The European CHMP criteria for influenza vaccine immunogenicity cannot be improved by the use of confidence intervals

In a well-know CHMP Note for Guidance criteria are given for evaluation the results of annual-update studies. Now and then it sometimes is suggested that these criteria can be improved by the use of confidence intervals. Here it is argued that this is based on a misinterpretation of what confidence intervals are. These intervals require that subjects are chosen at random from a population. In influenza studies this is never the case. (c) 2006 Elsevier Ltd. All rights reserved.