The general guidelines provided in Section 10.5.2 apply equally to oral presentations as to publications. It would rarely be appropriate to present results that are immature. An added difficulty with presentations is the need to submit abstracts several months in advance of the presentation date, and hence the need to predict whether the required number of events, say, will be reached by the time of the final pre-meeting analysis. Programme selection committees understandably would rather see actual results than the promise of data that 'will be presented'. However, if data are immature at the time of submission it is best to give data by randomized group only on those aspects of the trial which are complete (toxicity and compliance say), and to give overall event rates, or an estimate of the proportion of the required events seen so far. There will often be a compelling reason to present results at a particular meeting and so some flexibility is required. However, do bear in mind that the presentation of immature results runs two major risks; firstly that the results will be inconclusive, secondly that they may change with further follow-up. In both situations, any benefit gained by presenting early is lost. As a rule, aim to minimize the time between first presentation of trial results and submission for publication.

0 0

Post a comment