You can manage bookmarks using lists, please log in to your user account for this.
Media type:
E-Book
Title:
The time-dependent "cure-death" model investigating several endpoints simultaneously in trials treating high-risk patients with severe infections
University thesis:
Dissertation, Albert-Ludwigs-Universität Freiburg, 2017
Footnote:
Description:
Abstract: In clinical trials for the development of antibacterial drugs, diverse primary endpoints have been used and treatment effects are usually assessed at the end of follow-up which varies between studies. A highly patient-relevant statement would be an assessment over the entire follow-up period with cure and death as co-primary endpoints. We emphasise to examine the time-dependent multistate endpoint "get cured and stay alive over time", since this might be most relevant from the patients’ perspective and can capture different “cure patterns” over the treatment period. Such time-dynamic endpoints provide valuable additional information such that potentially hidden treatment effects can be revealed that might be overlooked when only presenting incidence proportions. Based on a "cure-death" multistate model, simple and sophisticated possibilities are introduced and compared to evaluate a treatment difference in probabilities to be cured and alive over time. As an example, non-inferiority is studied by means of one-sided confidence bands provided by a flexible resampling technique, or, an innovative regression method is used for a risk ratio of being cured and alive. These methods are further evaluated via a simulation study and applied to three topical data examples, a randomised controlled trial for the treatment of patients with hospital-acquired pneumonia, a randomised controlled trial for the prevention of recurrent Clostridium difficile infection, and a cohort study to investigate the effect of inadequate treatment for patients with ventilator-associated pneumonia due to the pathogen Pseudomonas aeruginosa. Multistate methodology, already entrenched in applied infection controlliterature for analysing observational data, is highly beneficial and easily applicable for clinical trials as well to examine patient-relevant endpoints