Censoring refers to a situation in survival analysis where the event of interest is not observed for some of the individuals under study.

In this Statistical Primer, we’ll define three types of censoring often seen in survival analysis studies.

Censoring occurs when the information on the survival time is incomplete or only partially observed.

Censoring can have a significant impact on the analysis and interpretation of survival data. It is essential to appropriately handle censoring in survival analysis to obtain accurate estimates of survival times, covariate effects, and other related parameters.

There are different types of censoring in survival analysis:

  • Right-censoring: This occurs when a participant is still alive or event-free at the end of the study period. In other words, the follow-up time for the participant ends before the event occurs. This is the most common type of censoring in survival analysis.
  • Left-censoring: This occurs when the true event time is known to be less than a certain time, but the exact time is unknown. For example, if an individual is diagnosed with a disease before the study begins but the date of onset of the disease is not known, we have left-censoring.
  • Interval-censoring: This occurs when the event time is known to fall within a certain interval, but the exact time of the event is unknown. For example, if a person develops glaucoma in between visits to the optician but the exact onset is unknown, we have interval censoring.

Latest Resources

Tutorials

Joint frailty models for recurrent and terminal events

In this post we’re going to take a look at joint frailty models, and how to fit them with our merlin command. Importantly, we’ll also discuss how to interpret the results. Joint frailty models An area of intense research in recent years is in the field of joint frailty models, which has become the commonly used name for […]
Read more

Tutorials

A user-defined/custom hazard model

This tutorial will illustrate some of the more advanced capabilities of merlin when modelling survival data, but with the aim of using an accessible example. During my PhD, Paul Lambert and I developed stgenreg in Stata for modelling survival data with a general user-specified hazard function, with the generality achieved by using numerical integration to calculate the cumulative hazard […]
Read more

Tutorials

Relative survival analysis

Relative survival models are predominantly used in population based cancer epidemiology (Dickman et al. 2004), where interest lies in modelling and quantifying the excess mortality in a population with a particular disease, compared to a reference population, appropriately matched on things like age, gender and calendar time. One of the benefits of the approach is […]
Read more

Tutorials

Simulating survival data with a continuous time-varying covariate…the right way

In this post we’ll take a look at how to simulate survival data with a continuous, time-varying covariate. The aim is to simulate from a data-generating mechanism appropriate for evaluating a joint longitudinal-survival model. We’ll use the survsim command to simulate the survival times, and the merlin command to fit the corresponding true model. Let’s assume a proportional hazards […]
Read more

Videos

Introduction to Epidemiological Study Designs

This video offers a comprehensive introduction to epidemiological study designs, emphasising their classification, key definitions, strengths, limitations, and practical applications. We will dive into the most commonly used study designs, exploring their structure, purpose, and the contexts in which they are most effective. Throughout the video, real-world case studies of landmark research will be used […]
Read more

Tutorials

An introduction to joint modelling of longitudinal and survival data

This post gives a gentle introduction to the joint longitudinal-survival model framework, and covers how to estimate them using our merlin command in Stata. A joint model consists of a continuous, repeatedly measured (longitudinal) outcome, and a time-to-event, with the two models linked by random effects, or functions of them. Let’s formally define everything we need. For […]
Read more

Tutorials

Probabilistic sensitivity analysis and survival models

Today we’re going to take a little look into probabilistic sensitivity analysis (PSA), and how it can be implemented within the context of survival analysis. Now PSA is used extensively in health economic modelling, where a particular parameter (or parameters) of interest, are altered or varied, to represent different scenarios and levels of variation. We […]
Read more

Videos

Multilevel (hierarchical) survival models: Estimation, prediction, interpretation

Hierarchical time-to-event data is common across various research domains. In the medical field, for instance, patients are often nested within hospitals and regions, while in education, students are nested within schools. In these settings, the outcome is typically measured at the individual level, with covariates recorded at any level of the hierarchy. This hierarchical structure […]
Read more

Specialist subjects

Real-World Evidence (RWE)

Real-World Evidence Real-world evidence (RWE) refers to data and information that, unlike data generated in clinical trials conducted in controlled environments, has been obtained from everyday clinical practice, patient registers, or other sources outside the clinical trial setting. RWE plays a crucial role in complementing traditional clinical trial data, providing insights into the safety, effectiveness, […]
Read more

Tutorials

Flexible parametric survival analysis with frailty

This example takes a look at incorporating a frailty, or random intercept, into a flexible parametric survival model, and how to fit them in Stata. First we’ll use merlin to estimate our model, and then the more user-friendly wrapper function stmixed. More details on these models can be found in the following papers: Crowther MJ, Look MP, Riley […]
Read more

Tutorials

multistate v4.4.0: semi-parametric multi-state modelling

The headlines: predictms now supports the Cox model as a transition model, estimated using merlin or stmerlin Predictions from a multi-state Cox model are implemented using a simulation approach Supported predictions from a multi-state Cox model include transition probabilities, probability, and length of stay, los Let’s take a look at what we can now do with multistate and in particular, the predictms command. We’ll […]
Read more

Statistical Primers

What is immortal time bias?

Immortal time bias Immortal time bias is a type of bias that can occur in observational research when the study design allows for a period of time during which the outcome of interest cannot occur, often referred to as “immortal time”. Simply put, immortal time bias occurs when information from a future event is incorporated into the […]
Read more
All Resources