Clinical Trials Visualizer Christina Bryant, Pranay Madan, Akhila Nair, Daniel Sylngstad The Peter F. Drucker and Masatoshi Ito Graduate School of Management and Claremont Graduate University Abstract Dashboarding Dashboard Views Our tool was created by merging databases to create a visualization dashboard for • The Clinical Trial Visualizer was built as a tool for rapid trend analysis using users interested in rapid trend analysis of clinical trials. The tool utilizes the Clinical Tableau 9.0 Trial Database, the Medical Subject Headings thesaurus, and Agency RSS Newsfeed. • Users can choose from a variety of filters including year, disease, sponsor, phase, The tool was created through Tableau to provide users with applicable visualization type of intervention, study type, etc. options to convey relevant trends and clinical trial topics. The dashboards are • Custom-made views have been built to ensure maximum usability for multiple particularly relevant for those interested in the participation options of studies or for type of users those interested in on going clinical trial and sponsors conducting them. The Applications complete tool gives users a look at diseases that are targeted by clinical trials with information on quantity, companies that conduct them, the location of these trials, the funding sources, the type of agencies running the trials, the phase of the trial, and much more to give users an effective dashboard for identifying trends to answer questions they may have regarding clinical trials. The user also have the ability to Analyze Current Trends in Clinical Research Identify Gaps in Clinical Research The dashboard can be used to analyze the current landscape of any disease. We are trying to identify gaps by analyzing trends in clinical trial data by focusing on certain key therapeutic areas. filter and cross-slice the underlying data using multiple parameters. Objectives Our group set out to combine clinical trial databases with a visualization platform to create a unique dashboard that acts as tool for displaying clinical trial trends. We felt this was important as there is a great amount of federal and industry spending on trials. The problem we see: Trend Analysis There is no tool currently available for rapid trend analysis Time Consuming Traditional techniques are time consuming and difficult Loss of Insights There are insights that can lead saving time and costs if identified Our Approach Relational Database Structure Some questions that can be asked are: • How many companies are conducting trials? • What is the status of these trials ? • By when can we expect results? • Are the trials largely government funded? An example trend that can be seen is: • Even though certain diseases have higher economic burden, there seem to be inconsistencies in the prioritization of clinical trials Individuals Can Easily Locate Trials and Enroll And for much more… Patients can easily visualize trials and make informed decisions about their health or the medication progress We can easily overlay more data and perform various other analysis References 1. Available from: http://www.clinicaltrials.gov/ct2/about-studies/learn 2. http://www.nlm.nih.gov/services/ctphases.html 3. National Institutes for Health, editor. Funding Facts - NIH Research [Internet]. [cited 2014 Apr 6]. Available from: http://report.nih.gov/fundingfacts/fundingfacts.aspx. 4. Roumiantseva D, Carini S, Sim I, Wagner TH. Sponsorship and design characteristics of trials registered in ClinicalTrials.gov. Contemp Clin Trials 2013;34(2):348–55. 5. Goswami ND, Pfeiffer CD, Horton JR, Chiswell K, Tasneem A, Tsalik EL. The State of Infectious Diseases Clinical Trials: A Systematic Review of ClinicalTrials.gov. PLoS • Initial data manipulation was performed in Microsoft Azure’s machine learning platform, conversions resulted in missing and flawed data which were then manually processed. • The data was later queried and transferred into Tableau • Database joins were performed to capture as much data as possible for the variables we felt were significant for user display ONE 2013;8(10):e77086. Acknowledgement We would like to thank Prof.Hilton and Prof. Tchalian for their constant support and guidance through out the project. We would also like to thank Sara and Sean for their help.