A predictive approach to selecting the size of a clinical trial, based on subjective clinical opinion
作者:
D. J. Spiegelhalter,
L. S. Freedman,
期刊:
Statistics in Medicine
(WILEY Available online 1986)
卷期:
Volume 5,
issue 1
页码: 1-13
ISSN:0277-6715
年代: 1986
DOI:10.1002/sim.4780050103
出版商: Wiley Subscription Services, Inc., A Wiley Company
关键词: Power;Prediction;Prior opinion;Pragmatic trial;Sample size
数据来源: WILEY
摘要:
AbstractThe ‘textbook’ approach to determining sample size in a clinical trial has some fundamental weaknesses which we discuss. We describe a new predictive method which takes account of prior clinical opinion about the treatment difference. The method adopts the point of clinical equivalence (determined by interviewing the clinical participants) as the null hypothesis. Decision rules at the end of the study are based on whether the interval estimate of the treatment difference (classical or Bayesian) includes the null hypothesis. The prior distribution is used to predict the probabilities of making the decisions to use one or other treatment or to reserve final judgement. It is recommended that sample size be chosen to control the predicted probability of the last of these decisions. An example is given from a multi‐centre trial of superficial bladder c
点击下载:
PDF
(826KB)
返 回