Limits...
A review and re-interpretation of a group-sequential approach to sample size re-estimation in two-stage trials.

Bowden J, Mander A - Pharm Stat (2014)

Bottom Line: In this paper, we review the adaptive design methodology of Li et al. (Biostatistics 3:277-287) for two-stage trials with mid-trial sample size adjustment.We argue that it is closer in principle to a group sequential design, in spite of its obvious adaptive element.Several extensions are proposed that aim to make it even more attractive and transparent alternative to a standard (fixed sample size) trial for funding bodies to consider.

View Article: PubMed Central - PubMed

Affiliation: MRC Biostatistics Unit, Cambridge, UK.

Show MeSH
Left: modified LSW design 2. Right: minimum conditional power guaranteed by design 2.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4288989&req=5

fig02: Left: modified LSW design 2. Right: minimum conditional power guaranteed by design 2.

Mentions: Adding in the extra constraint nmax = 90 to the remaining parameters of design 1 in Section 3.1, we calculate Cmax to be 1.936. This is listed as ‘design 2’ in Table 1, and we refer to capping the maximum sample size in this way as the modified LSW approach. Figure 2 (left) shows the total sample size of design 2 as a function of . Figure 2 (right) shows the minimum conditional power guaranteed by this design as a function of . It starts at close to 40% when and increases up to a maximum of 80% by . So, artificially constraining the sample size to not exceed a maximum value leads to some loss of power when is small. It is therefore important to assess this constraint's effect on both on the overall power and expected sample size of the adaptive trial.


A review and re-interpretation of a group-sequential approach to sample size re-estimation in two-stage trials.

Bowden J, Mander A - Pharm Stat (2014)

Left: modified LSW design 2. Right: minimum conditional power guaranteed by design 2.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4288989&req=5

fig02: Left: modified LSW design 2. Right: minimum conditional power guaranteed by design 2.
Mentions: Adding in the extra constraint nmax = 90 to the remaining parameters of design 1 in Section 3.1, we calculate Cmax to be 1.936. This is listed as ‘design 2’ in Table 1, and we refer to capping the maximum sample size in this way as the modified LSW approach. Figure 2 (left) shows the total sample size of design 2 as a function of . Figure 2 (right) shows the minimum conditional power guaranteed by this design as a function of . It starts at close to 40% when and increases up to a maximum of 80% by . So, artificially constraining the sample size to not exceed a maximum value leads to some loss of power when is small. It is therefore important to assess this constraint's effect on both on the overall power and expected sample size of the adaptive trial.

Bottom Line: In this paper, we review the adaptive design methodology of Li et al. (Biostatistics 3:277-287) for two-stage trials with mid-trial sample size adjustment.We argue that it is closer in principle to a group sequential design, in spite of its obvious adaptive element.Several extensions are proposed that aim to make it even more attractive and transparent alternative to a standard (fixed sample size) trial for funding bodies to consider.

View Article: PubMed Central - PubMed

Affiliation: MRC Biostatistics Unit, Cambridge, UK.

Show MeSH