Kaplan Meier And Cox Proportional Hazards Modeling: Hands On Survival Analysis

Non-parametric survival analysis techniques are often used in clinical and epidemiologic research to model time at risk until event without parametric assumptions. This workshop will walk through the concepts of follow-up time, event time, the hazard function, the cumulative distribution function, incomplete data, censoring, time dependencies or temporal biases, plotting of survival curves, testing the proportional hazards assumption, and model diagnostics. Using SAS ® system's PROC LIFETEST, Kaplan Meier curves along with the log rank and Wilcoxon tests will be investigated to establish statistical differences in survival times between two groups. From there we will use the SAS ® system's PROC PHREG to run a Cox regression to model time until event while simultaneously adjusting for influential covariates and accounting for problems such as attrition, delayed entry, and temporal biases. The workshop will conclude with using the baseline option to calculate survival function estimates for graphing the cumulative probability of event over the follow-up period. This workshop is aimed at intermediate level statisticians, epidemiologists, and data analysts