Analysis of pre-post ordinal data

Hi everybody,
I'm stuck in a problem with ordinal data analysis.
We have x patients treated, and two measurements of pain (before/after treatment) in a 10-point scale.
Which is the suitable test/tests to evaluate response to treatment?


Less is more. Stay pure. Stay poor.
Do you want to examine each measurement by itself? So pain score differences for each measurement?

Did you want to compare measurements? Did all patients get both measurement before/after treatment?

Please answer these questions and provide some more details, such as research question and hypothesis.