Surgery volume, quality of care and operative mortality in coronary artery bypass graft surgery: a re-examination using fixed-effects regression