what the moral of that paper um was what the most important findings were from that study. If there's any conflicts of interest, it's always interesting to see that, oh, we're talking about this new technology and actually the author is like a consultant for that company and then obviously that introduces bias into their potential um whatever it is that they're writing. And so I also wanted to know if there were any applications for my research project which was called Perfect. So this is just um two screenshots from articles that I read. And um I used to really prefer to write, sorry to read um a a printed out manuscript. Once you start doing any volume of research, you realize that that is not feasible to have printouts of everything and it's not good for the trees. So you have to get used to reading PDF S on screen. So again, I would highlight stuff in different colors. Speech bubbles are comments that I've dropped and added and been like, oh, this is a good idea or somebody else mentioned this, go back and check that da da, da da. I think it's really helpful for you to engage um with the study so that you're not just um believing verbatim what's been written, you need to really critically analyze it. Um So this is a, a helpful thing from um open academics. This is um just two examples of how to write an annotated bibliography. So this is a similar thing in how to capture your understanding or how to make notes about what you're writing. And they've just given two examples on, on how you can do that. So step five, you want to identify your key concepts, um create a reading grid, take notes as I've said. Um try and condense what you've read into the two major concepts. What were the two major things that this paper? Two or three major things that this paper has established and how is that going to inform um my research question and my research topic. Um So this is a um reading grid, I guess if you want to call it for um one of my reviews. So along the top in the yellow box, you can see the different categories of information that I was trying to extract from each paper. And then for each paper, I'm going to go through and try and populate that information. So that when I'm writing, I can make connections between the different papers um more easily. Now this is much more detailed for a systematic review. So this is just like a quarter of the systematic review um spreadsheet that I had, I had in the end, I think like 45 columns of data. And so we literally breaking down and you're dissecting down every little detail of the paper. And so basically, I did this extraction and then my colleague did this extraction and then we would combine them to make sure that we had agreed on um what the data that we was had extracted. And you can see straight away like for example, the of patients involved in the study, everybody mentions a di has a different way of reporting the age. Some people put the minimum age, some people put the maximum age, some people put the mean, some people wouldn't put it at all. So all of those things are the things that you need to extract out when you're doing a systematic review. So you wanna find your golden thread, you wanna identify relationships in the literature, connect them to your ideas, to your own ideas, think about what's missing. So when you're reading be like, oh OK, it'd be really interesting to know what they thought about X and then you realize that actually they never mentioned X and nobody ever mentions X. Why does nobody ever mention that? Is that something that I can research on? Is that something that I can look into? Um So one of the things um from my screening review when I was doing it, I actually realized that not a lot of people talk about the cost effectiveness of CT colonography because it's such complicated economic modeling. But it was something then that I could highlight in my write up to say we know a little bit about the cost effectiveness, but we actually we don't know a lot about it and that is an area of potential avenue for future research. So um you wanna work out your structure, the length of your literature review is going to depend on how many articles you've used um as your sources, a number of concepts that you're exploring the type of review that it is. Um And also the word limit of your, of your journal gonna start with an introduction, some sort of background, some sort of context. If you're doing a scoping or systematic review, you're gonna have your methods. Again, the level of detail will be dependent on which of those you're doing. The main body of your literature review is gonna include the topics, the key topics or the themes. Um again, identify gaps, controversies, you want to critically analyze, highlight any emerging trends and expand on them. Then you've got your discussion or conclusion and then your references. So um this is um from my narrative view, it just shows the different sections um in the paper, this does not have methods, it doesn't have results because we're just talking about the landscape, but we're breaking it down into logical sections. So when I read this, um I will have a good overview of the main aspects of ct colonography, for example. So just to finish up, I just wanted to very briefly mention some resources that I have found super helpful that I wish somebody told me about seven years ago. So end note, I've already um mentioned um use a reference manager. That's the only thing you take away from today's talk. Please, please please um use a reference manager, I think um end note is pretty much, I mean, I use mend, but I've used end note the most and um it's not the best looking, but it does what it's supposed to, you can rate um your um articles, give them a star rating for whatever you want. Um And not, you can also embed the PDF within the same window so it can open straight away, which is um really helpful. It's really nice to kind of curate your own database of stuff that you're interested in. Um remove all the duplicates, which is really important when you start writing up because if you search for a citation and you've got it in there twice invariably you will use one as a reference at one part of your document and then the second use the other one in a different part of your document and then you've cited the same thing twice in two different places and it will have different numbers in your reference list, which I learnt the hard way. Um ok. Word. Um please use it properly, really, really powerful if you know how to use it. If you're writing any sort of manuscript, um use a navigation tab. This is like the first thing I do when anybody sends me anything to review or to edit, change all of the title so that we can see the subheading so that now we can jump easily within the document. Don't just be writing anyhow and then put bold here underline there something something, something use the design tab and give yourself heading one heading, two heading, three subsections, especially if you're doing a longer piece of writing. Um Use track changes. Um If you ever send me anything to look at everything will be tracked. Um It's really helpful to see um how things are edited. It's really helpful to improve your writing. Um Also make good use of the comments features even just for yourself. So in my thesis throughout the whole thing, I've got comments on, I need to do this today, finish that. Remember to reread this, that paragraph needs rewriting link that to somewhere else. It's a really helpful way to just set yourself reminders or to communicate with your supervisor and open academics. I've already mentioned it, but this is a really good website um to, they've got lots of cool infographics and resources, really nice bite size stuff to kind of give you an overview of an, or an understanding of a topic. You can also follow them on Twitter. They've got millions of not millions, they've got loads of followers and posts um helpful academic tips, research rabbit I found really at the end. So I never really used it, which was disappointing. But this essentially um is a program that allows you to find all of the interconnected studies on a topic. So you can put a topic in um or an author in or a paper in and then it will basically show you how all the different references within that paper or that have been cited by that paper or that site, that paper are interconnected. So it's really good. It's kind of like falling down a rabbit hole. Hence the name of finding all of the interconnected studies about a particular topic, especially if you don't know where to start. And this can be really helpful. And um squint is another really core um program which is um like a mind mapping note taking program allows you to interconnect different cards together. Um It's searchable tag board, it's really good for planning literature reviews, papers, thesis. Um And I think I can um if you're interested, I can do an induction thing for you so you can get some money off. These are not affiliate links, but if you are interested, I can for sprint will, at least I think they've given me a thing I can give to you. And lastly one note um again, I use this more at the beginning of my research to take notes. A great place to mind dump. You can embed um links papers, documents. It's got a simple way to kind of categorize stuff. This is good also for like taking notes and meetings and things and because it's Microsoft, it's part of the whole suite. So you should have that um built in if you have office 365. OK. I intended to finish a bit earlier. I'm sorry. Um But hopefully we've got time for a few questions. Um Thank you very much. Thank you so much, doctor. Thank you so much. Um It's been a wonderful session within almost an hour. So um myself, I've learned a lot of things. I'm just gonna mention a few of them. Um First of all, we need to work with the target in mind having like target journal where we, where we hope to publish and that would help us shape our, our research or our publication in a way that would, that would be acceptable. Then we have to have the protocols, then do the data extractions, data comparisons and prosper is something I've not heard before. So that's one major takeaway for me today. And, um, when did you finally select the study, um, topic? You then go and, um, talk about how we should select topic of interest or of relevance and then do like literature search and refine our searches. Then the little, um, things that could help, like do A F and, um, a little asterisk there to kind of broaden our search. Then use of reference manager is something I don't think most people know so much about. Please. I beg you, that's the only thing you need to do after today. Don't write anything without a reference manager. Well, I mean, it's, it depends how much time you have and what your level of patience is, but I think it's unnecessary pain and because there are free ones now, um you can just, just pick one that's free and it will be very helpful. Yeah. And, and um um making notes, then identifying the key concepts, linking everything together, finding a code and thread, then having a good structure like introduction methodology. Then in the main body, then have a discussion and arrange your references and you did give us like loads of resources and notes, which is like your preferred. Um It was the one I used but might not be the best one. Yeah. Then um tracking changes, whatever you make on words, then open academics rich research, Rabbit script. This is my first time of hearing about that as well. And one note, I don't think I use that so much, but maybe going forward, it's something I would try to, to make more use of. It doesn't look like our participants have got like, um any particular interest or questions. There's a couple of questions. Let me see. So sorry, I can't. It's OK. Um, in a systematic review where you have more than one author, how do you decide who's the lead? And, and the so this is important and something that I didn't know this is a good question. So normally the person who does the most work is the first author in any, not just as any publication, then the final author is typically the supervisor of the work. So first author and final author are the the two prime spots. Now, second author is like the next best. But now nowadays, journals are allowing um you to have a joint first author. So lots of my papers um where my myself and my boss did a lot of work together or it was his idea that I executed, we are joint first authors. And so you'll see an asterix by both of our names and it will say these authors contributed equally, right? And then that allows the final author position to be given to one of the other supervisors. Yeah, anybody in the middle apart from 12 and final did whatever, right? So maybe the third person did the next amount of work. And then fourth, if you're like seventh in the middle, we know that you didn't do any heavy lifting generally, or you came with expertise, maybe you reviewed the manuscript when it was done, et cetera, but generally the 1st 23, maybe four, depending on, you know, how many people and the final person, the final person is generally the person who's responsible for the integrity of the work that's been done and that doesn't, that applies to all primary research publications. Yeah. Um So it's important in my opinion for you guys as you're starting out in academic work to understand where you will be in the place in, in the, the pecking order. If you like, you're not gonna be the final author. But if you're doing all of the work, all of the heavy lifting, you've written the first draft, blah, blah, blah, then you should be the first author or joint first author. Yeah. And it's easy to kind of get railroaded and to not, not get what you would do if you've done all of the work. If you didn't do a lot of the work, then you know, that's your business. Um So that, that's an important point which I kind of just had to work out by the by. Um So um Kane Da asked if you're not connected to an academic institution, is that barriers to getting access to? Yeah. So yes, if you don't have a, a, um, like a university login. It can be difficult to get access to um papers. Generally not even current or, or just in general so that you can use open Athens. I think you have an NHS password. You can get an open Athens account which will give you access to some stuff. Maybe not some of the more specialist um journals. Then also, um you can, if you have an education center, you can ask the library that I would really, I need these 10 papers for my audit or for my whatever. Can you please help me access them? Because I don't have free access. There used to be some free websites that were, I don't know if they were completely legal, but I don't know anything about that. So I can't comment on that. I don't know if and not is free but somebody, um, thank you. You've put the link, I don't think it's free. Um, but again, if you're affiliated with a university, then you should be able to get it as part of their software bundle if you are. But otherwise I wouldn't get hung up on that. I would just use any reference manager, Mende or S Toro. Um, I find the link for that. I'll put it in the chat. Um, and then tell me top er, maximum number of, so this is an interesting question about, is there a minimum, maximum number of articles that is ideal. No, there's no ideal number. However, if you do your searches and then you only have one article left at the end that is not a systematic review of the literature. Cos it's one paper. Yeah. So there is a kind of a sweet spot beyond which, um, if you're too low then it kind of becomes unreliable. But it really depends on the topic. And then if there isn't sufficient data on that topic, should you even be doing a systematic review? Maybe you could just do a scoping review and then you're highlighting the gaps or you just do a narrative review. So it's not every top, every subject that is systematic review worthy if that makes sense. Similarly, if you do a systematic review, do your searches and you're coming out with, even with the most strict search terms and criteria, there's like 100 and 50 papers left at the end. If you're doing it properly and you've got 44 columns of data that you have to extract per or thereabouts per, you're not gonna do that 100 and 50 times. So that means either the research question is too broad or you've gone wrong somewhere in your search terms. Do you know what I mean? So, I mean, I was 19 for one paper, 12 for 19 was a lot because it's a lot of work to do to extract all of that and it starts to care after a certain number 12 seem to be good for, for our systematic review. But yeah, it really depends on the um on the topic. Ok, thank you. And lastly I wanted to just put um um, I don't know if you guys know about radiant. So, um, radiant is a radiology, academic network for trainees. So depending on, um and they have centers, um, centers, not centers but, you know, reps across the country. So it's a good network to be part of. If you're interested in getting into academic work and often if they are like audits that need to be multicenter, then they'll do it through radiant and they're affiliated with lots of um of the other organizations. So they'll come to radiant to look for people that might be interested in doing some research at their institution um or their hospital. So it's just um good to be on their radar, especially if you're interested in getting into radiology, they might be able to help you with um getting access to some projects. Thank you very much um doctor for giving us your time and your wealth of knowledge to, to bless everyone who has been here. And um thanks for those who contributed to the questions because I think we got to learn more from, from the questions that were asked as well. So thank you everyone and see you guys in the next um in the next series because we, we'll have another one. I'm not sure Kenny. Um Do you have any, any um do you have the dates for the next event? So you're on mute? Yes. So I said the dates are not yet confirmed um but the next series will be, will be I'm inviting a professor of radiology to um who is also um the editor of a, a pediatric radiology journal uh should give us um some more tips on what um papers will be expecting. That's the journals themselves will be expecting. Um and this kind of quality and what we can do to ensure that um our research literature reviews are otherwise are um quality enough to be accepted. All right. Thank you. Thank you very much everyone for um attending. Thank you, Doctor Ana. Yeah, no problem. Thank you very much. Bye everyone. Thank you. Yeah.