Lower surgical risks in "contemporary" IBD cohorts: a reflection of better quality of care or statistics?