ABSTRACT
A clinical trial carried out during an early phase of the development of a treatment regimen is often an explanatory trial, described by Sackett (1983) as dealing with a question such as "Can a certain drug reduce cancer size?" or "Under what conditions does the drug reduce cancer size?" This type of question is also asked in earlier laboratory experiments with respect to efficacy. In such an explanatory trial or experiment, if it were discovered that a switch to the wrong treatment had occurred for an experimental unit, or if problems developed which precluded observing the planned outcome, that experimental unit would weaken the final comparison of active treatment and control. No evidence is provided on whether or not a drug reduces cancer size if that drug is not given or the tumor not measured after administration of the drug. Thus great care should be taken in an explanatory trial to avoid changes in treatment, or patient dropouts, whatever the reason, even though this leads to limitations in the scope of the trial. The design and analysis of explanatory trials are
discussed further in Section 2. See also the discussion of phase I trials in Chapter 1. General principles for the design and analysis of any clinical trial are also introduced in Section 2.
