Royal Statistical Society and NIHR Statistics Group meeting in Data Monitoring Committees and Simple Adaptive/Efficient Designs

Category: Laboratory Studies

Location: The Ridge Conference Centre, University of Sheffield | Date: 19 January 2016

Organised by Dr Dawn Teare, University of Sheffield

Agenda

10am-1pm: Data Monitoring and Ethics Committees (DMECs)

Speakers: Professor Thomas Jaki and Dr Lisa Hampson, Medical and Pharmaceutical Statistics Research Unit, Lancaster University

1.00-2pm Lunch

2pm-5pm: Introducing Practical Adaptive Designs.
Speakers
Professor Steven Julious, University of Sheffield (Slides)
Professor Sue Todd, University of Reading (Slides)
Dr Kevin Carroll, KJC Statistics

5.00-5.15 Closing summary

Brief Report

The meeting opened with Professor Thomas Jaki from Lancaster University who provided a comprehensive and practical overview of the role and responsibilities of a Data and Safety Monitoring Board (DSMB). The informative lecture provided useful recommendations which included ensuring DSMB members are supportive of the trial and structuring the study team to maintain blind during the construction and review of DSMB reports. This was followed by a helpful hands-on workshop led by Dr Lisa Hampson in which delegates got to act at DSMB members and had the opportunity to put into practice some of the issues highlighted in Professor Jaki’s lecture.

In the afternoon Professor Steven Julious from University of Sheffield gave a very entertaining and enlightening lecture of his experience in designing clinical trials utilising an adaptive approach. Professor Sue Todd from Reading University, through her enthusiasm, provided a useful introduction to confirmatory seamless adaptive designs introducing delegates to the difference between combination tests, conditional error function and Group sequential approaches with plenty of references highlighted for further reading. The one day meeting finished with a practical workshop led by Professor Julious where delegates discussed the logistical and statistical issues for a variety of proposed adaptive deigns.