|
Abstract:
|
In four placebo-controlled, double-blind, multi-dose hot flash studies, the placebo arm in each trial was divided into subgroups to match the amount of drug for each active arm. This approach was taken to reduce the pharmacy costs and medication distribution errors. But it raises statistical and design concerns that may undermine study validity. First, although patients wouldn’t know if they were assigned placebo or active drug, they know which dose group they were assigned. Second, pooling the placebo arms for analysis assumes a uniform effect within the placebo arm; however, if the effect is not uniform, the study agent efficacy could be diluted. Conflicting results regarding the placebo effect, dependent on placebo dose, have been reported. Thus, our aim is to investigate whether a placebo effect is dependent on placebo dose. By revisiting each of the four trials, hot flash frequency and severity scores within each arm will be quantified. A meta-data analysis across these studies is planned to b performed. We hope to understand whether a uniform hot flash relief due to the amount of placebo exists. The information will inform this design approach in future multi-dose studies.
|
