[Survival analysis techniques].

Abstract

Statistical methods known as survival analyses are useful for analyzing time-related events, in which time from a benchmark event to an endpoint is the focus of interest. Survival analysis describes not only patient survival statistics (as suggested by the name), but also other dichotomous outcomes such as time of remission, time of breastfeeding, etc. This paper discusses survival analysis techniques, commenting and comparing their utilization, especially in the field of oncology. It also presents and discusses types of epidemiological studies and data sources to which this type of analysis is applied. The authors take into account the difference between hospital-based or clinical series and population-based approaches. Interpretation of results is also discussed.