 Hello everyone and welcome to the webinar of the Rangiplos. This is an arsigny tool for presenting multiple outcome network victimized findings. Before I start, I would like to acknowledge the land of Toronto, which is known as Carunto, and is the traditional territory of many groups and nations. We would like to honour the elders, acknowledged skippers of this land, and acknowledge the harms of the past and present, and are grateful to have the opportunity to work on this land. Network meta-analysis allows the synthesis of data from a network of studies, facilitating the simultaneous comparison of many interventions for managing a specific clinical condition. An important advantage of network meta-analysis is that it can rank the effectiveness and safety of all the relevant interventions for a clinical condition. For example, in this systematic review with network meta-analysis, for acute diarrhea and gastroenteritis in children, we included 174 studies comprising approximately 32,000 children and 27 different interventions. We explored five outcomes, as you can see in those five different network plots, with a primary outcome, the diarrhea duration. So as you can see, its outcome builds a different network geometry, and its network may include different number of studies, different number of interventions, and a different number of patients in total. In this link table, we can see the results that we obtained from two outcomes, the diarrhea duration and stool frequency at day two. So we can see the driven effects and the confidence intervals for all the possible pairwise comparisons that we obtained from this network meta-analysis for the specific two outcomes. However, in this systematic review, we assessed five outcomes in total. So different link tables are obtained for each outcome, and therefore all this magnitude of results should be summarized in order to inform medical decision-making. Now the question is, how can we communicate all these network meta-analysis results in order to make an informed decision about what is the best or what are the best available treatments? Another example that we can see here is about the assessment of anti-apileptic drugs during pregnancy. In this systematic review with network meta-analysis, we included four congenital malformations and prenatal outcomes, and another five neurological developmental outcomes that makes in total nine outcomes that we included in this systematic review. And those nine outcomes included more than 100 studies, with more than 60,000 patients in total. So the question now is, how can we combine all this evidence and communicate best these results to the decision-makers? There are numerous ways to present network meta-analysis results, and this variation can also make the interpretation of these results difficult. We may have forest plots, we may have bubble plots, we may have rankograms, sucra plots, scattered plots. So we have all these different ways to present the network meta-analysis results and for its outcome separately. So this can become more challenging as the number of interventions and outcomes increases. In this paper, we presented a simple graphical approach to improve the presentation of results from the treatment ranking across multiple outcomes in network meta-analysis. So the ranking plot was first developed and published in 2016, and it allows the fast identification of the most likely past and the most likely worst interventions with respect to their effectiveness and safety in the outcomes that we examine. It can also identify interventions that have not been studied for a specific outcome in the past. So the ranking plot has already been cited in more than 120 publications, according to Google Scholar, and has been used in multiple areas, such as diuretrics, pediatrics, neurology, oncology, and so on. It is an easy way to interpret and actually read the results and communicate the results with the decision-makers. In this figure, for example, we can see the hierarchy for 15 treatments. So as you can see, we have 15 different radii, and with respect to the five outcomes, as you can see, we have five outcomes here. So its outcome is represented by a different circle, and we have presented the hierarchy of all these treatments across all the five outcomes with respect to their efficacy and safety. So the outer concentric circle represents the vomiting outcome, whereas the inner most concentric circle represents the arrhythmia outcome. Now its sector, as you can see, is colored according to the ranking of the particular treatment at the corresponding outcome, and the scale consists of the transformation of three colors. We have red, yellow, and green, ranging from the lowest to the highest value of the ranking statistic. Now in this plot, we can interpret in its sector only the color, so the area of the sector does not convey any intervention, any information. Now, uncolored sectors suggest that data on the outcome were not reported for the particular treatment in the trials that were included in the network meta-analysis. For example, here, the ranking statistic, as we can see, for granitron plus dexamethasone, we can see that the ranking statistic for the vomiting outcome, the sucra is equal to 84%. And it's actually higher than the nausea outcome where sucra is actually equal to 81%. So granitron plus dexamethasone is most likely the most effective for preventing nausea, vomiting, and PONV, but it performs the worst for arrhythmia. We developed an interactive tool, an interactive web application to produce the ranking plot when the study-level data for its outcome are available and without performing the network meta-analysis outside the tool. The ranking plot for signing up can be used in any type of discipline and disease using a network meta-analysis of multiple interventions compared in different studies obtained from any type of review, even if we have systematic reviews, rapid reviews, overviews of reviews, and so on. Now the plot has become popular and is currently part of other R-packages and R-signing apps, such as, for example, it is already part of the VSCOOP R-package, which provides visual ways of presenting the results on component network meta-analysis. Therefore, we developed the ranking plot R-signing app to help generate the plot easily and make it easily accessible to the users. The application runs directly the analysis and produces the ranking plot in one software, and we are going to explain this in the next few slides. Now, this is how the ranking plot web application looks like. So let us consider that we upload the data, which should be in an Excel file, and this Excel file will have different tabs, and its tab represents a different outcome data set. For example, here we have the contrast level data for the included studies, and we also have five different outcomes. So then we select the formats of the data, whether we present those data in arm level data or contrast level data, we also select the type of outcome, the effect size, and the model specifications in the next few options that we have, and then we hit submit once we make all these selections for all the different outcomes. So once we hit submit, the application runs the network meta-analysis and produces the plot. So the app also provides us with a table of the sugar results along with the plot as this table has been obtained from the analysis across all the outcomes. And of course, the user has the ability to download the plots in a PNZ format. It is also possible to adjust the font size of the values, the outcome names, according to their preference. And importantly, I would like to note here that no data is being collected during the process. Now, for example, using the outcomes for our systematic review with network meta-analysis about the comparative effectiveness and the risk of Britain birth of local treatments for cervical cancer by Athanasio Etal, and published in the Lancet of oncology in 2022. These data actually were contrast level data across all the studied outcomes. And once we actually uploaded those data on the ranking plots tool, we obtained this ranking plot as we can see here. So we can see the relevant sugar values where we present the hierarchy of the interventions across all the five outcomes, Britain birth and treatment failure, high grade treatment failure and so on. In this plot, we can see easily that CKC and laser colonization rank among the best interventions across all outcomes apart from Britain birth. In another example, in another systematic review, as an example by Trico Etal, using both continuous and binary arm level data across all these different outcomes in this excel file, we also can present the results altogether in one single plot, as you can see here. So we have all these multiple outcomes and all these multiple interventions in this network meta-analysis. Similarly, when we have time-to-event data such as in this example with COPD and exacerbations, we can easily obtain a ranking plot. So the type of data included in its outcome can be binary, continuous time-to-event, survival data. Despite all the outcome type of data, all of these can be modeled in this application in order to produce the ranking plot. Overall, the ranking plot can be used to quickly recognize what are the most likely best and most likely worst interventions with respect to their effectiveness and safety in any given outcome. It can also identify treatments that have not been studied for a given outcome and we can see those in uncolored sectors in the ranking plot. And currently, all the analysis, the ranking plot performs all the analysis in a frequent setting using the net meta R-pacodes, but we are currently extending the tool to incorporate Bayesian analysis as well as network analysis results according to the minimal clinically important differences. At this stage, I would like to acknowledge the software developers who have worked since the conception of this tool, as well as my team for their support on developing this tool. Thank you for listening this webinar.