1
$\begingroup$

I'm conducting a clinical trial with a treatment group and a placebo group and calculated the sample size based on a previous study. However, I would like to add two more new placebo groups with different characteristics, where there is no raw data to calculate the sample size and related values.

My question is how to derive the sample size needed for these two new placebo groups? How to verify the effectiveness?

$\endgroup$
5
  • $\begingroup$ Helpful link?. //With what other groups would proposed placebo groups be compared? Why different size from those? $\endgroup$
    – BruceET
    May 23, 2022 at 8:11
  • $\begingroup$ Do you mean we can just make the two new placebo groups have the same size as the original placebo group? $\endgroup$ May 23, 2022 at 8:57
  • 1
    $\begingroup$ How exactly will these new placebo groups be used in the analysis? Do you have reasons to suppose their characteristics will differ from the original assumptions you have made about a placebo group? If so, what do you know about those differences? $\endgroup$
    – whuber
    May 23, 2022 at 13:04
  • $\begingroup$ So there will be 3 placebo groups and 1 treatment group. People who are in charge of the trial decided to do that and I am only supposed to carry out some statistical results. The reason behind this is out of my knowledge scope. Does it make sense to add two more groups? How do I calculate the needed size of the two new groups without any information? $\endgroup$ May 23, 2022 at 13:18
  • $\begingroup$ In most RCTs, the purpose of a placebo group is to be identical to the treatment group in every way except receipt of treatment. Another purpose of the placebo group may be to represent current practice. If current practice is highly variable, perhaps additional placebo groups could be used, but I have never seen this in practice. Typically, additional groups in a trial are treatment groups whose treatment is measured against a single placebo group. Regarding sample size, is the trial is already underway and has someone looked at the data? $\endgroup$
    – Todd D
    May 25, 2022 at 4:08

0