2026-05-22T00:01:24Zhttps://repisalud.isciii.es/rest/oai/requestoai:repisalud.isciii.es:20.500.12105/235322024-11-29T07:52:36Zcom_20.500.12105_15322com_20.500.12105_2051col_20.500.12105_16967col_20.500.12105_16987
Repisalud
author
Rosselló, Xavier
author
González-Del-Hoyo, Maribel
2024-10-04T13:57:58Z
2024-10-04T13:57:58Z
2022-01
Rossello X, González-Del-Hoyo M. Survival analyses in cardiovascular research, part II: statistical methods in challenging situations. Rev Esp Cardiol (Engl Ed). 2022 Jan;75(1):77-85.
10.1016/j.rec.2021.07.001
1885-5857
Revista espanola de cardiologia (English ed.)
https://hdl.handle.net/20.500.13003/20154
34326022
L2014434156
2-s2.0-85114250277
https://hdl.handle.net/20.500.12105/23532
745030500012
This article is the second of a series of 2 educational articles. In the first article, we described the basic concepts of survival analysis, summarizing the common statistical methods and providing a set of recommendations to guide the strategy of survival analyses in randomized clinical trials and observational studies. Here, we introduce stratified Cox models and frailty models, as well as the immortal time bias arising from a poor assessment of time-dependent variables. To address the issue of multiplicity of outcomes, we provide several modelling strategies to deal with other types of time-to-event data analyses, such as competing risks, multistate models, and recurrent-event methods. This review is illustrated with examples from previous cardiovascular research publications, and each statistical method is discussed alongside its main strengths and limitations. Finally, we provide some general observations about alternative statistical methods with less restrictive assumptions, such as the win ratio method, the restrictive mean survival time, and accelerated failure time model.
eng
Survival analyses in cardiovascular research, part II: statistical methods in challenging situations
research article