19.5.2  Interim analysis and data monitoring

It is increasingly common practice for individual clinical trials to include a plan for interim analyses of the data, and to monitor safety. PMA offers a unique opportunity to perform these interim looks using the data contributed by all trials. The data may be pooled for this analysis, or looked at separately for each trial and the results then shared amongst the data monitoring committees of the participating trials.

 

The ability to perform interim analyses raises a number of ethical issues. Is it, for example, appropriate to continue randomization to ongoing studies after an overall benefit (in terms of the primary outcome, for example) of an intervention has been demonstrated? When results are not known in the subgroups of clinical interest, or for less common endpoints, should the investigators proceed with the study to obtain further information on overall net clinical benefit, for example, evidence of benefit for one outcome but not another, or evidence of harm.

 

If each trial has its own data monitoring committee, then communication among committees might be beneficial in this regard, as recommended by Hillman and Louis (Hillman 2003). The various committees would need to be aware of the other trials included within the PMA and their results, because these external considerations might influence the decisions made by a given monitoring committee: for example, whether or not to close a study early because of evidence of efficacy. Conversely, it might be argued that knowledge of emerging safety data from all participating trials might reduce the chances of spurious early stopping of an individual trial due to concerns about interim safety outcomes. It would be helpful, thus, for the various trial data safety monitoring committees to adopt a common understanding that individual trials should not be stopped until the goals of the PMA, with respect to subgroups and uncommon endpoints (or ‘net clinical benefit’), are achieved.

 

Another possible option might be to consider limiting enrolment in the continuing trials to patients in the subgroup(s) of interest if such a decision makes clinical and statistical sense. In any case, it might be appropriate to apply the concepts of sequential clinical trials methodology, such as the approach described by Whitehead (Whitehead 1997), to derive rigorous and stringent stopping rules for the PMA as individual trial results become available.