Clinical Relevance of Differences in Glomerular Filtration Rate Estimations in Frail Older People by Creatinine- vs. Cystatin C-Based Formulae.

Abstract

The risk of incorrect medication dosing is high in frail older people. Therefore, accurate assessment of the glomerular filtration rate is important. The objective of this study was to compare the estimated glomerular filtration rate using creatinine- and cystatin C-based formulae, the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations, in frail older people. We hypothesized that frailty determines the difference between the creatinine- and cystatin C-based formulae. The mean difference between CKD-EPI creatinine and cystatin C was determined using (cross-sectional) data of 55 patients (mean age 73years) admitted to a psychiatric ward for older adults. The level of agreement of these estimations was assessed by a Bland-Altman analysis. In all patients, the Rockwood's Frailty Index was derived and correlated with the mean difference between CKD-EPI creatinine and cystatin C. The mean difference between CKD-EPI creatinine (mean 71.2mL/min/1.73m2) and CKD-EPI cystatin C (mean 57.6mL/min/1.73m2) was 13.6mL/min/1.73m2 (p<0.0001). The two standard deviation limit in the Bland-Altman plot was large (43.2mL/min/1.73m2), which represents a low level of agreement. The Frailty Index did not correlate with the mean difference between the creatinine- and cystatin C-based glomerular filtration rate (Pearson correlation coefficient 0.182, p=0.184). There was a significant gap between a creatinine- and cystatin C-based estimation of glomerular filtration rate, irrespective of frailty. The range of differences between the commonly used estimated glomerular filtration rate formulae might result in clinically relevant differences in drug prescription and differences in chronic kidney disease staging.