15th August 2019
Data-driven Decision-making in a Dynamic Landscape
On Thursday, August 15th, in Mississauga, Ontario, Bayer Canada welcomed approximately 60 attendees to the PhUSE Single Day Event. The audience included representation across the pharmaceutical industry, regulatory agencies, local and US-based CROs and other healthcare-based data scientists. There was also academic presence, with students from the University of Waterloo, the University of Toronto and McMaster University.
The theme of the event was Data-driven Decision-making in a Dynamic Landscape. The presentations covered various aspects of how pharma collaborators are accelerating drug development in the industry. Some of the more technical tools, such as use of VBA, R Shiny and R Markdown, were highlighted, displaying their automation capabilities which provide immediate data and reporting to scientists and stakeholders. We also heard from those working on the FDA Clinical Development Design (CDD) and CDISC Data Exchange Standards – both which are developing a framework to help aid decision-making. The presentations also covered the use of synthetic control data and artificial intelligence, which sparked engaging discussions on how they help drive our decision-making abilities. In addition to the great presentations, the event allowed ample time for attendees to network with colleagues across organisations.
The hosts would like to acknowledge and thank all of our sponsors, presenters and participants for making this a successful event. We look forward to the next PhUSE Single Day Event in Canada!
Automatic Safety Slide Deck Generator – Felipe Fittipaldi, Bayer
Artificial Intelligence and Healthcare – Aman Bahl, Syneos Health & Touoweg Somda, EZManip Solutions
Data-driven Decision-making Through a Clinical Development Design (CDD) Framework – Hon-Sum Ko, FDA
Tools for Data-driven Decision-making – Sally Cassells, CDISC
Using Synthetic Control Databases to Accelerate Indication-specific Safety and Efficacy Evidence – Colin Neate, Roche
Mixture Hidden Markov Models for Multiple Types of Diseases – Yidan Shi, Student, University of Waterloo
Interactive Analytics and Automated Report Generation in Public Health Reporting – Matthew Kumar, Independent Consultant
Regulatory Report Automation with R – Shimeng Huang and Yuyao Song, Roche
View the event brochure.
Photos of the day:
Ankur Mathur, Bayer, welcomed attendees at the Mississauga SDE.
Shurjeel Choudhri, Senior Vice President and Head Medical and Scientific Affairs at Bayer HealthCare, kicked off the SDE with a discussion of the exponential increase in technology and the implications for improving health outcomes.
Touoweg Somda (Magus), Founder and CEO, EZManip Solutions Corp, gave an explanation of AI and ML.
Aman Bahl, Associate Director, Statistical Programming from Syneos Health, talked about improved decision-making in healthcare by utilising growth in computational power and advanced machine learning.
Hon-Sum Ko, Medical Officer, Division of Dermatology and Dental Products of the FDA, discussed a proposed Clinical Development Design (CDD) Framework developed from a design thinking approach to provide a data-driven information model for decision-making.
Sally Cassells, Sr. Director of Data Exchange Standards and Certification in the Data Science group explained that utilising the exchange standards may facilitate and improve an automated workflow through data-driven decision-making.
Colin Neate, Associate Director of Oncology Biostatistics at Roche, gave a talk about a pilot collaboration for metastatic breast cancer using synthetic control databases of recent clinical trial data to support drug development.
Yidan Shi, student from University of Waterloo, spoke about the advantages of using continuous-time Markov chains to determine disease types from data on an aging population of Nuns.
Matthew Kumar, an independent consultant, showed the attendees how to generate dynamic visualisations using R-Shiny to effectively analyse data.
Yuyao Song and Shimeng Huang, Roche Statistical Programming and Analysis, demonstrated how to take advantage of RMarkdown and RShiny to develop static and dynamic analytical output which can be used to increase the efficiency of the developmental workflow for faster filings.
The SDE concluded with an interesting panel discussion and some Q&A time with the audience.