[Monitoring clinical studies. Development, measures and consequences].

Abstract

BACKGROUND: Clinical trials play an important role in developing and establishing new therapeutic and diagnostic procedures. In the planning and execution of these trials procedures and measures which allow for continual observation, description and evaluation of the study progress are to be taken into account. This is due to ethical, scientific and economic considerations. Together these procedures and measures are termed "monitoring". Repeated evaluation of the main study question is one particular monitoring measure. Results from this sequential procedure may lead to an early termination of patient recruitment. In the last 3 decades methods for interim analyses were developed which take into account the increased chance of errors when evaluating repeatedly the same question. By adjustment they guarantee a prespecified level of significance in the end result of the trial. Even though statistical significance may be evident in an interim analysis, this has not always to result in early termination of the trial. The decision to end a trial early must include other considerations than the mere evaluation of the main study question. In particular, consequences of the decision such as credibility and transferability of the trial result to subsequent therapeutic routine application are to be discussed.