Different methods for assessing glomerular filtration rate in the elderly

OBJECTIVE: The objective of this study was to identify the best method to replace cystatin C in the evaluation of glomerular filtration in the elderly. METHODS: Individuals over 60 years of age from a primary care center were studied. Blood was collected to determine creatinine and cystatin C and 24-h urine. Three methods were compared to determine glomerular filtration: Creatinine clearance, Cocroft-Gault, modification of diet in renal disease, and Collaboration Epidemiology of Chronic Kidney Disease based on creatinine, considering as a reference the determination of glomerular filtration using the cystatin-based Chronic Kidney Disease Epidemiology Collaboration equation. The statistical methods used were linear regression, Bland-Altman curve, and receiver operating characteristic. RESULTS: A total of 180 elderly people were evaluated, but 14 patients were lost from the sample, resulting in a total of 166 patients. The average age of patients was 66.9±6.1 years, and 69.8% were females. Regarding the number of patients eligible for the study, there were 12 black, 108 brown, and 46 white, 42.77% hypertensive, and 38.3% diabetic. Glomerular filtration was less than 60 mL/min in 22.28% of patients. Regarding the evaluation of the different equations, the correlation coefficient was lower for creatinine clearance and progressively higher for Cocroft-Gault, modification of diet in renal disease, and Collaboration Epidemiology of Chronic Kidney Disease based on creatinine. The Bland-Altman diagram and the receiver operating characteristic curve showed similar performance to the correlation coefficient for the different equations evaluated. CONCLUSION: Collaboration Epidemiology of Chronic Kidney Disease based on creatinine presented the best performance. Creatinine debug had the worst performance, which reinforces the idea that 24-h urine collection is unnecessary in these patients.