[Analysis of Survival Time and Other Event Times - What Does the Surgeon Need to Know?]. / Überlebenszeitanalysen und andere Ereigniszeiten - was der Chirurg hierzu wissen sollte.

BACKGROUND: Survival time is an important parameter to investigate therapeutic measures. It plays a crucial role in study concepts, data analyses as well as publications. AIM: The aim of this study was to emphasise essential points, which need to be taken into account to (i) gain resilient results of survival time analysis and (ii) appropriately evaluate scientific reports. Corner Points/Main Statements: (i) The main analytical methods are Kaplan-Meier procedure to estimate survival time curves, the log rank test to compare two or more survival curves from independent samples and Cox regression for comparisons under simultaneous consideration of several influencing factors. (ii) Dependent relationships between survival and censoring probabilities may falsify these statistical procedures. (iii) For several end points, which need to be differentiated (such as death, progression etc.), and for interdependent sample elements, extended statistical procedures such as competing risk analyses or extended Cox regression models are available. CONCLUSION: Survival time analysis can be considered as being extraordinarily important for evaluation of data obtained in therapeutic studies. For the academic and publishing physician, in particular, for the clinical surgeon, a basic understanding of these methodological aspects in statistics is indispensable.

Time-to-event analysis with treatment arm selection at interim.

This paper discusses the application of an adaptive design for treatment arm selection in an oncology trial, with survival as the primary endpoint and disease progression as a key secondary endpoint. We carried out treatment arm selection at an interim analysis by using Bayesian predictive power combining evidence from the two endpoints. At the final analysis, we carried out a frequentist statistical test of efficacy on the survival endpoint. We investigated several approaches (Bonferroni approach, 'Dunnett-like' approach, a conditional error function approach and a combination p-value approach) with respect to their power and the precise conditions under which type I error control is attained.