Performance and comparison of the Cockcroft-Gault and simplified Modification of Diet in Renal Disease formulae in estimating glomerular filtration rate in a Chinese Type 2 diabetic population

Aims Our aim was to assess performances of the Cockcroft-Gault and simplified Modification of Diet in Renal Disease (MDRD) formulae in estimating glomerular filtration rate (GFR) in Chinese diabetic populations and their association with vascular risks. Methods A total of 1009 patients with Type 2 diabetes were categorized into low estimated GFR groups (GFR < 60 ml/min/1.73 m(2)) and control groups by the two equations. The performances of these formulae were assessed at different stages of kidney function. Carotid artery intima-media thickness (IMT) and the prevalence of diabetic retinopathy or albuminuria were compared among the groups. The ability of these formulae to identify established vascular risk markers using sensitivity, specificity, positive and negative predictive values were also compared. Results The prevalence of low estimated GFR was 32.7% with the Cockcroft-Gault formula and 5.2% with the MDRD formula, respectively. In low estimated GFR subjects by the MDRD formula, IMT was significantly thicker than those by the Cockcroft-Gault formula (1.2 mm vs. 1.0 mm; P < 0.05), with a higher prevalence of albuminuria (78.4 vs. 52.8%, P < 0.05) and diabetic retinopathy (46.5 vs. 30.5%; P < 0.05). The Cockcroft-Gault formula gave a specificity of 71.7% and a sensitivity of 37.0%, and the MDRD formula gave a specificity of 96.6% and a sensitivity of 7.9% in estimating low GFR relevant for established vascular risks. Conclusions These formulae performed differently in Chinese diabetic populations. The simplified MDRD formula is minimally superior to the Cockcroft-Gault formula for its high specificity and positive predictive values in estimating low GFR relevant for vascular risks.