NetCourse^{®} 631: Introduction to Survival Analysis Using Stata
 Content:

Learn how to effectively analyze survival data using Stata. We cover
censoring, truncation, hazard rates, and survival functions. Topics include
data preparation, descriptive statistics, life tables, Kaplan–Meier curves,
and semiparametric (Cox) regression and parametric regression. Discover how
to set the survivaltime characteristics of your dataset just once then apply
any of Stata's many estimators and statistics to that data.
Written for everyone who uses Stata, whether health researchers or social
scientists.
 Course leaders:
 Isabel Cañette, Senior Statistician at StataCorp
 Chuck Huber, Senior Statistician at StataCorp
 Jieyu Wang, Staff Statistician at StataCorp
 Course length:
 7 weeks (5 lectures)
 Dates:
 June 13–August 1, 2014 (details)
 Price:
 $295.00
Enroll
 Prerequisites:
 Stata 13, installed and working
 Course content of NetCourse 101 or equivalent knowledge
 Internet web browser, installed and working
(course is platform independent)

Course content
Lecture 1: Introduction to survival analysis
1 Introduction
2 The problem of survival analysis
2.1 The need for specific distributions
2.2 Answering specific kinds of questions
2.3 Censoring
2.3.1 Rightcensoring (withdrawal from study)
2.3.2 Leftcensoring
2.4 Truncation
2.4.1 Lefttruncation (delayed entry)
2.4.2 Righttruncation
2.5 Gaps
3 Survival analysis
3.1 The survivor and hazard functions
3.2 Hazard models
3.2.1 Parametric models
3.2.2 Semiparametric models
3.2.3 Nonparametric estimators
3.3 Analysis time (time at risk)
4 Summary
5 Exercises
6 References
Lecture 2: Setting and summarizing survival data
1 The purpose of the
stset command
1.1 The desired format—Introduction to
stset
1.2 (
st) Setting your data
1.3 The syntax of the
stset command
1.3.1 Specifying analysis time
1.3.2 Specifying what constitutes failure
1.3.3 Specifying when subjects exit from the analysis
1.3.4 Specifying when subjects enter the analysis
1.3.5 Specifying the subjectID variable
1.3.6 Handling gaps
2 After (
st) setting your data
2.1 Look at stset's output
2.2 Use stdescribe
2.3 Use stvary
2.4 Perhaps use stfill
3 Example: Hip fracture data
4 Appendices
4.1 Dates
4.2 Other formats
4.3 Convenience options
5 Exercises
6 References
Lecture 3: Setting and summarizing survival data
1 Nonparametric estimation
2 The Kaplan–Meier productlimit estimator of the survivor curve
2.1 Calculation of the Kaplan–Meier survivor curve
2.2 Censored observations
2.3 Delayed entry
2.4 Gaps
2.5 Properties of the Kaplan–Meier estimator
2.6 The sts graph command
2.7 The sts list command
2.8 The stsum command
3 The Nelson–Aalen estimator of the cumulative hazard
4 Alternative estimators of the survivor and cumulative hazard functions
5 Comparing survival experience
5.1 The logrank test
5.2 The Wilcoxon test
5.3 The Tarone–Ware test
5.4 The Peto–Peto–Prentice test
5.5 The Fleming–Harrington test
5.6 Test for trend across ordered groups
5.7 The Cox test
6 Exercises
7 References
Lecture 4: Regression models — Cox proportional hazards
1 Introduction
1.1 The Cox model has no intercept
1.2 Interpreting coefficients
1.3 The effect of units on coefficients
1.4 The baseline hazard and related functions
1.5 The effect of units on the baseline functions
1.6 Summary of stcox command
2 The calculation of results
2.1 No tied failures
2.2 Tied failures
2.2.1 The marginal calculation
2.2.2 The partial calculation
2.2.3 The Breslow approximation
2.2.4 The Efron approximation
2.2.5 Summary
3 Stratified analysis
3.1 Obtaining coefficient estimates
3.2 Obtaining the baseline functions
4 Modeling
4.1 Indicator variables
4.2 Categorical variables
4.3 Continuous variables
4.4 Interactions
4.5 Timevarying variables
4.5.1 Using stcox with option tvc()
4.5.2 Using stsplit
4.6 Testing the proportionalhazards assumption
4.6.1 Tests based on reestimation
4.6.2 Test based on Schoenfeld residuals
4.6.3 Graphical methods
4.7 Residuals
4.7.1 Determining functional form
4.7.2 Assessing goodness of fit
4.7.3 Finding outliers and influential points
5 References
6 Exercises
Lecture 5: Regression models — Parametric survival models
1 Introduction
2 Classes of parametric models
2.1 Parametric proportionalhazards models
2.2 Accelerated failuretime models
3 Maximum likelihood estimation for parametric models
4 A survey of parametric regression models in Stata
4.1 Exponential regression
4.1.1 Exponential regression in the PH formulation
4.1.2 Exponential regression in the AFT formulation
4.2 Weibull regression
4.2.1 Weibull regression in the PH formulation
4.2.2 Weibull regression in the AFT formulation
4.3 Gompertz regression (PH formulation)
4.4 Lognormal regression (AFT formulation)
4.5 Loglogistic regression (AFT formulation)
4.6 Generalized loggamma regression (AFT formulation)
5 Choosing among parametric models
4.1 Nested models
4.2 Nonnested models
6 Stratified models
7 Use of
predict after
streg
6.1 Predicting time of failure
6.2 Predicting the hazard and related functions
6.3 Calculating residuals
8 Use of
stcurve after
streg
9 Exercises
10 References
Enroll in NetCourse 631