| |
Stratification
Some variables can
influence the outcome of antiemetic studies. One way to control
for these, especially in small studies, is to stratify patients
based upon factors that are strongly associated with nausea and
vomiting. Among the large number of potential prognostic factors
studied, only patient age, gender, and any prior nausea/emesis
from chemotherapy have been reliably shown to affect nausea/emesis
probability from chemotherapy, beyond the emetogenicity of the
chemotherapy drugs themselves. The only real way to control for
prior nausea/emesis from chemotherapy is to study chemotherapy-naive
patients.
There is emerging support
for perhaps including susceptibility to motion sickness, patient
expectation, and alcohol history in the list. Stratification
can ensure a balance of known prognostic factors among the arms
of the study. The best way to ensure balance, of course, is to
design a trial with a large sample size. With a sufficiently large
sample size, imbalance of these factors is unlikely, due to the
inherently random nature of a large number of participants. In
addition, known and suspected prognostic factors should be monitored,
assessed, and evaluated in the analyses to further understand
their role and interaction in antiemetic trials.
|