1. Applied Bayesian Statistics in Drug Development, Bruno Boulanger, Arlenda.
Abstract: Over the last 15 years we have observed a growing use and application of Bayesian statistics in Biopharmaceutical research, development, manufacturing and sales. Whilst Bayesian statistics got mainly recognized in the industry through the concept of adaptive and innovative clinical trials, thanks to the availability of power computer and programming languages, Bayesian statistics are now more and more used in all areas from discovery to manufacturing and health economics for decision making.
The objective of this short course will be to introduce progressively the concepts common in Bayesian statistics, such as prior, posterior and predictive distributions through a wide variety of examples covering all the range of activities. In addition programming solutions in SAS Proc MCMC or STAN will be shown for each example to exemplify the ease and power of such languages.
In the light of the recent reproducibility crisis and questions raised about the unreasonable use of p-values, we also highlight the fact that the proper use of priors, posteriors and predictive distributions for modeling are way more relevant for the type of reality, constraints and decision making that is needed in biopharmaceutical industry. We will also show that it opens new opportunities for the patients, such as in orphan disease or quality of product.
Required: Basic SAS or STAN programming skills.
2. Clinical Graphs Using SAS, Sanjay Matange, SAS Institute Inc.
Abstract: Graphs used in the Health and Life Sciences domain and Clinical Research have special requirements for display of data in a clear and concise manner including raw data and derived statistics. Data needs to be displayed by treatment, visit and other classifiers along with related information such as Subjects at Risk aligned with the horizontal or vertical axis.
This presentation will introduce you to the key concepts of the SGPLOT and SGPANEL procedures including the layering of plot statements to create the clinical graph. We will review the process and features by creating graphs commonly used in the Pharmaceutical industry. These include Mean Change in QTc by Visit, Distribution of ASAT by Time and Treatment, Survival Plot, Forest Plots, Adverse Event Timeline, Waterfall Chart for Change in Tumor Size, Distribution of Maximum LFT Values by Treatment, Swimmer Plot, Panel of LFT Shifts by Treatment and Immunology Profile.
Audience: Graph programmers
Required: Basic SAS programming skills.