Hazard function for Weibull distribution

CathyVI — Fri, 02 Dec 2022 16:32:02 GMT

Hi,

How do I find the hazard function estimate for a categorical variable using weibull distribution.

My categorical variable is categorical variable is Tumt: 0=large,1=squamous,2=small,3=adeno

I want hazard function for patients with large tumors (large =0).

I see that proc lifereg can be used to calculate weibull distribution but I don't know how to get the hazard function for a specific categorical variable or can I use proc lifetest? if yes how do I code it pls.

This is my code:

proc lifereg data=Cancer ;
class tumt ;
model survdays*dead(0) = treat tumt / dist=Weibull;
probplot;
inset;
run;

Re: Hazard function for Weibull distribution

webart999ARM — Wed, 07 Dec 2022 02:05:23 GMT

To calculate the hazard function estimate for a categorical variable using the Weibull distribution in SAS, you can use the PROC LIFEREG procedure. The PROC LIFEREG procedure allows you to fit a parametric survival model to your data, and then use the fitted model to estimate the hazard function for different groups of subjects.

In your case, you want to calculate the hazard function for patients with large tumors (tumt = 0). To do this, you can use the PROC LIFEREG statement to fit a Weibull model to your data, and then use the HAZARD option in the MODEL statement to request that the procedure estimate the hazard function for each level of the tumt variable. Here is an example of how you might do this in SAS:

PROC LIFEREG DATA=Cancer;
  CLASS tumt;
  MODEL survdays*dead(0) = treat tumt / DIST=WEIBULL HAZARD;
  PROBPLOT;
  INSET;
RUN;

In this example, the PROC LIFEREG statement specifies that you want to fit a Weibull model to the data in the Cancer dataset. The CLASS statement indicates that the tumt variable is a categorical variable, and the MODEL statement specifies the model that you want to fit. The HAZARD option in the MODEL statement tells SAS to estimate the hazard function for each level of the tumt variable.

After running this code, you can use the PROBPLOT and INSET statements to create a probability plot that shows the estimated hazard functions for each level of the tumt variable. You can then use this plot to compare the hazard functions for different tumor types and identify any differences in the hazard of death for patients with different tumor types.

Alternatively, you can use the PROC LIFETEST procedure to calculate the hazard function for a specific group of subjects. For example, if you want to calculate the hazard function for patients with large tumors (tumt = 0), you could use the following code:

PROC LIFETEST DATA=Cancer;
  TIME survdays*dead(0);
  STRATA tumt;
  TEST / HAZARD;
RUN;

In this code, the PROC LIFETEST statement specifies that you want to perform a survival analysis on the data in the Cancer dataset. The TIME statement indicates the time-to-event variable and the event variable, and the STRATA statement indicates that you want to stratify the data by the tumt variable. The TEST statement specifies that you want to calculate the hazard function for each stratum (i.e. for each level of the tumt variable).

I hope this helps! Let me know if you have any other questions.

topic Re: Hazard function for Weibull distribution in Statistical Procedures

Hazard function for Weibull distribution

Re: Hazard function for Weibull distribution