PhUSE EU Connect 2018 took place in Germany’s financial capital Frankfurt, 4th - 7th November and brought together a range of experts to tackle the most pressing issues facing statistical programmers today. The agenda was superb with 143 presentations in 16 different streams and nearly 30 posters. This year’s event theme ‘Future Forward’ did not disappoint and there were some very thought-provoking talks on the drug development industry's challenges and what we can do in the future to meet these challenges. Additional hot topics were: Analytical Risk Based Monitoring, Machine Learning, and Data Standards and Governance. We found this year's event informative and well attended.
In this blog, we share the contributed posters and presentations from our Statistical Programmers and summarize some of the particular highlights from the sessions and posters that our team members attended.
Contributed Sessions Featuring Cytel Programmers
Cytel statistical programmers once again made important contributions at PhUSE EU Connect:
Affairs in Medical Affairs – Poster
The “CDISC Stupidario” (the CDISC Nonsense) – Poster
SDTM Legacy Data Conversion – Paper
Laura Phelan & Angelo Tinazzi
An Efficient Way for Statistical Review Using the RShiny Application – Presentation
Adarsh Nagare & Sameer Bamnote
The ‘Unveiled’ Secrets of ‘Windows Batch Scripting’ – Paper
Nicolas Rouille & Laura Phelan
Adverse Event Data – gRaphically – Presentation
Lina Rajput & Prajakta Chitale
Click the button below to access their slides for a limited time.
Serious Adverse Events Identification: A Machine Learning Approach
This session was of interest because Serious Adverse Event (SAE) reporting is a critical component of patient care and drug safety profile development during a clinical trial. The presentation compared various machine learning classification models to identify SAEs, utilizing a data set of more than 1 million adverse event records from 1,818 completed trials on Medidata's platform.
Adarsh found the idea of usual neural networks for analysis of SAEs with a large amount of data has the advantage of needing less prior modeling and it had high performance rates compared with other more onerous model-based approaches.
Automating an Extended SDTM Mapping in Clinical Trials Using Machine Learning
Nassim Sleiman, Associate Director, Statistical Programming, attended this session with the view to learning how machine learning could be used to map data.
The Food and Drug Administration mandates mapping of all clinical trial data sets to the Standard Data Tabulation Model (SDTM). The information in the presentation delivered a concrete application of how machine learning can be used to simplify and automate SDTM mapping.
Key take away
Unfortunately SDTM mapping is a manual and time-consuming activity. The session demonstrated that machine learning can be used to standardize legacy data mapping to SDTM format. The advantage of doing this internally could speed up this activity and allow programmers to focus on the non-standard data collected in a study.
From Machine-readable CDISC Standard Specifications to the e-Protocol
Angelo Tinazzi, Director Statistical Programming attended this session to learn more about the projects his friend Jozef Aerts, a well-known guru of CDISC, was working on.
CDISC standards are published as PDF/HTML documents, so not machine- readable. In this presentation, he discussed a number of projects where he is currently involved; all these projects have the objectives of improving the way we use standards and how, by better using standards and available technology. One day, perhaps we may see Alexa for SDTM.
The paper also won the prize as best paper in the Standards Implementations stream.
Key take away
Jozef is always years ahead of the industry and brings a very futuristic perspective. It is always very engaging to hear some of his ideas, but his projects are also very practical and Angelo is very interested in an ongoing project at PhUSE: SDRG in XML.
Meet the team at PhUSE US Connect 2019
We are heading to PhUSE US Connect 2019 in Baltimore, February 24-27. Why not arrange a meeting with us to talk about any projects you're working on? Schedule an expert consultation at PhUSE US or stop by Cytel's booth to speak with one of our expert Statistical Programmers.