114

ORIGINAL ARTICLE

Acta Medica Indonesiana - The Indonesian Journal of Internal Medicine

Predictive Value of Different Estimated Glomerular Filtration Rates on Hospital Adverse Events Following Acute Myocardial Infarction

Anggoro B. Hartopo1,2, Budi Y. Setianto1,2, Putrika P.R. Gharini1,2

1 Department of Cardiology and Vascular Medicine, Faculty of Medicine, Gadjah Mada University, Yogyakarta, Indonesia. 2 Department of Internal Medicine, Faculty of Medicine, Gadjah Mada University, Yogyakarta, Indonesia.

Correspondence mail: Jl. Pakel Baru Selatan no. 50 Umbulharjo, Yogyakarta 55162 , Indonesia. email: anggorobudih@yahoo.com.

ABSTRAKTujuan: untuk menilai apakah berbagai rumus perhitungan laju filtrasi glomerulus (LFG) mempunyai

nilai prediktif yang berbeda terhadap kejadian yang tidak diinginkan selama perawatan di rumah sakit pada pasien dengan infark miokard akut. Metode: desain penelitian adalah potong lintang. Data dari pasien dengan infark miokard akut yang direkrut secara berurutan dianalisis. Tiga persamaan estimasi LFG yang berbeda yaitu Cockroft-Gault (LFGC-G), MDRD (LFGMDRD) and CKD-EPI (LFGCKD-EPI) dihitung. Kejadian yang tidak diinginkan selama perawatan di rumah sakit pada pasien penelitian direkam. Nilai-nilai prediksi dari LFG ini terhadap kejadian yang tidak diinginkan selama perawatan rumah sakit dibandingkan dengan kurva ROC. Analisis univariat dan multivariat dilakukan untuk menilai LFG mana sebagai prediktor independen kejadian yang tidak diinginkan selama perawatan rumah sakit. Hasil: dari 103 pasien penelitian, 49 pasien (47.6%) mengalami kejadian yang tidak diinginkan selama perawatan rumah sakit. Proporsi kejadian yang tidak diinginkan secara bermakna berhubungan dengan LFGMDRD (p<0.01) and LFGCKD-EPI (p=0.02), namun tidak dengan LFGC-G (p=0.10). Kejadian buruk perawatan rumah sakit diprediksi lebih baik oleh LFGMDRD dibandingkan oleh LFGCKD-EPI (AUC, 0.698; 95%CI: 0.596-0.800, p<0.01 versus AUC, 0.693; 95%CI: 0.591-0.796, p<0.01). Analisis multivariat menunjukkan disfungsi ginjal sedang (adjusted OR 3.50; 95%CI: 1.38-8.85, p<0.01) dan berat (adjusted OR 8.13, 95%CI: 1.38-47.91, p=0.02) berdasarkan LFGMDRD merupakan prediktor independen untuk kejadian yang tidak diinginkan perawatan rumah sakit. Kesimpulan: LFG berdasarkan MDRD mempunyai nilai prediktif yang lebih baik dibandingkan LFG berdasarkan CKD-EPI untuk kejadian yang tidak diinginkan perawatan rumah sakit pada infark miokard akut. Disfungsi ginjal sedang dan berat berdasarkan LFGMDRD merupakan prediktor independen untuk kejadian buruk perawatan rumah sakit pada infark miokard akut.

Kata kunci: laju filtrasi glomerulus, MDRD, CKD-EPI, infark miokard akut.

ABSTRACTAim: to assess whether different glomerular filtration rate (GFR) equations render different predictive value

on hospital adverse events in patients hospitalised due to acute myocardial infarction. Methods: the study design is cross-sectional. Data from consecutive patients with acute myocardial infarction were analyzed. Three different estimated GFR equations, i.e. Cockroft-Gault (eGFRC-G), MDRD (eGFRMDRD) and CKD-EPI (eGFRCKD-EPI) were calculated. Hospital adverse events in these study patients were recorded. The predictive values of these eGFRs on hospital adverse events were compared with ROC curve. Univariate and multivariable analysis to assess

Vol 45 • Number 2 • April 2013 Predictive value of different estimated glomerular filtration rates

115

which GFR equation as independent predictor for hospital adverse events were performed. Results: among 103 study patients, 49 patients (47.6%) experienced hospital adverse events. Proportion of hospital adverse events were significantly associated with eGFRMDRD (p<0.01) and eGFRCKD-EPI (p=0.02), but not with eGFRC-G (p=0.10). Hospital adverse events were better predicted by eGFRMDRD than by eGFRCKD-EPI (AUC, 0.698; 95%CI: 0.596-0.800, p<0.01 versus AUC, 0.693; 95%CI: 0.591-0.796, p<0.01). Multivariable analysis showed moderate (adjusted OR 3.50; 95%CI: 1.38-8.85, p<0.01) and severe (adjusted OR 8.13, 95%CI: 1.38-47.91, p=0.02) kidney dysfunctions based on eGFRMDRD were independent predictors for hospital adverse events. Conclusion: an eGFR based on MDRD gave better predictive value than eGFR based on CKD-EPI on hospital adverse events among acute myocardial infarction. Moderate and severe kidney dysfunctions based on eGFRMDRD were independent predictors for hospital adverse events following acute myocardial infarction.

Key words: glomerular filtration rate, MDRD, CKD-EPI, acute myocardial infarction.

INTRODUCTIONPatients suffering from kidney diseases have

greater probability to acquire cardiovascular disease.1,2 End-stage kidney disease patients contribute to nearly 60% mortality rate following hospitalization due to acute myocardial infarction.3 Furthermore, moderate and severe kidney dysfunctions are independent predictors for in-hospital mortality in patients with acute myocardial infarction.4,5

The level of kidney function is best characterized by the estimated glomerular filtration rate (eGFR) which gives valid measure of the filtering capacity of the kidneys.6 Several eGFR equations have been developed to accurately estimate glomerular filtration function in general population and in diseased patients. In the acute clinical setting, such as acute myocardial infarction, measuring kidney function should be prompt and rapidly give accurate information to assist clinical decision. An eGFR based on Cockroft-Gault (C-G) equation requires body weight measurement that sometimes is difficult to obtain rapidly in acute clinical setting. Other eGFR equations, the modification of diet in renal disease (MDRD) and the chronic kidney disease epidemiology collaboration (CKD-EPI), have been recently developed and validated in population and clinical setting replacing Cockroft-Gault.7-9 Although both equations require ethnic data adjustment which differentiate African-American race among others, Asian multiethnic population do not need such adjustment in eGFR calculation.10

Calculating eGFR based on MDRD or CKD-

EPI has not been routinely performed in acute myocardial infarction patients in our coronary care unit, since the cardiologists focused on other predictive scores such as TIMI, GRACE or Pursuit scores. This is in reality, especially when serum creatinine concentration is normal or only moderately increased. Furthermore, automatic eGFR value from laboratory datasheet has not been available currently in our laboratory hospital. In the present study, we evaluate the role of eGFR based on MDRD, CKD-EPI and C-G equations in predicting hospital adverse events among acute myocardial infarction patients.

METHODS

Study Patients

The study design is cross-sectional. This study is based on a secondary analysis from a cohort of patients which we have previously published.11 Patients with acute myocardial infarction (AMI) admitted to intensive coronary care unit (ICCU) of Dr. Sardjito Hospital, Yogyakarta, Indonesia, were enrolled. The time of study enrollment between September 2008 and May 2009. The inclusion criteria were (1) male and female with age between 35 and 80 years old and (2) commencement of anginal pain less than 24 hours before admission. The exclusion criteria were (1) patients with known chronic diseases, such as chronic kidney disease stage V, chronic heart failure with NYHA class more than II and known hepatic cirrhosis, (2) patients with known valvular heart disease, (3) patients with known malignancy, (4) patients presented

Anggoro B. Hartopo Acta Med Indones-Indones J Intern Med

116

with acute comorbidity such as acute stroke, acute infection and sepsis and (5) patients with venous thromboembolism.

The diagnosis of AMI consisted of ST elevation myocardial infarction (STEMI) and non STEMI (NSTEMI). STEMI was diagnosed based on anginal pain lasting more than 20 minutes, electrocardiography examination revealed ST segment elevation in two or more adjacent leads and elevated cardiac enzyme (troponin I ≥0.6 ng/mL). NSTEMI was diagnosed based on angina pain lasting more than 20 minutes, electrocardiography examination revealed no ST segment elevation and elevated cardiac enzyme (troponin I ≥ 0.6 ng/mL).

During hospitalization, we collected data on demography and medical backgrounds. Medical history related to cardiovascular risk factors, i.e. diabetes mellitus, hypertension, previous ischemic heart disease (IHD) and smoking status were documented. On-duty physicians assessed clinical presentation on admission. Informed consent was obtained from patients or their families. The study was approved by Ethics Committee of Faculty of Medicine Universitas Gadjah Mada, Yogyakarta, Indonesia

Laboratory ExaminationPeripheral venous blood samples were

drawn from study patients on hospital arrival, before commencing any intravenous therapies. EDTA blood sample was used for blood cell count and serum was used for blood chemistry measurement. Serum creatinine level was measured according to Jaffe reaction method using SYNCHRON Cx Calibrator on SYNCHRON CX Systems (Beckman Coulter, California, USA). Laboratory measurements were conducted in our hospital central laboratory.

Estimated glomerular filtration rate (eGFR) was calculated based on these formulas: (1) eGFR based on MDRD (eGFRMDRD) [mL/min per 1.73 m2] = 175 × (serum creatinine [mg/dL])−1.154 × age [years]−0.203 × (0.742 if female) × (1.212 if black),7 (2) eGFR based on CKD-EPI (eGFRCKD-EPI) (mL/min per 1.73 m2) = 141 × (minimum serum creatinine [mg/dL]/κ or 1)α × (maximum serum creatinine [mg/dL]/κ or 1)−1.209 × 0.993 age × (1.018 if female) × (1.159 if black), where κ is 0.7 for

female and 0.9 for male and α is −0.329 for female and −0.411 for male,8 and (3) eGFR based on Cockroft-Gault (eGFRC-G) [mL/min per 1.73 m2] = (140 – age [years] x (body weight [kg] x (0.85 if female) / (72 x serum creatinine [mg/dL]).12

The eGFRMDRD, eGFRCKD-EPI and eGFRC-G were divided into three categories, i.e. <30, 30-60, and ≥60 mL/min per 1.73 m2.6

Study OutcomesWe followed the study patients during

hospitalization. Hospital adverse events were the outcomes in this study, which comprised in hospital mortality and complications, i.e. acute heart failure, shock cardiogenic and ongoing pain during hospitalization. Hospital death was determined as any cause of death following AMI. Acute heart failure was diagnosed based on at least one of the clinical signs and symptoms of pulmonary congestion in physical examination or on chest radiography, the use of intravenous loop diuretics (furosemide) to treat presumed pulmonary congestion or the use of positive inotropic agents (dobutamin or dopamine). Shock cardiogenic was diagnosed based on systolic blood pressure <90 mmHg and the use of vasopressor agents (dopamine or norepinephrin or both). Ongoing pain was determined as the complaint of pain lasted >24 hours of admission despite adequate treatment, therefore more doses of analgetic drugs or invasive procedure needed.

Statistical AnalysisFor continuous data, the Kolmogorov-

Smirnov test was performed to analyze the normality of the data. The normally distributed data were presented as mean±SD and compared with student T test or analysis of variance (ANOVA) test. Categorical data were presented as proportion and compared with chi-squared test. Bland-Altman plot was constructed to analyze agreement between eGFR calculations. Receiver operator characteristics (ROC) curve was created to analyze predictive value between eGFR on hospital adverse events. Univariate and multivariable analysis were performed to investigate the association between predictors and study outcomes using forward stepwise logistic regression analysis. Statistical analysis

Vol 45 • Number 2 • April 2013 Predictive value of different estimated glomerular filtration rates

117

was performed using SPSS version 13.0 (SPSS, Chicago, IL, USA). A statistically significant association was considered when p<0.05.

RESULTSPatient’s characteristics were shown in Table

1. As many as 103 patients were analyzed. All patients were eligible for eGFR calculation, except for eGFR based on Cockcroft–Gault, which was eligible only for 51 patients. Our patients were mostly Javanese in ethnicity, which represents half of population in Indonesia and currently showed high prevalence in chronic kidney disease.13,14 Hospital adverse events occurred in 49 patients (47.6%). Table 1 showed that patients with hospital adverse events were more likely to have diabetes mellitus (p=0.02), higher heart rate (p=0.02), higher serum creatinine level (p=0.01), higher serum glucose

Table 1. Patient’s characteristics based on hospital adverse events

Variables Adverse events (n=49) No adverse events (n=54) Total (n=103) p value*

Age (years), mean±SD 58±9.4 57±9.6 58±9.5 0.32

Gender, n(%)

Male 38 (78) 46(85) 84 (82) 0.32

Female 11 (22) 8 (15) 19 (18)

Comorbidity, n(%)

Hypertension 23 (47) 22 (41) 45 (44) 0.53

Diabetes mellitus 13 (27) 5 (9) 18 (17) 0.02

Previous IHD 18 (38) 15 (28) 33 (32) 0.33

Current smoking 23 (47) 23 (43) 46 (45) 0.66

Clinical presentation, mean±SD

Systolic BP (mmHg) 121±23.7 126±24.3 123±24.2 0.27

Diastolic BP (mmHg) 75±17.6 77±15.4 76±16.4 0.55

Heart rate (x/min) 88±24.3 79±15.7 83±20.7 0.02

Laboratory result, mean±SD

White blood cells (x103/mm3)

12.3±4.2 11.1±3.2 11.6±3.7 0.09

Creatinine (mg/dL) 1.6±0.7 1.3±0.5 1.4±0.6 0.01

Glucose (mg/dL) 189.7±107.6 147.2±79.7 167.4±95.8 0.03

eGFR, mean±SD

Cockroft-Gault ** 51.3±17.9 62.9±21.6 56.9±20.5 0.04

MDRD 49.4±18.7 62.7±22.3 56.4±21.6 <0.01

CKD-EPI 52.2±20.1 65.9±21.2 59.4±21.8 <0.01

Diagnosis, n(%)

STEMI 35 (71) 35 (65) 70 (68) 0.47

NSTEMI 14 (29) 19 (35) 33 (32)

* between adverse events vs. no adverse events; ** eligibly calculated for 51 patients

level (p=0.03) and lower eGFRC-G (p=0.04), eGFRMDRD (p<0.01) as well as eGFRCKD-EPI (p<0.01). Stratification according to eGFR value, proportion of hospital adverse events were significantly associated with level of eGFRMDRD (p<0.01) and eGFRCKD-EPI (p=0.02), but not with eGFRC-G (p=0.10) (Figure 1).

The agreement between eGFRMDRD and eGFRCKD-EPI was better in lower eGFR value than in higher eGFR value. An eGFRCKD-EPI tended to reallocate eGFR into higher value as compared with eGFRMDRD, however some patients with higher eGFR showed the contrary tendency toward higher eGFRMDRD (Figure 2).

ROC curve showed higher AUC value in eGFRMDRD as compared to eGFRCKD-EPI (Figure 3). Furthermore, hospital adverse events were slightly better predicted by eGFRMDRD

Anggoro B. Hartopo Acta Med Indones-Indones J Intern Med

118

than by eGFRCKD-EPI (AUC, 0.698; 95%CI: 0.596-0.800, p<0.01 versus AUC, 0.693; 95%CI: 0.591-0.796, p<0.01).

Patient’s characteristics according to eGFRMDRD were presented in Table 2. With

the exception of creatinine and eGFR values, overall characteristics were comparable among eGFR categories.

To investigate the predictors for hospital adverse events, we included all variables into univariate analysis. For continuous data, we

Table 2. Patient’s characteristics according to eGFRMDRD

Variables eGFRMDRD ≥60 (n=47)

eGFRMDRD 60-30 (n=47)

eGFRMDRD <30 (n=9) p value

Years of age, mean±SD 56±9.3 58±9.7 64±5.8 0.60

Gender, n(%)

Male 38 (81) 38 (81) 8 (89) 0.84

Female 9 (19) 9 (19) 1 (11)

Comorbidity, n(%)

Hypertension 17 (36) 25 (53) 3 (33) 0.20

Diabetes mellitus 8 (17) 9 (19) 1 (11) 0.84

Previous IHD 12 (25) 16 (34) 5 (56) 0.19

Current smoking 25 (53) 17 (36) 4 (44) 0.25

Clinical presentation, mean±SD

Systolic blood pressure (mmHg) 124±21.9 125±25.3 113±29.3 0.44

Diastolic blood pressure (mmHg) 77±14.9 77±14.2 66±29.5 0.15

Heart rate (x/min) 78±19.5 88±19.8 86±26.8 0.05

Laboratory result, mean±SD

White blood cells (x103/mm3) 10.8±3.4 12.6±3.5 11.2±5.4 0.07

Creatinine (mg/dL) 1.0±0.2 1.6± 0.3 2.8± 0.6 <0.001

Glucose (mg/dL) 167.9±99.8 169.3±99.3 154.9±54.4 0.92

eGFR, mean±SD

Cockroft-Gault* 69.1±18.5 50.0±14.5 21.7±6.3 <0.001

MDRD 74.3±16.3 45.0±8.9 22.6±4.1 <0.001

CKD-EPI 78.3±13.0 47.4±10.1 22.7±4.2 <0.001

Diagnosis

STEMI 35 (75) 31 (66) 4 (44) 0.19

NSTEMI 12 (26) 16 (34) 5 (56)

* eligibly calculated for 51 patients

Figure 2. Bland-Altman plot to measure agreement between eGFRMDRD and eGFRCKD-EPI showed better agreement in lower eGFR value, whereas in higher eGFR value the bias was more apparent.

Figure 1. Proportion of hospital adverse events according to eGFR level showed proportional changes of events along with severity of kidney function in all eGFR equation. Significant difference of proportion was observed in eGFRMDRD and eGFRCKD-EPI.

Vol 45 • Number 2 • April 2013 Predictive value of different estimated glomerular filtration rates

119

converted it into categorical data, i.e. age >60 years, systolic blood pressure <90 mmHg, heart rate >100 x/min, white blood count (WBC) >11x103/mm3, creatinine >1.5 mg/dL and glucose >150 mg/dL) and entered them into a univariate model. An eGFRMDRD less than 60 predicted hospital adverse events, i.e. eGFRMDRD 30-60 (OR 4.21; 95% CI 1.77-10.04, p<0.01) and eGFRMDRD <30 (OR 9.15; 95% CI 1.68-49.93, p=0.01). Furthermore, an eGFRCKD-EPI <30 also predicted hospital adverse events (OR 7.00; 95% CI 1.35-36.23, p=0.02). However, an eGFRC-G did not significantly predict hospital adverse events (Table 3).

For multivariate analysis (Table 3), eGFRMDRD and eGFRCKD-EPI were adjusted with other variables which had p ≤0.20 on

Table 3. eGFR prediction for hospital adverse events by logistic regression analysis

VariablesUnivariate analysis Multivariable analysis

Unadjusted OR (95%CI) p value Unadjusted OR

(95%CI) p value

Age>60 1.42(0.65-3.08) 0.38 NA** NA

Male gender 1.54(0.71-3.34) 0.28 NA NA

Hypertension 1.29(0.59-2.81) 0.53 NA NA

Diabetes mellitus* 3.54(1.16-10.82) 0.03 - NS***

Previous IHD 1.51(0.66-3.47) 0.33 NA NA

Current smoking 1.21(0.56-2.63) 0.62 NA NA

Systolic <90 mmHg 1.11(0.15-8.17) 0.92 NA NA

Heart rate >100 x/mins* 8.62(2.68-27.69) <0.01 7.25(2.15-24.39) <0.01

WBC > 11x103/mm3 1.33(0.61-2.90) 0.47 NA NA

Creatinine >1.5 mg/dL* 1.95(0.86-4.43) 0.11 - NS***

Glucose >150 mg/dL* 3.16(1.41-7.10) <0.01 - NS***

eGFRMDRD*

≥60 1 - 1 -

30-60 4.21(1.77-10.04) <0.01 3.50(1.38-8.85) <0.01

<30 9.15(1.68-49.93) 0.01 8.13(1.38-47.91) 0.02

eGFRCKD-EPI*

≥60 1 - 1 -

30-60 2.16(0.93-5.02) 0.07 - NS

<30 7.00(1.35-36.23) 0.02 - NS

eGFRC-G

≥60 1 - NA NA

30-60 2.32(0.70-7.71) 0.37 NA NA

<30 2.87(0.57-14.5) 0.31 NA NA

STEMI diagnosis 1.36(0.59-3.13) 0.47 NA NA

* variables entered in multivariable analysis, ** NA, not applied in multivariable analysis, ***NS, not significant

Figure 3. Area under the curve of eGFRMDRD and eGFRCKD-EPI showed eGFRMDRD had slightly better predictive value for hospital adverse events than eGFRCKD-EPI.

Anggoro B. Hartopo Acta Med Indones-Indones J Intern Med

120

univariate analysis, i.e. diabetes mellitus, heart rate >100 x/mins, creatinine >1.5 mg/dL, and glucose >150 mg/dL. Using forward stepwise logistic regression analysis, we found that eGFRMDRD 30-60 (adjusted OR 3.50; 95%CI: 1.38-8.85, p<0.01) and eGFRMDRD <30 (adjusted OR 8.13, 95%CI: 1.38-47.91, p=0.02) were independent predictors for hospital adverse events, whereas eGFRCKD-EPI were not.

DISCUSSIONMeasuring kidney function in acute

myocardial infarction patients is of paramount importance. Not only does it give information on kidney performance, it also provides therapeutic and prognostic implications. Patients with decreased kidney function have greater propensity to side effects from medications, such as hemorrhage from anticoagulants, kidney failure from contrast-angiography and reduced kidney function from angiotensin-converting enzyme inhibitors. Such drawbacks lead to suboptimal therapy due to the withholding of certain medications in patients who in fact need more aggressive theurapeutic modalities due to their illness.

Current scoring systems for acute myocardial infarction, either STEMI or NSTEMI, such as TIMI, GRACE and PURSUIT scores do not incorporate eGFR value into their scoring system.15 GRACE score integrates creatinine value in the scoring systems, however it does not recommend measuring level of kidney function be included in the score. A GFR equation is so far the best indicator for kidney function, an eGFR incorporated creatinine level along with age, race, sex and body size.6 Single creatinine value for predicting hospital events can be misleading, because it is influenced by muscle mass, which is represented by age, race and sex in the eGFR.

In acute myocardial infarction patients hospitalized in the coronary care unit, a significant proportion of patients had moderate to severe kidney dysfunction.16,17 Our study similarly showed more than 50% study patients had moderate to severe kidney dysfunction based on MDRD at admission. We excluded patients with chronic kidney disease stage V who underwent hemodyalisis. Characteristics

of our study populations showed no significant difference among eGFR value based on MDRD.

Our study result showed moderate (eGFR 30-60 mL/min per 1.73 m2) and severe (eGFR <30 mL/min per 1.73 m2) kidney dysfunction, estimated with MDRD, was independently associated with hospital adverse events. Many studies have shown deleterious impacts of moderate to severe kidney dysfunction in hospital mortality and complication following acute myocardial infarction.1,16-21 These studies used eGFR based on MDRD, which is widely approved to measure kidney function as a substitute for the previous Cockroft-Gault equation.7 However, a new equation was developed because eGFR based on MDRD contained systematic bias toward underestimating measured GFR at higher value, which lead to overdiagnosis of chronic kidney disease.8 CKD-EPI was developed and gave more accurate eGFR value than MDRD, especially in general population.22 In an acute illness, such as acute myocardial infarction, several studies had demonstrated higher predictive value of eGFR based on CKD-EPI over MDRD to predict long term outcome,23,24 although other study found different result.25 Few studies evaluated the performance of eGFR based on CKD-EPI as compared to MDRD in the setting of acute myocardial infarction, especially in the Asian population.

Comparing the prediction value between eGFRMDRD and eGFRCKD-EPI using area under curve value, we found that hospital adverse events were better predicted by eGFRMDRD than by eGFRCKD-EPI. Furthermore, in multivariable analysis, moderate and severe kidney dysfunction based on MDRD independently predicted hospital adverse events. Al Faleh et al. (2012) showed different results which favor for eGFR based on CKD-EPI for predicting hospital outcome in acute coronary syndrome.26 Different patient population and severity of disease spectrum may account for this difference.

A small sample size was our main limitation in this study, as a secondary analysis from our previously published study.11 However, our study provided pilot data regarding the predictive value of eGFR based on both MDRD and CKD-EPI on hospital adverse events among acute myocardial

Vol 45 • Number 2 • April 2013 Predictive value of different estimated glomerular filtration rates

121

infarction, at least in our ethnic population.

CONCLUSION MDRD gives better predictive value than

eGFR based on CKD-EPI on hospital adverse events among acute myocardial infarction. An eGFR based on MDRD independently predicted hospital adverse events.

REFERENCES1. Wright RS, Reeder GS, Herzog CA, et al. Acute

myocardial infarction and renal dysfunction: a high-risk combination. Ann Intern Med. 2002;137:563-70.

2. Al Suwaidi J, Reddan DN, Williams K, et al. Prognostic implications of abnormalities in renal function in patients with acute coronary syndromes. Circulation. 2002;106:974-80.

3. Herzog CA, Ma JZ, Collins AJ. Poor long-term survival after acute myocardial infarction among patients on long-term dialysis. N Engl J Med. 1998;339:799–805.

4. Masoudi FA, Plomondon ME, Magid DJ, et al. Renal insufficiency and mortality from acute coronary syndromes. Am Heart J. 2004;147:623-9.

5. Pitsavos C, Kourlaba G, Panagiotakos DB, et al. Association of creatinine clearance and in-hospital mortality in patients with acute coronary syndromes: the GREECS study. Circ J. 2007;71:9-14.

6. National Kidney Foundation. Kidney disease outcome quality initiative (K/DOQI) clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis. 2002;39:Suppl 1:S1-S266.

7. Levey AS, Coresh J, Greene T, et al. Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med. 2006;145:247-54.

8. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150:604-12.

9. Matsushita K, Tonelli M, Lloyd A, et al. Clinical risk implications of the CKD epidemiology collaboration (CKD-EPI) equation compared with the modification of diet in renal disease (MDRD) study equation for estimated GFR. Am J Kidney Dis. 2012;60:241-9.

10. Teo BW, Xu H, Wang D, et al. GFR estimating equations in a multiethnic Asian population. Am J Kidney Dis. 2011;58:56-63.

11. Hartopo AB, Setianto BY, Gharini PPR, et al. On arrival high blood glucose level is associated with detrimental and fatal hospitalization outcomes for acute coronary syndrome. Cardiol Res. 2011;2:160-7.

12. Åkerblom A, Wallentin L, Siegbahn A, et al. Cystatin C and estimated glomerular filtration rate as predictors for adverse outcome in patients with ST-elevation and non–ST-elevation acute coronary syndromes: results

from the platelet inhibition and patient outcomes study. Clin Chem. 2012;58:190–9.

13. Danawati CW, Nagata M, Moriyama H, et al. A possible association of Pro12Ala polymorphism in peroxisome proliferator-activated receptor gamma 2 gene with obesity in native Javanese in Indonesia. Diabetes Metab Res Rev. 2005;21:465-9.

14. Prodjosudjadi W, Suhardjono, Suwitra K, et al. Detection and prevention of chronic kidney disease in Indonesia: initial community screening. Nephrology (Carlton). 2009;14:669-74.

15. Khalill R, Han L, Jing C, et al. The use of risk scores for stratification of non-ST elevation acute coronary syndrome patients. Exp Clin Cardiol. 2009;14:e25-30.

16. Anavekar NS, McMurray JJ, Velazquez EJ, et al. Relation between renal dysfunction and cardiovascular outcomes after myocardial infarction. N Engl J Med. 2004;351:1285-95.

17. Dwipa L, Soelaeman R, Roesli RMA, et al. Cardiometabolic risk factors and acute kidney injury based on urinary neutrophil gelatinase associated lipocalin (NGALu) in acute coronary syndrome patients. Acta Med Indones. 2012;44:3-9.

18. Santolucito PA, Tighe DA, McManus DD, et al. Management and outcomes of renal disease and acute myocardial infarction. Am J Med. 2010;123:847-55.

19. Beattie JN, Soman SS, Sandberg KR, et al. Determinants of mortality after myocardial infarction in patients with advanced renal dysfunction. Am J Kidney Dis. 2001;37:1191-200.

20. Menon V, Sarnak MJ, Lessard D, et al. Recent trends in hospital management practices and prognosis after acute myocardial infarction in patients with kidney disease. Am J Cardiol. 2004;94:1290-3.

21. Goto M, Oda E, Matsushita H, et al. Renal dysfunction was an independent predictor of in-hospital death and ventricular rupture in patients with acute myocardial infarction. Cardiol Res. 2012;3(3):123-32.

22. White SL, Polkinghorne KR, Atkins RC, et al. Comparison of the prevalence and mortality risk of CKD in Australia using the CKD epidemiology collaboration (CKD-EPI) and modification of diet in renal disease (MDRD) study GFR estimating equations: the Aus Diab (Australian diabetes, obesity and lifestyle) study. Am J Kidney Dis. 2010;55:660-70.

23. Choi JS, Kim CS, Bae EH, et al. Predicting outcomes after myocardial infarction by using the chronic kidney disease epidemiology collaboration equation in comparison with the modification of diet in renal disease study equation: results from the Korea acute myocardial infarction registry. Nephrol Dial Transplant. 2012;27:3868-74.

24. Skali H, Uno H, Levey AS, et al. Prognostic assessment of estimated glomerular filtration rate by the new chronic kidney disease epidemiology collaboration equation in comparison with the modification of diet in renal disease study equation. Am Heart J. 2011;162:548-54.

Anggoro B. Hartopo Acta Med Indones-Indones J Intern Med

122

25. Tomaszuk-Kazberuk A, Kozuch M, Malyszko J, et al. Which method of GFR estimation has the best prognostic value in patients treated with primary PCI: Cockcroft-Gault formula, MDRD, or CKD-EPI equation?--A 6-year follow-up. Ren Fail. 2011;33:983-9.

26. Al Faleh HF, Alsuwaida AO, Ullah A, et al. Glomerular filtration rate estimated by the CKD-EPI formula is a powerful predictor of in-hospital adverse clinical outcomes after an acute coronary syndrome. Angiology. 2012;63:119-26.