Computer generated transcript

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

Good evening everyone. We will be three minutes just uh like a little before the session. I hope you are doing well and safe. Uh Today, we will continue uh I will take 1st 10 minutes to continue what I was doing uh last week. Uh We need just to see uh two systematic review and meta analysis that was published uh And uh to practice a bit uh about the best steps that we saw them before and then we are going to move uh to the module two. We are gonna to learn how to, to, to, to manage all the list of articles that will be included, how to deal with them, how to deal with each article, how to do the data extraction and how to do the quality assessment, how to evaluate the quality and the risk of bias in each article of the included articles and then to move and to collect them all and to prepare them for the next modules. Module number three, it's about to decide if we can, we can go for meta analysis or just to do a descriptive qualitative summary or we can go to meta analysis and do a quantitative summary. Yeah, I will repeat very briefly about what we saw in the module one and the module one. First, we saw how to define our research question, how to find, how to find, how to find the gap in the literature. Then to establish our team and then to begin to write our protocols, our protocols will be will be composed of three main categories, three main parts. The first part about our first strategy. What what are the key words that we are going to use them? What are the first strategies that we are gonna put on it on multiple database, at least three or four databases to define our selection criteria, our eligible inclusion and exclusion criteria based on them. We are going to select the articles that will be included in our meta analysis and then to define what are the information that we are going to collect them from each article and to build our data collection for that's a collection form that, that will be composed of three main parts. The first part concerning all the sociodemographics criteria of the article and the participant that are present in this article. Second part, it's concerning the study quality criteria that I will going to discuss it today, how to assess the quality of each article, how to assess the risk of bodies that are present in each article. And the third part, it's concerning the outcomes, show results with ades can ratio A can the confidence interval ades can regression coffi all the outcome statistical outcomes and how they measure these outcomes will be mentioned. In this part nura we write our protocol we registered on Prospero. We run our search on the database. We collect all the articles we remove duplicate. Then by the screening part two part two step, first step concerning the screening of title and abstract, we read all the title and abstract of all the articles that you find them on all databases based on title and abstract field. You got to exclude the articles that are not related to our topic and to include the articles that are related to the next screening step. 1000 articles are a title and abstract of those of those 1000 articles and 800 are not related to our topic are excluded 202 100 screening of full text you are going to do to, to read uh the full text of those 200 based on the full text. And uh by applying our inclusion and exclusion criteria, we are going to select the final list of included articles in our mechanisms. What are, what I'm going to discuss today? I will go, I will go to today and tomorrow to today and tomorrow I will discuss after we collect or after we, we put our final list of included articles, how to deal with those articles, how to first extract, extract the information in the data extraction form and for sure should be at least uh done by two independent authors. Then how to evaluate the study quality and the risk of buys that are present in each article. And finally to prepare our data to decide if it's gonna go for meta analysis or to do a qualitative summary of that. Before I begin in the module two, I will go briefly for those two just to see, I'm sorry to see what r the steps that were done in those article and do meta analysis. First, I will begin in this meta analysis concerning cognitive treatment or cognitive training in Parkinson disease. A systematic review and meta analysis as I mentioned last week, it's very important and one of the criteria of rima guideline to mention in our title that this paper, it is a systematic review and meta analysis. As I as we can see here, the first put domain or title and then a systematic review and meta also here. Cognitive behavior, CBT for depression and anxiety of Parkinson's disease. Then they mentioned a systematic review and meta. It's important. It's mandatory to mention it is a systematic review and meta then they put the abstract and you skip it and go for the method section in the method section. As you can see here. First, they mentioned this review adheres to the Prisma guideline, preferred reporting items for systematic review and meta analysis. We are going to see to date very briefly. What are Prisma guidelines? What are those guidelines that we should follow with them during our uh working and our writing in any system at meta analysis? And also expend to mention this in our uh paper, we should mention this review adhere to Prisma guidelines. Then they mentioned the code number of registration, the ID number of registration on Prospero. They mentioned first eligibility criteria. What are the inclusion and exclusion criteria based on what they select the articles? We included RCT SA randomized clinical trials, examining behavioral therapy observation, clinical trials had masal uh including had training. Uh We we included only those that clinical trials are differentiating condition between the study groups and where clinical trials compromised at least 50% of the intervention and they define the inclusion and exclusion criteria of the study all later and search strategy and selection. How they do the selection of the articles. First, we systematically search midline and base in senile and central for the terms Parkinson in combination with wide use term describing cognitive based intervention. Yeah, they mentioned what are the database that was used? They used four, database, five, database and base fours and central. What are the terms they use all the terms concerning Parkinson and cognitive based intervention? As you can see the appendix, we can see the appendix in the appendix. They put the search strategy in details what are the search terms and what was the first strategy that was used also they mentioned for uh for for, for which date they arrived, they inspect 6 November 2014. Again the last date of uh articles that are included no limitation of publication. This whole ado the screening class, one reviewers performed initial eligibility screening by assessing that and abstract of all the results. Following initial screenings, two independent reviewers, they review the full texting as you can see here and H A limitation and one reviewer did did the title and abstract Analys at least two reviewers to get to, to do with the target and abstract screening. But and no, they mentioned a two step first target and a screening. Then two independent reviewers do the full version screening after the inclusion and the three main parts, first inclusion exclusion criteria. Second part, it's about search strategy and include and selection. Third part, it's concerning data extraction form, data extraction form and coding. Uh The two independent reviewers coded each outcome measure into one cognitive or behavior domain based on the cation provided intact attack. Uh OK. OK. They didn't mention it because they put it uh in the table. No, the table E one outcomes but they mentioned and know what are the information that was used in the data? Had a risk of. OK. I will start on you concerning the next module and concerning risk of risk of bias and study appraisal uh today and tomorrow just I will show you this prisma flow chart will go to show us how they select their articles first. They identified 1400 articles in all the database. Yeah, they search for all the related to Parkinson in five database, five database, 1400 articles. They remove the duplicate, duplicate articles that are present in many database. At the same time, they remove the duplicate. The A 300 are removed. The 1100 they did the screening of the of the title and abstract of those 1100 articles, 1100 articles and uh 25 articles. 95 article, sorry, 2000 Rao based on and abstract are excluded. And we have only 95 per 9, 95 articles that are included in the next step of screening full text. Yeah, they read the full text of 95 articles. 95 articles after they uh read the full the full the full text as we can see here, 88 were ex were excluded and seven were included in the final list in the meta analysis. And in the system home improve the prisma flow chart mandatory to put it. Now we select our final list of included articles and it is mandatory to mention the reason behind the exclusion and the full text here in this, in a they mentioned the reason just a 91,000 or excluded based on part abstract. However, it's mandatory to put the reason in the full text screening step. Yeah. Here it's mandatory all and 88 wi exclude them. It's mandatory and 23 not cognitive training and nine no, no target population, one author did not provide raw data. Uh It's mandatory to mention the reason behind uh the exclusion and the full text. I will go also uh just to show you also the data extraction table, it's a bit summarized but I will show you they mention here each study number of males mean age of participant uh T TT design. OK. It's not a very well uh developed table pass. And now we can I will show you another paper here. Uh Try to come in. I will go briefly briefly about the method they mentioned it is conducted via Prisma guidelines, study selection inclusion exclusion criteria. As you can see here, they mentioned our inclusion criteria. OK? Literature search. They mentioned two reviewers independently, independently research by two reviewers database that was used. What are the keywords? We use the keywords, text to cognitive CBT CBT, behavioral therapy, psychotherapy, Parkinson, Parkinsonism, depression, parkinsonian anxiety, data extraction, data extraction. That extraction can show what uh what infos they extract from each article, they extract author year, sample size and design mean age sex situation, intervention outcome and measure of included articles. And after this, they mentioned that quality assessment that we are going to discuss it today. And so analysis also we can see here the Prisma guideline records that was identified duplicate after removing screen screen screening based on an abstract screening based on the full text included in the meta analysis was five H 10,000 screening data title and abstract. They exclude based on title and abstract 5219 or 19. They read the full text based on the full text. They exclude 12 and seven. As you can see here, it's mandatory to mention the reason why we exclude in the full text. Uh OK. Her data extraction table as you can see here first, also a name count, mean age of participant, number of control and intervention, female, male versus male intervention number and duration of sessions and outcome. How they measure the depression and they measure, they put the scale that was used to measure the outcome. They measured the depression in Hamilton scale, uh depression part anxiety and Hamilton scale anxiety part in this study, they measured the depression use uh P HQ nine ph DPH Q nine, I guess 101 and anxiety uh using the generalized anxiety score. Uh OK. It's enough for this part. I will I go to our module for today? Any questions so far? Uh Ma honey OK. OK. Concerning the module of, of today. OK. Prism of flow chart. Uh After we build our prisma flow chart, it's important to contact expert, to contact expert, to contact authors who are expert in the field. To identify any or any ongoing or missing study or to find unpublished but relevant AFA to, to, to contact expert in the domain. We still research head domain and our art. We can, we can, we can miss a data, we can miss a study. Uh ma man a published no present, present target database. Our hot films and journals man on man listed and pop our our base our database but we can some sometimes for you not to contact expert not driving mobile step number 14, search for additional references. After we run our search on database, we collect the articles and remove the FF and we didn't miss any articles we can if you for relevant studies that are mentioned, how to, to, to, to, to, to review the reference reference list of previous systematic review that are related to our systematically. It's important to see other systematic review that are related to our topic, our topic and related topic to review the reference list and to try to find any articles that are missing any article abstract, screening the end of an article last night. It's important to be done by two independent author or it's it's important to, to, to, to do, to review the reference list of previous system and preview of previous articles. Uh It's important, it's important and all articles that are included in the comment to review are sometimes the reference list, make the final selection list and to draw the flow chart. Uh We draw the flow chart of PRIZ guidelines. It's important to mention number of studies excluded based on and abstract number of full text screening, number of studies excluded after full text. And it's important to mention the number of studies excluded after full text. And also with the reason for exclusion, it's important to mention the reason of exclusion and the number of final number of study that would be included in the systematically come man hurt your head. Has this flow chart record that I identified by each database record. After removing duplicate flow chart record screening based on abstract additional exclusion. They go for full text eligibility for based on type and abstract, remove, remove uh remove, based on full text, so on full, based on full text. And the reason behind uh why we exclude them based on full text. And finally, the number of uh final articles that will include included in our uh OK. I will go briefly now in those two tables concerning the Prisma flow chart. OK. II, I'm sorry, the same uh doctor. Just give me one minute in this Prisma flow chart in this Prisma guideline or in, in this Prisma checklist. I will uh I will go briefly and summarize checklist. We came in an expanded checklist on in this checklist, improve in each part of our system. I give you what we should put first in the title section we should mention in the title. It is a systematic review and and and meta analysis as I mentioned before. And in the title we should mention it is a systematic review and meta analysis identify the report as a system in the abstract. See the Prisma. OK. If you check it for the abstract, I will go introduction, build introduction to mention the objective. Why this review? It is important in the in the context of existing not it in the methods section for first eligibility criteria and eligibility criteria, inclusion and exclusion. What are the inclusion and exclusion criteria, information sources search strategy. What was the database that are used? And what are the keyword search strategy? What was the database that were used? And what are the keywords and the register and the register on prospero selection process? How the selection process was done. It's important to mention that uh how many reviewers screened each record and each report, whether they worked independently or no uh detail of automation was used in the process. You know, now we use dry and we should mention, you know, we use Ryan, how we use dry. And it is uh it is made by two independent author. It's it's important they mentioned the name and the name of of of each author uh screening of the abstract authors la uh screening of the full text. Uh They mentioned it's, it's they, they worked independently. It's important to mention all those details. It's it's mandatory to mention them data collection. What was the information uh that are collected during uh using beta data extraction form? And finally, and finally, data items, all the outcomes to list and define all the outcomes to list and define all the variables. Today, we are going to discuss about the risk of bias and quality study effect measure of statistics, statistics, the method section about the results section, how we should uh how we should report and how we should write each section. We should write our results section. First, we should here as you can see so that to do a flow flow diagram to show our selection criteria, then study characteristic and data extraction table. Then the table about quality assessment, then the result of each individual studies on our statistical work. And um first block one, also it's important to mention the heterogeneity studies but discussion also the discussion first general interpretation. Any limitation. Uh how would any limitation in the review process to uh to discuss the implication of the result in the practice and the future research? But as you can see here, hey, there again, it's a summarized checklist. Uh If you also the be an expended checklist, it all can also in each uh in details in each part, what you should mention in each section on each part of each section, what we should mention now, after we draw our final list of include of included articles, we apply our data extraction form in person. And at least we should it should be done by two independent reviewers to extract independently the data from each article. Close attention to the data extraction process will yield an an initial understanding and description of the sure characteristics of the body of evidence and will pave the way for di and interactive process of sentences to follow. It is important, it is it is it is very important that extraction for it is one of the very dedicate and very specific step in the in the meta analysis. Because any uh any any fold in this data extraction for all the information, all the numbers, statistics, meta analysis. No, OK. I am not able to do this. Each individual that all the details, nothing is missing. Uh All the information are very organized. Also uh I hate this carefully convert all the data using the same unit and uh uh mean age ma she had to be used, she had to be, she have to, we should convert them all to the same unit. Uh We need to convert each one in D OK. Now how uh how, how can a step uh it is an important step in the meta analysis? He the quality assessment and risk of. Yeah. No, no. Uh after we include, after we, we decide what are the included studies in our meta analysis and systematic review. OK. Now for a first the screening of start and abstract screening of the full text, we define list, the final list of included articles, hi articles, but that one that extraction. So that's not that ex extraction. Sorry, we expect the main information from them. Now, we need to evaluate the study quality or the quality of those study and to assess what were the bias that uh are present in those articles. Now, we need to assess the quality of each article that will be included in, in our systematically. L is a normal heck first she is normal he analysis. Our general results evidence accurate. Just random papers run low quality result. It's a good result. Yeah, it is a 12 meal a or a not much is more sensitivity analysis analysis. It's done to based on the quality of the study and I finish the sensitivity analysis to general result. Then after general result, all I want to assess, I want to get the results of only high quality study and, and the ana ham study trip to general meta analysis to a mean difference. 0.2. Uh by now a high quality study and a significant mean difference. And I can have a significant so be no significant sensitivity analysis. And when I include only the high quality studies, we didn't uh we, we lost our significant result, our significant result, the result, no significant evidence. As I mentioned before, evidence and results should be interpreted in the light of the quality of the included study. Sensitivity analysis. Our me sensitivity analysis, the quality of the research and how a study has been conducted. Uh quality of the study quality of the study is the methodology study in a what were the inclusion criteria? Comparability? Uh Selection is to measure the outcome. We need to assess each paper based on the study design. Pay for control and to pay for the specific to pay for specific tool, pay for no clinical trials. Also specific tool pay for clinical trial, also specific tool. Hello, Robins Robins one Robbins two clinical trials. Uh now concerning the observational study and I usually uh uh Rob and Newcastle Ottawa scale. Uh it's an easy scale clear scale. The Newcastle Ottawa Scale here for observational studies and that version in Newcastle Ottawa scale version for cohort version for case control and version for cross sectional. There are very similar difference by usually it's a score and a highly highly scale criteria. By applying those criteria. A total score had a total score become uh maximum by nine stars. Over nine was an 8/9. It's high quality 5/9. It's a moderate quality 3/9. It's a quality score out of no more quality and high quality. Uh This is scale, this is scale uh based a sh a uh what what what examined Hida scale had scale examined three main parts, our sheet selection of participant. Uh He judge how the selection of participant was done in this article based on the part with comparability, a cases will control similar one or two stars based on the outcome and exposure, how they measure the outcome and that can accurate. And the outcome and statistics can be up to three point selection domain. They select a participant comparability domain outcome domain, the domains maximum studies that achieve a score of nine stars are judged to be at low risk of bias. OK. OK. And a high quality, high quality and a low risk of bias in study kill ma risk of bias, kill mccain, high quality risk of bias. OK. We, we have many type of ways we have information bias, we have recall bias. We have selection bias. We have many type of ways this type of study type of the design in each how ma a selection bias, a selection bias man. And I know how I select my participant. He uh ma ma uh effect masal effect of pollution and lung cancer. My participant ct scan or lung lung uh lung disease, lung disease. I have a selection because uh because I select out uh one type of bi bi and case control. Even in two years or three years, the type of bias kill, kill fashion HH PMA, low risk of bias and height. What uh Abel example, one of the scale had the Newcastle Ottawa scale assessment tool for case control. OK. It's yeah. Uh Each question should be given one star and maximum of two stars can be given in the comparability domain. Yeah, see as I mentioned an atic domain, it's a selection domain. Comparability domain and exposure domain. The selection domain. How would they select their participant? We should judge to give 1234, maximum four stars in this domain based all how they select their participant and four question but still go to apply them on each article. Yeah, remember. OK. Concerning the cases is the case definition. Adequate. Yes. With independent validation, we select all the cases with depression when and psychiatrist it is a very case definition. Adequate. One star, depression, one we have to zero. It is there are representative, depression, depression, which all the preference may be made with depression. Bi selection bias, 01 star how they selected the control is a control care hospital is not not mentioned. They should mention and control were from were recruited from community community and the one star definition of control. They should section the whole article to me section. They mentioned the meal section and the cases were masal ands and physician s based the evidence criteria. Not one point. There are represent case one point control can on the community. The hospital uh came in one point they should mention under control. There were free from any illness or from any disease. Uh depression controlled were recruited from community and and there were they were free from any depression uh for diagnosis before the study before starts on certain comparability cases will control a case control. OK. Cases will control a match. Cases will control the social demographics criteria. In fact, when I need to compare to, let me mafi, the only difference between them was and AMBI y uh I want to start imaon the effect of smoking on lung cancer, the cases was control. I hope it s and smoking will age the bias and our results because a big effect of confounder the whole age with the only difference who will outcome I need to the exposure, I need to measure very similar and female and the control very similar BM I very similar age. High power similar. The only difference here and how did they smoke? How they didn't smoke? Yeah, power at similar. And the only difference it is the exposure that I need to stop concerning the exposure and outcome if the exposure are measured and if, if I assess a cigarette smoking, I should interviews I just randomly uh with this interview record and IP in the hospital, same method was used in the cases. Uncontrolled cases controls the same protocol was used for cases under control rate for both the group. Hello, an example to be to study the study, all the articles that are included for each article we kill domain for S and the selection domain, comparability domain and outcome domain. As you can see here had a total score will power at the high low quality of uh that's just like two words, prepare the data for the analysis. We should collect all the data extraction form, get information the Excel sheet uh and how to do the analysis analysis. Qualitative as descriptive, quantitative meta analysis, description summary, our qualitative description, quantitative summary, quantitative for delo to me analysis uh analysis. It's important paragraph and the result we should mention and how they how we select our paper. And a and after title and abstract screening and after full text screening uh to summarize the main sociodemographics criteria of the included articles and a mean age. The articles, a total percentage II would mean of total percentage of females. Yeah. And the main characteristics if the meta meta analysis is not feasible. So I thought we should each article, each article uh to mention to to summarize each article results to direction and size of effect and consistency across studies. Before I thank you the analysis for on statistics but our results a small description, a small quantitative summary after removing duplicate at the and one OK. 1000 were are removed based on abstract 25 95% in the full text. Masan uh ma come to and provide information. He also ma w did not respond. Uh ma ma session lengths, muscle between 30 90 minutes total uh training time. Can addition you a brief summary, brief descriptive summary then by the descriptive A statistics, a review to describe narratively, qualitatively and uh the results, but you a descriptive qualitative summary about our process. And thank you for today. Any question? Yeah. OK. Let's we just a minute like this. Uh Maria, yes, don't worry, we are going to send uh also all the references uh that I use them. Also, this is my email, any information or any any any questions you can ask me and send me an email. Uh Don't worry Christopher, I will send the documents that I use them uh as you can find them all in a website and you can find them by Google, but I will send them all in the quality assessment can just you can put on some quality assessment, you can put new auto quality assessment pool on Google or you can fight them. Ok? I will send to you network. Thank you very much guys and have a safe and good evening and see you tomorrow. Oh, yes, see you Sunday and have be safe for Sunday. Good night.