Several tests have been recently implemented in the nonparametric comparison of current status survival data. However, they are not suited for the situation of crossing hazards. In this setting, we propose a new test specifically designed for crossing hazards alternatives. The proposed test is compared to classical implemented tests through simulations mimicking crossing hazards situations with various schemes of censoring. The results show that the proposed test has a correct type I error and generally outperforms the existing methods. The application of the proposed test on a real dataset on immunogenicity of interferon-β among multiple sclerosis patients highlights the interest of the proposed test.
Keywords: crossing hazards; current status data; immunogenicity; permutation test; weighted logrank.
Copyright © 2017 John Wiley & Sons, Ltd.