Literature DB >> 19340815

Blinded assessment of treatment effects utilizing information about the randomization block length.

Frank Miller1, Tim Friede, Meinhard Kieser.   

Abstract

It is essential for the integrity of double-blind clinical trials that during the study course the individual treatment allocations of the patients as well as the treatment effect remain unknown to any involved person. Recently, methods have been proposed for which it was claimed that they would allow reliable estimation of the treatment effect based on blinded data by using information about the block length of the randomization procedure. If this would hold true, it would be difficult to preserve blindness without taking further measures. The suggested procedures apply to continuous data. We investigate the properties of these methods thoroughly by repeated simulations per scenario. Furthermore, a method for blinded treatment effect estimation in case of binary data is proposed, and blinded tests for treatment group differences are developed both for continuous and binary data. We report results of comprehensive simulation studies that investigate the features of these procedures. It is shown that for sample sizes and treatment effects which are typical in clinical trials, no reliable inference can be made on the treatment group difference which is due to the bias and imprecision of the blinded estimates. (c) 2009 John Wiley & Sons, Ltd.

Entities:  

Mesh:

Year:  2009        PMID: 19340815     DOI: 10.1002/sim.3576

Source DB:  PubMed          Journal:  Stat Med        ISSN: 0277-6715            Impact factor:   2.373


  2 in total

1.  Maximum type I error rate inflation from sample size reassessment when investigators are blind to treatment labels.

Authors:  Magdalena Żebrowska; Martin Posch; Dominic Magirr
Journal:  Stat Med       Date:  2015-12-23       Impact factor: 2.373

2.  Blinded versus unblinded estimation of a correlation coefficient to inform interim design adaptations.

Authors:  Cornelia U Kunz; Nigel Stallard; Nicholas Parsons; Susan Todd; Tim Friede
Journal:  Biom J       Date:  2016-11-25       Impact factor: 2.207

  2 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.