Estimativa do ritmo de filtração glomerular em pacientes obesos: análise comparativa entre fórmulas

Abstract

Introduction: Obesity and Chronic Kidney Disease (CKD) are two relevant problems for public health worldwide. Obesity is a recognized risk factor for the development and progression of CKD. The assessment of renal function by estimating the Glomerular Filtration Rate (GFR) is a challenge in patients with obesity, as there are conflicting data on the accuracy of the formulas currently available in this population. Objective: Compare different formulas used to estimate the GFR in obese individuals in different ranges of Body Mass Index (BMI) and renal function, evaluate the influence of ideal body weight and body surface area (BSA) on them and compare the results obtained by the calculations with 24-hour urine creatinine clearance. Methods: Cross-sectional analytical study carried out in a single nephrology center, with individuals over 18 years old, who had a diagnosis of obesity and different degrees of renal dysfunction. Direct measurement of GFR by 24-hour urine creatinine clearance and estimated GFR from plasma creatinine by various formulas were evaluated, using current and ideal body weight and calculated body surface area. Pearson's correlation test, Bland-Altman graphs, ROC curves were used. Measures of bias, relative bias (error %), sensitivity, specificity, positive and negative predictive values, and accuracy were calculated for all formulas. It was considered significant p < 0.05. Analyzes were performed using the SAS 9.4 program. Results: The final sample consisted of 45 patients, 51.11% men, with a mean age of 55.93 ± 11.39 years; 66.67% hypertensive and 40% diabetic. The mean BMI was 35.87 ± 4.62 kg/m2 , and the majority was composed of grade I obese individuals (48.89%). The GFR measured by 24-hour urine had a mean of 73.38 ± 35.92 ml/min/1.73m2 . The formula that showed greater accuracy for patients with grade I obesity was SalazarPI (% error = 2.30 ± 33.92), followed by Cockroft-GaultPI (% error = -2.84 ± 32.76). For patients with grade II and III obesity the most accurate was CKDEPI (% error = 3.84 ± 41.79) followed by MDRD (% error = 4.00 ± 38.43). When using the ideal body weight in the Cockroft-Gault, Sobh, and Salazar formulas, the mean GFR was closer to the 24-hour urine creatinine clearance. In grade I obesity Cockroft-Gault adjusted for ideal body weight and calculated AUC showed less bias than the original formula. In obesity grades II and III, the formula with the least bias was SalazarPI. In the range of RFG < 60 ml/min, the most accurate formula was Salazar. When using the ideal weight in the formulas there was na increase in the Pearson correlation. Conclusion: The results demonstrate that in this sample the use of ideal body weight improved Cockroft-Gault performance, especially in grade I obesity. In more advanced grades of obesity CKD-EPI and MDRD showed the best performance.