Assessment of Renal Function and Predictive Performances of GFR Estimating Equations among Nigerian Diabetic Patients

Aims: To assess the renal functions in Nigerian diabetic patients and to examine the predictive performances of Glomerular Filtration Rate (GFR) estimating equations.
Study Design: A case-control study.
Place and Duration of Study: Department of Physiology and University College Hospital, University of Ibadan, Ibadan, Nigeria. May-August, 2009.
Methodology: One hundred and nine volunteers comprising 58 diabetic patients receiving treatments and 51 healthy individuals. Measured GFR (mGFR) was by creatinine clearance and the equations includes Cockcroft and Gault, CG; Modification of Diet in Renal Disease, MDRD study equation; Chronic Kidney Disease and Epidemiological study group, CKD-EPI and Mayo Clinic Quadratic, Q equation. Ethnicity factor was administered as appropriate. Performances were determined by mean bias, precision and accuracy.
Results: mGFR was significantly (P=.05) reduced among the diabetic when compared with the non-diabetic though within the recommended range for normal renal function. Among the diabetics, CG equation has the least bias when compared with the mGFR but overestimated the GFR by 2.42ml/min/1.73m2 while Q has the highest bias. When the bias of other equations where compared with that of CG, the CKD/EPI formula significantly underestimated the GFR (P=.05) and the Q significantly overestimated GFR (P=.05). The highest precision was by CG and the least was found in the CKD/EPI though not significantly. The highest accuracy in this group was by CKD/EPI. In the non-diabetics, the least bias was recorded in the MDRD when compared with the mGFR while the highest was recorded in the CKD/EPI, the bias when compared with that of CG, the CKD significantly underestimated GFR by up to 7.54ml/min/1.73m2 (P=.001). Precision was highest in the Q though, not significant while its accuracy was significantly lower (P=.05) when compared with the CG. Adjustment for the ethnicity factor significantly overestimated GFR in our two study groups.
Conclusion: Creatinine-based predictive equations are useful in estimating renal functions but the CG as well as the MDRD equations are more superior in their predictive ability among Nigerians and the use of the ethnicity factor is not recommended in Nigerian African as there is overestimation when used with the relevant equations.