Estimación del filtrado glomerular en personas mayores de 85 años: comparación de las ecuaciones CKD-EPI, MDRD-IDMS y BIS1

Discrepancies among equations to estimate the glomerular filtration rate for drug dosing decision making in aged patients: a cross sectional study

BACKGROUND

Patients often require adjustments to drug doses due to impaired renal function. Glomerular filtration rate (GFR) estimation using various equations can result in discrepancies, potentially leading to different dose adjustment recommendations.

AIM

To determine the clinical significance of discrepancies observed between different equations used to estimate GFR for drug dose adjustments in a real-world group of patients over 65 years in primary care.

METHOD

The Cockcroft-Gault (CG), Modification of Diet in Renal Disease (MDRD), Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI), and Berlin Initiative Study 1 equations were applied to estimate GFR in a group of patients over 65 years old attending a primary care center. Results were compared using Bland-Altman plots, and limits of agreement (LoA) and overall bias were calculated. Regression analyses were conducted to identify the null difference GFR and the slope of differences for each pairwise comparison.

RESULTS

A total of 1886 patients were analyzed. Differences between patient-adjusted and body surface area (BSA)-normalized versions of the equations were not clinically relevant for dose adjustments, with LoAs below 20 mL/min. However, discrepancies among the original versions of several equations presented LoAs over 30 mL/min. Greater differences were found between CG and MDRD or CKD-EPI equations.

CONCLUSION

Clinically relevant differences in GFR estimation were observed among different equations, potentially impacting drug dose adjustments. However, discrepancies were not considered significant when comparing patient-adjusted and BSA-normalized versions of the equations, particularly for patients with BSA close to the average.

Does Estimated Glomerular Filtration Rate Predict In-Hospital Mortality in Acutely Unwell Hospitalized Oldest Old?

Globally the population of older adults is the fastest growing age group. Estimated glomerular filtration rate (eGFR) is an estimation of true kidney function with lower eGFR associated with higher mortality. However, few studies explore eGFR's prognostic value in the nonagenarian. We investigated the association between eGFR on admission and mortality among the nonagenarians hospitalised with acute illness. A retrospective analysis of a prospective cohort study included patients aged ≥ 90 admitted into three acute medical assessment units or acute geriatric wards in England and Scotland between November 2008 and January 2009. Association between eGFR and all-cause mortality was evaluated using the Cox proportional hazard models controlling for potential confounders including frailty. 392 patients with mean (SD) 93.0 ± 2.6 years (68.45% women) were included. The median (IQR) eGFR was 26.61 (18.41-40.41) mL/min/1.73². 63 died in in hospital. Low eGFR was not associated with mortality (Hazard ratio (HR) 1.00 (95% CI 0.98-1.02) overall or in sub-group analysis by frailty (HR 0.96 (0.92-1.01)) or by eGFR of ≤30 (HR 1.01 (0.95-1.06). We found no evidence of prognostic value of eGFR in predicting in-hospital mortality in the acutely unwell hospitalised nonagenarians.

[Estimation of glomerular filtration rate in the elderly patient: Clinical implications in the use of antibiotics]

INTRODUCTION

In elderly, impaired kidney function may lead to relevant clinical implications, especially in drug dosage. There is no consensus on the best formula to estimate glomerular filtration rate (GFR) in this context. While the Cockcroft-Gault (CG) equation is the most used one in drug development, the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation is the most employed by clinical laboratories. The objective of this study is to analyze the differences and implications of using one or another equation when adjusting the dose of antibiotics.

MATERIAL AND METHODS

Cross-sectional study performed in patients older than 80 years admitted to an acute geriatric unit for infectious disease. Socio-demographic, functional, clinic, assistance and laboratory variables were gathered, calculating GFR according to CG and CKD-EPI. A concordance study between both equations, bivariate analysis and multiple linear regression were carried out.

RESULTS

100 patients were recruited, with mean age 88.5±4.2 years, and 54% women. A relative overestimation of 13.63mL/min was observed in GFR according to CKD-EPI, with a low-moderate concordance between both formulas. Greater need for dose adjustment of antibiotics with GC than with CKD-EPI was detected (38% vs. 23%; P<.001), as well as dosage discrepancies, especially with levofloxacin.

CONCLUSIONS

Both equations are not interchangeable in the elderly. When used to adjust antibiotic dose, CKD-EPI may overestimate GFR and may cause an increase in adverse reactions. Further studies should be carried out with the aim of reaching a consensus on the most appropriate equation for the elderly patient.

[Concordance between the equations "Chronic Kidney Disease Epidemiological Collaboration" and "Modification of Diet in Renal Disease" with the "Berlin Initiative Study" to estimate renal function in the elderly]

OBJECTIVE

Estimated glomerular filtration rate (eGFR) is calculated routinely using creatinine-based formulas, but their reliability in the elderly is limited. The aim of this study was to analyse the concordance between the BIS1 equation which is specific for the elderly, and the usual CKD-EPI and MDRD-IDMS in a large population over 70 years of age.

MATERIAL AND METHODS

Retrospective cross-sectional study in which the eGFR was calculated using BIS1, CKD-EPI and MDRD-IDMS equations based on gender, age, and creatinine data of 85,089 subjects (58.5% women, mean age 78 years [IQR 73-83]).The following statistics were carried out: Wilcoxon test, Bland-Altman graphic analysis, study of the concordance using the intraclass correlation coefficient (ICC), and comparison tables for the classification of CKD.

RESULTS

The median of the eGFRs using BIS1 was 58mL/min/1.73m² (IQR 48-70), using CKD-EPI was 68mL/min/1.73m² (IQR 53-84), and using MDRD it was 68mL/min/1.73m² (IQR 53-82). The concordance between BIS1 and CKD-EPI (intraclass correlation coefficient =0.87) was found to be acceptable. It was lower with MDRD (intraclass correlation coefficient =0.81). A mean difference of 8mL/min/1.73m² (SD 2.6-18) was found BIS1 vs. CKD-EPI, and 10mL/min/1.73m² (SD 6-27) with BIS1 vs. MDRD, which was maintained when stratifying by gender and age groups.

CONCLUSIONS

Despite the acceptable statistical agreement, the eGFR obtained with the BIS1 equation is not interchangeable with CKD-EPI or with MDRD-IDMS. The BIS1 equation gives lower values than CKD-EPI, and classifies patients into a higher level of CKD, mainly when the eGFR is above 30mL/min/1.73 m².

[Estimated glomerular filtration rate and post-hospital mortality after discharge from an acute care for the elderly unit]

OBJECTIVE

To determine the relationship between estimated glomerular filtration rate (eGFR) and mortality in a retrospective cohort of older adults admitted to an acute care for the elderly (ACE) unit.

MATERIALS AND METHODS

The study included 1,678 patients aged 60 years and over admitted to an AEC, in Cali, Colombia, from 2012 to 2015, and followed- up until 2016. The primary outcome was mortality. Renal function (eGFR) was estimated using Modification of Diet in Renal Disease Study (MDRD-4) equation. The renal function was grouped according to the eGFR (ml/min/1.73m²) as follows: slightly decreased (≥60), moderately decreased (30-59), and severely decreased (<30). Bivariate survival and multivariate Cox regression analyses were performed.

RESULTS

In the univariate analysis, patients with severely decreased eGFR had higher mortality than those with a higher eGFR (P=.046). In the group with severely decreased eGFR, survival was lower in the functionally dependent group (Barthel index [IB]<60) than in the independent group (IB≥60) (log rank test; P=.001). In the multivariate analysis, there was a significant increase in the risk of death in the elderly with severely decreased eGFR (<30) compared with slightly decreased eGFR (≥60) (hazard ratio [HR], 1.44; 95% confidence interval [CI]; 1.02-2.05, P=.039). There was also a significant increase in the risk of death in the dependent elderly compared to the independent ones [HR 1.72; 95% CI; 1.26-2.34, P=.000], those who had the high morbidity (≥4) with low albumin (<3.2g/dL) compared with those with low morbidity (0-3) and high albumin (≥3.2) [HR 1.77; 95% CI; 1.18-2.65, P=.005], and in those with a high (16-102mg/dL) C-reactive protein (CRP) compared with those with low CRP (0-15) [HR 1.42; 95% CI; 1.01-2.01, P=.043].

CONCLUSIONS

The risk of mortality after hospital admission to an AEC unit is greater in patients with eGFR<30. Poor functional status performance, high comorbidity, low plasma albumin, and increased inflammation markers are additional prognostic factors to be taken into account. The improvement in the functional status could improve the survival after hospitalisation.