You are previewing Survival Analysis Using SAS.
O'Reilly logo
Survival Analysis Using SAS

Book Description

Biomedical and social science researchers who want to analyze survival data with SAS will find just what they need with this easy-to-read and comprehensive guide. Teaches many aspects of data input and manipulation. Numerous examples of SAS code and output make this an eminently practical book, completely updated for SASĀ®9.

Table of Contents

  1. Copyright
  2. PREFACE
  3. 1. Introduction
    1. WHAT IS SURVIVAL ANALYSIS?
    2. WHAT IS SURVIVAL DATA?
    3. WHY USE SURVIVAL ANALYSIS?
    4. APPROACHES TO SURVIVAL ANALYSIS
    5. WHAT YOU NEED TO KNOW
    6. COMPUTING NOTES
  4. 2. Basic Concepts of Survival Analysis
    1. INTRODUCTION
    2. CENSORING
    3. DESCRIBING SURVIVAL DISTRIBUTIONS
      1. Cumulative Distribution Function
      2. Probability Density Function
      3. Hazard Function
    4. INTERPRETATIONS OF THE HAZARD FUNCTION
    5. SOME SIMPLE HAZARD MODELS
    6. THE ORIGIN OF TIME
    7. DATA STRUCTURE
  5. 3. Estimating and Comparing Survival Curves with PROC LIFETEST
    1. INTRODUCTION
    2. THE KAPLAN-MEIER METHOD
    3. TESTING FOR DIFFERENCES IN SURVIVOR FUNCTIONS
    4. THE LIFE-TABLE METHOD
    5. LIFE TABLES FROM GROUPED DATA
    6. TESTING FOR EFFECTS OF COVARIATES
    7. LOG SURVIVAL AND SMOOTHED HAZARD PLOTS
    8. CONCLUSION
  6. 4. Estimating Parametric Regression Models with PROC LIFEREG
    1. INTRODUCTION
    2. THE ACCELERATED FAILURE TIME MODEL
    3. ALTERNATIVE DISTRIBUTIONS
      1. The Exponential Model
      2. The Weibull Model
      3. The Log-Normal Model
      4. The Log-Logistic Model
      5. The Gamma Model
    4. CATEGORICAL VARIABLES AND THE CLASS STATEMENT
    5. MAXIMUM LIKELIHOOD ESTIMATION
      1. Maximum Likelihood Estimation: Mathematics
      2. Maximum Likelihood Estimation: Practical Details
    6. HYPOTHESIS TESTS
    7. GOODNESS-OF-FIT TESTS WITH THE LIKELIHOOD-RATIO STATISTIC
    8. GRAPHICAL METHODS FOR EVALUATING MODEL FIT
    9. LEFT CENSORING AND INTERVAL CENSORING
    10. GENERATING PREDICTIONS AND HAZARD FUNCTIONS
    11. THE PIECEWISE EXPONENTIAL MODEL
    12. BAYESIAN ESTIMATION AND TESTING
    13. CONCLUSION
  7. 5. Estimating Cox Regression Models with PROC PHREG
    1. INTRODUCTION
    2. THE PROPORTIONAL HAZARDS MODEL
    3. PARTIAL LIKELIHOOD
      1. Partial Likelihood: Examples
      2. Partial Likelihood: Mathematical and Computational Details
    4. TIED DATA
      1. The EXACT Method
      2. The DISCRETE Method
      3. Comparison of Methods
    5. TIME-DEPENDENT COVARIATES
      1. Heart Transplant Example
      2. Construction of the Partial Likelihood with Time-Dependent Covariates
      3. Covariates Representing Alternative Time Origins
      4. Time-Dependent Covariates Measured at Regular Intervals
      5. Ad-Hoc Estimates of Time-Dependent Covariates
      6. Time-Dependent Covariates That Change at Irregular Intervals
    6. COX MODELS WITH NONPROPORTIONAL HAZARDS
      1. Testing the Proportionality Assumption with the ASSESS Statement
    7. INTERACTIONS WITH TIME AS TIME-DEPENDENT COVARIATES
    8. NONPROPORTIONALITY VIA STRATIFICATION
    9. LEFT TRUNCATION AND LATE ENTRY INTO THE RISK SET
    10. ESTIMATING SURVIVOR FUNCTIONS
    11. TESTING LINEAR HYPOTHESES WITH CONTRAST OR TEST STATEMENTS
    12. CUSTOMIZED HAZARD RATIOS
    13. BAYESIAN ESTIMATION AND TESTING
    14. CONCLUSION
  8. 6. Competing Risks
    1. INTRODUCTION
    2. TYPE-SPECIFIC HAZARDS
    3. TIME IN POWER FOR LEADERS OF COUNTRIES: EXAMPLE
    4. ESTIMATES AND TESTS WITHOUT COVARIATES
    5. COVARIATE EFFECTS VIA COX MODELS
    6. ACCELERATED FAILURE TIME MODELS
    7. ALTERNATIVE APPROACHES TO MULTIPLE EVENT TYPES
      1. Conditional Processes
      2. Cumulative Incidence Functions
    8. CONCLUSION
  9. 7. Analysis of Tied or Discrete Data with PROC LOGISTIC
    1. INTRODUCTION
    2. THE LOGIT MODEL FOR DISCRETE TIME
    3. THE COMPLEMENTARY LOG-LOG MODEL FOR CONTINUOUS-TIME PROCESSES
    4. DATA WITH TIME-DEPENDENT COVARIATES
    5. ISSUES AND EXTENSIONS
      1. Dependence Among the Observations?
      2. Handling Large Numbers of Observations
      3. Unequal Intervals
      4. Empty Intervals
      5. Left Truncation
      6. Competing Risks
    6. CONCLUSION
  10. 8. Heterogeneity, Repeated Events, and Other Topics
    1. INTRODUCTION
    2. UNOBSERVED HETEROGENEITY
    3. REPEATED EVENTS
      1. Problems with Conventional Methods
    4. Robust Standard Errors
      1. Random-Effects Models
      2. Fixed-Effects Models
      3. Specifying a Common Origin for All Events
      4. Repeated Events for Discrete-Time Maximum Likelihood
    5. GENERALIZED R2
    6. SENSITIVITY ANALYSIS FOR INFORMATIVE CENSORING
  11. 9. A Guide for the Perplexed
    1. HOW TO CHOOSE A METHOD
      1. Make Cox Regression Your Default Method
      2. Is the Sample Large with Heavily Tied Event Times?
      3. Do You Want to Study the Shape of the Hazard Function?
      4. Do You Want to Generate Predicted Event Times or Survival Probabilities?
      5. Do You Have Left-Censored Data?
    2. CONCLUSION
  12. 1. Macro Programs
    1. INTRODUCTION
    2. THE LIFEHAZ MACRO
      1. Parameters
      2. Program
    3. THE PREDICT MACRO
      1. Parameters
      2. Program
  13. 2. Data Sets
    1. INTRODUCTION
    2. THE MYEL DATA SET: MYELOMATOSIS PATIENTS
    3. THE RECID DATA SET: ARREST TIMES FOR RELEASED PRISONERS
    4. THE STAN DATA SET: STANFORD HEART TRANSPLANT PATIENTS
    5. THE BREAST DATA SET: SURVIVAL DATA FOR BREAST CANCER PATIENTS
    6. THE JOBDUR DATA SET: DURATIONS OF JOBS
    7. THE ALCO DATA SET: SURVIVAL OF CIRRHOSIS PATIENTS
    8. THE LEADERS DATA SET: TIME IN POWER FOR LEADERS OF COUNTRIES
    9. THE RANK DATA SET: PROMOTIONS IN RANK FOR BIOCHEMISTS
    10. THE JOBMULT DATA SET: REPEATED JOB CHANGES
  14. REFERENCES