When I just integrate it from 0 to 2pi and make all the subs I get:

lil' of this, lil' of that and I get:

and only one is in the contour (via Mathematica Abs command) giving a residue of

then I get:

which agrees with the numerical result. Yours is the negative of that.