Survival Analysis

Survival Analysis

Survival Analysis Table

• time: Time of event/death
• at_risk: Number of people remaining in the study
• censored: Number of people removed from the study for reason other than event/death
• observed: Number of people for which the event/death was observed
• removed: censored + observed
• entrance: Number of people in the beginning of the study
• KMestimate: Kaplan-Meier Estimate for the event
• lower_0.95: Lower limit of 95% confidence interval of the KM estimate
• upper_0.95: Upper limit of 95% confidence interval of the KM estimate

The Survival Analysis Table is the concatenation of each of the above variables for all the different conditions.

Cox Proportional Hazard Table

The Cox Proportional Hazard takes into account how covariates affect the survival probability function.

• partial log-likelihood: log likelihood of the fitted coefficients. The higher the value, the better the model.
• Concordance: Evaluates the predictions made by the CPH model. A higher value represents better predictions from the model and vice versa.
• Partial AIC: AIC value at the fitted coefficients. AIC estimates the quality of each model, relative to each of the other models. A good model is the one that has minimum AIC among all the other models.
• log-likelihood ratio test: Measures the goodness of fit for a model. Compares the model to the trivial model of no covariates. Higher value represents a better fitted model. Value between -inf to inf
• log rank test: Compares different survival distributions. Thus a P.value(log rank test) < 0.05 indicates a significant difference between the different population survival curves

The bottom table shows the results of the fitted model.

• coef: Coefficient value fitted to the covariate
• exp(coef): Hazard ratio for the covariate
• se(coef): Standard deviation of the coefficient
• coef lower 95%: Lower limit for the confidence interval of the covariate coefficient
• coef upper 95%: Upper limit for the confidence interval of the covariate coefficient
• exp(coef) lower 95%: Lower limit for the confidence interval of the covariate hazard ratio
• exp(coef) upper 95%: Upper limit for the confidence interval of the covariate hazard ratio
• z: Z-score
• p: p value

More information and explanations can be found here