Webinar recording and presentation slides
By Prof. Ian Marschner
Professor Ian Marschner, Professor and Co-Director of Biostatistics in the NHMRC Clinical Trials Centre presented a discussion of the statistical principles that underlie the activities of data and safety monitoring boards for clinical trials.
The webinar covered the potential problems that can arise when treatment comparisons are made repeatedly over the course of an ongoing trial; the potential for multiple analyses to inflate the chance of a false positive conclusion; the concept of group sequential statistical methods that allow valid statistical analysis of accumulating data while controlling the chance of false-positive conclusions; commonly used early stopping rules within the group sequential framework and the general concept of alpha-spending; other approaches to statistical monitoring including conditional power and Bayesian approaches.
Ian Marschner has over 30 years of experience as a biostatistician working on clinical trials and epidemiological research in cardiovascular disease, oncology and HIV/AIDS. His research also focuses on the development of new statistical methodology, particularly for designing and analysing clinical trials. He has over 100 peer-reviewed research publications and has been a Chief Investigator on research grants totalling over $20 million, including the Australian Trials Methodology Research Network (AusTriM), an NHMRC Centre of Research Excellence.