Presenting our methodology research at ISCB 2016

01 Sep 2016

This week, the MRC Clinical Trials Unit at UCL presented some of our latest methodological research at the International Society for Clinical Biostatistics (ISCB) annual conference in Birmingham, UK.

The ISCB conference was attended by over 600 biostatisticians from around the world, with sessions spanning important areas of biostatistics. Keynote talks came from Diego Kuonen on big data and data science in pharmaceutical development, and David Spiegelhalter, who spoke about sex statistics, and is used to this talk being a platform to make statistics more interesting; in this setting the challenge was to make sex more interesting!

Researchers from the Unit presented work on:

  • adaptive trial design
  • cluster-randomised trial design
  • network meta-analysis
  • missing data and sensitivity analysis
  • longitudinal data analysis and non-inferiority trials

as well as a presentation (and contribution to a panel) at the mini-symposium on designing trials in smaller populations. Slides and posters presenting our work are below, for those interested.

The Unit is grateful to the organisers for putting together such a stimulating programme.





How do you design trials in smaller populations? A framework

Tim Morris

Mini-symposium 1: Statistical methods in rare diseases and small populations

Design and implementation of trials for biomarker-driven questions

Matthew Sydes

Invited paper: Adaptive designs for oncology clinical trials

Reference-based sensitivity analysis for clinical trials with missing data via multiple imputation: Investigation of Rubin's variance estimate

Suzie Cro

Contributed session: Missing data

Inconsistency in network meta-analysis: a critical comparison of methods

Suzanne Freeman

Contributed session: Meta-analysis and medical decision making

Accounting for uncertainty in the timing of seroconversion in combined models for pre- and posttreatment CD4 counts in HIV-patients

Oliver Stirrup

Contributed session: Longitudinal data analysis

Optimal design of cluster randomised trials with baseline data

Andrew Copas

Contributed session: Cluster randomised and stepped wedge trials

Non-inferiority trials: inconsistent guidelines, inconsistent reporting? A review of selected journals

Sunita Rehal

Poster session

Further information: