Programming the statistical procedures from SAS

hazard ratio from regression coefficients in an interaction term

Reply
Contributor
Posts: 29

hazard ratio from regression coefficients in an interaction term

[ Edited ]

I am using proc phreg to estimate hazard ratios, my model is as follows: 

 

time*event(0) = week|avg education race;

random zipcode;

 

I would like to get the hazard ratio for 'week'. I was originally doing this by hand by summing the estimate from week and the interaction term, but assume there is a way to do this using estimate/contrast? 

 

Thank you!

Super User
Posts: 23,705

Re: hazard ratio from regression coefficients in an interaction term

Try a HAZARDRATIO statement instead. Much simpler syntax. 

Regular Contributor
Posts: 159

Re: hazard ratio from regression coefficients in an interaction term

youdont have week as a main factor in the model, just in the interaction? you'll need to use 'at' in the hazardratio statement @Reeza suggested, see here: https://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewer.htm#statug_phreg_sect0...

--------------
blog: papersandprograms.com
Contributor
Posts: 29

Re: hazard ratio from regression coefficients in an interaction term

Posted in reply to PaulBrownPhD

When I use the hazardratio statement I cannot recreate what I get when I am summing the coefficients by hand. This is why I wasn't using it initially - is there a specification that needs to be made?

 

Example:

 

 EstimateExp summed estimateHazard Ratio from SAS
average0.0001960.9997636280.995
week-0.0002121  
week*average-0.0000243  
Regular Contributor
Posts: 159

Re: hazard ratio from regression coefficients in an interaction term

are you doing it by hand or in a data step? it could be rounding error, calculate the hr in a data step as exp(estimate). Also did you use 'at' in the hazard ratio statement? if you just sum them up then it's at average=1?

--------------
blog: papersandprograms.com
Super User
Posts: 23,705

Re: hazard ratio from regression coefficients in an interaction term

99% of the time the error is with the hand calculation. 

 

Do it with a data step so you can't get any typo's or rounding errors and post your code if it still doesn't match.

 

 

Ask a Question
Discussion stats
  • 5 replies
  • 115 views
  • 0 likes
  • 3 in conversation