Stratification vs. Adjustment

[djbateman]

This is more of a general statistical question rather than SAS specific.

We run a lot of survival curves on clinical trial data, and we get hazard ratios to accompany those curves.  When we make no changes to the data, we get an unadjusted hazard ratio.  I have been asked to run stratified and adjusted hazard ratios as well.  I know how to code these in SAS, but I have a hard time explaining them.  I have asked others in the company, but I'm not getting too far.  Can someone clearly explain the difference between stratified and adjusted?

I believe stratified is running a bunch of hazard ratios based on different strata (e.g., race, gender, etc.).  Then SAS pools those to give a more precise hazard ratio than the unadjusted hazard ratio.  Is that correct?  If so, great, but I have no idea how to explain adjusted analyses.