Dear All
I am designing a study to compare the mean costs of 2 treatments and
have decided to use an adaptive design as I have no data on the expected
mean difference or the standard deviation of the mean costs on which to
base a sample size justification.
I have decided to go for a 2-look design where after an interim analysis
I will estimate the final sample size required to detect a difference in
the mean costs, on the basis of the mean difference and the standard
deviation of the costs observed at the interim look.
My question is this, when estimating the mean difference at the interim
look and when combining the data from the "2" trials, i.e. patients to
1st look, patients from 1st look to final patient, is it valid to use an
adjusted estimate of the mean difference and standard error, rather than
the unadjusted estimates? My preference is to use the adjusted estimate
to improve sensitivity and to reduce confounding bias..
Any thoughts and does anyone know any references where this approach has
been used?
Many Thanks
Trevor
Trevor Mole PhD
Statistics Manager
Smith & Nephew Wound Management Division,
101 Hessle Road, Hull, HU3 2BN
Smith & Nephew plc
[log in to unmask]
T Office +44 (0)1482 673003
T Home +44 (0)1904 658793
F Office +44 (0)1482 673110
Confidentiality.
This electronic transmission is strictly confidential to Smith & Nephew and intended solely for the addressee. It may contain information which is covered by legal, professional or other privilege. If you are not the intended addressee, or someone authorised by the intended addressee to receive transmissions on behalf of the addressee, you must not retain, disclose in any form, copy or take any action in reliance on this transmission. If you have received this transmission in error, please notify the sender as soon as possible and destroy this message.
Smith & Nephew Medical Limited
Registered in England and Wales No.605496 with registered office at PO Box 81, 101 Hessle Road, Hull HU3 2BN
|