Computer generated transcript

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

Hey guys. Um I think we'll just make a start now. Um, but so just to introduce myself. So I'm aha, I'm one of the fifth years and so I'm gonna be doing your talk on ICA one. just for some context. I did um my PSE in farm last year. So if I just share my slides, yeah. Ok. Hopefully everyone can see that. Ok. Um, if at any point I'm going too quickly or you guys have any questions just, um, pop it in the chat. Um, ok. So, um to start, what are we gonna cover today? So the first thing we're gonna do is just go over IC one. So what's expected of you, um, what you have to do and then we're gonna go over how to read a paper because I feel like they don't really teach you that in med school. Um, especially if you haven't had um, prior experience with research. It can be quite daunting because you don't know quite exactly where to start. We're then gonna go, um, more specifically on to IC one. So how to structure your report. Um The bulk of this is gonna be about how to critique literature because that's the main thing you're gonna be assessed on and then also just other resources we can use and then some questions at the end. Um So just make sure you've kept an ear out um for the questions. OK. So to start, so you guys are right at the beginning right now. So we're gonna cover, I say one which is only worth 4.5 percent. So it's not actually worth too much. So it's about 12% of module one. And so in terms of what you're expected to do, so it's a 1000 word written report. And so either it's going to be a less to the editor or a commentary. So um I know for Gastro last year it was a commentary and then for Farm, it was like a less to the editor, but I think generally they do tend to be the letter of the editors. Um But they're more or less the same thing. It's just in terms of formatting that it might be slightly different depending on which one you've been given. Um It's an assessment of your ability to evaluate and critique your papers. So um that's kind of where you have to. So we will go over this anyways, but it's to do with like what are the strengths limitations comparing it to wider literature as well? And then you need to make sure that the way you structure your report and we will go over this anyways. I do have examples but you need to make sure it mirrors um like a journal in the lancet commentary style. Um But yeah, so either you're gonna be given a paper to critique or you're going to choose a paper yourself. So that just depends on what your BSE prefers. And so I'm just very briefly sorry to interrupt. We can't, we can't see, are you in presenting mode at the moment? Um We can't see. Oh, I think there we go. We can't see. Um we can see the slides moving along. I think it's OK. Now, um uh what side is it on now? It's on six, but before it's frozen, I think it might be ok now. Um but I'll let, I'll let you know if that anymore. Um OK, I know. All right, thanks. Yeah. OK. Sorry about that. Um Basically I was just talking about what um the IC is about. So it's to do with, so you have to write 1000 word uh report. It's an assessment of um your abilities to critique the papers and it has to be in the style of the lancet commentary. So that's all the information you've missed. Um If you couldn't see this slide before um in terms of choosing a paper. So depending on your BSE, you will either be given a paper or you'll have to choose one. Most people will be given one, but I'll just very briefly, um, go over how to pick a paper if that's what you're supposed to do. So you should have like some sort of Q and A with your BSC lead anyways at some point. So, um, that would be a good time to ask them more specifically about this, but they'll give you some guidance about where they want you to take the paper from. So it might be that, um, they need it from the last five years or the last 10 years. And then there might be some more guidelines about which types of journals you can select your paper from. But if not just feel free to pick any paper, but just make sure it's something that firstly, you find interesting because it can be quite hard to motivate yourself through reading a paper anyways. So picking something interesting, just helps kind of speed things along a little bit. Um something that's easy to understand and then something that's easy to critique. So there's two ways to go about it. So you could pick something that's a little bit more broad because then that gives you more things to talk about because if it's um something that's quite well understood, there's probably more literature about it. So you'll be able to draw up more papers to compare it to. Um or you could pick something that's a little bit more niche because um then you just have a clearer focus on what exactly it is that um you need to be talking about? OK, so how to read a paper? So are my slides? Ok. Now, um can everyone see this? Ok. No, II don't know if it moved. Oh, there you go. No, no, it's through now but we can, we can see the notes as well. II it's fine that we can see the notes but just um um ok, so you can't just see the slide. Yeah, so you can't just see the slide. It has the, it has the notes as well. Um But I mean it, it's, it's fine if you want to continue as is um is the slides moving along when I press it? So is it, has it moved to a different slide now? No. Um ok. Um It might be delayed but it's on, so now it's on eight slide eight. Oh no. And now it's back on seven 94 a cane. Um Fine. Uh Just kind of uh if you just give me a second then. No, that's fine. Uh uh Sorry guys. OK. Uh I think I'm just gonna have to do it like this but you guys can see the slides now. Yeah, we can see the, the slides as you move along, I think. Now um it's on how to read a paper now. Seven, not eight. All right. OK. Fine, cool. Um So, OK, the s probably won't work then, but it's fine. So essentially you um I was just talking about the main points of how to pick a paper. So um just follow the guidance given by b make sure it's something that you understand and something that you have a lot of literature to compare it to as well. Um in terms of reading the paper, so you, you can read it top to bottom. It's just about preference and what you find easiest. But this is the structure that I think I think works be best for me. Um So the title to start with. So that's going to give you a good overview of what the aim of the paper is and also talk to you about maybe some like the main findings. So it might just be phrased differently. So it might have been phrased like um oh so this drug reduces the risk of XYZ in this patient demographic. Then you go to the conclusion. So the thing with the conclusions is um it will essentially just tell you what the main findings are, which is kind of the main thing you need to know. Um Anyways. So it will kind of just help you gauge what the relevance of the paper is um to whatever you're trying to compare it to as well. Um And then you wanna go over the abstract. So the abstract is a very a nice kind of summary or um basically the main findings of the paper. So in about two lines um per section. So it will have like two lines on the introduction, two lines on the, um, the methods, two lines on the results, two lines on the conclusions. Um So that's gonna give you like a good foundation of understanding the paper in the first place then, um, you'll have your introduction. So you don't have to initially read through the entire introduction. So the first and the last, um, paragraphs are probably the most important. So the first paragraph is what's going to tell you, um kind of like what the context behind the topic is. So it's gonna tell you a little bit more about the disease. It will tell you a little bit more about the drug itself. And then the last paragraph will kind of tell you where the gaps in the literature currently are. And it will tell you um what the paper is going to do to address this and the p pretty much the purpose of the paper and their rationale, then you've got the results. So with the results, you, um it can be quite daunting to try and look at because a lot of the time it would just be lots of different numbers and like, um it will tell you significant values and it can be quite difficult to understand exactly what it means and what it's trying to tell you. Um So what's good about this is it will essentially just kind of summarize the main binding. So it will address all of the outcomes of primary and the secondary. Um And then you can kind of got get like um a good overview of everything and then you've got your discussion. So your discussion is quite important in terms of where you're going to draw your points from. So a lot of the times your paper will self evaluate anyway, so it will tell you what the main strengths are, what the main limitations are. Um And it will also kind of compare it to the current literature as well. But at the time, so bear that in mind. So if this paper is like 10 years old, it's not going to have anything from the last 10 years um within the paper. And also papers take quite a while to get published anyways. Um So it might be that it hasn't referenced literature from the last 20 years rather than just the last 10 years. But that will give you a good foundation of where to pick your papers from as well. But just be careful with the reference is because often papers will reference something that either doesn't exist or they've referenced it incorrectly. And so you can't entirely rely on what they've used and then you can read through it again, top to bottom and then go over everything else as well. Ok. So does anyone have any questions so far on how to read the paper? Oh, a week? Oh, ok. No, ok. Fine. So one other thing that I want to say is what I find quite useful is just printing out your paper and highlighting the different um points you're going to make. So you have a physical copy just because it's easier to digest the information. Um And then next to each paragraph, in pretty much every single section, what you want to do is you want to write like half a phrase or a sentence next to it to basically just summarize what that paragraph is telling you. Because when you're writing your report, it can be quite easy to lose track of exactly where you are. And so even just copying and pasting sections of your paper into chat GPT so that you can kind of just understand what that specific paragraph is telling. You might just help you a little bit. OK? Cool. So more specifically about your ICA, so this I've taken directly from MEDLINE and so it's what imperial deemed to be features of a good report. So I've highlighted the main points. So in depth critical analysis. So the way they kind of differentiate like the people that get first and the people that get to one is by um how you critique. So it's not enough to just talk about the um strength and limitations of your paper alone, you need to compare it to current literature. So the way you do that is try and find similar papers that are investigating the same thing in a different way or they've um, investigated, um, something different but still ha that has clinical relevance to what your paper is talking about. Um So we will go over that anyways in, in more detail, how to do that thoroughly. But, um, that's essentially where you need to try. And, um, that's basically how you get your high marks. You also want to look at different types of information. So you probably remember from like first year, second year LM map, you've got that pyramid of like um the strength of like the type of re research that you're looking at. So you've got like observational studies towards the bottom and then uh systematic reviews and meta analyses towards the top then says use high quality uh primary research articles. So that's like case studies and things discuss your key strengths, limitations and controversies in the field. So that's kind of what I was talking earlier, uh talking about earlier with the critical analysis but um where I want to kind of draw your attention to is the controversies in the field. So try and um see if the findings from your paper kind of contradict anything that's standard practice. Um So look at nice guidelines. Is there anything in the guidelines that it might challenge? Um what might that mean for standard of care, things like that? So you're really kind of um tying together your clinical relevance here then um talking about so the next few bullet points are to uh to do with how you're writing. So your writing style, so communicating effectively and ccin um being mindful of the level of your audience and making sure your thoughts are well organized and um they're easy to follow. So all of these things you can kind of address just by um swapping your drops with people. Um not just in your BSE S but between BSE S and I think that's probably better as well because for certain BSE S, everyone will be getting the same paper. So they're not gonna accurately gauge whether or not you've communicated in a way that makes sense and is easy to follow if they know the paper just as well as you do. So um try and find your friends that um do different BSS or like even friends that do like Biomed, try and get you um get them to read your paper and your um commentaries to see whether they have any comments that um to make about your paper and things you can tweak basically um then moving on. So making sure that you're um engaging convincing professional references. So the reason I've kind of like put a box around this is because some BSE is not, all of them will prefer certain referencing styles. Um Most of them won't mind but just be mindful that some might have a preference and then also referencing your own paper. So a lot of people will just reference their paper at the start and then not again, later on, um any time you refer to anything in like the methods or um the results, you need to make sure you're rerefer and then the final thing and probably one of the more important things. So making sure that you're expressing your criticisms and disagreements professionally. So, um the underlying tone of your commentary should be positive, but a lot of what you're saying is going to be negative things about your paper just because that's what you're kind of expected to do because of the nature of your IC. So you're going to be picking out things that they haven't done well and things that they could have done better and saying that another papers done it better. So you just need to be mindful that the way you're addressing it is in a polite and respectful manner and we will go over some examples on like how to do this more specifically anyways. Um But yeah, that's to do with the general structure. And then, so the reason I've kind of brought this up is because so this is a mark scheme and when I saw this, I don't think it was that great at guiding how I should be writing. Um It's quite difficult to as a psa student where you haven't actually looked at anyone's work before to gauge what the difference between excellent and outstanding is. Um So I think this is good in terms of the brief bullet points of what and like a skeleton of what you need to make sure your report looks like. But I think that's as far as it goes. So I wouldn't rely on this too heavily. Um, the reason I didn't want to bring this up is because I don't know if anyone's spoken to you about how they mark your ICA S. So we don't get told until a couple of months in. Um, but basically you can only be given a mark that's within the numbers in the brackets here. So you can't get given like 89 or 86. And so you have two assessors and they'll both give you one of these uh numbers and then they'll take an average of that. And um the other thing was that I feel like as Imperial med students, like it's so easy to wanna aim to get as close to 100% as possible. Um Don't be disheartened if you end up in like the good or very good um bracket instead because firstly, this is like your first IC, you haven't had much experience, it's only worth 4.5 percent. And also, generally speaking, it is very difficult. Um I don't think I know very many people at all that got outstanding in this ICA. Um But yeah, so in terms of the general structure of your ICA, so you can kind of divide it into three and we'll go over each of them anyways, um more specifically. But so firstly, you have your introduction so that should follow the same conventions as the Lancet Journal. So if you're unsure about um the general structure and how to introduce your work, then just have a look at that and just reference some other articles that are already in there. Then what you wanna do is you wanna identify the article and you wanna name, drop the author as well, congratulate them. Um Talk about both your rationale for writing and commentary as well as the authors, rationale for writing a paper. So you give a brief summary of what they did, what they found et cetera efficiently super long and then your overall impression. Um Do you think it was good? What were the main limitations? Your critiques should then um be the bulk of it? So you mention your point, then you have like a counter argument. So your counter argument might um stem directly from just general points you can make about um the method something that was weak, maybe like sample size. Um But the way you kind of develop that further is by comparing it to other literature. So try and find, like I said, another paper that's done the exact same thing, but maybe they have different outcomes and um explore why they have different outcomes. Is it something in their methods? Was there a technique that they use that, that literature deems to be the more superior technique. And then you want to also talk about improvement. So how could they have improved? Um Again, you might find a paper that um has done that improvement and it's shown to be more effective and you can tell that by the nature of their results, then you've got your conclusion. So your conclusion is basically an echo of your introduction, but it's more focusing on what do you think main limitation was and what is the improvement? And then something that you add here that you shouldn't add in your introduction is the clinical relevance bit. So either it's going to be that your research is going to be a stepping stone for a more developed project or it's something that's already directly relevant. So it contraindicates something that's already mentioning the guidelines, for example. So this is an example of an introduction. So I'll give you guys a few seconds to have a read. Um But essentially there's like it's the four main parts that we spoke about. So it needs to follow the conventions of the lancet. You need to identify the article quite early on name, drop the author, identify the rationale. Um You want to state what your argument and your overall opinion is and then also summarize what the study is talking about what they found, et cetera. OK. Um So I'll give you guys about 20 seconds just to have a read through that um as well. OK. And then I've got another example here. And so this is taken from someone last year. And so essentially, we're applying the same things we just learned about what to improve in the introduction over here. So here we've got the um the conventions of the Lancet. So this is a little bit more difficult because um because of like the guidance in the handbook, there are certain ways that you need to present. Um So like having the title and the author in the text box, you'll see an idea. We also go in here, but I omitted that for confidentiality. Um Then you can see here that they've identified the article and name dropped the author quite early on in the first sentence. Um You wanna make sure they've also explored the rationale of the author. Why did they write the paper? So here you can see it's a little bit more lengthy than it should be. The other thing is they didn't give that overall impression and talk about the main limitations, but they did give a very brief summary about what the paper is talking about at the end. So just some other things. So we spoke about this already, but uh very briefly about having, making sure that the underlying tone is quite positive. So using words, like we commend the author, we congratulate the author, they've raised interesting points or thought provoking ideas about XYZ. Um and then summarizing the aim and hypothesis. So like I said this probably isn't as distinct as it could be. Um And also over here they didn't need to do this, but they've named another paper in the introduction. You can do this, but I probably wouldn't recommend it. And then the final sentence. So, having subtle and um subtle comments about what you think about the uh, paper itself. So, was it well controlled? Um, was it anything that they could have done better? Um And like I said, just these key words, like come in, congratulate interesting ideas, um Things like that. OK. So does anyone have any questions about the introduction at all? Uh I think this is being recorded by the way, um Just for those asking? OK, cool. Um Next. So critiquing your paper. So essentially um you can divide this into three. So there's different ways that you can go about critiquing your paper and applying your critique. Um And this is like the simpler of the two. So splitting into three. So the first thing is um looking at your methods. So here you're kind of talking about is the aim clear, is it specific for what they're trying to explore? And then how appropriate are the methods that they've used? Is there a better method that they could have used? And again, like I said, compare it to other papers and what other papers have done. Um And then you could be a little bit more specific as well. So it might be a specific um device or tool that they've used to get, get their results. But other people might have said that another tool is better um things like that. Um So you can go very detailed in terms of your critique when you've got the results. So the validity of the results. So um how accurate are they um talk about the stats analysis as well. So there are like protocols anyways depending on like the data type that you have. Um But some and analysis types are a little bit more robust than others. So I know that there's some, I can't remember the name and it's off the top of my head, but some will like eliminate for confounding some will um kind of um standardize their database off the sample size, things like that. And then the main thing is just um what was the outcome? Is it different to what the current literature suggests? And why is it different? Why is it not different? Um things like that? And then you've got um implications. So you've got clinical relevance basically. And the what, what is the value of this paper? So is it contributing to another research project or is it directly already um challenging something that's already in place? OK. So that's kind of like the general, this is how you would do it. But there are other ways that are a little bit more systematic and more thorough. Um So there's this like very weird. Um And very long acronym that you can use. And so essentially, we'll just go through every single one of these um in a little bit more detail. Um So the first part is question and relevance. So what is the author trying to do here? What is their main aim? What is the relevance? Um Why are they doing it? What's their rationale, are they addressing any gaps in the literature or is there something else? A lot of the times what will happen is your author will have already done some research and they're doing even more research that kind of relates to that first project, but they publish just two different papers. So a good tool you can use is something called research rabbit. So I've listed this at the end anyways, but essentially if you copy and paste the name of your paper into research Rabbit, it'll come up with two things. So it will come up with other papers that the authors done that might be relevant. Um And most of the time they are. And the second thing is other papers that are as similar as you can get basically to your paper. Um and it creates it in like a nice little mind map just to uh to help you understand which ones are the best papers to kind of pick out. Then you've got um P so population intervention control and outcomes. So your population will be like your participants essentially. So who do they include, who did they exclude? And so this is quite an important point because what a lot of people won't realize is like there'll be two papers that have done the exact same thing. But because their inclusion exclusion criteria are slightly different in terms of demographic, it might be something to do with comorbidities, then that can skew their results. And that's another critique point that you can make. So remember that like certain countries might have a higher prevalence of the disease they're looking at or even rural areas in the UK, they might have less complex disease because people are being sent to secondary care facilities that are in bigger cities like London instead, then you've got your intervention. So talk about what was their primary intervention? Um What are, what have other papers done instead of this intervention? And it might be very, very subtle things that make a difference. So things like let's say you're giving someone medication, uh they might be giving it twice a day instead of once a day. Um The route of access might be different, things like that. Uh And then you look at your control, so your control variables, how were they controlled? Were they controlled effectively? Could they have done something better? Then you look at your outcomes? So what was the primary outcome? Was it relevant? Um for what they wanted to get out of their paper? And what is the, where the clinical relevance. Were there any secondary outcomes? And did they address all of the necessary outcomes for their paper as well? Ok. Um So then what I wanted to cover kind of quickly now is just bias. So, um I feel like a lot of these overlap and it can be quite difficult to distinguish between them. So I just thought I'd quickly go over them. You probably have heard of some of these before anyways. So you've got observer bias. So that's to do with, um So it's not actually to do with the participant, it's to do with the person interviewing the person. So it's a bias created by the interviewer essentially. Um And so I've kind of defined them over here at the bottom. Sorry. So it's when the opinions of the investigator have kind of influenced the outcomes, then you've got attrition bias. So that's to do with your drop out rate. Um So the way you kind of reduce that is by um this method called intention to treat. So a lot of people won't randomize and what they'll do is they'll assign people to groups like the placebo versus the treatment arms based off of the severity of the disease and who they think should be treated. So, um the intention to treat is kind of the opposite of that. So you randomize instead, then you've got confirmation via. So essentially you have pre, pre existing conception and like beliefs and you're going to basically pick and choose based off of what is going to confirm what you really believe you've got sampling bias. So it's when your sample doesn't represent, and it's not generalis to like the general demographic that you're trying to look up selection bias. So, um that's to do with who you pick and choose out of your participants and you kind of reduce that with randomization question ordered bias. So that's just to do with, it might be that you ask a certain question before that makes the um participant think that they need to answer in a certain way. It's quite similar to language bias. Um which is just when you use maybe like positive or negative reinforcement within your questions, that kind of skew what they say. Um So saying stuff like, oh you don't drink, do you or you don't smoke anymore? Do you um something like that? Then you've got response detection bias. So response bias is just um you just answer with what you think they want to hear or in a way that makes you look better and then detection bias. So you reduce that by blinding. So it's just when the way that you kind of collect your data is going to be different depending on what groups you're looking at. So it's a subconscious V OK. Then you've got um RNA. So part of Rambo. So you've got recruitment location, so you can do basically recruit in different ways. So consecutive versus non consecutive, you can look at different centers rather than just selecting participants from one. And the best combination is basically consecutive and multicentre. Um And so these are points that you can just make in your critique and then also looking at locations. So we kind of already discuss this anyways. But um more rural areas might be different compared to um more heavily populated centers like London and then also just between countries as well. And then you've got allocations. So you've got randomisation. So there's different types. Um So block and cluster randomisation. So block randomisation is when you have a center that um oh sorry, cluster randomization is when you have a center that will give an equal amount of each arm. So it might be that you um have people with ad disease and with without the CS and they'll give 20 of each and then another center will do four and four. So um block is just overall across all the centers, each arm is equal. So cluster tends to be the more superior one and then you've got blinding as well. So there's three different types, right? So I go over this a little bit more here anyways. But um so with blinding, you've got a single blinders. So when the participant doesn't know double when the physician doesn't know triple when um the outcome extractor also doesn't know. Um you also have something called blinded outcome adjudication. So um it's something that's a little bit more difficult um to explain. But essentially, it's just um a scenario where you can't necessarily blind. So um for something where the treatment is like surgical, you're not necessarily going to be able to um uh do double blinding or triple blinding. So, um with maintenance, so this is just talking about maintaining your participants. So just making sure there's adherence to the treatment that might be something you can talk about also attrition virus. So we kind of really discuss this anyway, but to do with um patients dropping out. So if they think that they're in the placebo group, they're more likely to drop out. OK. And then in terms of statistics, so there's different things you can talk about here. So power is the main thing. So people will talk about the power of the study. So a study with higher power will be one that has a high sample size. That's generally the inclusion you can make, you can look at stats models which I already spoke about. So there's some that and more suitable for different data types, look at effect size. So what is the strength of the relationship you can look at absolute and relative risk reduction and also how appropriate are the conclusions that they've drawn from the results? One thing that comes up quite a lot is statistical significance. So um you've probably already been spoken to about this, but statistical significance doesn't necessarily mean it, it's not the exact same thing as clinical significance. Something can be statistically insignificant but still have clinical relevance. And so that's an important point you make in your critique and that's usually something that you would say in your conclusion as well. Um OK, then you've got R MP. So resource availability and population representativeness. So that's the same as like generalisability. So resource availability. So are they using any specialist equipment within their methods that you're not necessarily going to get in a typical hospital? And so that will affect um the relevance of what you're talking about? And then also just in general was what they were saying cost effective. The final part is probably the easiest part. So funding ethics and conflicts. So um this is easier because your paper will have a dedicated section towards this anyways where they talk about who they were funded by any ethical implications and conflicts of interest. So it's quite easy to draw those out. But um there are things that you would probably mention last because they're probably the weaker points you can make. Um So I'll just wait like 10 seconds just in case anyone has any questions about any of these. So that kind of just goes over the general critique. OK. Sign. So OK, the next point is um just picking your point. So it can be quite overwhelming when you've got so many different things. You can say it's quite a long acronym. Um So the way you kind of decide what to include and what not to include is two things. So firstly, what's your strongest point? So your strongest point is going to be the one that's going to not necessarily devalue the paper, but just um it, it's going to be the biggest limitation of the paper. Um So it's gonna affect how robust the methods and the results and things are. And then also what point do you have the most number of papers to compare it to? So what I did is I pick like three or four points and put it on the table and then under each heading, I would basically link papers that I can refer to that kind of relate back to my point. Um So I've already discussed this but um there are like different A I sites you can use like research rabbit that will help you do that. And um so you don't have to spend hours and hours looking for papers and there's also one called mese. So essentially it's like a search engine but um you ask it something medical and then it will reply with an answer, but the answer will be based off of what the current literature says. So it will link um the papers that it's taken those conclusions from as well. So it's very quick and easy. Um We've really kind of discuss this anyways. Um as part of the slide about what MEDLINE says, uh you should do but just make sure that you're being polite with the way you kind of phrase your criticisms because they can very heavily mark you down for that. Um And then remember that for your future I CS as well, like when you're peer assessing and things, um because that is something that often comes up quite a lot. OK, fine, then you've got your conclusion. So like I said, your conclusion is basically an echo of your introduction, but you're talking a little bit more about your clinical relevance here. Um So you want to um in positive reinforcement, talk about your overall impression as well. What was the main limitation? How can you address that main limitation? And then also what is the clinical significance here? Um Just looking at the chat about questions. So first question, do we have to criticize the other papers we were referencing? So you kind of use the other papers to criticize your main study. So, um let's say you have uh a drug that you're investigating, it's a new drug about um treating diabetes. And then there's another paper that's used the exact same drug, but they've given it twice a day rather than once a day. And there's also research on that drug about um comparing the effectiveness of the drug if you give it once rather than twice and they think that giving it twice is better. So what you'd say is you would say this papers um obviously in a better way than this. But um you would say this papers um given the drug once, whereas the other papers given it twice and other um literature and then you'd link that literature uh be like um you'd say something like um that literature deems um a twice daily dose to be more superior than a once daily dose. And therefore, this might have impacted the outcome. So you're not real like your critique is more this paper did this instead. Um That's better because so you're not primarily just critiquing the other papers. It, it does work in um the opposite way though. So like it might be that your papers done something that's better than another paper. So you can say it's the exact same thing and then in terms of websites and resources and then there was a question about that. Um I've linked a slide at the end uh with all the list of like a list of all the resources that I'm mentioning anyways. So I'll go through that at the end. Um But yeah, any questions about the conclusion or are we happy to move on? OK. So yeah, so this is the list of the resources. So I've kind of split this up, but in terms of finding other papers, there's three that I recommend. So I use Research Rabbit and me with Research Rabbit. I think you need like um a subscription for using more than I think 10 papers. So you can do 10 searches. Um, medi research is a very good one though. Um, because it will, yeah. And you, the thing is you can ask it, follow up questions as well. So it's like chat GPT but it will link in, um, the literature that it's using to find its answers for consensus. I haven't used, but I've heard that it's quite good and it's quite similar to many research as well. And then you've got the other four. So ask your PDF, perplexity, chat GT and Cord. So ask your PDF it, basically, you download a paper as a PDF and resubmit it into this um search engine and it will basically just summarize what it's tell what, what the PDF is basically saying. So it will do it by section. Then you've got complexity. So, perplexity is the same as Chat GPT, but it's a lot more accurate and a lot more reliable. Um But that said, I know a lot of people that would use Chat GPT. The one thing I don't recommend is just drawing your ideas from here because it's not gonna be enough. Um And also you would just get caught for plagiarism. A lot of people would also copy and paste sections of their writing into Chat GPT just to check like grammar and things also wouldn't recommend because that will come up as page and then um even with clause. So Claude is quite good in terms of summarizing things. I think it does it in a better way than chat GPT but obviously personal preference. Um, don't ask it for papers because I know people that would do that. And um, it, sometimes it will just make up papers rather than link actual ones that exist. And then obviously you get locked down for that and then casp here. So that's basically this tool that exists and it's like different checklists for different things. So um it might be one more specific to systematic reviews or there might be one more specific to um meta analyses. And so essentially, it's just a little bit more guidance on how to structure or critique and different points that you can make. Um But yeah, does anyone have any questions about anything that I said so far or are we all OK to move on? No. OK, fine. Um If you have any questions, just put it in the chat and then I'll answer them as soon as I see them. Um So just some quick questions. Uh so just feel free to put it in the chat, but otherwise I'll wait about 10 seconds, just have a think about what you think the answer might be. Um So the first one is what does po stand for as part of the structure of your analysis? So I'm giving you four different options. So just wait a few seconds and that the answer is d so population intervention, control and outcome. So population in terms of your inclusion and exclusion criteria intervention. So what was the primary intervention? Um control, what were your control variables? How did they control them and outcome? So, what was the primary outcome? What was the secondary outcomes? Were there any outcomes that they missed out? Were they were relevant and necessary? Then you've got OK, this is difficult because my elevations aren't working but um something that influences the power of your study. So the answer was sample size. Um This is a point that comes up quite a lot and it's quite an easy thing that you can comment on. Um But it is quite generic as well. So there's not really that much you can say about it, but it is still a valid point you can bring on. So a higher sample size would result in a higher power. Um This one which of the following is an example of observer bias. So this one is a little trickier. Um So a researcher unconsciously records data that aligns with their expectations, a participant changes their behavior because they know they're being observed, a researcher accurately records all their data regardless of their personal expectations. And a participant gives answers that they think are socially acceptable rather than truthful. And so the answer is a, so remember that observer bias is to do with the um person interviewing rather than participant themselves. OK. And then the next question again, you can see the answer anyways, but the way you reduce attrition bias is to do with intention to treat. So just making sure you're randomizing rather than allocating people based off of whether you think they should be in the treatment group or the control and person on. OK, then you've got detection bias. So the way you kind of reduce that is just by blinding, um so that the physician doesn't have any subconscious bias towards the participant. And so it doesn't collect the information in a different way. So which out of blocking cluster randomization makes a more robust study design. So you've got um cluster. So remember that cluster is when each center will give an equal proportion of participants per per arm. So treatment and placebo rather than block where the overall number of participants will be equal in either arm. And the reason for that is because let's say you have um a more advanced center and they've given 20 of one and one of the other, then you can't really like the the amount of analysis you can do is very limited. Um because standard of care is different between centers as well. And then which out of the following is an example of blinded outcome adjudication. So the first one, a study where the participants and researchers both know the treatment group assignments but the outcome assessor do not. The second one is a trial in which the participants researchers and outcome assessors are all aware of treatment assignments. The se the third one, a study where the outcome assessors are unaware of which treatment each participant receives to prevent bias from judging the outcomes and da trial in which only the researchers are aware of the treatment assignments, but the participants and outcome assessors are not. And the answer is to. So A and D are both examples of double blinding. So D is very uncommon. Uh Most of the time it will just be the um participants and their physicians rather than the participants in the outcome assessors. And then D is just an example of um while there's no blinding, everyone knows um who's in the treatment and who's in the placebo group. OK. And then the final one. So which of the following A I sites can be used to find other papers and the answer was just maybe such. So code is used to kind of summarize things same with per and then C is that checklist guidelines? Um OK, fine. So that's the end of the presentation. Does anyone have any questions before you finish? Ok, fine. Um So the QR code is just for some feedback, sorry guys that the slides weren't working properly. Um But all of the slides and the information can be sent to you anyways afterwards. And if you do have any other questions, just feel free to email me. My email is on the uh slides as well. Um Yeah, great. Thank you so much. Aha for that. Um That was really useful. I know for myself. Um Hopefully for everyone else as well. Um Like I said, so, um the feedback, this is useful for um aha as well as us um for her birthday again and stuff. So if you just have a, a couple of minutes, it's very, very quick just to fill out the feedback. That'd be fantastic. Um Otherwise, um like you said, send her an email if you have any questions, um or pop them in the chat just now before we wrap things up. Um But otherwise I guess everyone thanks for your attention um and have a nice day, rest of your evening. Ok. Thanks guys. Thanks hea.