Excellent Tutorial: do you use the same method if you want to check the interaction with a covariate that has 3 or more categories (e.g. interaction with smoking: 1)current smoker; 2)former smoker; 3)never smoker. How would you manage to test for this kind on interaction. Thank you so much in advance for your precious help.
Thank you for the video. Something is wrong with taking the hazard ratio of group 1 (.365) and dividing by the hazard ratio for group 2 (.879), which is equal to .4152, not the interaction ratio of .399. The numbers are close, but not exact. I am not sure if the relationship holds like that.
So when the assumption of Cox proportion regression is not met, what test should we use instead of it? Or we don't need to worry about the assumption? Because when you plot the K-M curve, it clearly crossed over, but you conducted the cox regression anyway. So, is it not a big deal?
Cox regression computes the average effect of a treatment over time, even if KM plots cross-over, Cox computes the average effect. So it is your call to consider it is a big deal or not with computing an average effect when KM crossover. You can compute HR for each time-interval where effect is relatively stable, for example, compute HR before 7 years, and another HR for the time after 7 years.
I have one doubt.. If not used categorical value it (spss) will consider last (1-surgery) as reference value. Then how can u tell 23% reduction in CBAG. Kindly clarify me
I usually put 0/1 bianary exposure variable in Covariate box, instead of putting it in a Factor box in SPSS. This way, CABG=0 becomes a reference level thus HR=0.77 refer 23% risk reduction in CABG.
I love your examples. I don't use SPSS but that doesn't matter. You are a great teacher.
You are always impressive in teaching the spss including statistics. I really do appreciate.
This tutorial is very clear, simple and easy to understand. I got valuable knowledge and helps from that video. I really appreciate you. Thanks a lot!
Thank you for your excellent tutorials
THANK YOU FOR YOUR EXPLANATION. YOU ARE SO ANSOLUYLY EXCELENT
Nice explanation ❤. Thanks .
Excellent Tutorial: do you use the same method if you want to check the interaction with a covariate that has 3 or more categories (e.g. interaction with smoking: 1)current smoker; 2)former smoker; 3)never smoker. How would you manage to test for this kind on interaction. Thank you so much in advance for your precious help.
Thanks for your tutorial.
One thing, I’m wondering how I can check that the variables fulfil the proportional hazard assumption.
Please see this video. ua-cam.com/video/s_tUi7mAJ44/v-deo.html see this video.
tqvm
Thank you for the video. Something is wrong with taking the hazard ratio of group 1 (.365) and dividing by the hazard ratio for group 2 (.879), which is equal to .4152, not the interaction ratio of .399. The numbers are close, but not exact. I am not sure if the relationship holds like that.
InstaBlaster...
Thinks
So when the assumption of Cox proportion regression is not met, what test should we use instead of it? Or we don't need to worry about the assumption? Because when you plot the K-M curve, it clearly crossed over, but you conducted the cox regression anyway. So, is it not a big deal?
Cox regression computes the average effect of a treatment over time, even if KM plots cross-over, Cox computes the average effect. So it is your call to consider it is a big deal or not with computing an average effect when KM crossover. You can compute HR for each time-interval where effect is relatively stable, for example, compute HR before 7 years, and another HR for the time after 7 years.
@@ayumishintani7044 Thank you sooo much!
I have one doubt.. If not used categorical value it (spss) will consider last (1-surgery) as reference value.
Then how can u tell 23% reduction in CBAG. Kindly clarify me
I usually put 0/1 bianary exposure variable in Covariate box, instead of putting it in a Factor box in SPSS. This way, CABG=0 becomes a reference level thus HR=0.77 refer 23% risk reduction in CABG.