336 PART 6 Analyzing Survival Data

You may not be sure which variables in your data to include as predictors in the

regression. We provide advice on model-building in Chapter 17.

After you assemble and properly code the data, you execute the regression in sta-

tistical software using a similar approach as you use when doing ordinary least-

squares or logistic regression. You need to specify the variables in the regression

model:

1.

Specify the two outcome variables.

• ^{The event status variable}

• ^{The time-to-event variable}

2.

Specify the predictor variables.

Make sure you are careful when you include categorical predictors, especially

indicator variables. All the predictors you introduce should make sense

together in the model.

Most software also lets you specify calculations you want to see on the output. You

should always request at least the following:»

» ^{Coefficients table, including HRs and their confidence intervals»}

» ^{Tests of whether the hazard proportionality assumption is valid}

You may also want to request the following output:»

» ^{Summary descriptive statistics about the data. These can include number of}

censored and uncensored observations, median survival time, and mean and

standard deviation for each predictor variable in the model»

» ^{One or more measures of goodness-of-fit for the model»}

» ^{Baseline survival function, which outputs as a table of values and a survival}

curve»

» ^{Baseline hazard function values, which output as a table and graph}

After you specify all the input to the program, execute the code, retrieve the out-

put, and interpret the results.