Immortal time bias in oncological survival studies with homeopathic add-on
Andreas Gleiß, PhD
Clinical Biometrics, Center for Medical Statistics, Informatics and Intelligent Systems, Medical University Vienna
In oncological studies of patients' survival times homeopathy is often offered as add-on therapy in addition to the conventional treatment. If the study is observational, i.e. the allocation to the add-on is not randomized at cancer diagnosis, then patients with short survival time are under-represented in the homeopathy group for the investigation of the effectiveness of hoemopathy. This talk will raise the awareness for the resulting so-called immortal time bias and summarize adequate methods to deal with it.
Andreas Gleiss has a PhD degree in mathematics and a position as biostatistician and lecturer at the Section for Clinical Biometrics within the Center for Medical Statistics, Informatics, and Intelligent Systems, Medical University of Vienna, Austria. Since 2003 he was primarily engaged with the statistical support of clinical studies in various fields from neuropsychology or pediatrics to oncology. He was involved in planning clinical, pre-clinical and epidemiologic studies, randomization, statistical analysis, and publication of clinical and pre-clinical studies. Recent methodological publications are in the field of omics data and explained variation. During recent years he found special interest in the statistical and design aspects of studies in the field of complementary and alternative medicine.