Survival Analysis for Medical and Health Professionals

This course provides a practical introduction to the analysis of data in the form of time-to-event, or survival times. Such data is frequently highly skewed and times may be censored. These features, together with clinical questions in a survival context, require dedicated statistical techniques. This course begins with an overview and continues to cover the following topics: summary statistics and exploratory graphics, simple hypothesis testing, regression modelling using the Cox model and some extensions to this model. R, SAS, SPSS or Stata may be used for practical work.

Course Information:

Course Date 25th September 2019 - 26th September 2019
Course Code S3RISAMHP
Course Leader Dankmar Böhning

Course Description

Overview

Survival data arise in many medical areas. Examples include time to death after an operation, time to recovery from an accident, and duration of pain relief.

One particular aspect of time-to-event data is censoring, where the time to an event is not known exactly. This course focuses on handling right censoring, where a time is only known to be greater than a certain value. The methods of analysis for survival data fully encompass the issue of censoring.

The course is a basic practical introduction to some of the commonly-used tools for analysing survival data involving right censored values. Statistical theory underlying the different approaches is kept to a minimum, and emphasis is placed on how to summarise data and how to interpret common hypothesis tests. The course also introduces and explains the concept of modelling survival data based on the widely-use Cox regression model.

All practical work may be done using R, SAS, SPSS or Stata. Each participant may choose one of these statistical packages, according to their preference, and carry out all practical exercises using that package. Examples used will be drawn from a variety of applications in medicine and health.

Who Should Attend?

Medical and health professionals who need analytical tools for making inferences from survival data. Participants will be assumed to have some knowledge of elementary statistical techniques (e.g. hypothesis testing, standard errors and confidence intervals) and linear regression (e.g. concept of a statistical model, comparing models).

How You Will Benefit

You will acquire practical experience in the use of commonly-used techniques for the analysis of survival data, and an appreciation of more complex methods.

What Do We Cover?

• Survival data: properties, examples, issues of censoring, right censoring
• Summary statistics and graphics: survival curves and the Kaplan-Meier estimate
• Comparison of two groups: common hypothesis tests
• The concept of a hazard of an event
• The Cox proportional hazards model
• Comparison of Cox regression models
• Predictions from the Cox model
• Stratified Cox regression model
• Time varying covariates.

Choice of Software

Practical work may be done in one of the following statistical packages for the duration of the course: R, SPSS, SAS or Stata.