Skip to main content

Survival Score Tool Icon Survival Score Tool

The Survival Score tool provides both the estimated relative risk and restricted mean survival time based on a Cox proportional hazards model, which can be estimated using the Survival Analysis tool.

The estimated relative risk (returned in the output field "Risk") provides the expected probability the unit described in a record will not survive to the next time period relative to a unit that possessed the mean characteristics in the sample used to estimate the model. For example, if the returned value for a particular record is 2.4, it means that the entity described by that record is 2.4 times more likely not to survive to the next period than an entity that possessed the mean of the predictors in the data used to estimate the model.

The restricted mean survival time is the mean number of time periods that a unit described in a record will survive over a user-specified (longer) time period and is returned in the output field "RMST_<time periods> where <time periods> is the number of time periods on which the mean is based. For instance, if the time measure is in days, and the specified time period is 365 days (a year), then the returned values (which will appear in the field RMST_365) will be between 0 and 365. A value of 249 would indicate the entity described in a record is expected to survive 249 days out of the 365-day time period.

Important

This tool is not automatically installed with Alteryx Designer or the R tools. To use this tool, download it from the Alteryx Community Gallery.

Connect Inputs

  • A Cox proportional hazards model object produced by the Survival Analysis tool.

  • A data stream that is consistent with the associated Cox proportional hazards model in terms of field names and types

Configure the Tool

  • The number of time units to use as the basis for the restricted mean survival time: The number of time units is an integer value.

  • The number of records to score at a time: The tool can break the input data into chunks, scoring a chunk at a time, and thereby avoid R's in-memory processing limitation. This option controls the number of records contained in each chunk of data.

View the Output

The original data stream with the relative risk and restricted mean values based on the Cox proportional hazards model appended to the data.