Event

Multi-arm multi-stage trials improve efficiency by allowing promising treatments to be selected at interim analyses. But once selection is driven by favourable interim results, ordinary treatment-effect estimates can be too optimistic. In this talk, I will discuss penalised likelihood estimation after treatment selection in MAMS trials. The method places the usual maximum likelihood estimator and the fully conditional maximum likelihood estimator on a single tuning-parameter path, creating a continuum between ignoring selection and fully conditioning on it.

Andrew Martin is the inaugural Professor of Innovative Clinical Trials at the University of Queensland, where he leads the biostatistics group within the cLinical Trials cApability program, ULTRA. He is also Honorary Professor in Biostatistics at the NHMRC Clinical Trials Centre, University of Sydney. Before academia, Andrew held senior biostatistics roles at Pfizer and Roche. His work spans clinical-trial methodology and the design, conduct, analysis and reporting of investigator-led trials, with international impact on policy, practice and research.

This seminar will be held via Zoom and will be recorded.

Date and time:

Wednesday 29 July, 1-2pm AEST

Zoom:

Please click this URL to start or join. https://monash.zoom.us/j/86965965236?pwd=2BTCs4QvlF1zR3hgROQzKaqbjvTfBv.1
Or, go to https://monash.zoom.us/join and enter meeting ID: 869 6596 5236 and passcode: 013223