Objective: Chronic kidney disease is a global public health concern. Glomerular filtration rate (GFR) prediction based on serum creatinine is used to assess renal function in the elderly. The Cockcroft-Gault (CG) formula, based on body surface area (CG/BSA formula), and the Modified Diet and Renal Disease formula (MDRD formula) are commonly used in assessing renal function in clinical practice. The objective of this study was to investigate the agreement between the GFR estimate, CG/BSA formula, and the MDRD formula in elderly outpatients.
Methodology: An outpatient chart review was conducted on consecutive elderly patients aged ≥ 65 years over a 9-month period. Data regarding age, gender, cognitive status, clock drawing, weight, height, and serum creatinine were collected. Pearson's correlation coefficient, Bland-Altman plot, and kappa statistics were used for statistical analysis.
Results: Of the 170 patients who participated in the study, 71% were cognitively impaired or had dementia. Using the CG/BSA formula, 79% of the patients had stage 3 renal disease (GFR< 60 mL/min); only 56% were diagnosed as such using the MDRD formula. There was a high correlation between the CG/BSA and MDRD formulas (Pearson correlation coefficient, 0.88; P < 0.0001). However, the kappa statistic was 0.47, indicating low agreement between the 2 formulas.
Conclusion: The diagnosis of having stage 3 chronic kidney disease depended on whether the CG/BSA or MDRD formula was used. A 17% discordance rate (9.61 mL/min) was seen between the MDRD and CG/BSA formulas for estimating GFR, and there was low agreement between the 2 formulas. Further studies are needed to assess which predictive formula is appropriate for elderly patients.