A short while ago I requested help on Power calculations for Equivalence in
Survival Analysis, many thanks to those who responded. Your help was really
appreciated. Below is a summary of those replies.
Cheers
Paul
=========================================================
----- Original Message -----
From: Pauly <[log in to unmask]>
To: <[log in to unmask]>
Sent: 25 November 1999 19:02
Subject: Query: Power for Equivalence in Survival Analysis
> Dear Allstat,
>
> I have been asked to provide sample size calculations for a study to show
> equivalence where the outcome of interest is a time to event variable.
>
> Has anyone of you experience or pointers, references that may help me?
>
=========================================================
From: arjun roy
Hi Paul,
Skaff Pamela([log in to unmask]) has published SAS macro for euivalence
in survival data. You may request her to send you a copy or check SUGI,
1999.
Thanks,
Arjun Roy
Note:It was actually an award winning paper form SUGI 1998.
Pamela was also very helpful.
=========================================================
From: Hamid Pezeshk
Dear Paul
you may look at
1. Sample Size Methodology; Desu and Raghavarao (1990); Academic Press
(or if you are doing sample size determination for suvival analysis
problem
in some clinical trials)
2. The design and analysis of sequential clinical trials; Whitehead(1997);
John Wiley.
I hope these help.
Good luck
Hamid
=========================================================
From: Paul Seed
Try:
Parmar MKB, Machin D. Sample size for survival studies
In Parmar MKB, Machin D (1995) Survival Analysis - a practical approach. Ch.
10, pages 193-207; Wiley, New York,
=========================================================
From: "Dr. Robert Nemeth"
I am not sure, whether it helps: I remember only an old paper of S. Wellek
(Biometrics, 49, 877, 1993) concerning equivalence within the PH model. He
did also some simulation on the power, but not provided any closed
expressions for it
Kind regards
Robert
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|