PhD Studentship in Epidemiology and Statistics/Clinical Trials Unit, Warwick Medical School

Project: Adaptive designs for a clinical trial to compare treatments for ankle sprain

A recent innovation in clinical trial design is the use of adaptive designs, in which the data from the trial are analysed before the end of the trial at one or more interim analyses. The results of these analyses can then be used to modify or adapt some element of the design of the remainder of the trial, including for example the chance to stop the study early, to select or refine the treatment(s) or dose(s) being assessed or to modify the sample size for the remainder of the trial. The opportunity for the final trial design to reflect the emerging data can lead to improved efficiency, but can also lead to a number of statistical and logistical challenges. 

Of much recent interest has been the development of adaptive seamless designs (ASDs) that allow for selection of the most promising treatment(s) at one or more interim analyses. This enables less promising treatments to be dropped, allowing the conclusion of the trial to be reached more rapidly without a loss in scientific integrity. 

This project will explore the use of Adaptive Seamless Designs in the setting of a trial of treatments for ankle sprain. The application will be based on the CAST study conducted by Professor Sallie Lamb at Warwick Clinical Trials Unit. This trial compared recovery times for four different methods of mechanical supports for the ankle following a sprain using a conventional fixed sample size design. This project will involve the retrospective evaluation of the use of an ASD for this trial. This evaluation will include the quantification of any savings in total sample size, assessment of the logistical requirements for implementation of such a design and extensions to the existing methodology to ensure that it is applicable in this setting. 

A particular methodological challenge will be the use of an interval-censored time-to-event endpoint. When a trial is conducted in a number of stages, this can lead to data from patients recruited in one stage only becoming available at a later stage, and hence to correlation between stage-wise test statistics that must be allowed for in any final analysis. The project will use a range of analytic and computational techniques, including data simulation, to evaluate the statistical properties of any methods proposed. 

Supervisors: 
Professor Nigel Stallard, Professor of Medical Statistics and Head, Statistics and Epidemiology and Professor Sallie Lamb, Professor of Rehabilitation and Director of Warwick Clinical Trials Unit, Division of Health Sciences, Warwick Medical School, University of Warwick 

Eligibility criteria and details of award: 
This studentship is fully funded by a Medical Research Council (MRC) Capacity Building Studentship in Mathematics and Statistics and Warwick Medical School. It includes full tuition fees for a UK/EU student and an annual stipend at the standard MRC rate (currently £13,590 per year) for three years. 

The student will be based at Warwick Medical School (WMS) and will enrol on the WMS postgraduate student training programme. They will be encouraged to interact with other WMS statisticians and to participate in meetings of the WMS Statistics and Epidemiology Group, and will also benefit from relevant seminars and training in the Department of Statistics. Registration will initially be for an MPhil degree. Upgrade to PhD registration will follow presentation of a satisfactory report at the end of the first year. 

Candidates should have a good honours degree (2.1 or above) and/or higher degree including a substantial component of mathematics and/or statistics and must satisfy the MRC Studentship eligibility requirements (see http://www.mrc.ac.uk/Fundingopportunities/Applicanthandbook/Studentships/Eligibility/index.htm). An interest in clinical trials is also essential. 

Project duration: 3 years 

Application Deadline: 30 September 2012

Applications can be made via http://www2.warwick.ac.uk/services/its/servicessupport/studentadmin/postgrad/

You may leave the list at any time by sending the command

SIGNOFF allstat

to [log in to unmask], leaving the subject line blank.