Survival analyses in cardiovascular research, part I: the essentials

Abstract

This review provides a practical guide to the essentials of survival analysis and their reporting in cardiovascular studies, although most of its key content can be extrapolated to other medical fields. This is the first in a series of 2 educational articles laying the groundwork to address the most relevant statistical issues in survival analyses, which will smoothly drive the reader from the most basic analyses to the most complex situations. The focus will be on the type and shape of survival data, and the most common statistical methods, such as nonparametric, parametric and semiparametric models. Their adequacy, interpretation, advantages and disadvantages are illustrated by examples from the field of cardiovascular research. This article ends with a set of recommendations to guide the strategy of survival analyses for a randomized clinical trial and observational studies. Other topics, such as competing risks, multistate models and recurrent-event methods will be addressed in the second article.