Does the ckd epidemiology collaboration (CKD-EPI) equation perform better than the modification of diet in renal disease (MDRD) study equation for estimating GFR levels abos 60 mL/min/1.73m(2) in subjects with diabetes?