Introduction to Systematic Reviews & Meta-Analysis | Practical Hands-on Session | Session 1

追加: 2026-05-18

[00:02:05]Please let me know if I'm audible. Uh yes, you're audible. Dr. Nimil, please speak a bit uh in loud volume so that uh everyone can uh hear it clearly. Thank you. Okay. Okay. But I think this is not being live stream now. Is it Is that right? Because >> uh it is live stream. It is live stream.

[00:02:32]Okay. Okay. Okay. Okay. I think it's it's okay now.

[00:02:41]Okay. Okay. Okay.

[00:02:55]Rahman.

[00:02:57]So today we will be covering basics of systematic review and meta analysis and uh I hope you guys have heard of that.

[00:03:08]It's okay now I think.

[00:03:20]It's okay. It's okay. It's totally fine.

[00:03:26]Okay.

[00:03:32]Hello everyone. Uh a warm welcome to the first session of our much awaited metaphor metaphor all hands-on learning session. Today we have the honor of starting this journey with an exceptional speaker who brings a wealth of knowledge, experience and passion for research.

[00:03:50]Please join me in welcoming our distinguish distinguished speaker for session one. Dr. Muhammad Nabil Sadir, final year MBA student at the prestigious King Edward Medical University, Lahore, founder of Nibbras Research Academy, a highly accomplished researcher with 50 plus publications.

[00:04:12]His abstracts have been accepted at numerous prestigious international conferences included including ESMO, ESCO, SH, EC and many more.

[00:04:23]His work his work has earned acceptance in highly reputed journals including the Lancet and Jamba. He is skilled in in a wide range of research tools. Zotterero, Reman, Meta Excel and many more. Uh Nabil is not just a researcher. He is a mentor, innovator and passionate educator who has helped countless students navigate the world of systematic review and meta analysis.

[00:04:55]Today he will deliver a practical hands-on session on introduction to systematic reviews and meta analysis. He will cover essential topics including introduction to systematic review and meta analysis uh and much more through live practical demonstration. Dr. Nabil, the stake is all yours. We are all excited to learn from you. Inshallah.

[00:05:26]Thank you so much for so brief introduction and I'm truly humbled to be here among you guys. Uh assalamu alaykum and uh thank you everyone for joining us today. We'll be covering basics of introduction to systematic review and meta analysis. We'll be covering uh what are the different components steps in doing meta analysis. What are the resources that you need to learn meta analysis as an early career researcher who is just buing in the research world?

[00:06:03]So we'll be making this an interactive session. You guys can post your queries on the YouTube and I will be answering side by side as well and we'll be having dedicated time for the queries as well.

[00:06:19]So uh we'll be continuing in English but if you guys are comfortable with the mix of udu and English we'll be uh I'll be happy with that as well. So let's begin without any further delays.

[00:06:36]So today we'll be discussing systematic reviews and meta analysis.

[00:06:49]So how many of you guys have heard of systematic Q meta analysis or how many of you have or ever published or happen to work on any systematic Q and meta analysis just let me know in the chat uh I have opened the live chat so I'll be seeing your messages from the YouTube and I'll be answering here there might be a bit uh 10 to 20 second delays but I will be entertaining the delay as well.

[00:07:17]So, how many of you guys have ever published or worked or heard of systematic view and meta analysis?

[00:07:27]Just let me know in the chat. Uh be active and that will be not just passive listening. So, I want your active participation. Uh just give me thumbs up or hearts when when you like anything. If you have any questions, you can post your questions as well. So I'll be continuing Rahman and Rahin. So this session is all about introduction to systematic review and meta analysis.

[00:07:55]So majority of our participants have heard of meta analysis. Some of them have worked on meta analysis. So that's good start. I'm Muhammad Nabil as uh introduced by the worthy panel. So that's hierarchy of evidence. You can see the meta analysis sits at the top of evidence hierarchy. That's due to its features, its qualities. We have the RCTs, cohorts, case controls, case series and case report. So, so that's uh a mean from evolutionary point of view that meta analysis sits at the top of evidence hierarchy.

[00:08:39]a bit about me that is already so we'll be learning and understanding the basics the purpose and significance of systematic reviews differentiating between systematic reviews and meta analysis recognizing the key steps involved in conducting systematic reviews key tools soft and soft commonly used in the meta analysis and we'll be also discussing uh the search strategy the search string for the pop and the clinical trials and we'll try to cover as well. So what is a systematic review?

[00:09:26]So a systematic review is a structured and comprehensive synthesis of the research for a specific question that is the textbook definition and uh what are its key features that make it uh so much like hyped? What are the different features of systematic TV that makes it so much accomplished and it sits at the top of evidence hierarchy that it answers the focused clinical research questions like our questions are not broad uh just we answer in the essays in the blog posts and if we discuss in the research the the narrative reviews they discuss a very broad topic uh they they have no defined methodology they have no predefined inclusion excluding criteria what I mean by predefined including excluding criteria means that we first define which types of study we'll be including which type of population we will be including the type of treatment we will be including so basically everything that we are going to include we define it first and then we include And it's comp we have the comprehensive search strategy like everything is given in such a manner that if someone uses our methodology he will be able to exactly generate the same strength uh of the evidence that we generated he will get the same answers he will get the same results. show systematic review that is that answers a focused research question. We have defined including excluding criterias prior to starting the review. We have a complete search strategy for all databases for all the keywords that we use and we have the transparent and reproducible process. So if anyone wants to do to redo or to confirm our findings, he'll simply use our methodology and he'll get the same results. So the transparency and the reproducibility makes the systematic and meta analysis strong enough. For example, I have an example of efficacy and safety of impeccing renal outcomes in heart failure patients with reduced ejection fraction. So you guys can see that looks very uh long and very complicated but it's easy can see how how focused the question is and how everything is explicitly given even in the title of the systematic review. So you can see we have the patients with the heart failure with reduced ejection fraction. Then we have our intervention or treatment that is impacting. we have the placebo control and then we are focusing on the outcomes that address the heart function and other adverse effects. So everything in a systematic loop that is predefined and explicitly given and it addresses a focused research question. So this is an example of a focused research question.

[00:12:56]So what is a meta analysis? uh that basically a statistical method that combines data from multiple studies to estimate the overall effects. So if we combine the numbers from all the included studies and perform an analysis in to put in a simple words so that becomes a meta analysis. If we have uh like the individual studies are having uh small number of participants and we combine four or five studies so we get a meta analysis with greater number of participants. If we have the greater population it increases our statistical power. Uh we can resolve the uncertaintities from conflicting studies like there might be some trials on same drug in same population that uh that says that the treatment is uh associated with the significant outcomes and on the same like manner the same drug can have uh different results in a different region or on a different population like the results could be different in different ethnicities, the results could be different in different geographical uh areas. So it becomes difficult uh for a clinician for policy makers to exactly like reach a conclusion. Here comes our role as a researcher. We put everything and we we do uh statistical analysis on that that becomes a meta analysis and we provide the consensus on the conflicting studies. It provides the quantitative summary of the evidence.

[00:14:40]So we have responses This is an evidence pyramid. You can see uh we have different study designs that are uh in different order. So you can see the systematic reviews and meta analysis sits at the top of the evidence hierarchy because there are reasons for that.

[00:15:14]So we have the guidelines are the personal opinion because they have no study designs. They are at the bottom. Then we have the qualitative and the quantitative studies that may include the single institution or the single method use studies like the case controls like the case series like the case studies and then we have the experimental studies case controls and cohorts.

[00:15:41]And then we have randomized control trials. As we move upwards, uh there is increasing aggregation of data size or depth of the study and generalizability. So if we combine uh three or four trials that become the systematic review and meta analysis and definitely it carries more weight than the primary studies because the information is filtered. So if there is any physician or researcher that is like studying an issue and there are lot of studies so he'll not simply go to the primary studies he'll just simply read the latest systematic review and meta analysis about that and he will get get to know everything about that because the information is filtered the analysis is done and everything is critically done.

[00:16:36]So there are different types of reviews.

[00:16:39]We have the traditional reviews. We call them the literature reviews. And then we have the systematic reviews. The reviews that are focused on the specific research questions. They have explicitly explained predefined research methodology. These are transparent.

[00:16:59]These are reproducible. Then we have the meta analysis. So these are systematic reviews. The category, the types of systematic reviews that includes the stats, basically the statistical analysis, the aggregation or pooling of the data using some softwares.

[00:17:20]It's same uh the literature review and the narrative review is the same thing.

[00:17:27]And uh we have the meta narrative review said. So Aribba uh there is a lot of difference. There is no explicit meth methodology for the narrative review.

[00:17:39]The research question for the narrative review is very broad. There there are no search strategy for that. There is no account for the risk of management risk of bias assessment. There is no account of hetrogenity of the included studies.

[00:17:54]Basically that's just like an essay. You you gather article of your choice and you make a narrative that becomes a narrative review. But in systematic review or meta analysis you define everything first and then you build your study on that. So that's reproduce reproduc reproducible that is transparent that has explicit methodology it accounts for the hypogenity it accounts for everything.

[00:18:20]So meta analysis it's a quantitative synthesis and narrative review is just you can say an essay or or just a qualitative way.

[00:18:40]Then uh so systematic review this is just like a funnel where we put a lot of studies and then we use the systematic review process and we filter and that's the results of that process.

[00:18:57]So systematic review is just like a funnel. We put a lot of studies the studies that fulfill our criteria. These are included. We synthesize information from that and that becomes a systematic view. So everything is arranged and reproducible.

[00:19:14]Uh here is a slide for the difference between different types of studies.

[00:19:20]Uh we have the literature review we call narrative review as well. So it summarizes a topic that is broad like a cancer treatment. It's qualitative. Uh the sources may be biased because it's uh up to the uh it is as per the reviewer like you you like to have to publish the positive results or a single aspect of a topic will be doing a narrative review or a literature review on that and you'll be publishing only that. It does not define any types of included studies. There is no explicit methodology. There is no search strategy for that. Then we have the systematic reviews that answers specific clinical questions. You can see the difference between the question of a narrative review and a question of a systematic review or for example is vitamin C or chemotherapy a better treatment option in patients with age of with patients over the age of 40. So you can see how explicit explicit and specific clinical question is that. Then it defines specific search strategy. We list all the databases that we use. We list included and inclusion and exclusion criterias and uh we can perform a meta analysis if the studies are uniform and there there are enough quantitative data to to pull. So uh a meta analysis is a systematic review that includes uh the the analysis part, the quantitative part. So the quant the quantitative part or the analysis part, the statistical analysis part. It gives us the quantitative value of the overall effects the difference between the studies and uh everything in a quantitative way in the form of numbers. Uh here is another slide showing the difference between the systematic review and meta analysis.

[00:21:21]Systematic review could be qualitative or quantitative. On the other hand, the meta analysis is quantitative. We explore the hetrogenity and uh there is no mention no account of hetrogenity and so I have a question for the participants. How many clinical trials are published every year? So that we can we can have an idea why we are just focusing on systematic review and meta analysis so much in today's session. So uh how many randomized clinical trials are published every day?

[00:22:38]We have answer from Shabir in 30 clinical trials. No, no. I'm talking about overall in the world. How many clinical trials are published? Any rough idea about that?

[00:22:56]so that we can focus on how why we need randomized control trials when we have clinical trials we have other types of study bilal has answered that in thousands so yes how many thousands thousand that could be 100 thousand that could be even more that any any raw figure in thousands Okay.

[00:23:44]So we have over 19,000 randomized control trials that are published every year in different databases. So you can see you can imagine how many randomized control trials are published. The letters to editor case controls cohorts cross-sectional studies and other experimental study designs are uh are not included in that. So we see a boom in the number of papers being published every year. So the number is increasing exponentially. So uh in the upcoming years we'll have thousands of like the double or triple of that number the clinical only clinical trials being published in different databases. So the scientific data is basically exploding.

[00:24:35]So in early 90s the cockran collaboration came up with the idea to summarizing the randomized control trials so that uh the boom could be like there is lot of articles. So how will we get the articles of our interest as a physician as a scientist or any other academician. So I will have I will be having no time to studying like if I have uh 20 trials on a on on on a disease or on a drug and I want to prescribe to my patients. So it would be difficult. So here comes the role of researchers the systematic reviewers and meta analysis. So that's why we do systematic reviews and meta analysis because the scientific data is increasing exponentially and we need to make reviews we make we should we have to make it easy for the policy makers to the physicians and to the other researchers so that we can present the data in the contact form.

[00:25:37]So here is another question for you. We have three clinical trials. Trial one, two and three. I will be having uh question for the participants. Which trial is better and why it's better?

[00:25:56]So we have randomization blinding the consort guidelines.

[00:26:04]So which trial is better? We have two drugs, drug A and B. So which trial is better and why do you think so?

[00:26:45]Okay.

[00:26:59]So, Zahash said trial 2 is better. I would like to know your reason as well why trial 2 is better.

[00:27:22]So Raf said trial three is better said trial two is better. So why trial two is better? Why trial three is better? Why are you guys saying which trial is better and why? I would like to know the reason.

[00:27:39]Zahara said there is less bias, there is randomization, blinding and two populations multi-op.

[00:27:46]So, so there is a confusion in the participants when we have only three clinical trials. So, what if every trial is randomized?

[00:27:59]It's blinded and it follows the consort guidelines. So which trial is better and which drug is effective now.

[00:28:26]So now I have made the common grounds like we have all the trials are randomized they are double blinded they follow the consort guidelines. So which trial enter now and why which drug should be given to the patients? They are patients with the same disease but we have the two different drugs. So which drug we'll be choosing and why?

[00:28:58]So they say we'll be using trial three but it has only 100 patients. It includes only black population but it's latest.

[00:29:08]We have 1,000 participants, 500 participants, two ethnicities and uh it's published like in 2022.

[00:29:24]So yes, but what if we keep on increasing the number of trials uh number of patients including different ethnicities. So that would be much difficult to understand which drug is in fact better for the patients as we were divided in just three clinical trials.

[00:29:46]So if we combine all these trials we can per that would be called meta analysis and we can quantify which drug is better and why that drug is better. So that's why we do clinical we do the meta analysis. I think it's clear now. Give me thumbs up if you if you get this point. So we'll be continuing with the steps of meta analysis. Just let me know in the chat box and if you're having any queries we can take three questions for now.

[00:30:41]So bilal if we combine all these three trials the audience were divided whether drug A is better whether drug B is better so what if we combine all these three clinical trials and we perform a meta analysis we will be having 1,00 plus 500 plus 100 patients we'll be having Asians Asians and blacks both ethnicities we'll be having studies published from 2005 to 2023. So we will be having a complete all the studies that are published so far. We'll combine all these trials and we'll perform data analysis to see which drug is actually helpful for the patients.

[00:31:44]So okay.

[00:31:49]So why we conduct systematic review or meta analysis? So to basically compile the information we'll get single line answers whether drug A is better or drug B is better for that population. we resolve the conflicting results as we had three clinical trials, two drugs. So it was difficult for the for us to determine which drug is in fact helpful for the patients. So if we perform a meta analysis, we get answers to the conflicting results that is the basis of evidence-based medicine that leads to change in of the policy, the treatments and also we identify the research gaps.

[00:32:33]What if both drugs are not helpful in managing the patient's symptoms? It means that we are missing something. We need to design new treatments that are much better in treating that disorder.

[00:32:47]So what are different key features of systematic review and meta analysis? As we discussed earlier, a meta analysis clearly states the objectives with predefined eligibility criteria. The eligibility criteria mean the inclusion and the exclusion criteria means which studies will be including and which studies will be excluding and why >> and we have the explicit or reproducible methodology for that. The systematic review attempts to include all the studies that meet the eligibility criteria and there are assessment tools to of findings. The risk of bias in the study so that we can assess the validity of the findings, how valid our findings are and the systematic presentation and synthesis of of the findings of the included studies.

[00:33:54]What are benefits? So uh it increases statistical power. It improve precision uh ability to address inconsistencies and ability to answer new questions. So these are different benefits of conducting systematic review and meta analysis. So as we have a lot uh a lot of benefits of systematic use of meta analysis. So there are limitations as well. So what are limitations and what do you get from this picture?

[00:35:03]So shabi said it's hydrogenity. Yes, it implies hetrogenity. So uh we have the garbage in and garbage out. If we include the bad quality studies, we'll get a bad quality meta analysis and if we put a good quality study, we'll get good quality meta analysis. And if we combine different studies with that heterogenous results, the results would not be generalizable. So it would be difficult for us to make conclusions.

[00:35:39]Then we have the hetrogenity in the studies. These are the three different three main limitations in meta analysis.

[00:35:48]If we include the poor quality and the bad quality studies, the meta analysis would be of bad quality and if there's there are publication bias the things that we cannot address because studies usually tend to publish the positive findings and hydrogenity the variation in the different studies. The studies could have different participants different ethnicities conducted conducted in different uh uh regional regions. So these these may include different variations that might uh be difficult to interpret the results and generalize the findings. So these are the main limitations of a systematic review and meta analysis. I think I have instructions from the management. I will be taking uh a break now. So hand it's over to the management.

[00:36:45]We'll be taking a 5 minute break and we'll be joining again.

[00:37:16]Uh okay. We will send you uh we'll get again and we will join from there. Okay.

[00:37:23]Let's uh leave this meeting. Okay. Thank you so much.

[00:43:33]We have four 20 uh 4 minutes and 20 seconds now. Uh, is there any So far we have discussed about uh the systematic review and the meta analysis the features that make it distinguished and put it on the top of evidence hierarchy. So what are different steps if you want to conduct a systematic review and meta analysis on your own. So first we have to define a research question that would be according to the pico framework. We'll be discussing this framework shortly.

[00:45:14]Then the second step is to develop a protocol and register it to the probi important part. Then we conduct a comprehensive literature search on different databases. We check if there is if it's already published or not and how many studies that would be possibly included in that. Then we from inclion and exclusion criteria. We select these studies and then we assess the quality of these included studies. We extract the data and analyze. We synthesize the findings in the form of manuscript. We interpret and we write the review and and publish. So these are seven to eight rough steps in conducting a systematic review and meta analysis.

[00:46:10]Let me join the live stream from YouTube. If you guys have any queries so it's clear now and uh yes so how will be select the topic? The topic selection is the most single crucial step in a systematic review and meta analysis because everything starts around the topic selection starts around a reasonable systematic review and meta analysis.

[00:46:55]So the ch what are different qualities of that topic for systematic review and meta analysis. So the topic should be novel, should be new, it should be interesting and clinically relevant to the current discussions.

[00:47:23]So I have a question for the participants.

[00:47:28]You are an editor of high impact journal. For example, answer and you receive a paper. The title of the paper is smoking as risk factor for lung carcinoma a systematic review and meta analysis. So it's strategy data extraction analysis write up everything is perfect. Would you like to publish that article or not? And if not, why not?

[00:48:39]So Bilal has rejected that topic because it's not novel.

[00:48:45]Anyone else? Everything is good. The the methodology is highly reproducible. The search strategy, the data extraction, the analysis and everything that is up to the mark that is perfect.

[00:49:27]So almost everyone has rejected this novel meta analysis because they say it lacks any no it it lacks any knowity. So definitely if I were to the editor of the journals I would have rejected that article because uh besides of the methods the most important thing in the topic selection that it should be novel. We know from like two or three decades ago that smoking is a risk factor for lung carcinoma. So it's established factor. So if something is established fact there is no need to conduct any meta analysis. So it's waste of time and efforts. So that will be not published. So you have to come up with the novel interesting and clinically relevant articles. So move forward. So I will be formulating the research question that is the framework. We'll be using the pico or the pico framework in which we'll be defining our population problem are the patients who are the patients for example in the first example that we discussed that was the patients with heart failure with reduced ejection fraction. So the population of our that systematic review was the patients with reduced heart failure with reduced ejection fraction. So that was our population. Then interventional RX expo what we are doing to them. So we were giving them the impeilin that is SGLT2 inhibitor. So the intervention is the drug the surgical procedure are the expired that we are giving to the participants to the population. then definitely we are comparing it with something that might be a placebo that could be another drug or intervention or any other procedure. So we always have a comparison comparator in systematic review and then what outcomes are we measuring the improvement in the heart failure parameters the adverse effect.

[00:51:38]So all these things that we are measuring these are the outcomes. So every research question for a systematic review should have all these components.

[00:51:48]It should have a defined explicit pipo.

[00:51:52]Uh everything should be defined. Who are the patients? What what is the problem?

[00:51:57]What we what what do we have to do with them? And uh that could be an intervention, it could be a drug, it could be a medical therapy, that could be a surgical procedure, that could be any relevant thing. Then we have the comparator. What happens to that? And what is the outcome of that? Another example of the type of criteria the patients population or the problem. Then we have the intervention. Then we have the comparator outcome and sometime it could be time bound like one year outcomes two years outcomes or short short-term and long-term. So if there is any mention of that the time that that is also sometimes part of the pipo criteria so I'll be selecting a topic these are the general principles So first we'll be selecting the area of interest because the medical field or if we move towards the sub specialities these are way much broader to select any physible topic when we'll be selecting a disease. is that there is active discussion on that drug that disease. Then we will read all the available treatments that are available in the markets, the latest clinical trials that are being published on that. Then we will select a treatment. We'll find the gap. We'll go to the databases. We'll check how many trials are being published on the specific drug. Then we'll formulate our research question to check the feasibility. If it's not feasible, we'll start from here. And if it's feasible, we'll do the literature search. We'll search the database and we will form our search strategies and we'll go with that.

[00:54:22]So what do I mean by the area of interest? That the area of interest is something that align with your research interest that is of clinical significance and uh there is relevance to the stakeholders like it is important to the patients, it is important to the clinicians and it is important to the policy makers. The mentor is available readily to you and the field you want to pursue and there is a lot of new data available in that. For example, you are uh that could be cardiology, neurology, gastroenterology, hematology and oncology. So there are some examples of the areas of interest. First you will select an area of interest that is of clinical significance that is relevant to the stakeholders. The mentor is available. the field you want to pursue in future. For example, you want to become a cardiologist, you will be doing research in cardiology and the lot of lot of new clinical data is available.

[00:55:26]For example, the cardiology hot field, the hematology, no hot field. So, uh and the rheumatologist field lot of new clinical trials are being published I think on the daily basis if I'm not wrong.

[00:55:42]So for example, you want to become a cardiologist and you select cardiology and you got to know different diseases that are common in cardiology. We have the coronary heart disease, we have arrhythmias, valvular disorders, we have the aotic disorders, uh we have the heart failure, sclerosis etc. So I am interested in heart failure and I get to know all the treatment options for the heart failure. I have ACE herbs and uh we have the beta blockers, SLT2 inhibitors, diuretics, deoxxyin etc. So these are different treatment options that we give. So I am interesting in interested in SLT2 inhibitors and I read about that and I found that we have a lot of new drugs of that class uh and there are clinical trials being published in different databases. So I'll form my research question and see.

[00:56:42]So I have a research question that impact closing in patients with heart failure with reduced ejection fraction systematic review and meta analysis. So this is how I form my my topics and then then we'll check the physibility on the databases. This is how we narrow down starting from a field and then coming up with the novel treatment for a disorder that is fra and a lot of new clinical data is published about that.

[00:57:17]And then we have an uh the example we already explained this example efficacy and safety of impact reprogramming in improving the renal outcomes in the heart palian patients with reduced ejection fraction. So that was an example in the cardiology. The next step that is developing a protocol. So I'll be make developing the protocols.

[00:57:44]uh why we develop a protocol so that we ensure the transparency it serves as a road map you figure out everything the analysis plan the strategy the outcomes that you'll be including the participants or the co-authors you will be including you register on the prospect so it's a complete road map for the review uh what are the different key elements of that protocol that we call it synopsis or the research question uh cyopsis in our like context uh and uh we'll call it a protocol like we get cyopsis from the university uh for a cross-sectional study or any other original study. So there is the same idea for systematic review and meta analysis. We register a protocol. We get a number identifier so that we are the only working on that meta analysis and that would be published because we have registered that meta analysis prospectively.

[00:58:58]I'll show you guys how to check the physibility of the topic. Uh you can guys post some of the examples of the topics that you want me to check the physibility. Uh please post in the chat box uh the research questions are the topics that you want me to check the physibility and I'll be checking for you if they are feasible to do a systematic view or meta analysis or not. Just uh put it in the chat box.

[00:59:29]So we register sis prosperos on uh our protocols on prosper that's a website of international prospective register of systematic and previous and meta analysis. Uh you can guys search the website and that that would be the first uh website here we can you can uh register your systematic protocol of your systematic review and meta analysis. So the next step is the literature search. These are different databases that we search. Uh we have the pubad, we have prain, we have scopus, mbase, clinical trial, web of science, we have Google scholar that is search engine not a database. So these are different databases that we search for already published literature so that we can find the studies to be included in our review. So uh the format is free. The cockrine is free. The clinical trials is free.

[01:00:31]All these databases, Scopus, MBS and web of science, these are the paid databases.

[01:00:37]And we have the Google Scholar that is the search engine. Uh it's not a database. There is no explicit search strategy for that. So uh it's ideal to search at least two different databases and report their search strategy at least in your supplementary files that will be published with your review. You have to like we have this and these three clinical three databases. So make sure to search at least two databases.

[01:01:14]Then the next step is to identify the articles first like gather article from different databases. So it's possible there might be a lot of uh duplicates.

[01:01:26]So screen these articles in the first step will remove the duplicates. That step is called the dduplication. Then we screen on the basis of just titles and abstracts. We don't need the full text.

[01:01:41]We just skim through the title and abstract and if it's relevant the relevant keywords are present in the title and the abstract we include it and then we do a full screening on the on that article. Let's suppose I got 500 articles from the PubMed and then I got 600 articles from the the cockrine and I got 400 articles from the clinical trials. So all these articles are related to my topic. Then it's possible the same articles could be indexed in all three databases. So I will remove all the duplicates. The first step will be removing the duplicates. So you can do it manually.

[01:02:27]You can do it on Zotterero. You can do remove the duplicates on And then we have the next step.

[01:02:50]Just wait a minute.

[01:03:03]We gather the full tests. We do the screening on the basis of the predefined inclusion and exclusion criteria that we have defined before starting the review.

[01:03:14]So and then we include the studies.

[01:03:18]There are some tips for reducing the bias. Always use two independent reviewers. Two people should be ideally doing the screening and everything and the third reviewer should be uh checking upon all the things so that there is minimum waste. For example, I gathered all these articles from from different databases. Then I removed the duplicates and then I uh for for example I have 1,500 articles total and then I removed like there are 500 duplicates. So I have now 1,000 articles then I will be screening them on the basis of their titles and abstract. That is called primary screening. Examples 700 articles are removed. So these are irrelevant. Then I have 300 remaining articles. I will be screening of their full text. I will be finding their full text. And at the end for example I have 20 articles. They have same population, same treatment, same outcomes. So I will be including these studies at the end in my systematic review and meta analysis.

[01:04:36]So what are different tools for screening? Uh there are some popular tools. Ryan, Zutero, we have Endnote, we have Google Sheets and the Microsoft Sheets. Uh this is my rating for the different tools. I prefer Ryan for primary screening. It's very efficient first and uh free tool. Then there is a zotterero and end note at the endote stage. So I usually don't use that. The Google sheet uh is an excellent tool for doing secondary screening then the process then there's a process of data extraction. So what are uh for example we are like moving forward with the question from the start impact in art with reduced ejection fraction. So what components we'll be extracting in the data extraction I'll uh I'll classify into three different subcategories. We have study information uh like the identifiers the name of authors the year of publication journal and the DOI. Then I'll extract the data from the baselines or the study characteristics like the patients in the treatment, the patients in the control or anything related to that. That would be my baselines and then I'll be extracting data for the adverse effects for this for the efficacy outcomes like improvements in the heart failure parameters. So I have uh I will make two sheets. One will be about the baseline characteristics and one will be about the outcomes. We'll extract the data in the same format that we'll be putting into the analysis software.

[01:07:05]So this is an example of data extraction sheet.

[01:07:13]This is how I extract my data. You can see there are columns for the study ID.

[01:07:20]We have the name of all the authors. We have year of publication. We have the study design and the location of these trials. Then the information about the number of participants in both groups.

[01:07:33]So these are the baseline information prior to starting the treatment.

[01:07:46]and the medical history of the patients, the BMI and the baseline. All these things, these are the study characteristics. Then all the lab values, these all all things come in uh the study information and baselines.

[01:08:04]Then I have extracted the data in exactly that format that will be used for the data analysis in the RAVMAN.

[01:08:13]This is for the dichotomus outcomes. You can see I have the events and the total for the treatment and the control. I have same studies IDs, number of participants in each and the doses for the subgroup analysis. And this is for the continuous outcome. You can see we have the mean, standard deviation and number of participants for all different studies.

[01:08:42]So this is about so how which tools we'll be using for the data extraction.

[01:08:47]These are different tools that I prefer.

[01:08:49]This writing is solely based on my experience with this tool. So I use Google Sheets because it allows you to interact and uh work as a team.

[01:09:19]How will be assessing the quality of the different uh studies? We have different tools for different study designs included. If we have uh clinical trials, randomized control trials included in our meta analysis, we'll be using the risk of risk of bias 2.0. zero. If we have observational studies like case controls and cohorts in our meta analysis, we'll be using the Newcastle tapa scale. If we have non-randomized study, we can use Robins I or E based on intervention or explorer. They ensure the reliability the results how reliable are the results.

[01:10:00]And this is an example of uh rob 2.0 zero that this is called traffic light plot and that is the table that is Newcastle scale rating table. You can see we have different three subcategories and the different domains of the these. So we critically assess the quality of the included studies. If the quality of included studies is good, the meta analysis is of good quality, the results are generalizable.

[01:10:33]So I'll be performing the meta analysis.

[01:10:36]Uh we have different types of effect measures that we use depending upon the type of uh the outcomes. If we have dichotmous outcomes like the past fail, the mortality, the relapse that are given in the form of events and total. So these are usually that usually called dichotoous outcome and we'll be using odds ratio or risk ratio a risk difference for that. But if we had the continuous outcomes for example the mean length the mean value of a alt props ejection friction GFR. So anything that has a value that is ranging that could be minus that could be zero that could be any value uh that is the the the range of value that is called the continuous outcome. So I'll be using main difference or standardized mean difference for that. There are two two models random effects model and the fixed effect model. We always prefer random effects model because the fixed effect model is used when we have the the trials are the studies conducted conducted on the same participants by the same uh researchers in the same region. So that's less common. So we'll normally prefer the random effect model.

[01:11:59]Odds ratio is used for the observational studies and the risk ratio and risk difference. These are used for the randomized control trials. We use the mean difference when we have the data in the same units and we use the standardized mean difference when we have the data in different units. So these are this is the forest plot. You might be familiar with that. Uh there are different components of that. We have the study or subgroup here in in the left side. uh this express the effect estimates of the information. You can see the totals and the events in the treatment and the control and that is the effect measure. The risk ratio of the odds ratio these are called the effect measure. Uh these are the overall statistics that is about the hetrogenity that is the hetrogenity. Then we have the p value that refers to the significance of the study. Then the one side the left side that is about the treatment. And then we have the other side that is about the placebo. So we have these this implicate the size of the study the box and this this line expresses the confidence interval. Uh which means okay uh where actually our the for example if we have uh large number of individuals or participants will have the large blue box and we'll have the smaller confidence interval. It means that there are less chances of error uh and we are sure that our results lie in this range.

[01:13:45]The our true effect lie within this range. So if we have this long long uh line it means that the effect could be here the effect could be here that could be here that could be here. So there is a greater margin of error in our true effect.

[01:14:04]Again uh that is about the forest plots.

[01:14:07]It refers to the effect measure. Then we have uh the model that we are using the confidence interval limit that is the line implicates the 95% confidence interval. Then we have the odds ratio.

[01:14:25]Uh that's the diamond that all we need.

[01:14:29]It it gives us the idea about the overall odds ratio. There are different tools that we use for meta analysis. We have the refman the most uh famous tool for that. We have open meta analyst, we have meta excel that is extension of excel. These are free tools. We have jamovi, we have art studio and we have another great tool that is called comprehensive meta analysis. So all these tools they have their different benefits. uh we use it for different types of meta analysis. Uh how we'll be reporting our findings. We'll be using the Prisma checklist and we can also use the grade assessment. What are different components of the manuscript that we'll be doing? These are the abstract, introduction, methods, discussion and the references.

[01:15:24]So the next step when you are done with the meta analysis would be choosing a journal. So where would you publish?

[01:15:32]It's better to publish in open access journals than uh the hybrid and the closed ones. We'll use the journal finders for example elsewhere spring and the other libraries to find the relevant journals based on the abstract and the title relevance of that uh things to consider in submission where the strength of your study, the similar studies that published in the journals, the impact factor or the indexation.

[01:16:46]uh we'll be quickly going through that and then we'll be starting the practical demonstration about the topic and the search strategy and we'll be impact of the journals the acceptance rates and the time for the first decision. Then there are some suggestions from my side. The books if you are a book guy uh we have different books. Uh if you want to learn more about the meta analysis that is introduction to systematic review and meta analysis. Introduction to meta analysis we have the handbook of systematic reviews of interventions and finding what works in healthcare. These are the uh guide books for systematic review and mater uh which social medias to use as a researcher. So I'll advise you to use LinkedIn research gate, Google Scholar and LinkedIn uh and X formerly Twitter.

[01:17:46]All of these they are very helpful. What are different courses that I would recommend you to strengthen your understanding of systematic review, meta analysis and scientific writing? Uh that is introduction to systematic view and meta analysis by John Hopkins University on Corsera. Then the writing in size is the perfect course if you want to learn each and every aspect of systematic review and meta analysis.

[01:18:13]And then we have understanding the basics of clinical research and the I programming if someone is interested in learning the R and advanced bio stacks.

[01:18:23]So there are some people to follow on socials that they usually post on social media about research about research designs about AI. So hope you get connected to these people you'll learn a lot of new things as well.

[01:18:40]So if you have any questions otherwise we'll be uh having a demonstration about how we'll be searching different databases and we'll be checking the visibility of topic as well. So I'll first answer the questions and we'll then we'll start with the demonstration.

[01:20:15]So far we have no questions.

[01:20:40]Please post your queries.

[01:20:54]So uh Do you have any topic in your mind that you want me to check the visibility?

[01:21:02]I'll be showing you. Uh we have a topic.

[01:21:05]So this is how I check the visibility of topic. I go to the Google Scholar. As we know this is not a database. Let me share my screen.

[01:21:34]So I go to the Google Scholar and uh we have protein nodastic protein and we'll be using and whenever we will be combining two different terms uh that are not synonyms of each other we will be using and and if there are synonyms uh the same words for the same different words for the same disease we'll be using or so these are different boolean operators that we'll be using so we have the early uh routine is decompression So I have my intervention I have combined that is the treatment arm that is the control arm and that is this treatment decompression.

[01:22:49]And I have that are two different uh like treatments that I have combined uh treatment and comparator. Then I have elective colctal surgery.

[01:23:10]So this is my population. Then I have my intervention and comparator. And I'll click search.

[01:23:17]So you can see we have 7,000 results. Uh on the top we have met uh paper a meta analysis published in 1995 on selective versus proteinogastic decompression after elective lepardy.

[01:23:33]It's not about the the CRC elective colon and rectile surgery without nogastric decompression a prospective trial that is published in 1980. So I'll be like the public the results published in the last years I'll be is routine as a listing necessary following the emergency show I think there are uh I'll be searching trials for that so I'll be adding and and RC for that if I add RCT it will search all related to that topic.

[01:24:25]Click search.

[01:24:40]So I'll click on anytime.

[01:25:09]So we have elective colon rectile surgery without nopastic decompression.

[01:25:14]We have one trial, we have another trial. So so we have some of the trials but we have a meta analysis that is published. I think that topic is already searched. There are trials that are like 30, 35 or 40 years ago. I don't think it's a good topic because there are less there are few new trials on that. So we'll move forward with another we have to do and forism and I will be searching randomized control tri for that. So uh we have the clinical trial published in neurology in 2025 that is rational for that. So I think there are very few trials on that. It means that's a new drug. We need at least two trials to conduct a meaningful systematic review and meta analysis. The the studies published these are phase one trials. So we cannot combine phase one with two. So I think there is only one clinical trial on that others are rational and development of that and then we have the the phase one trial summary and the other analysis I don't think it's a good topic to conduct a meta analysis on that as we have only single clinical trial I can like search the other pages and uh I don't see there is another trial but is related to another disease so there is only one clinical trials on that drug.

[01:26:53]We cannot do a meta analysis on that single drug. So it's not a feasible topic. So this is how we check the feasibility. The Google Scholar is the best thing so far to check the feasibility of the topic. You can also like if you want to see if there is any systematic review on that topic, you can add the systematic word with that.

[01:27:19]So uh it's only clinical trial to burden as adjunctive therapy. We have the fixed dose. There are clinical other clinical trials but there is no systematic review uh focusing on that. We have the phase one trial, phase two trial. So we need to study all available studies. We'll select the studies that are highly relevant. We'll screen them. We'll try to find their full text and we'll find if we have at least two studies with two or three common outcomes in all the studies so that we can pull them and perform a meaningful meta analysis. So this is how we'll be performing a meaningful meta analysis. So this how we'll be searching that topic on clinical trials and making search strategy. So I'll be searching clinical trials then the pop for that. And I need to open the mesh as well.

[01:28:26]Oh yeah, I know it's like start doing everything. I think it's better to build the base the basic knowledge the baseline and then the next the next session will be helping you guys so much. So I think this the slides these concepts this uh is not a waste of time rather it builds your background knowledge the bas baseline knowledge that is necessary for you to understand the upcoming sessions.

[01:29:40]So let's search this topic on the clinical trials as well. So I am putting this in clinical trials. Uh I'm copying the drug. Simply go to the clinical trials.

[01:31:32]So the clinical trial is open. Now I will paste uh the drug in the intervention or the treatment.

[01:31:41]Then I'll copy the disase and I'll paste the conditional disease. So that's Parkinson's disease. The clinical trials I think is one of the simplest to do the search strategy.

[01:32:00]It will show us the results of the all the studies. So we have three studies.

[01:32:06]One is open label trial that is not double blinded. Then we have the two trials that is completed. So I think uh if we open the trials and see either they are same or different. So I'll simply download the file with all the default information and will always try to export the file in RAIS format not in the rest format. So let me reexport into the rais format the research format that will be uh that can could be later on opened using zotterero and not rion then uh that is for for that I'll simply let me create a document for that so that you can know how we form the database uh Search strategy.

[01:33:26]So I'll be creating a word document.

[01:33:30]I'll name it search strategy.

[01:34:36]So we create a document. Uh this is our uh the title of the study. So what is our intervention? We have the drug that is our intervention.

[01:34:49]Uh the population is the patients with the Parkinson's disease and the comparator that is we are considering placebo or any other standard treatment for that.

[01:35:10]So first we have the search strategy for the clinical trials.

[01:35:25]We search clinical trials and we found three results and which such strategy we used when we were searching clinical trials we'll be using copying and pasting here. Exactly. So uh that was my search strategy. I'll be copying all these terms.

[01:35:50]Copy here and just paste it here. So So this is my search strategy for the clinical trials.

[01:36:35]This is how I will be searching clinical trials and you guys will be reporting your search strategy here like how many clinical trials were there when you search the database and which terms did you search and even you can paste the time date and time here so that uh if someone uh else searches and he'll get the same results. So this is why schematic views and meta analysis are reproducible. So uh let me move forward.

[01:37:04]Same to the advanced mode.

[01:37:37]And in the advanced mode here we will copy and paste all the terms that are related to our intervention. And then we'll be copying all the terms that are related to our population. So let me uh add an example.

[01:37:58]So this is an example of search strategy. We'll be using that to save time because we have just left few minutes. So this is an example of a good search strategy. You can see uh this will go in the supplementary files.

[01:38:15]uh we have the topic efficacy and safety of the drug in patients with type 2 diabetes systematic review and meta analysis. So we search the pubet. How we'll be searching the pubet and then the clinical trials the same that I showed you. We'll be searching clinical trials as well. And then in the same way we'll be searching the mase. We'll be searching the block. First uh we'll go to the mesh library. The mesh is currently not working. Uh let me search it again.

[01:38:50]The library contain all the words and their different.

[01:38:57]So first we'll get all the synonyms. For example, I have type one diabetes.

[01:39:05]Type one diabetes. So this is my population. I will click or and I will add other synonyms like insulin dependent diabetes. So this is another synonym of the population.

[01:39:27]I will add using or. And if I have another term for example uh I have stable diabetes that is another term that is used for the type 1 diabetes. I will add with or here is another term that is adult onset diabetes that is also synonym type 1 diabetes I'll add using or so I will first create a string using or related to all the search terms that are related to the that first string that is related to the population. So first I will create a string for the population using using the R and I will click on the search. So I'll click on the search and it will show all the results that includes type 1 diabetes in any of the synonyms. So you can see we have 3 lakhs 46,400 uh 46,424 results. So I will again go to advanced.

[01:40:42]Now I will uh similarly add all the synonyms for the metformin. So metformin has different names. Metformin hydrochloride glucage HCl metformin we have dimethile buanide. So all these are different synonyms that I took from the mesh website. So mesh website is currently not working on my device that's why I'm using that. First I will be adding the mesh as it is. I'll be copying and pasting it here in the advanced search and I will add using or the synonyms of intervention or uh the population all these will be added through or So I'm combining all the synonyms.

[01:41:33]I'll limit to three because we're short of time. So I'll click on search.

[01:41:42]So I will be having all the results that are related to we have 37 results on the metform. So now I got the first for the population and the second string for the intervention. Now I will combine these two. First I click on the three dots. I'll click on add query. This will be added in the query box. Then I'll click on add. So how I will be combining this? As I have intervention and and population I will be combining using and whenever you have two different like intervention or population you will be always adding with the and if you have synonyms of the same either intervention or population you will be using or boolean operator. So I'll click on search and you you can see I got the fewer results than uh the previously I was getting results in lags. So I'll cop select all the copy this search strategy copy in my document.

[01:42:47]I'll write the format how many results I got from the format here and then the search strategy that I used for the format. So this is how it works. Uh all these information that we are using that is refer but we will be making it neat and clean just like the document that I showed you. you'll be making in that form. You'll be using for that all available terms and you will also record the number of articles that we searched uh and the other databases for example clinical trials then we have the results all the results from the mase as well. And this is a paid currently we don't have any access but it automatically combines all the synonyms. So we don't have add it manually. Uh the same way we add it in the format. It works similar uh the coin works in the similar way. We'll be first adding all the terms related to the population. Then we will be adding Just give me a minute. I'm joining from my computer.

[01:51:20]with your body.

[01:52:26]Can you please guys make me let me share my screen. So so let's continue. We disc we were discussing about the search strategy. First I'll be explaining uh I will be explaining about the search strategy and how I'll be building our search strategy for PubMat and for the clinical trials. So uh this is an example for Can you guys see my screen? Please let me know.

[01:53:23]Dr. Bal and Dr. the the admin, can you please see my screen?

[01:53:30]>> Yeah. Yeah, it's easy.

[01:53:32]>> Okay. Okay. So, this is an example of a certain strategy. the document we'll be making the similar for that we'll be using the same topic because it's easy uh the search strategy goes to the supplementary files and it is published along with the manuscript so what are different components to break down first we have the database and then we have the number of articles that we got when we search the terms and I have also an listed all the terms that we searched.

[01:54:07]So how we get all these terms and uh how we'll be searching? So first we need to go to the mesh.

[01:54:17]We'll open the mesh website and we will open the pub. So we'll go to the advanced search in the pub.

[01:55:38]you search mesh library and then you will open the advanced search in the popat. So uh open the mesh that is mesh.ncba.nlm NH nlm.

[01:56:04]So our population for example in that case is type 1 diabetes.

[01:56:10]So I will be first searching type two diabetes. So I'll be searching type 2 diabetes.

[01:56:17]Just copy and paste your population and click search.

[01:56:25]So you can see we have two different uh classes. The mesh library that is basically uh medical subject headings that is vocabulary about the diseases.

[01:56:37]So you'll simply click on the first that is of our interest. You can see there is a definition of the whole definition the year of introduction to the terms and the mesh unique ID the entry term. So you can see these are all the entry terms that are being used for the type 1 diabetes. You can see these are like 10 to 20 different synonyms. These are all these are signifying a single disease that is type 1 diabetes. So we will click on add to search builder. First we can see uh diabetes and mesh is selected. So we'll go to the advanced search in the format.

[01:57:24]First you will enter the mesh term and click add and click add with power. So if we have different synonyms uh or the different keywords for a same intervention or the treatment uh or the treatment first we'll combine all the terms or all the synonyms related to the intervention with or then we'll make another string for the treatment uh for the population where we will be combining all different synonyms of the population and then we'll get two strings one is for the intervention and another is for the population. Then we'll combine these two with the end. So let me show you.

[01:58:06]Now I will add all these terms one by one to this advanced builder and you can see this is adding up all the terms. It means I am combining all different synonyms of a single that so that I am not missing any articles. It is quite possible some articles published years ago could have uh used uh another words that currently are obsolete but they remain in the databases so that we are not missing any articles. So I'm just combining few. You will combine all these and you will click on the search and you will make a string.

[01:58:55]You can see we got uh 22 237,000 articles related to that. So this these are all the articles on the pub about type two diabetes that either contain type 1 diabetes that either contain this term. So all of these are synonym. If one of the synony one of the synonyms is present the article is listed here. Then again I will go to the advanced and similarly I will be searching my treatment that is metformin.

[01:59:38]So there is mesh term for metformat as well.

[01:59:43]I will click on the metform. You can see the examples. You can everything see everything related to the red So I have the definition of metformin. I have all the terms that are alternatively being used for the metformin. So first I will clear this.

[02:00:37]I'll click add on the search builder. I will copy the metformin mesh term and I'll go to advanced. I'll click on add with or and then I will add all other terms related to that with or. So as I'm combining these terms you will be combining one by one and you can see the string is being built.

[02:01:09]So you will add all the terms. You can see the search string is being built.

[02:01:13]You will click on search and it will show you all the articles on the contain any of the synonyms of the metformin.

[02:01:23]So you can see we have 30 37,000 articles related to the metformin on the format. So I'll click on the advanced again. Now I have two queries, two strings. One is for the intervention and another is for the population. So I will combine these two. I'll click on the three dots, click on add query. So this is added here in the query box. Now I want to add all these queries by clicking on add with and so these are like from from two different uh like population and intervention. So I will be combining with the and and I will click on search.

[02:02:05]It will show me the articles that contain both of the terms metformin for type two diabetes. So you can see I have 14,114 articles. So this is your search search strategy. The your search string you will click copy and you will paste in the document. Similarly I have pasted.

[02:02:24]So this will be my search strategy for theat with the search strategy. You will click to send to click on citation manager and all the results and you will export all the results. That file will be downloaded and that will be used for the screening purposes. And similar file that we used from the clinical trials.

[02:02:46]We'll combine all these files in a single folder. Then we'll upload to Ryan, Zotterero or Endnote or any other software where we want to do the screening and we'll be using for the screening. So I think uh this is what I could deliver in that short time. I'll be happy if the another session is given to me in the next or or after that weekend. So I'll be continuing the detailed demonstration of such strategy and I think the practical besides the practical aspects the theory is essential uh for starting because you guys have a lot of uh high intensity sessions ahead. So I wish you very best of luck.

[02:03:37]Thank you so much for staying with me and I'm thankful to research consortium for arranging these sessions free of cost for the students. Best of luck to all of you and thank you so much for the opportunity.

[02:03:49]>> Thank you so much Dr. Thank you so much.

[02:03:58]>> Uh hello uh I will continue the session.

[02:04:02]Dr. Nabil has uh did search strategy and how to import the articles. Now I will continue uh from screening that how we screen. Uh let me share my screen.

[02:04:25]Uh please confirm that my screen is visible.

[02:04:40]Okay, it's it's so long session but uh later on uh people can uh watch it in 2x and uh revise it. Okay. Now if we have made search strings and imported uh the uh the and exported the risk files like this for example if I search advanc or I search like this uh dismenoria for example primary dismenoria I have just searched.

[02:05:12]Okay.

[02:05:14]Uh it it has 2582 results and I search quality of life with it.

[02:05:22]Quality of life of life. Okay.

[02:05:32]I will have very much results on this. I will go to advance.

[02:05:38]I can combine these two queries.

[02:05:41]uh you can see like this uh there is something like in in this uh string but I will delete it. I will delete it.

[02:05:49]Okay, I am confirming it and let's combine these two. Okay, how can I confirm it? I will confirm it like this uh one. This is the first one and I will add end in the end.

[02:06:09]Okay. and the second one like this.

[02:06:16]Okay. And then I will search combinely.

[02:06:21]Now you see there are 423 results on this. Okay. If I go to advance and I look into this, I will be having this much results on combining these two. Now what these two means? What are these details? So you can see the details like this here. Here are the details of my search. Okay. Here are the search details on which the uh the search has been run. Okay. Now uh Dr. Nabil has also taught you that how to import these files.

[02:06:53]Okay.

[02:06:55]Now like this I will uh I will click save.

[02:06:59]Okay. I will uh not not on this page. I will uh all the results I will uh try to export it or save or create file not summary I will create PubMet okay or CSV we can download in both I will uh create file and it will be downloaded here okay now I will open Ryan this is Ryan you will have to create your account and you can also invite members to it okay now I am creating another review I will create review I will uh Title this menu go here.

[02:07:40]Yeah.

[02:07:42]Quality of life.

[02:07:49]Quality of life.

[02:07:52]Okay. I I have added it is bio uh it is systematic review and the domain is bio biome. Okay. Uh I I do not have to uh add description because it's optional. Uh just to make the things in a bit hurry.

[02:08:11]Okay. Now it is uh saying that upload articles. I already have uh I already have uh downloaded it. So I will upload it here.

[02:08:27]and go to downloads and I will upload it here. Okay, continue.

[02:08:39]It will take some time but it will be uploaded to it. Okay, meanwhile I will also uh take some help from the Google Scholar because I will have to do some dduplication too.

[02:08:58]Google Scholar.

[02:09:02]Uh I will go to my library. My library I I have uh this library and I will go to this menu.

[02:09:10]This menu export. Okay. Uh this manuria and I will export in reference manager format.

[02:09:21]Okay. Export all on this page. Export all with this label.

[02:09:26]save the export.

[02:09:32]Now these are been uh added. Okay. And I will also up uh upload these citation a bit later but in on uh in another way.

[02:09:41]Okay. Now it is saying that uh invite members. Okay. You can add collaborators with you like I am adding my friend uh Dr. Noman here. Okay. and his role as will be a reviewer. If he is screening with me uh it should be collaborator but collaborator collaborator is in paid version to say message that's why I I will put a reviewer and if I have to add some message here I can also add okay I will invite him for the screening and after this I will uh now I have 423 reserve s as I have added. Now I also want to add further references I have uh downloaded from the Google uh from the Google Scholar. I will click here add references. This is other way. Okay.

[02:10:39]From my device and I will select files from my computer. Okay. And I will open it here. And now I will click continue.

[02:10:51]And it will take some time but it will be uploaded.

[02:10:56]Okay. I will click finish.

[02:10:59]And now you see.

[02:11:04]Okay. Now what will I do is I will do detect duplicates. Okay. I will do dduplication. Dduplication. It will uh it will take some time but it will be done soon. Okay.

[02:11:20]Now there are 37 duplicates. Now I will try to resolve duplicates as uh we do not need duplicates. Uh detect detection has already been done. I will start resolving it. Okay.

[02:11:38]Now they are saying that uh it is 96% match. Okay. The title is matched, the authors are matched, the year of publication and other things are match.

[02:11:47]So I will look into the details and if it is matching so I will keep one. Okay.

[02:11:54]In this one there is no abstract and in this one there is abstract. But I know this is the same study. One is from Google Scholar and other one is from PubMed. So I will keep one. Okay. Not two also this one this is a bit uh this these are different. These these studies are different. So what I will do is I will do is keep both articles. Okay. I will keep both and uh like this I will do all the studies. Okay, for the sake of uh time I I am not uh time is contracted now. So I will do a little it not much.

[02:12:30]I am doing it in a hurry. I'm not looking into it. Okay. I'm just keeping the left or keeping the right. This is the decision based on my reading not on like I am doing now. I I I should read it but I'm not doing it because I I need to focus on my uh on my uh what I am talking on my talks not on the screen.

[02:12:55]So I from here I can click keep both both articles. From here I can keep keep left. Okay. And here right. I will have to read these. Okay. with focus.

[02:13:34]These are heavy software. So uh telling uh sometimes that keep refreshing and refreshing. So I have refreshed it.

[02:13:43]Now it is zero duplicates. Now I have resolved it. You can see here uh uh one is not duplicate. 18s are deleted and 18s are resolved. Okay. Now the next step is you can uh you can see here we have the updated figures. We will have the up updated uh figures. Now you can see that uh 18 are uh resolved, 18 are deleted and unresolved are zero.

[02:14:11]So now go to the screening we have uh we have now 431 articles we imported.

[02:14:21]How much we imported? 449. Uh but we detected 37 duplicates in this. We resolved 18 and now we have uh we have 431. Okay. Now you see here these all are undecided. Okay. If I have uh if I done screening I will have uh decided undecided and other things like that. So for this I can uh exclude an article by clicking exclude and I can include an article by clicking include and if I am confused about some study I can add maybe. Okay, this is the short if I have to do with the reason if somebody is not agreeing with me. So I can also put a reason here from here wrong drug wrong things like okay now uh here you can see there is a blind mode on okay Dr. Noman my colleague uh I have added his email in this cleaning also he will not see my decisions in this stage. If I turn it off I will uh if I turn it off he will see my decision from now onwards. Let me show you some review my active review uh in which I have of the like this one.

[02:15:48]Okay, in this review we had uh 1445 articles in which we had 6 uh 96 duplicates and we resolved it uh like this. And now let's go to the screening.

[02:16:02]Now we have undecided zero. We have uh undecided zero. We excluded 975 studies.

[02:16:10]We have zero maybe because we have resolved it and we have included 31 studies. You can see now here I have included it and my colleague Dr. Fose has included. Okay. Uh it's in short about the primary screening. We have screened it by title and abstract.

[02:16:29]Okay. We have screened it by title and abstract. Okay. And what are the complex? These are the complex. You can see the conflicts here. I have included this study but Dr. Feros has not included the study because he had told me the reason not full text available clinical trial no result posted. Okay.

[02:16:48]So I will agree with him. I will also exclude him. Okay. So the conflict will be resolved. The conflict will be resolved. But just for the uh sake of record I am keeping it like this because we have to put reasons for why we excluded this. Okay. to primary screening. Now how we do secondary screening? Secondary screening just look into this sheet and all the things will be clear to you. Okay.

[02:17:42]Okay. Now this is my screening. This is secondary screening. Okay. Now I have all the studies with value labeled. I have all the studies I have labeled on the number. Okay. Let's see like this. I have this. This is the article. Okay.

[02:18:05]Now if I open the articles now you will see this is study 14. This is study 1 2 3 4 5 and up till 31. Okay we will have up till 31. These are 31 studies full PDFs. I will have to uh look into it in very detail and after this I will add my decision in this sheet in this sheet.

[02:18:29]Now on the study number one, I have looked into the full text and I have decided that I will include it that I will include it.

[02:18:41]Uh check here I have made the decision of inclusion and also my colleague had made the decision of exclusion. Okay. So we have included the green highlighted study full text. The second one is also included. The third the third, fourth, fifth and sixth. These are being excluded. Okay. Why these are being excluded? The the here are the reasons.

[02:19:03]This is oral presentation. This is not the full text PDF. Okay. And this is uh here we had the wrong intervention. Uh we had the wrong intervention here also.

[02:19:13]This is the oral presentation and many things like that. So this is all about the screening process. Lastly, I can build a Prisma flowchart and uh tell that uh this is this is like this. Okay.

[02:19:27]Now let me show you one study in which uh we have done it.

[02:19:35]Okay. Let me show you one study uh Prisma flow chart. Okay. Prisma flow chart. We have uh the publish systematic review and meta analysis. Uh below where chronic graph versus host disease.

[02:20:05]Yeah, here it is. Now you can see here how we do it. This is a published study in the Indian Journal of Hematology and you will see it here.

[02:20:33]It is taking some. Yeah. Now it is the topic was safety and efficacy of bilateral foride resistant chronic graph versus host disease uh a systematic review. We haven't done meta analysis on this because meta analysis of quantitative synthesis. Okay. And I was the first author in this and you can see now the uh Prisma flowchart of it. Okay.

[02:20:56]Where is the Prisma flowchart of it? Now this is the Prisma flowchart of it.

[02:21:00]Okay. What we do in the Prisma flowchart? We first we uh do a search string. Okay, we apply a search string and we extract articles like in PubMet and uh we do not use Google Scholar here. But the other databases we have used. Okay, after the dduplication or duplicates, we removed 455 and we screened this much articles. We excluded this much. We uh sort retrieved uh 47 for secondary screening. We assess this for eligibility and we have uh put on the reasons full text not available 11 uh some studies there were wrong design and in some studies we had wrong intervention. Okay. And lastly we included six studies and now see this study uh we have concluded these six studies like this. Okay. Now you can see this is the first, this is the second, this is the third, fourth, fifth and sixth. Now we have concluded it. If you want to read it, I can also provide you the full text. But uh it's of no use because it uh uh you can read the topic of your interest not my topic or someone else topic. Uh in which field you have interest read analysis. Okay. If you have any question, you can ask me in the chats.

[02:22:25]Okay. Yeah, Ryan is free and uh it is not downloaded, but you can use it online on your Chrome. Okay. Now, how to find topic? Topic recent systematic reviews analysis on it. Okay. Andification like for example they have included IBD inflammatory bowel disease or chronic Pediatric and adults pediatric adults and elderly because this thing is new. You are narrow down the things you are add.

[02:24:04]If you have any questions you can ask me.

[02:24:13]Special thanks to all of you for joining us today.

[02:24:17]We'll see you tomorrow inshah at 9.

[02:24:20]>> I'm waiting for your questions guys.

[02:24:22]>> Uh at 900 p.m.

[02:24:27]We'll continue this session for a week inshallah.

[02:24:54]Okay, I think we are ending. There are no questions. Okay, so guys, thank you so much for your time and inshallah this session will be available for you. You can click the link and there will it will be available. Thank you so much.