|
In this section, we
identify the more important practical issues that arise when multicenter
trials are planned, and offer recommendations that may allow multicenter
trials to be undertaken more effectively. It is important to emphasize
that these types of trials introduce few new issues or components
beyond those that must be considered when undertaking any randomized
control trial. The exceptions to this broad statement may be related
to:
- Data
management and data analysis;
- Confidentiality
of data; and
- The inevitable layers
of complexity added to data analysis when large trials are conducted
in multiple clinical settings.
Rather than introducing
considerations that do not arise in single center studies, multicenter
trials seem inevitably to intensify virtually every aspect and
every stage of the multicenter trial with potential complexities.
The burden of experience
and the conventional wisdom in the field seems to reflect that
problems encountered when multicenter clinical trials are undertaken
almost always flow from errors of omission, principally failure
to anticipate the need for communication at every stage and/or
failure to communicate with the most relevant or responsible individuals
at each stage.
The "bottom line"
for a multicenter clinical trial is that all participating centers
and all participating investigators must agree to follow common
protocols -- not only commonly agreed upon protocols for clinical
treatments, but commonly agreed upon and uniformly executed protocols
for recruitment, enrollment, randomization, and all aspects of
data collection, as well as maintaining confidence in data integrity
and protection of the rights of human subjects (Friedman
et al., 1998b).
Several of these matters
are given more attention in the pages that follow.
|