2-stage design with interim sample size estimation [Two-Stage / GS Designs]
Dear d_labes,
this is quite a complex question to which I will add some more questions.
The first and most important question in my mind is why not consider a single stage design with a sample size review at interim? So basically not allowing for the study to stop early (besides the sponsors constraint).
The second choice one needs to make is if the sample size review is going to be blinded or not (note that I am ignoring any knowledge about a possible shift in point estimates). Ekkehard Glimm has (very recently) done some work on the type-I error inflation in blinded sample size reviews for non-inferiority trials and from memory of a presentation I saw the conclusion was that:
a) the inflation is negligible for all practical purposes.
b) that the usual sample size review does in fact make use of the treatment allocation internally.
The trouble is that the results are not yet published as far as I know.
As for, how blinded sample size reestimation is done in non-inferiority trials see for example Biom J. 2007 Dec;49(6):903-16.
Hope that helps
this is quite a complex question to which I will add some more questions.
The first and most important question in my mind is why not consider a single stage design with a sample size review at interim? So basically not allowing for the study to stop early (besides the sponsors constraint).
The second choice one needs to make is if the sample size review is going to be blinded or not (note that I am ignoring any knowledge about a possible shift in point estimates). Ekkehard Glimm has (very recently) done some work on the type-I error inflation in blinded sample size reviews for non-inferiority trials and from memory of a presentation I saw the conclusion was that:
a) the inflation is negligible for all practical purposes.
b) that the usual sample size review does in fact make use of the treatment allocation internally.
The trouble is that the results are not yet published as far as I know.
As for, how blinded sample size reestimation is done in non-inferiority trials see for example Biom J. 2007 Dec;49(6):903-16.
Hope that helps
Complete thread:
- 2-stage design with interim sample size estimation d_labes 2010-10-05 15:09
- 2-stage design with interim sample size estimationJack 2010-10-25 18:20
- Null = bioINequivalence Helmut 2010-10-25 19:16
- 2-stage design with interim sample size estimation d_labes 2010-10-26 13:28
- 2-stage design with interim sample size estimation Jack 2010-10-26 18:26
- Blind or not, Indecisions or Decisions d_labes 2010-10-27 16:01
- Blind or not, Indecisions or Decisions Jack 2010-10-27 18:14
- Blind ? Ohlbe 2010-10-27 23:37
- Blind ? Jack 2010-10-28 10:24
- Blind 2Stage Astea 2019-05-30 10:37
- Blind 2Stage Helmut 2019-05-30 12:56
- Blind 2Stage and an elephant Astea 2019-05-31 14:36
- Which elephant? Helmut 2019-06-02 20:14
- Blind interim in parallel design? Astea 2019-06-03 01:47
- Blind men d_labes 2019-06-05 12:23
- The Blind Leading the Blind Astea 2019-06-06 00:55
- Blind men d_labes 2019-06-05 12:23
- Blind interim in parallel design? Astea 2019-06-03 01:47
- Which elephant? Helmut 2019-06-02 20:14
- Blind 2Stage and an elephant Astea 2019-05-31 14:36
- Blind 2Stage Helmut 2019-05-30 12:56
- Blind 2Stage Astea 2019-05-30 10:37
- Blind ? Jack 2010-10-28 10:24
- Blind ? Ohlbe 2010-10-27 23:37
- Blind or not, Indecisions or Decisions Jack 2010-10-27 18:14
- Blind or not, Indecisions or Decisions d_labes 2010-10-27 16:01
- 2-stage design with interim sample size estimation Jack 2010-10-26 18:26
- 2-stage design with interim sample size estimationJack 2010-10-25 18:20