โ Scribed by BROWN; FRIEDMAN; TEMPLE; COLLINS; SIMON; FEINSTEIN; CHALMERS; BEGG; POCOCK; FREEDMAN; MEIER; GEORGE; DEMETS; ELLENBERG; CLEMENS; BERENDES; LEE
No coin nor oath required. For personal study only.
Dr. Brown: I would like to get at some points where there really is some disagreement. For example, we debated whether new information should be used to change the trial's statistical stopping guidelines. Is there disagreement? Imagine the most extreme case. You have monitoring guidelines set up, and a new, very influential study with purportedly clear results is published. All of a sudden there is a question of whether the guidelines should be changed or the trial stopped early.
We are couching the issue too narrowly by talking about statistical approaches to stopping or not stopping. What we should be talking about is using information to alter a protocol or not alter a protocol, whether it is stopping the randomization to one of the groups, stopping the entire study, or changing entry criteria. There are a whole variety of options that data monitoring committees have available to them. Let me use the example of the recent results on beta carotene and vitamin E supplementation from the trial of male smokers in Finland.' Every group that is sponsoring other studies with beta carotene had to reconsider them. Many issues arose. Do we have to change the informed consent? What do we tell the patients in our studies? Do we have to alter the entry criteria? Do we have to drop certain patients from the study? One very important consideration obviously is to maintain the size and length of the study or extend it or shorten it.
Dr. Temple: It is hard to see why meta-analysis should be distinguished from the results of a single large trial in this regard.
๐ SIMILAR VOLUMES
**Henry Green considered Concluding the finest of all his books** Concluding--set in a single summer day--has at its heart old Mr. Rock, a famous retired scientist: he lives in a cottage on the grounds of a girl's boarding school. Living with him is Elizabeth, his somewhat unstrung granddaughter; h