Create a data table

From the Welcome or New Table dialog, choose to create a multiple variables data table. If you’re just getting started with Cox proportional hazards regression, and don’t have your own data, you can choose to start with a sample data set.

Enter the data

As with other analyses performed from the multiple variables data table, the structure of the data must follow a common structure. Each row of the data table represents a different observation, individual, animal, participant, etc. Each column of the data table represents a different variable.

For Cox proportional hazards regression, at least three variables are absolutely required:

1.Elapsed time: the duration of time for each individual/observation until the event of interest or time of censoring. This elapsed time can be in any convenient units, for example days, weeks, or months. The values entered for elapsed time do not need to be rounded to an integer value, and can be entered as a decimal. Currently, it’s not possible to enter starting time and ending time for a single observation and have Prism calculate the duration

2.Event/Censor classifier: a variable whose values indicate if the individual/observation experienced the event of interest or was censored. This variable can be either a continuous or categorical variable. Often, this information is coded as a continuous variable with values of "1" representing individuals that experienced the event of interest and values of "0" representing individuals that were censored. However, a categorical variable with levels of "Died" and "Censored" would work just as well. In either case, controls on the model tab of the analysis parameters dialog will allow you to specify which value (or level) represents events and which represents censored observations. Additional controls will also allow you to indicate how to handle any other values in the selected variable (they can be treated as missing, treated as events, or treated as censored observations)

3.At least one predictor variable: additional variables (categorical or continuous) can be included in Cox regression, and each variable should be entered into its own column. Note that there is no need for you to encode categorical variables manually. Instead, you can simply enter text labels to identify the level of the categorical variable that the observation belongs to. If entering data this way, be sure that the variable type in the data table is set to categorical. Note that these additional predictor variables include

Finally, note that there is no need to code for interactions in the model manually. In addition to specifying the predictor variables to include in the model, Prism will also allow you to specify two-way and three-way interactions if you choose to include them.