Background and Importance Glomerular filtration rate (GFR) is usually estimated by using renal markers like creatinine (cr) or cystatin C (cysC), but results are not always overlapping.

Aim and Objectives Evaluate the effect of using Cockroft-Gault (CGcr) and Chronic Kidney Disease-Epidemiology Collaboration (EPIcr, EPIcysC and EPIcr/cysC) equations in vancomycin monitoring.

Material and Methods Data from the last 5 years were collected retrospectively. All patients (n=34) who had simultaneously cr, cysC and observed vancomycin concentrations (Cobs) obtained within a range of 48h (n=47), were included. Pharmacokinetic Bayesian estimation was performed with PKS – Abbott^®. For each pair GFRs/Cobs, the predicted concentration (Cp) and the daily dose required to obtain a maximum and minimum concentration of 25 and 15 µg/ml, respectively, were determined. The absolute error (E), E=Cobs-Cp=0, was used as an indicator of the adequacy of the equations used.

Results Estimated GFR showed statistically significant differences (mean ± standard deviation): CGcr=110.6 ± 76.5, EPIcr=97.5 ± 36.3, EPIcysC=42.8 ± 18.6 and EPIcr/cysC=64.3 ± 25.2 ml/min/1.73 m² (p<<0.05).

CGcr, EPIcr and EPIcr/cysC equations overestimated (E>0) renal function: E=1.50 ± 1.53 (95% confidence interval [CI]: 1.05 to 1.95), E=1.62 ± 1.35 (95% CI: 1.22 to 2.02) and E=0.47 ± 1.14 (95% CI: 0.14 to 0.81) µg/ml, respectively. Renal function was underestimated (E<0) with EPIcysC, E=-1.06 ± 1.54 (95% CI: -1.51 to -0.60) µg/ml.

The estimated differences in daily doses ranged from 100 to 1600 mg/70Kg/day, considering CGcr equation as reference.

Conclusion and Relevance The overestimation of GFR with equations dependent on cr, CGcr, EPIcr and, to a lesser extent, EPIcr/cysC, was marked in patients with abnormally low cr. Conversely, with EPIcysC equation, which depends on cysC, a biomarker independent of muscle mass, GFR was underestimated. This may be due to factors that increase cysC, without renal function impairment, such as hypertension, corticosteroid therapy and malignancy, all common in hospitalised patients, but poor data did not allow to explore this association.

The differences in the GFR estimates are clinically relevant on dosing adequacy, being suggestive that in the presence of abnormally low cr, equations with cysC are preferred.

Studies are needed to identify the variables responsible for the observed variability, in order to previously select the most appropriate equation for each case.

References and/or Acknowledgements 1. Teaford HR, et al. Cystatin C: A Primer for Pharmacists. Pharmacy. 2020 Mar; 8(1):35

Conflict of Interest No conflict of interest.