[R] Summary : Whitehead's group sequential procedures

Emmanuel Charpentier emmanuel.charpentier at sap.ap-hop-paris.fr
Tue Oct 28 15:16:03 CET 2003


Dear List,

I recently asked about any R implementations of Whitehead's methods for 
sequential clinical trials. Here's the summary of answers so far :

1) There is no R public implementation of those methods

2) There exists an interest for such a package, which does things quite 
different from Lan-deMets paradigm (shortly : Whitehead's methods allows 
for unplanned interim analyses in any number, while Lan-deMets plans a 
fixed number of interim analyses at fixed (information) times).

3) I know how to implement simple things : design (with approximations), 
  monitoring and stopping rule. I'm still hacking my way through 
Whitehead's book to understand his procedures for the distribution of 
the total number of patients included (which is a random variable) and 
the final analysis (after study termination).

4) It has been suggested that I contact Scott Emerson, author of an S 
package called seqTrials. Does someone has an address for him ?

I'll start implementing the simple stuff, trying to create a framework 
flexible enough to allow generalization. I'll probably plea for help at 
some time in the future ...

However, don't hold your breath : I'm doing this in my spare time, on 
top of a busy schedule ...

					Emmanuel Charpentier

-- 
Emmanuel Charpentier			Tel : +33-(0)1 40 27 35 98
Secrétariat Scientifique du CEDIT	Fax : +33-(0)1 40 27 55 65
Assistance Publique - Hôpitaux de Paris
3, Avenue Victoria, F-75100 Paris RP - France




More information about the R-help mailing list