On Thursday 16 December, the RSS South Wales Local Group organised an event named ‘RSS South Wales: Centre for Trials Research Showcase’. Tim Pickles, NIHR Doctoral Fellow at the Centre for Trials Research (CTR) and member of the local group committee, organised the event, which showcased the wide-ranging methodological applied work of statisticians at the CTR. The CTR is a UKCRC registered clinical trials unit at Cardiff University.
CTR Director Kerry Hood chaired the event held in the Hadyn Ellis Lecture Theatre at Cardiff University.
Dr Rebecca Playle, senior lecturer in medical statistics and deputy director of statistics at CTR began with a presentation on ‘Designing Studies for individual participant data (IPD) meta-analysis’. The BRAINTRAIN study, which aimed to utilise neurofeedback from real-time MRI brain scans to improve patient outcomes in the field of alcohol dependence, anxiety, PTSD, autism and binge eating, was the basis of this work. There are differences between meta-analysis and IPD meta-analysis, which were highlighted, along with the need for IPD meta-analysis in this case. The study contained multiple different designs including feasibility trials and single arm studies, so care needed to be taken to ensure the analyses accurately reflected the effects of the neurofeedback.
Mandy Lau, Research Associate in Statistics at CTR, then presented on ‘Psychometric testing for validating a cancer symptom recognition measure’. The ABACus3 trial used a patient reported outcome measure (PROM) on about ‘awareness and beliefs about cancer symptoms’ to determine its effectiveness. This PROM had been through multiple adaptations from its original development and thus new psychometric testing was required. Analyses showed that this version functioned as two separate measures. When used in effectiveness analyses, one of these separate measures suggested the intervention was successful whilst the other showed no evidence of a difference.
Dr Charlotte Wilhelm-Benartzi, Research Fellow in Statistics at CTR, gave the audience an insight into ‘Surrogate Biomarkers and Surrogacy in novel clinical trial designs’. This methodological work has been rigorously designed but has yet to be implemented. Clinical trials validating the use of surrogate biomarkers are needed. Standard designs can be re-purposed for this used including the randomised withdrawal design, the seamless design, the repeated measures design and the multiple response design. Future simulation work will provide answers on the best trial setting for each design
Professor Adrian Mander, Professor of Medical Statistics and Director of Statistics at CTR, gave the final presentation on ‘Master Protocols and Platform Trials in Type 1 Diabetes’. Whilst there are many examples in the field of cancer research, master protocols are still a novel concept. There are lots of new terms and therefore little consistency. A master protocol will be implemented in the INNODIA network. Here, data will be collected in both observational cohorts and multiple trials with numerous different designs. The master protocol allows for operational efficiency across the network.
The presentations from this event are available in .pptx format:
INNODIA master protocol
Surrogate Biomarkers and Surrogacy in novel clinical trial designs
Designing studies for IPD meta analysis
Psychometric testing for validating a cancer symptom recognition
Watch a recording of this event.