Survival analysis using Stata®


Last update: July 5, 2009

Organizer: Metrika Consulting

Date: Saturday November 7, 2009

Place: Stockholm

Language: English

Deadline: Sign up before November 1, 2009

Cancellation policy: If you have signed up and need to cancel, you do not need to pay any fee if you cancel more than 11 days before the course starts (the dead-line for signing up). You will need to pay half the fee if you cancel less than 11 days before the course starts.

Price: 3000 SEK, excl. VAT.

Registration: Just write an email to Metrika Consulting

Instructor:

Nicola Orsini
The National Institute of Environmental Medicine,
Karolinska Institutet, Stockholm, Sweden.
Phone: +46 8 52487837

Course description

This course introduces participants to statistical methods for the analysis of studies where the endpoint is time to an event. We will study methods for estimating patient survival (life table and Kaplan-Meier methods), comparing patient subgroups (log-rank test), and modelling survival (primarily Poisson regression and the Cox proportional hazards model). Emphasis will be placed on describing how epidemiological cohort studies can be analysed in the framework of survival analysis.

Participants

The course is intended for anyone who wishes or needs to use Stata. Basic knowledge of Stata and statistics is ideal. Familiarity with Microsoft Windows will be assumed. Students will gain the familiarity with Stata by practicing.

Preliminary outline of the course

Morning session

  • Central concepts is survival analysis: censoring, truncation, survivor function, hazard function
  • Estimating survival (or cumulative incidence) using the actuarial and Kaplan-Meier methods
  • Testing for differences in survival between groups using the log-rank test
  • Estimating rates and modelling them using Poisson regression
  • Point and interval estimates based on the regression model
  • Graph the rates and ratios of rates
  • Testing hypothesis (Wald and likelihood based)

Afternoon session

  • Cox proportional hazards model
  • Modeling categorical and continuous predictors
  • Evaluate confounding effects
  • Test for interaction
  • Diagnostics for the Cox model
  • Comparison of the Cox and Poisson regression models
Schedule

  9:00 - 12:00 Morning Session
12:00 - 13:00 Lunch
13:00 - 16:00 Afternoon Session

The students will have their hands on a keyboard with Stata running troughout the learning sessions. Each learning session will give equal weight to tutorials and practical exercises.

Course material

Lecture notes, datasets, exercises, and solutions will be available over the internet for the students.

{Home | Author | Presentations | Publications | Software | Courses | Links}