Sample Size Determination in Clinical Trials
2 days 11-12 March
Location Reading, UK
Presenters Anne Whitehead and Patrick Kelly
Cost (6 weeks or more before start of course)
Student £360
Public sector/academic £450
Other £510
(less than 6 weeks before start of course
All participants £600
At the planning stage of a clinical trial a key question is "How many
patients do we need?" For a superiority trial the main objective is to
provide evidence of superior efficacy for the new therapy relative to the
control. This requires a sample size that ensures sufficient statistical
power to detect a clinically relevant improvement. Equivalence or
non-inferiority trials require sample sizes that establish with sufficient
confidence that the new therapy is respectively equivalent to or no worse
than the control.
This course provides a framework for sample size determination in
superiority, non-inferiority and equivalence trials. It considers a general
parametric approach applicable to binary, ordinal, time to event or
normally distributed patient responses. Comparisons are made with other
commonly used approximate and exact methods. Extension to more than two
treatments, including a factorial design, is considered. Practical aspects
such as allowance for drop-outs, protocol violations and mid-trial sample
size reviews are addressed. The course includes practical sessions
involving the use of hand calculators and software packages such as nQuery
Advisor(r) and PASS. Use of other packages is discussed.
Programme
* A general parametric approach to sample size calculations
* Binary, ordinal, time to event and normal data
* A unified approach to superiority, equivalence and
non-inferiority trials
* Parallel group and cross-over designs
* Designs with more than two treatments
* Practical and ethical considerations
* Software
For further details please contact:
Barbara Dodds
MPS Research Unit
The University of Reading
PO Box 240 Earley Gate
Reading RG6 6FN
UK
Tel: + 44 (0) 118 931 6662
Fax: + 44 (0) 118 975 3169
e-mail: [log in to unmask]
|