Prognostic Comparison of the Estimations of Renal Function in Patients With Acute Heart Failure

Abstract

Background: The prognostic significance of the eGFR calculated by either the four-level Race Chronic Kidney Disease-Epidemiology Collaboration study equation (CKD-EPI4R) or the Chinese-modified Modification of Diet in Renal Disease equation (cMDRD) has not been compared in Asian populations with acute heart failure (AHF).

Methods and Results: A total of 3,044 patients hospitalized for AHF were enrolled. The National Death Registry was linked to identify deaths within a 5-year follow-up. Net reclassification improvement (NRI) was calculated to compare the prognostic value of either eGFR equation. During a median follow-up of 23.3 months, 1,424 (47%) patients died. Both eGFR_cMDRD and eGFR_CKD-EPI4R were independently predictive of death in the total study population (hazard ratio and 95% confidence intervals per 1-SD: 0.76, 0.71–0.81 and 0.74, 0.70–0.79, respectively), and in the subgroups of either reduced (HFrEF) or preserved (HFpEF) ejection fraction, after accounting for important confounders. With reference to eGFR_cMDRD, eGFR_CKD-EPI4R may improve the NRI by 2.0% (0.8–3.2%) for the prediction of death. The prognostic value of the CKD stages categorized by eGFR_CKD-EPI4R significantly outperformed eGFR_cMDRD with a categorical NRI of 9.5% (4.7–14.3%) in the total study population, 11.5% in HFrEF, and 8.3% in HFpEF.

Conclusions: Both eGFR_cMDRD and eGFR_CKD-EPI4R were independently associated with long-term survival in patients with AHF. However, the CKD stages derived from eGFR_CKD-EPI4R improved the risk stratification of death, compared with eGFR_cMDRD.

Impaired renal function has prevailed in patients hospitalized for heart failure (HF), and it is associated with poor clinical outcomes.^1,2 The identification of renal dysfunction in patients with HF is thus crucial for risk stratification and subsequent tailored therapy. Accurate measurement of the glomerular filtration rate (GFR) by assessing the clearance of exogenous markers, such as inulin, is time-consuming and impracticable. Levey et al therefore developed an estimating equation for GFR that the Modification of Diet in Renal Disease (MDRD) study has adopted in clinical practice worldwide.³ Given racial discrepancies,⁴ Ma et al⁵ and Kong et al⁶ further proposed and validated a Chinese-modified MDRD (cMDRD) equation for better estimation of GFR in Chinese populations. Nowadays, the estimated GFR (eGFR) is the most used surrogate of renal function for the diagnosis of chronic kidney disease (CKD) and an indicator of the prognosis in various populations.

Given the cohort for development of the MDRD equation was existing CKD, the Chronic Kidney Disease-Epidemiology Collaboration (CKD-EPI) equation was established from a general population to demonstrate a better estimation of GFR.⁷ Next, racial coefficients were introduced to improve the performance of the original CKD-EPI equation for 4-level race (CKD-EPI_4R, including Black, Asian, Native American and Hispanic, White and other).⁸

Although the CKD-EPI equation may have outperformed the MDRD equation in HF patients in Western countries,^9,10 the prognostic differences of these 2 equations have not been evaluated in Asian populations with HF. We therefore investigated the prognostic value of eGFR, derived from either the cMDRD or CKD-EPI_4R equation, in an Asian cohort of acute HF (AHF). We further disclosed the dissimilarity of eGFR between the phenotypes of HF.

Methods

Patients who had been hospitalized during October 2003 to December 2012 for new or exacerbated symptoms and signs of HF,^11,12 were enrolled into an intramural registry. Patients with severe liver cirrhosis (Child-Pugh score B or C), sepsis, or endstage renal disease requiring hemodialysis was excluded from this analysis. In brief, 3,182 patients were eligible for the registry, but 138 were excluded for lack of follow-up (12 patients), missing creatinine data (94 patients), or undergoing hemodialysis (32 patients). The associated medical records, including morbidities, prescriptions, and hematological and biochemistry data were retrieved from a web-based electronic recording system. The Review Committee of Taipei Veterans General Hospital approved the use of the registry data for research purposes.

The left ventricular ejection fraction (LVEF) was obtained from the 2D-guided M-mode echocardiography in accordance with the recommendations of the American Society of Echocardiography.¹³ HF with reduced EF (HFrEF) or with preserved EF (HFpEF) was defined as either LVEF<50% or ≥50%, respectively. The ratio of peak mitral inflow velocities at early (E) and late diastole (A), and right ventricular systolic pressure were also measured.

On admission, blood was sampled for hematology and biochemistry, including N-terminal pro-B-type natriuretic peptide (NT-proBNP) levels. The serum creatinine level was measured by a modification of the kinetic Jaffe reaction described by Levey³ with alkaline picrate as the reactant. Because commercial measurement of NT-proBNP (Roche Diagnostics, Basel, Switzerland) was only available after 2009, there were missing values for NT-proBNP in this analysis.

Calculation of the eGFR

The cMDRD equation⁵ of eGFR is:

eGFR_cMDRD=186×(Scr)^−1.154×(age)^−0.203×0.742 [if female]×1.233, where Scr is serum creatinine.

The CKD-EPI_4R formula can be expressed as a single equation⁸ as following:

eGFR_CKD-EPI4R=141×min(Scr/κ,1)^σ×max(Scr/κ,1)^−1.209×0.993^Age×1.018 [if female]×1.052, where κ is 0.7 for females and 0.9 for males, σ is −0.329 for females and −0.411 for males, min indicates the minimum of Scr/k or 1, and max indicates the maximum of Scr/k or 1.

The stages of CKD were classified by eGFR according to the clinical guidelines of the National Kidney Foundation Kidney Disease Outcomes Quality Initiative: ≥60 mL/min/1.73 m² (Stage 1 and Stage 2); 59–45 mL/min/1.73 m² (Stage 3a); 44–30 mL/min/1.73 m² (Stage 3b); 29–15 mL/min/1.73 m² (Stage 4); and <15 mL/min/1.73 m² (Stage 5).¹⁴ Patients with eGFR <60 mL/min/1.73 m² were defined as having CKD.

Follow-up

We linked the database to the National Death Registry for the clinical outcome of death. Patients were followed for up to 5 years. The National Death Registry database registers valid information according to the International Classification of Disease, 9th Revision (ICD-9). The ICD-9 codes for cardiovascular death were 390–459.¹⁵

Statistical Analysis

Data are presented as the mean±standard deviation (SD) for continuous variables, and as number and percentages for categorical variables. Student’s t-test or the χ² test was used to conduct the comparisons. The NT-proBNP level was taken into natural logarithm transformation before the statistical analyses because of the skewed distribution. Intraclass correlation coefficients and Cohen’s κ index were used to evaluate the agreements of eGFR or the stages of CKD, defined by eGFR_cMDRD or eGFR_CKD-EPI4R. The mean differences between the 2 eGFR equations were analyzed using the Bland-Altman method.¹⁶ Receiver-operating characteristic (ROC) curve analysis described the predictive values of eGFR_cMDRD or eGFR_CKD-EPI4R with 5-year mortality and the c-statistic test was used to compare the differences of areas under the curves.¹⁷ Cox proportional hazard models were used to evaluate eGFR in the prediction of 5-year death after accounting for confounders. The model performance with eGFR_CKD-EPI4R compared with eGFR_cMDRD was evaluated by using net reclassification improvement (NRI).^18,19 We quantified the degree of correct classification according to the outcome by estimating the categorical NRI using cross-categories of eGFR for both formulas.²⁰ Kaplan-Meier survival curve analysis was conducted to compare the prognoses of different CKD stages. The two-sided differences were considered statistically significant at the 0.05 significance level. Statistical analyses were performed using IBM SPSS software version 21.0 (SPSS Inc., Chicago, IL, USA) and SAS software version 9.4 (SAS Inc., Carey, NC, USA).

Results

A total of 3,044 patients (75.6±13.1 years, 68% male, 40% HFrEF) constituted this study. The baseline characteristics are shown in Table 1. Compared with HFpEF, patients with HFrEF were younger and more likely to be men, and had lower admission systolic blood pressure, higher heart rate, less hypertension and more coronary artery disease. In addition, patients with HFrEF had better renal function, in terms of lower creatinine level, and higher eGFR_cMDRD and eGFR_CKD-EPI4R. The prevalence of diabetes or atrial fibrillation, and levels of blood urea nitrogen, sodium and potassium were comparable between the 2 groups. NT-proBNP was higher in patients with HFrEF than in those with HFpEF, and patients with HFrEF also had higher E/A ratio and right ventricular systolic pressure.

Table 1. Baseline Characteristics of the Study Population by Phenotype of Heart Failure

Variables	Total (n=3,044)	HFrEF (n=1,208)	HFpEF (n=1,836)	P value
Age, years	75.6±13.1	72.5±14.7	77.8±11.5	<0.001
Men, n (%)	2,073 (68)	914 (76)	1,159 (63)	<0.001
SBP, mmHg	140±32	133±30	144±33	<0.001
Heart rate, beats/min	88±23	92±24	86±22	<0.001
Morbidities, n (%)
Hypertension	1,860 (61)	652 (54)	1,208 (66)	<0.001
Diabetes	1,124 (37)	427 (35)	697 (38)	0.144
CAD	994 (33)	481 (40)	513 (28)	<0.001
AF	884 (29)	350 (29)	534 (29)	0.838
ACS	384 (13)	189 (16)	195 (11)	<0.001
Biochemistry
BUN, mg/dL	36±23	35±23	36±24	0.224
Creatinine, mg/dL	1.9±1.6	1.8±1.4	2.0±1.6	0.004
eGFRcMDRD, mL/min	62±35	65±34	60±35	0.001
eGFRCKD-EPI4R, mL/min	48±26	52±27	46±26	<0.001
Sodium, mmol/L	139±5	139±4	139±5	0.442
Potassium, mmol/L	4.0±0.7	4.0±0.7	4.0±0.7	0.445
NT-proBNP, pg/mL (n=1,120)†	8.6±1.4	9.0±1.2	8.4±1.5	<0.001
Echocardiography
LVEF, %	54±20	35±13	67±10	<0.001
E/A ratio	1.1±0.7	1.4±0.9	1.0±0.6	<0.001
RVSP, mmHg	44±17	45±17	43±17	0.001
Medications, n (%)
β-blockers	1,347 (44)	607 (50)	740 (40)	<0.001
RAS inhibitors	1,970 (65)	791 (66)	1,179 (64)	0.390
MRA	1,201 (40)	603 (50)	598 (33)	<0.001
Loop diuretics	2,036 (67)	838 (69)	1,198 (65)	0.020

^†Natural logarithm transformation. ACS, acute coronary syndrome; AF, atrial fibrillation; BUN, blood urea nitrogen; CAD, coronary artery disease; CKD-EPI_4R, Four-level Race Chronic Kidney Disease-Epidemiology Collaboration Group; cMDRD, Chinese-modified Modification of Diet in Renal Disease; E/A ratio, the ratio of peak mitral inflow velocities at early (E) and late diastole (A); eGFR, estimated glomerular filtration rate; HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; LVEF, left ventricular ejection fraction; MRA, mineralocorticoid receptor antagonist; NT-proBNP, N-terminal pro-brain natriuretic peptide; RAS inhibitors include angiotensin-converting enzyme inhibitor and angiotensin-receptor antagonist; RVSP, right ventricular systolic pressure; SBP, systolic blood pressure.

Agreement of eGFR Measurements

The Bland-Altman analysis showed the correlations and discrepancies between eGFR_cMDRD and eGFR_CKD-EPI4R (Figure 1). The value of eGFR_CKD-EPI4R was generally lower than that of eGFR_cMDRD, with a mean difference (95% confidence interval) of 13.5 (7.9, −34.9) mL/min/1.73 m² (Figure 1B). The intraclass correlation coefficient of eGFR_cMDRD and eGFR_CKD-EPI4R was 0.968 (0.965, 0.970) (Figure 1A). When 51% of the study population (47% of HFrEF and 53% of HFpEF) was categorized as having CKD by eGFR_cMDRD, 68% (64% of HFrEF and 70% of HFpEF) would have CKD as defined by eGFR_CKD-EPI4R (Figure 2). The κ value of the 5 categories derived from the 2 formulas was 0.659 (0.634–0.684), which indicated a fair strength of agreement (Supplementary Table).

Figure 1.

Scatterplots of (A) the estimated glomerular filtration rate (eGFR) calculated by cMDRD and CKD-EPI_4R equations and (B) agreement of eGFR between the cMDRD and CKD-EPI_4R equations by Bland-Altman analysis. cMDRD, Chinese-modified Modification of Diet in Renal Disease; CKD-EPI_4R, 4-level Race Chronic Kidney Disease-Epidemiology Collaboration Group.

Figure 2.

Distribution of chronic kidney disease (CKD) defined by the eGFR using the cMDRD and CKD-EPI_4R equations in patients with heart failure and reduced ejection fraction (HFrEF) or preserved ejection fraction (HFpEF). *P<0.05, **P<0.001 by McNemar’s test. Abbreviations as in Figure 1.

Prognostic Impact of eGFR

During a median follow-up of 23.3 months (interquartile range 8.1–45.3 months), 1,424 (47%) patients died. The Kaplan-Meier survival curve analyses demonstrated that severity of CKD, stratified by either eGFR_cMDRD or eGFR_CKD-EPI4R, was related to long-term mortality (Figure 3). The survival curves, stratified by the stages of CKD, were better separated by eGFR_CKD-EPI4R than by eGFR_cMDRD (χ²=137 and 125, respectively). Although 1,538 (50.5%) patients were classified as having CKD and 988 (32.5%) subjects were not by both equations, 518 (17%) patients were categorized as CKD by eGFR_CKD-EPI4R but not by eGFR_cMDRD. The Kaplan-Meier survival curve analysis suggested that subjects with conflicting classifications of CKD were more associated with significantly higher mortality than were those without CKD according to both equations (Figure 4).

Figure 3.

Kaplan-Meier survival curve analysis related to the stage of CKD, defined by cMDRD equation (A) or CKD-EPI_4R equation (B). Abbreviations as in Figures 1,2.

Figure 4.

Kaplan-Meier survival curve analysis related to the presence of CKD, defined by either or both the cMDRD and CKD-EPI_4R equations. Abbreviations as in Figures 1,2.

After accounting for age, sex, LVEF, serum sodium levels and comorbidities, both eGFR_cMDRD and eGFR_CKD-EPI4R were independently associated with mortality (hazard ratio and 95% confidence intervals per 1-SD: 0.760, 0.713–0.811 and 0.743, 0.698–0.791, respectively) (Table 2). eGFR_CKD-EPI4R was more accurate in predicting prognosis than eGFR_cMDRD, as evaluated by a higher X² value (Table 2). In the ROC curve analysis for all-cause death, the area under the curve was numerically higher but not significantly with eGFR_CKD-EPI4R than with eGFR_cMDRD (Table 3). The model performance for the prediction of long-term survival was improved by eGFR_CKD-EPI4R compared with eGFR_cMDRD, with a NRI of 2.0% (0.8–3.2%) (Table 3).

Table 2. Cox Regression Analysis of 5-Year Total Mortality Rate*

eGFR equation	HR per 1 SD† (95% CI)	X2	P value
Total population
cMDRD	0.760 (0.713~0.811)	69	<0.001
CKD-EPI4R	0.743 (0.698~0.791)	86	<0.001
HFrEF
cMDRD	0.690 (0.621~0.767)	48	<0.001
CKD-EPI4R	0.694 (0.628~0.766)	52	<0.001
HFpEF
cMDRD	0.800 (0.737~0.869)	28	<0.001
CKD-EPI4R	0.770 (0.710~0.836)	39	<0.001

*Adjusted for age, sex, LVEF, sodium, hypertension, diabetes mellitus, and coronary artery disease. ^†The standard deviation (SD) of cMDRD and CKD-EPI_4R was 34.6 and 26.3 mg/dL, respectively. HR, hazard ratio; CI, confidence interval. Other abbreviations as in Table 1.

Table 3. Performance of the Cox Proportional Hazard Models

	cMDRD	CKD-EPI4R	P value
Total population
AUC (95% CI)	0.608 (0.588~0.628)	0.615 (0.595~0.635)	0.620
NRI	Ref.	2.0% (0.8~3.2%)	0.001
Categorical NRI	Ref.	9.5% (4.7~14.3%)	<0.001
HFrEF population
AUC (95% CI)	0.646 (0.615~0.677)	0.656 (0.625~0.686)	0.659
NRI	Ref.	1.0% (−0.0~1.9%)	0.054
Categorical NRI	Ref.	11.5% (4.0~19.0%)	0.003
HFpEF population
AUC (95% CI)	0.587 (0.561~0.614)	0.593 (0.567~0.619)	0.744
NRI	Ref.	1.9% (−0.1~4.0%)	0.066
Categorical NRI	Ref.	8.3% (2.0~14.6%)	0.010

AUC, area under the receiver-operating characteristic curve; NRI, net reclassification improvement. Other abbreviations as in Table 1.

With reference to patients with eGFR ≥60 mL/min/1.73 m², the age-, sex-, LVEF- and serum sodium level-adjusted hazard ratio for 5-year mortality increased along with progression of CKD stage in the study population (Figure 5A,B), and in patients with HFrEF or HFpEF (Figure 5C–F). A significant increment of the prognostic impact together with worsened CKD stage was observed with classification by eGFR_CKD-EPI4R rather than by eGFR_cMDRD (Figure 5). Compared with eGFR_cMDRD, eGFR_CKD-EPI4R improved the risk stratification related to progression of CKD stage. The categorical NRI was 9.5% (4.7–14.3%) in the total study population, 11.5% (4.0–19.0%) in patients with HFrEF, and 8.3% (2.0–14.6%) in patients with HFpEF (Table 3, Supplementary Table).

Figure 5.

Adjusted hazard ratio (HR) and 95% confidence interval (CI) for death across the stages of CKD, after accounting for age, sex, left ventricular ejection fraction, and sodium in the study population (A,B), and patients with heart failure and reduced ejection fraction (C,D) or preserved ejection fraction (E,F). Abbreviations as in Figures 1,2.

Discussion

Main Findings

The study revealed that the major discrepancies of eGFR derived from either the CKD-EPI_4R or cMDRD equation were in patients with preserved renal function. More patients would be categorized with CKD by eGFR_CKD-EPI4R than by eGFR_cMDRD. However, both eGFR_CKD-EPI4R and eGFR_cMDRD were independently correlated with the clinical outcomes of AHF. In addition, eGFR_CKD-EPI4R may outperform eGFR_cMDRD in predicting long-term survival in the homogenous Chinese population. The risk of death related to CKD stage was better predicted by using the CKD-EPI_4R equation than by the cMDRD equation. The prognostic superiority of eGFR_CKD-EPI4R, compared with eGFR_cMDRD, remained consistent regardless of the phenotype of HF.

Development of the eGFR Equations

The measurement of renal function is recommended to recognize renal impairment and monitor kidney function for risk stratification and tailored therapy.¹⁴ GFR is an essential risk factor for clinical outcome in various populations.^1,21 Although actual renal function can be obtained by calculating the renal clearance of ¹²⁵I-iothalamate, ^{99 m}Tc-DTPA, or inulin,²² the generalizability is poor. Although serum creatinine level is widely used as a surrogate of renal function, its representativeness is of doubt when age, sex and body weight also influence the creatinine level. Therefore, Levey et al first developed the MDRD equation to better quantify renal function.³ Given the MDRD equation may underestimate the GFR of individuals with normal renal function,^23,24 the CKD-EPI equation was developed from a more heterogeneous population.⁷ The CKD-EPI equation was then validated to be more accurate than the MDRD equation, especially in patients with preserved renal function.^7,25 The National Kidney Foundation has recommended reporting eGFR by using the CKD-EPI equation with every serum creatinine measurement.²⁶

In addition, race is also a major determinant of renal function because of different muscle mass and dietary habits.²⁷ The MDRD or CKD-EPI equation did not perform well in the estimation of GFR in Asians,^4,8 and the modified equations for either Chinese or Asians have improved the bias of estimation.^5,6,8

Discrepancy Between Different eGFR Equations

Given the present study did not measure the true values of GFR by assessing the clearance of exogenous markers, we were not able to evaluate the accuracy of either eGFR equation. However, Al-Wakeel et al reported that the CKD-EPI equation was more accurate than the MDRD equation for the estimation of GFR in a small population, compared with inulin clearance.²⁸ In our study, eGFR_CKD-EPI4R was lower than eGFR_cMDRD in all subjects by a mean difference of 13.5 mL/min, and the difference was more prominent when eGFR was greater. Therefore, the CKD-EPI_4R equation would categorize a greater proportion of the study population as having CKD, compared with the cMDRD equation (68% vs. 51%). The result was in line with the findings of McAlister et al in their meta-analysis of 20,754 patients with HF that the CKD-EPI equation demonstrated higher estimation of renal dysfunction than the MDRD equation.²⁹ Our study may be the first to show that the CKD-EPI_4R equation also recognized more cases of CKD than did the cMDRD equation in Asian patients with HF.

Prognostic Value of eGFR

Matsushita et al demonstrated that the CKD-EPI equation was more accurate for risk classification for death than the MDRD equation in patients with CKD or known risks of CKD.³⁰ McAlister et al in a pooled analysis of 20,754 patients with HF further showed the advantage of the CKD-EPI equation over the MDRD equation in the prediction of adverse clinical events.²⁹ In the present study, both eGFR_CKD-EPI4R and eGFR_cMDRD were predictive of long-term survival, independent of age, sex, LVEF, sodium level and comorbidities. Although the c-statistics for the prediction of 5-year death by eGFR_cMDRD (0.608) and eGFR_CKD-EPI4R (0.615) were not significantly different, ROC curve analysis was not meant to evaluate prognostic discrepancies. On the other hand, NRI might be a better measure for comparing the predictive ability for survival. We also clarified that the CKD-EPI_4R equation was a better prognostic tool than the cMDRD equation by showing better risk stratification and significant NRIs. The findings remained true in patients with either HFrEF or HFpEF. In addition, subjects who were defined as having CKD by eGFR_CKD-EPI4R but not eGFR_cMDRD actually carried a higher mortality risk, compared with those without CKD by both equations.

Study Limitations

Given that the present study was a single-center registry and an observational study, there were some biases arising from internal and external validities. Because the baseline characteristics were similar to other population-based HF cohorts,³¹ and we adjusted for all the observed confounders, the study results might be able to be extrapolated to other populations. Second, we did not apply the Taiwanese modified MDRD and CKD-EPI equations because of their popularity and the generalizability of these equations.³² Third, the medications prescribed before the index hospitalization were not recorded in the registry. We therefore were unable to analyze the influence of drugs on eGFR. Fourth, only NT-proBNP but not BNP data were available in the registry and NT-proBNP would be more likely affected by renal dysfunction. Given that the values of NT-proBNP were only available for one-third of the study population, we did not include NT-proBNP in the multivariate Cox regression analysis. Moreover, we did not measure the true GFR in this study for the comparisons. Further studies are warranted to fairly judge the accuracy of these 2 Asian-modified equations. Lastly, we did not comprehensively measure the indices of LV diastolic function; the E/A ratio is not an appropriate indicator of diastolic function.

Conclusions

Renal dysfunction is common in patients with HF and associated with worse outcome. In addition, there are critical criteria involving renal function for the prescription of HF medications, including renin-angiotensin system blockers and mineralocorticoid antagonists. Therefore, it is necessary to have an accurate estimation of renal function. The present study has shown that the CKD-EPI_4R equation generally had lower estimates of GFR and defined more patients with CKD, compared with the cMDRD equation. The categorized CKD stages by CKD-EPI_4R improved risk stratification in the prediction of death of patients with AHF, regardless of the HF phenotype. The study suggested mandatory calculation of eGFR by the CKD-EPI_4R equation with measurement of serum creatinine for early identification of at-risk patients requiring tailored therapy.

Acknowledgments

The study was supported by Taipei Veterans General Hospital (V100C-145, V101C-092, V102C-119, V103B-017, V104C-172), Ministry of science and technology (MOST 102-2314-B-010-052, 103-2314-B-010-050-MY2), and Ministry of Health and Welfare, Taiwan with grant (MOHW-105-TDU-B-211-133017, MOHW106-TDU-B-211-113001, MOHW107-TDU-B-211-123001) and the death registry.

Conflicts of Interest

None declared.

Supplementary Files

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-18-1013

References

• 1.   Smith GL, Lichtman JH, Bracken MB, Shlipak MG, Phillips CO, DiCapua P, et al. Renal impairment and outcomes in heart failure: Systematic review and meta-analysis. J Am Coll Cardiol 2006; 47: 1987–1996.
• 2.   de Silva R, Nikitin NP, Witte KK, Rigby AS, Goode K, Bhandari S, et al. Incidence of renal dysfunction over 6 months in patients with chronic heart failure due to left ventricular systolic dysfunction: Contributing factors and relationship to prognosis. Eur Heart J 2006; 27: 569–581.
• 3.   Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: A new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 1999; 130: 461–470.
• 4.   Zuo L, Ma YC, Zhou YH, Wang M, Xu GB, Wang HY. Application of GFR-estimating equations in Chinese patients with chronic kidney disease. Am J Kidney Dis 2005; 45: 463–472.
• 5.   Ma YC, Zuo L, Chen JH, Luo Q, Yu XQ, Li Y, et al. Modified glomerular filtration rate estimating equation for Chinese patients with chronic kidney disease. J Am Soc Nephrol 2006; 17: 2937–2944.
• 6.   Kong X, Ma Y, Chen J, Luo Q, Yu X, Li Y, et al. Evaluation of the Chronic Kidney Disease Epidemiology Collaboration equation for estimating glomerular filtration rate in the Chinese population. Nephrol Dial Transplant 2013; 28: 641–651.
• 7.   Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009; 150: 604–612.
• 8.   Stevens LA, Claybon MA, Schmid CH, Chen J, Horio M, Imai E, et al. Evaluation of the Chronic Kidney Disease Epidemiology Collaboration equation for estimating the glomerular filtration rate in multiple ethnicities. Kidney Int 2011; 79: 555–562.
• 9.   Valente MA, Hillege HL, Navis G, Voors AA, Dunselman PH, van Veldhuisen DJ, et al. The Chronic Kidney Disease Epidemiology Collaboration equation outperforms the Modification of Diet in Renal Disease equation for estimating glomerular filtration rate in chronic systolic heart failure. Eur J Heart Fail 2014; 16: 86–94.
• 10.   Manzano-Fernandez S, Flores-Blanco PJ, Perez-Calvo JI, Ruiz-Ruiz FJ, Carrasco-Sanchez FJ, Morales-Rull JL, et al. Comparison of risk prediction with the CKD-EPI and MDRD equations in acute decompensated heart failure. J Card Fail 2013; 19: 583–591.
• 11.   Ho KK, Anderson KM, Kannel WB, Grossman W, Levy D. Survival after the onset of congestive heart failure in Framingham Heart Study subjects. Circulation 1993; 88: 107–115.
• 12.   McMurray JJ, Adamopoulos S, Anker SD, Auricchio A, Bohm M, Dickstein K, et al. ESC guidelines for the diagnosis and treatment of acute and chronic heart failure 2012: The Task Force for the Diagnosis and Treatment of Acute and Chronic Heart Failure 2012 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association (HFA) of the ESC. Eur J Heart Fail 2012; 14: 803–869.
• 13.   Sahn DJ, DeMaria A, Kisslo J, Weyman A. Recommendations regarding quantitation in M-mode echocardiography: Results of a survey of echocardiographic measurements. Circulation 1978; 58: 1072–1083.
• 14.   Levey AS. A decade after the KDOQI CKD guidelines. Am J Kidney Dis 2012; 60: 683–685.
• 15.   Sung SH, Cheng HM, Wang KL, Yu WC, Chuang SY, Ting CT, et al. White coat hypertension is more risky than prehypertension: Important role of arterial wave reflections. Hypertension 2013; 61: 1346–1353.
• 16.   Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307–310.
• 17.   DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: A nonparametric approach. Biometrics 1988; 44: 837–845.
• 18.   Kerr KF, McClelland RL, Brown ER, Lumley T. Evaluating the incremental value of new biomarkers with integrated discrimination improvement. Am J Epidemiol 2011; 174: 364–374.
• 19.   Leening MJ, Vedder MM, Witteman JC, Pencina MJ, Steyerberg EW. Net reclassification improvement: Computation, interpretation, and controversies: A literature review and clinician’s guide. Ann Intern Med 2014; 160: 122–131.
• 20.   Pencina MJ, D’Agostino RB Sr, Steyerberg EW. Extensions of net reclassification improvement calculations to measure usefulness of new biomarkers. Stat Med 2011; 30: 11–21.
• 21.   Tsagalis G, Akrivos T, Alevizaki M, Manios E, Stamatellopoulos K, Laggouranis A, et al. Renal dysfunction in acute stroke: An independent predictor of long-term all combined vascular events and overall mortality. Nephrol Dial Transplant 2009; 24: 194–200.
• 22.   Perrone RD, Steinman TI, Beck GJ, Skibinski CI, Royal HD, Lawlor M, et al. Utility of radioisotopic filtration markers in chronic renal insufficiency: Simultaneous comparison of 125I-iothalamate, 169Yb-DTPA, 99 mTc-DTPA, and inulin. The Modification of Diet in Renal Disease Study. Am J Kidney Dis 1990; 16: 224–235.
• 23.   Lin J, Knight EL, Hogan ML, Singh AK. A comparison of prediction equations for estimating glomerular filtration rate in adults without kidney disease. J Am Soc Nephrol 2003; 14: 2573–2580.
• 24.   Rule AD, Gussak HM, Pond GR, Bergstralh EJ, Stegall MD, Cosio FG, et al. Measured and estimated GFR in healthy potential kidney donors. Am J Kidney Dis 2004; 43: 112–119.
• 25.   Stevens LA, Schmid CH, Zhang YL, Coresh J, Manzi J, Landis R, et al. Development and validation of GFR-estimating equations using diabetes, transplant and weight. Nephrol Dial Transplant 2010; 25: 449–457.
• 26.   Inker LA, Astor BC, Fox CH, Isakova T, Lash JP, Peralta CA, et al. KDOQI US commentary on the 2012 KDIGO clinical practice guideline for the evaluation and management of CKD. Am J Kidney Dis 2014; 63: 713–735.
• 27.   Baxmann AC, Ahmed MS, Marques NC, Menon VB, Pereira AB, Kirsztajn GM, et al. Influence of muscle mass and physical activity on serum and urinary creatinine and serum cystatin C. Clin J Am Soc Nephrol 2008; 3: 348–354.
• 28.   Al-Wakeel JS. Accuracy and precision of the CKD-EPI and MDRD predictive equations compared with glomerular filtration rate measured by inulin clearance in a Saudi population. Ann Saudi Med 2016; 36: 128–134.
• 29.   McAlister FA, Ezekowitz J, Tarantini L, Squire I, Komajda M, Bayes-Genis A, et al. Renal dysfunction in patients with heart failure with preserved versus reduced ejection fraction: Impact of the new Chronic Kidney Disease-Epidemiology Collaboration Group formula. Circ Heart Fail 2012; 5: 309–314.
• 30.   Matsushita K, Mahmoodi BK, Woodward M, Emberson JR, Jafar TH, Jee SH, et al. Comparison of risk prediction using the CKD-EPI equation and the MDRD study equation for estimated glomerular filtration rate. JAMA 2012; 307: 1941–1951.
• 31.   Krum H, Gilbert RE. Demographics and concomitant disorders in heart failure. Lancet 2003; 362: 147–158.
• 32.   Chen LI, Guh JY, Wu KD, Chen YM, Kuo MC, Hwang SJ, et al. Modification of diet in renal disease (MDRD) study and CKD epidemiology collaboration (CKD-EPI) equations for Taiwanese adults. PLoS One 2014; 9: e99645.