Survival analysis

Survival analysis is a set of statistical methods for studying the time until an event occurs, including death, failure, or recovery. It accounts for observations whose event time is not fully known.

Censoring (statistics): Censoring occurs when an observation’s event time is only partially known, often because follow-up ends first. Right censoring lets survival analysis include people whose event has not occurred by the end of observation.
Time-to-event data: Time-to-event data record durations from a defined starting point until a specified event or the end of observation. Survival analysis models this combination of durations and event indicators.
Clinical trial: A clinical trial is a study that prospectively assigns or evaluates interventions in human participants. Trials use time-to-event outcomes to compare treatments on endpoints such as relapse or death.
Binary logistic regression: Binary logistic regression models the probability of a binary outcome as a function of explanatory variables. Unlike survival analysis, it usually ignores when an event occurred and how long follow-up lasted.
Kaplan–Meier estimator: The Kaplan–Meier estimator estimates a time-to-event distribution from observed event times and censored observations. It produces a stepwise estimate of survival probability without assuming a specific distribution.
Kaplan–Meier curve: A Kaplan–Meier curve is a step function estimating the probability of remaining event-free over time. Its steps show estimated survival, while flat stretches reflect periods without observed events.
Reliability engineering: Reliability engineering uses engineering methods to ensure systems perform their required functions over time. Its component-lifetime studies apply survival methods to failure times and censored tests.
Parametric survival model: A parametric survival model specifies a probability distribution for event times and estimates its parameters. It offers distribution-based predictions where the Cox model leaves baseline hazard unspecified.
Hazard function: The hazard function is the instantaneous event rate at a given time, conditional on the event not having occurred earlier. It describes how event risk changes over time among those still event-free.
Survival function: A survival function gives the probability that a time-to-event variable exceeds a specified time. It is the central quantity estimated by many survival methods.

Linked from 21 pages