Relative risk and risk factors

[iressa13]

When calculating relative risk, I adjusted for a known risk factor but the relative risk did not change from my crude. My instructor confirmed that my code was correct so does this just mean that the known risk factor isn’t actually a risk fact? I’m a bit confused and sorry that this isn’t SAS programming related but any help would be appreciated thank you!

[ballardw]

It never hurts to share the code you used. Depending on procedure used and "adjustment" method we might be able to provide some targeted things to examine. Right now all I would say is "something in the data". If your adjustment is used typically on a population as a whole such as smoking status in health studies it may not still be appropriate if your sample is all lung cancer positive individuals.

If there were any messages in the log other than x records read from data set lib.data and execution time you may want to include that as well. 

[Reeza]

---

@iressa13 wrote:

When calculating relative risk, I adjusted for a known risk factor but the relative risk did not change from my crude. My instructor confirmed that my code was correct so does this just mean that the known risk factor isn’t actually a risk fact? I’m a bit confused and sorry that this isn’t SAS programming related but any help would be appreciated thank you!

---

The code may be correct, but there's a few other reasons this can occur and still be wrong, a few are:

• data didn't change correctly - ie didn't point at new dataset
• selection method was used - effect wasn't included in model overall
• user error - you looked at the wrong output
• missing data - new factor is missing data