Premium
Understanding on the Pooled Test for Controlled Clinical Trialsn
Author(s) -
Lu Hong Laura,
Huque Mohammad
Publication year - 2001
Publication title -
biometrical journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.108
H-Index - 63
eISSN - 1521-4036
pISSN - 0323-3847
DOI - 10.1002/1521-4036(200111)43:7<909::aid-bimj909>3.0.co;2-o
Subject(s) - type i and type ii errors , statistics , statistical power , post hoc , statistical hypothesis testing , clinical trial , post hoc analysis , test (biology) , econometrics , homogeneity (statistics) , medicine , meta analysis , computer science , mathematics , biology , paleontology
Several independent clinical trials are usually conducted to demonstrate and support the evidence of the efficacy of a new drug. When not all the trials demonstrate a treatment effect because of a lack of statistical significant finding, the sponsor sometimes conducts a post hoc pooled test and uses the pooled result as extra statistical evidence. In this paper, we study the extent of type I error rate inflation with the post hoc pooled analysis and the power of interaction test in assessing the homogeneity of the trials with respect to treatment effect size. We also compare the power of several test procedures with or without pooled test involved and discuss the appropriateness of pooled tests under different alternative hypotheses.