 Γεια σας, my name is Dadosavrenovil and today I am presenting to you another COVID-19 sign application. That is a web application for a leading meta-analysis of COVID-19 trials. That COVID-19 was created in the concept of the COVID-19 anime initiative. That is an international project which performs a leading systematic review of COVID-19 trials. COVID-19 anime was created since the early days of the COVID-19 pandemic with three main pillars. First of all, we continuously collected, critically apprised all the available evidence related to COVID-19 interventions. Then to synthesise the available evidence through meta-analysis and finally to make the results publicly available through the COVID-19 anime platform. For almost three years now, the COVID-19 anime data extraction team was identifying on a daily basis new trials through the LOVE and the COHRAN COVID-19 study registered platforms. All this continuous work has led to a huge amount of available information in our database. Up to March 2023, the COVID-19 anime database was consisted by 572 trials that were investigating 332 treatment comparisons and 63 trials investigating the efficacy or safety of 27 vaccines used against COVID-19. All the data for the different per-wise compilations in our database are pulled through meta-analysis and are summarised with forest plots that are a standard way to summarize the analytical results. These forest plots are then uploaded in the COVID-19 anime platform. The analysis takes place using some very specific settings that are defined in the COVID-19 anime protocol. For example, that we use a random effect inverse variance model to synthesise our data or that we use the restricted maximum likelihood method to estimate the heterogeneic parameter. These specific analysis settings create a consistent environment in order to do the analysis, but on the other hand it is creating our results to be known and amenable. And we know that with such a huge amount of available data, spatial analysis topics will inevitably arise. For example, we know that across all the trials with hospitalised patients, the risk for serious adverse events is only 8%. This refers to the trials in the COVID-19 anime database. So this implies that serious adverse events is a rare outcome and we know that in such cases the inverse variance model can be problematic. In fact, from the literals we know that in such cases alternative models should be used instead of the inverse variance. Beyond the different statistical issues that might arise, we were in the situation where stakeholders and different guideline developers that were using the COVID-19 anime platform wanted additionally to investigate our data. They wanted, for example, to investigate the impact of the trial characteristics in the results and they wanted to produce their preferred evidence summaries. Overall though, accommodating the spatial needs through solely the COVID-19 anime platform can be challenging as it will render the output in the platform very confusing for the different users. To accommodate the spatial needs we launched and made freely available the Metacovid application. That is an R-sign application for which the calculations are mostly based to the metaphor R-package. This application allows for a real-time analysis of the COVID-19 trials for both treatments and vaccines based on the COVID-19 anime database. Through the application, the users can modify the analysis in several different ways. For example, they can modify the type of the meta-analytical model, they can modify the methods for estimating the heterogeneity, and they have also some other options. The output of the application is again a forest plot that the user can download and use for their own personal needs or their own scientific projects. Let's have a look now and see how the users can use Metacovid in practice. So once the users enter in the application, they are introduced to the home page that we usually use in order to do some announcements related to the functionality of the application or the COVID-19 anime database. The two basic analysis options in Metacovid are connected to the menus related to the COVID-19 treatments or the COVID-19 vaccines. I will focus today on the COVID-19 treatments because the analysis options in the COVID-19 vaccines are almost the same like the treatments. So here are the different options that are available in this panel. The second thing that the users are asked to do here is to make a choice of the treatment comparison of interest. By clicking this menu we see that we have different and several available comparisons which are organized in terms of their treatment type that is stated here with bold letters at the top. Today I'm going to choose the comparison between corticosteroids versus either standard care or placebo. One thing that the users are asked to do here is to make a choice for an outcome of interest. In total Metacovid can contain data for in total 13 outcomes of interest and today I'm going to choose the outcome of mortality at day 28 as I think that this is a very interesting outcome. So by choosing the treatment comparison and the outcome of interest, the Metacovid application is producing forest put based on those two choices. Here we can see that this is a forest put that contains all the basic elements that we are used to finding forest put that summarizing analysis results. For example, we see the study specific data, and of course we see at the bottom the overall summary result printed in the form of a diamond. However, the forest put in the Metacovid application is also enhanced with some additional characteristics. For example, we can see that the studies are also grouped according to the severity of the included patients. And we have the subgroup of the studies that contain patients of mixed severity and the subgroup of patients, studies that contain patients of critical severity. A subgroup analysis takes place within those subgroups and the subgroup specific diamond is also printed within the forest. Additionally, all Metacovid forest put contain a risk of bias assessment that is related to each one of the studies that can be found in our forest put. The assessment of the risk of bias is done by a team of experts working with the COVID-19 project and using the Cochrane Rob2 tool. Additionally, we can see that at the bottom of the forest put, we have also some reports about the heterogeneity that refers to the overall analysis. And whenever we have more than one subgroups in the forest put, we have also the test for the subgroup differences posted at the bottom. Until now, the analysis is done only by choosing a treatment comparison and an outcome of interest and by retaining the rest of the options in the default form as they are in predefined. The users can modify these predefined options and they can bring the analysis in the form that they want. For example, they can change the heterogeneity estimator and use another one. For example, they can use the Sydney German estimate in order to do this analysis and you can see that immediately the output for the heterogeneity report changes. Personally, they can change, for example, the subgroup criterion and deviate from the default choice of grouping the studies according to their severity. Here, I will choose, for example, to do the analysis based on the conflict of interest status of its study and you can see that the forest put again updates and groups the studies according to the reported conflict of interest status. You can see that now we have a diamond for each respective subgroup. In this forest put, we can see additionally that we have some studies reporting high risk of bias, and of course some users may want to exclude those studies from the analysis. This can be done easily by clicking the respective pattern and we can see that again the forest put immediately updates. There are several other options that we demonstrate another one. We can use MetaCovid in order to deviate from the default choice of the inverse variance model and use some alternative models by clicking this button that refers to a sensitivity analysis using different models. When we do that, the results are summarized in the form of a table and we can see that here we have the results in terms of alternative models such as the Mantelhansel or the PTO method. So this is how the MetaCovid looks like and how it can be used in practice. I'm going to go back to the presentation now and summarize. So as I told you before MetaCovid is an application that allows all the users of the COVID enemy platform to interact with our database. The primary analysis can be modified in many different ways through the application and the users can download the forest tools and in order to use them for their own projects. Of course there are disadvantages on the use of the application, for example the users cannot upload their own data. Hence of course all these user-friendly environment does not guarantee the proper interpretation of the findings. Overall MetaCovid appeared to be a helpful application. Many stakeholders and catlin developers have declared engagement with the application and this translates into numbers because according to the Matomo tracker the application has an average of 140 unique users per month. Before I close I need to mention that some more details can be found in our open access publication at research methods. Thank you very much for your patience, I am happy to take your questions now.