Suggestions on the use of statistical methodologies in studies of the European Group for Blood and Marrow Transplantation.

摘要

This document explains some of the relevant methodological issues involved in planning a clinical study using survival and time-to-event outcome data, particularly in the field of haematopoietic stem cell transplantation, and indicates the appropriate statistical methods to use for the analysis. As the majority of these methods are commonly used in survival and event-history analysis, this document discusses their potential and limitations with reference to common SCT research situations. Some reference is given to methods, recently appearing in the literature that may be capable of handling complex investigations. These guidelines also address various practical issues, such as recoding or transforming variables in regression models or reporting results.