THE PRIMARY HEALTH CARE SPECIAL INTEREST GROUP MEETING
Thursday 30th June 2016
Royal Statistical Society, 12 Errol Street, London, EC1Y 8LX 2-5pm
All welcome. No pre-registration is necessary
THEME: Internal Pilot Studies and Early Phase Designs
Speakers:
David Gillespie Cardiff University
Title: Overview of Internal Pilot Studies It is now commonplace for studies funded by the various streams of the National Institute for Health Research (NIHR) to require internal pilot studies. These involve running through all procedures that are planned for the main trial, with data from the internal pilot permitted to contribute to the final analysis. Clear progression rules are required on key outcomes thought to be important to the feasibility of the study (e.g. recruitment, retention, compliance with the study protocol), with the findings from the internal pilot considered by funders before deciding whether or not the main trial should go ahead.
As a statistician working in a trials unit, I have been involved in several internal pilot studies. I have also sat on independent committees of projects that have required these, and have peer-reviewed reports submitted to the NIHR. This talk will be based on my reflections of these three very different types of experiences I have had with internal pilot studies, focusing particularly on the types of outcomes that tend to be looked at, the different progression criteria that I have used or seen used, and communication between project team, independent committees, and funders.
Nigel Stallard University of Warwick
Title: Optimal sample sizes for pilot studies to obtain evidence of efficacy There is little consensus on the design of early phase clinical trials or pilot studies, in part due to the fact that such studies are conducted for a number of reasons. An important role of early trials in drug development and other clinical research areas is to obtain early indication of efficacy prior to committing to a larger definitive study. Considering studies with this aim, this talk will explore the problem of determination of an appropriate sample size using a Bayesian decision-theoretic perspective, leading to designs that minimise the average number of patients required per successfully identified effective therapy.
Gareth McCray Lancaster University
Title: Sample size estimation after an internal pilot in paired comparative diagnostic accuracy studies with a binary response
The sample size required to power a study to a nominal level in a paired comparative diagnostic accuracy study, i.e. studies in which the diagnostic accuracy of two testing procedures are compared relative to a gold standard, depends on the correlation between the two tests - the lower the correlation the greater the sample size required. A priori, we usually do not know the correlation between the two tests and thus cannot determine the exact sample size required. One option is to use the implied sample size for the maximal negative correlation, giving the largest possible
sample size. However, this is potentially wasteful of resources and unnecessarily burdensome on study participants as the study is likely to be overpowered. A more accurate estimate of the sample size can be determined after an internal pilot has been completed and we have some knowledge of the true value of correlation between tests. This talk discusses a sample size estimation and re-estimation method based on the maximum likelihood estimates, under an implied multinomial model, of the observed values of correlation between the two tests. The method is illustrated by comparing the accuracy of two procedures for the detection of pancreatic cancer, one
procedure using the standard battery of tests, and the other using the standard battery with the addition of a PET/CT scan all relative to the gold standard of a cell biopsy. Simulation of the proposed method illustrates its robustness under various conditions.
To be followed by Group Discussion
Further details from www.rss.org.uk
Meeting organiser: Maurice Marchant ([log in to unmask])
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|