P8400 A PPLIED D ATA A NALYSIS IN E PIDEMIOLOGY LAB 8 C OX P ROPORTIONAL H AZARDS R EGRESSION A NSWER K EY DATASET TO BE USED: HIVHepC.sas.7bdat Please note: This lab uses the same dataset as in Lab 7.The data dictionary is provided below. Objectives: 1. To perform Cox proportional hazards regression 2. To interpret effect estimates from the model 3. To test for mediation and interaction using Cox proportional hazards models 4. To assess the proportional hazards assumption using two methods In this data set, the following variables are provided: Table 1: Data Dictionary of the HIVHepC dataset Variable Description Range patient Identifier for the Amsterdam Cohort Studies participant Random number MSM Men who have sex with men 0=No 1=Yes IVDrugUse History of IV Drug Use 0=No 1=Yes ART_init Has ART been initiated? 0=No 1=Yes HepC Is the patient HepC positive? 0=No 1=Yes cohort_date Date of enrollment into cohort anonymous (changed to 1st day of month) – HIV diagnosis dd-mm-yyyy death Has the subject died? 0=No 1=Yes last_alive_date Date last alive anonymous (changed to 1st day of month) dd-mm-yyyy death_date Date of death anonymous (changed to 1st day of month) dd-mm-yyyy age Age of patient at the last clinical visit 99 (years) HIV_diag_1996 Was HIV diagnosed before 1996? 0=No 1=Yes
P ART I: U NIVARIABLE C OX R EGRESSION 1.What is a hazard ratio? 2.Run a Cox regression model to investigate the association between survival and having a Hepatitis C diagnosis among HIV+ males and interpret the results. (You’ll need the pyears variable you created in Lab 6 - the code follows in case you forgot to save it.) ; ;
You've reached the end of your free preview.
Want to read all 12 pages?
- Fall '09
- Survival analysis, Hazard ratio, Proportional hazards models, msm