a b s t r a c t

Background: Brain natriuretic peptide (BNP) is well established in detecting acute decompensation of heart failure (ADHF). The role of BNP at discharge in predicting mortality is less established. Accumulating evidence suggests that inflammatory cytokines play an important role in the development of heart failure. We aimed to examine the contribution of BNP, interleukin 6, and procalcitonin to mortality in ADHF.

Methods: A cohort of 33 patients with ADHF was identified between March 2009 and June 2010 at Rambam Health Care Campus, Haifa, Israel. The cohort was followed up for all-cause mortality during 6 months after hospital discharge. Cox proportional hazard model was used to assess the association between BNP, interleukin-6 and procalcitonin and all-cause mortality.

Results: As compared to BNP at admission, BNP at discharge was more predictive for all-cause mortality. The area under the curve for BNP at admission and discharge was 0.810 (P = .004) and 0.864 (P = .001) respectively. Eleven patients (33.3%) patients who died during the follow-up period had higher BNP levels, median 2031.4 (IQR, 1173.4-2707.2), than those who survived; median 692.5 (IQR, 309.9-1159.9), (P = .001). On multivariate analysis, BNP remained an independent predictor for 6 month all-cause mortality HR 9.58 (95% CI, 2.0-45.89) for levels above the median compared to lower levels, (P = .005). Albumin, procalcitonin and interleukin 6 were not associated with all-cause mortality.

Conclusions: BNP at discharge is an independent predictor for all-cause mortality in patients with ADHF. Compared with BNP at admission, BNP at discharge has slightly higher predictive accuracy with regard to 6- month all-cause mortality.

(C) 2013

Introduction

heart failure is the leading cause for hospitalization among patients older than 65 years of age [1]. Clinical trials and observational studies have identified multiple prognostic factors for mortality in patients admitted with acute decompensated heart failure, including renal function, hyponatremia, brain natriuretic peptide (BNP), and cardiac troponins [2].

Accumulating evidence suggests that inflammatory cytokines play an important role in the development of heart failure. Recent in vitro

? This Work was Supported in part by The Rappaport Family Institute for Research in the Medical Sciences, Technion, Israel Institute of Technology, Haifa, Israel.

* Corresponding author. Department of Internal Medicine “B”, Rambam Health Care Campus, P.O. Box 9602, Haifa, 31096, Israel. Tel.: +972 4 8542676; fax: +972 4

8543252.

E-mail address: z_azzam@rambam.health.gov.il (Z.S. Azzam).

¹ These authors contributed equally to this work.

and in vivo studies have indicated the importance of interleukin 6 (IL- 6) in the development of hypertrophy and the prevention of apoptosis in cardiac myocytes through the activation of their common receptor, gp130 [3]. Previous clinical studies showed that the plasma level of IL- 6 is elevated in patients with advanced CHF and that such high levels are associated with a poor prognosis for CHF patients [3].

Few studies have investigated the role of procalcitonin in patients with acute heart failure, although there was a slightly increased level of procalcitonin in patients with severe acute heart failure, the results were not conclusive. These results support the hypothesis that Inflammatory process has a role in heart failure [4].

BNP is a well-established biomarker for the diagnosis of acute decompensation in patients with heart failure and when combined with clinical assessment, as compared with clinical assessment alone, it has been found useful in assessing the Severity of symptoms and predicting adverse outcomes related to heart failure [5-7]. Interest- ingly, it has been shown that a shorter time to BNP measurement led

0735-6757/$ – see front matter (C) 2013 http://dx.doi.org/10.1016/j.ajem.2013.10.002

Six-month mortality“>to a shorter time to diuretic administration and concomitant lower rates of mortality [8]. In contrast to the role for BNP in the diagnosis of acute decompensation in patients with heart failure, the role of BNP at discharge as a predictor of subsequent complications has not been widely examined. Logeart et al showed that high pre-discharge BNP levels are strong, independent marker of death or re-admission in patients with systolic heart failure and ejection fraction of less than 45% [9].

In the present study, we evaluated the contribution of BNP, IL-6 and procalcitonin to mortality in patients with acute decompensated heart failure as defined by mortality within 6 months from hospital discharge.

Methods

Patient population

The study was performed at Rambam Health Care Campus, Haifa, Israel and included patients who were admitted to internal medicine department with the primary diagnosis of acute decompensated heart failure between March 29, 2009 and June 5, 2010. Notably, after the initiation of the study, due to logistic problems, recruitment of patients in the study was withhold and continued after few months. ADHF was diagnosed according to the European Society of Cardiology criteria including a brain natriuretic peptide (BNP)

level N 400 pg/mL [10].

Patients were excluded if they had any of the following: age under 18 years and pregnant women.

Data that were collected included: (1) Demographic and clinical characteristics (age, gender, and duration of hospitalizations); (2) co- morbidities (ischemic heart disease, hypertension, diabetes mellitus, hyperlipidemia and atrial fibrillation); (3) laboratory variables; at least 2 sets of blood samples were obtained from all participants in the study (at the day of admission and at the day of discharge). Hemoglobin levels, mean corpuscular volume and red cell distribution width (RDW) were measured using the Advia 120 Hematology Analyzer (Siemens Healthcare Diagnostics, Deerfield, IL, USA). Glucose, Blood urea nitrogen and Creatinine levels were measured using the “Dimension” (Siemens Healthcare Diagnostics). IL-6 levels were determined using the “Quantikine human Interleukin 6 Elisa assay” (R&D Systems, MN, USA). procalcitonin levels were analyzed using an automated microparticle immunoassay for Procal- citonin (Liaison Brahms PCT; Brahms Diagnostics, Berlin, Germany). BNP levels were determined using the BNP microparticle enzyme immunoassay using an AXSYM system (Abott Laboratories, Green Oaks, IL; USA).

Study endpoint

All patients were followed up for 6 months after hospital discharge. The primary end point of the study was all-cause mortality. Mortality data were retrieved from the database of our hospital and the ministry of health.

Statistical analyses

Continuous variables are presented as means and standard deviations along with medians and interquartile range as some variable were not normally distributed. Categorical variable are presented as proportions. Comparisons of continuous variables between two groups were studied with the Mann-Whitney test, and comparisons of proportions between categorical variables were studied with the chi-square test or the Fisher exact test as appropriate. Subjects were followed up for 6 months from their discharge, and time to death was calculated from their discharge until death. The distribution of time to death is presented by Kaplan-Meier curves and compared with log-rank test. The association between time to death

and potential predictors of mortality were analyzed by the univariate Cox proportional hazards model. Variables with P <= .1 in the univariate model were included in the multivariate Cox proportional hazards model using the backward selection. The results of univariate and multivariate analyses are summarized by Hazard ratios (HRs) along with 95% confidence interval.

The predictive value of BNP at admission and BNP at discharge for 6-month all-cause mortality was assessed calculating the area under the curve of the receiver operating characteristic curve.

For all analyses, P b .05 for 2-tailed tests was considered statistically significant. All statistical analyses were performed using SPSS 18.0 (SPSS Inc, Chicago, IL).

The study was conducted in accordance with the principles of the Declaration of Helsinki and approved by The Rambam Hospital Institutional Review Board.

Results

Study population characteristics

Overall 33 patients were recruited in the study out of 125 patients who were admitted with the primary diagnosis of acute decom- pensated heart failure during the study period. 15 (45.5%) were females, and the median age was 80.0 years (IQR, 75.5-88.0). Using the median cut point of BNP levels at the time of discharge, we categorized the study patients into 2 groups: subjects with BNP below the median (<=992.6 pg/mL) and subjects with BNP above the median (N 992.6 pg/mL). Baseline and laboratory characteristics according to BNP level are depicted in Tables 1A and 1B, respectively. Compared to patients with BNP below the median, patients with BNP levels above the median were more likely to be males and have ischemic heart disease (P = 0.022 and 0.019, respectively) (Table 1A).

Six-month mortality

Eleven patients (33.3%) patients died during the follow-up period. Baseline and laboratory characteristics of subjects according to the outcome (non-survivors compared to survivors) are depicted in Table 2A and 2B. Subjects who died were more likely to have ischemic heart disease and atrial fibrillation (P = .040 and .004, respectively) (Table 2A).

Subjects who expired, have higher BNP levels, median 2031.4 (IQR, 1173.4-2707.2), than those who stayed alive at the end of follow-up, median 692.5 (IQR, 309.9-1159.9), (P = .001). Subjects who died had also higher BUN and troponin levels, and lower Albumin levels (Table 2B).

The distribution of time death according to the level of BNP at discharge is shown in Fig. 1. BNP, albumin, BUN, creatinine, the presence of atrial fibrillation, and a prior history of congestive heart failure were the only variables found to be significantly associated with 6-month all-cause mortality on univariate Cox proportional hazard analysis (Table 3). On multivariate analysis, BNP remained independently associated with six month mortality; HR 9.58 (95% CI, 2.0-45.89) for levels N 992.6 pg/mL compared to lower levels, (P =

.005) (Table 3). In addition to BNP, atrial fibrillation was also associated with six month mortality in multivariate analysis. The Spearman’s correlation coefficient between BNP and atrial fibrillation was very low; r = -0.090 (P = .624). Notably, albumin did not reach statistical significance on multivariate analysis, even when it was forced into the model. In addition BNP remained an independent predictor of 6-month mortality when albumin was forced in the model; HR 6.48 (95% CI, 1.28-32.72).

Procalcitonin and IL-6 did not reach statistical significance on univariate Cox proportional hazards model, neither when they were forced into the multivariate model.

Table 1A

Baseline characteristics of patients with decompensated heart failure as primary diagnosis at discharge according to BNP levels (median cut points)

Table 1B

Laboratory characteristics of patients with decompensated heart failure as primary diagnosis at discharge according to BNP levels (median cut points)

Variables	BNP	BNP	P		Variables	BNP	BNP	P
	<=992.6 pg/mL	N 992.6 pg/mL				<=992.6 pg/mL	<=992.6 pg/mL
	(n = 17)	(n = 16)				(n = 17)	(n = 16)
Age (y)			.80		Sodium (mEq/L)			.169
Mean +- SD	79.7 +- 11.6	79.0 +- 10.9			Mean +- SD	140.0 +- 4.7	142.5 +- 3.4
Median (IQR)	80.0 (76.0-88.5)	81.0 (74.2-87.2)			Median (IQR)	141.0 (139.0-142.5)	141.5 (140.0-145.0)
Gender (%)			.022		Albumin (g/dL)			.086
Females	11 (64.7%)	4 (20.5%)			Mean +- SD	3.31 +- 0.43	2.92 +- 0.65
Males	6 (33.3%	12 (75%)			Median (IQR)	3.30 (3.0-3.75)	3.05 (2.80-3.45)
Smoking (%)	5 (29.4%)	8 (50.0%)	.296		Interleukin-6 (pg/mL)			.666
BMI? (kg/m2)	26.3 +- 4.4	27.5 +- 4.9	.670		Mean +- SD	32.8 +-19.9	44.0 +- 39.0
Mean +- SD	25.5 (22.9-29.5)	25.6 (24.3-32.7)			Median (IQR)	26.7 (21.1-42.7)	27.8 (17.6-54.0)
Median (IQR)					Troponin			.174
Duration of hospitalization	5.5 +- 4.9	5.8 +- 3.9	.443		Mean +- SD	0.09 +- 0.21	0.12 +- 0.13
Mean +- SD	5.0 (2.0-6.0)	5.5 (3.0-7.0)			Median (IQR)	0.02 (0.01-0.11)	0.09 (0.01-0.15)
Median (IQR)					Procalcitonin (ng/mL)			.650
Re-hospitalization for CHF	7 (41.2%)	6 (37.5%)	.829		Mean +- SD	0.48 +- 1.47	0.22 +- 0.23
Ejection fraction b50% (%)	7 (41.2%)	8 (61.5%)	.269		Median (IQR)	0.1 (b0.1-0.1)	0.1 (b 0.1-0.13)
Left ventricular end					Creatinine (mg/dL)			.171
diastolic volume?? (mL)					Mean +- SD	1.43 +- 0.63	1.77 +- 0.75
Mean +- SD	162.7 +- 42.6	158.9 +- 102.6	.229		Median (IQR)	1.20 (1.0-2.0)	1.70 (1.18-2.08)
Median (IQR)	160.0	138.0			BUN (mg/dL)			.052
	(118.5-194.0)	(102.6-174.0)			Mean +- SD	32.60 +- 19.0	48.7 +- 27.5
Left ventricular end			.563		Median (IQR)	28.0 (20.5-43.0)	48.5 (26.0-60.5)
systolic volume??? (mL)					Hemoglobin (g/dL)			.652
Mean +- SD	75.7 +- 56.0	102.6 +- 99.6			Mean +- SD	11.2 +- 1.9	11.7 +-1.3
Median (IQR)	60.5 (35.0-116.0)	79.5 (30.5-137.5)			Median (IQR)	11.8 (9.8-12.8)	11.6 (10.8-12.9)
Heart rate (per min)			.339		RDW			.639
Mean +- SD	62.5 +- 15.0	72.9 +- 13.6			Mean +- SD	15.6 +- 1.9	16.4 +- 3.4
Median (IQR)	76.0 (69.0-80.5)	72.5 (63.2-79.0)			Median (IQR)	15.5 (13.8-16.5)	15.8 (14.3-16.9)
Systolic blood pressure			1.0
Mean +- SD	128.6 +- 28.3	123.7 +- 18.0

Median (IQR) 118.0

(107.5-149.0)

Diastolic blood pressure Mean +- SD	62.5 +- 15.0	63.7 +- 14.1	.588
Median (IQR)	61.0 (49.5-72.0)	68.0 (51.0-75.2)
QRS duration N 120 (ms)	0 (0%)	3 (18.8%)	.103
Mechanical ventilation (%)	0 (0%)	2 (12.5%)	.227
Cardiopulmonary resuscitation (%)	0 (0%)	2 (12.5%)	.227

126.5

(107.2-136.0)

Discussion

The present study shows that higher BNP levels at discharge are associated with increased 6 month all-cause mortality. As compared with BNP at the time of admission, the level of BNP at discharge showed slightly higher prediction accuracy for 6-month all-cause

Defibrillation (%) 0 (0%) 0 (0%) –

Cardiac catheterization (%) 0 (0%) 1 (6.3%) .485

Dialysis 0 (0%) 1 (6.3%) .485

Prior history of CHF (%) 10 (58.8%) 12 (75.0%) .465

Ischemic heart disease (%) 10 (58.8%) 15 (93.8%) .019

Hypertension 12 (70.6) 13 (81.3) .688

Diabetes mellitus (%) 7 (41.2%) 6 (37.5%) .830

Peripheral vascular disease 1 (5.9%) 4 (25.0%) .175

Chronic lung disease (%)	1 (5.9%)	2 (12.5%)	.601
Hyperlipidemia (%)	9 (52..9%)	8 (50.0%)	.866
Atrial fibrillation (%)	10 (58.8%)	8 (50.0%)	.732

mortality in patients with acute decompensated heart failure.

As heart failure continues to be one of the leading causes for hospitalization among patients older than 65 years of age [1], endless efforts are invested in discovering better predictors for its related mortality and complications. Nowadays, BNP is considered one of the most important biomarkers in the diagnosis of acute decompensated heart failure [11]. Recently, it was reported that BNP is reliable as a follow up marker for monitoring Treatment response [12-14]. The role of BNP levels at the time of discharge as a predictor of mortality has

Loop diuretics (%)	14 (82.4%)	15 (93.8%)	.316	not been sufficiently examined.
?-Blockers (%)	14 (82.4%)	13 (81.3%)	.932	Our findings are consistent, in part; with the findings of Chen at el
Spirinolactone (%)	4 (23.5%)	2 (12.5%)	.412	that BNP level before discharge is a significant predictor of

ACE-inhibitors/ARBs (%)	15 (88.2%)	13 (81.3%)	.576
Digoxin (%)	1 (5.9%)	1 (6.3%)	.996

* Data for BMI was available in 10 (30.3%) patient.

?? Data for left ventricular end diastolic volume was available in 24 (72.7%) patients.

??? Data for left ventricular end systolic volume was available in 16 (48.5%) patients.

The predictive accuracy values, for 6-month all-cause mortality, of BNP level at the time of admission and the time of discharge were compared using receiver operating characteristic curve analyses. The area under the curve was 0.810 (P = .004) for BNP level at admission and 0. 864 (P = .001) for BNP level at discharge (Fig. 2).

Using BNP threshold of 992.6 pg/mL at discharge, the positive predictive value and the negative predictive value (NPV) for 6-month all-cause mortality were 56.3% and 88.2%, respectively. Notably, the threshold of 690 pg/mL for BNP level has an NPV of 100% for 6-month all-cause mortality.

rehospitalization, but not mortality [15]. Notably, in disagreement with our study, Chen et al have shown that blood urea nitrogen (BUN) was a significant predictor of rehospitalization and mortality. This discrepancy might be related to the small number of our cohort.

BNP levels at discharge were the best predictor of 1-year mortality and/or rehospitalization among older patients hospitalized with heart failure in the OPTIMIZE-HF study [16]. Logeart et al, also, showed that high pre-discharge BNP levels are strong, independent marker of death or re-admission in patients with systolic heart failure and ejection fraction of less than 45% [9].

It should be emphasized that when a threshold of 690 pg/mL for BNP level was used; the NPV for 6-month all-cause mortality was 100% in our study. Cheng et al found that low (b 430 pg/mL) BNP level at the time of discharge was a strong negative predictor of hospital re- admission [17].

Interestingly the time required to achieve clinical improvement was not different between subjects with high and low BNP levels as evidence by the similar duration of hospitalization of the two groups

Table 2A

Baseline characteristics of patients with decompensated heart failure as primary diagnosis at discharge according to the outcome (Survivors vs. non-survivors)

Table 2B

Laboratory characteristics of patients with decompensated heart failure as primary diagnosis at discharge according to the outcome (survivors vs. non-survivors)

Variables	Survivors	Non-survivors	P		Variables	Survivors	Non-survivors	P
Age (years)					Sodium (mEq/L)			.077
Mean +- SD	81.8 +- 9.2	74.6 +- 12.8			Mean +- SD	140.3 +- 4.2	143.1 +- 3.9
Median (IQR)	82.5 (76.0-89.0)	76.0 (59.0-85.0)			Median (IQR)	140.5 (139.0-143.2)	142.0 (141.0-145.0)
Gender (%)					Albumin (g/dL)			.003
Females	12 (54.5%)	3 (27.3%)	.116		Mean +- SD	3.34 +- 0.37	2.68 +- 0.66
Males	10 (45.5%)	8 (72.7%)	.266		Median (IQR)	3.30 (3.05-3.72)	2.80 (2.40-3.10)
Smoking (%)	8 (36.4%)	11 (100%)	.714		Interleukin-6 (pg/mL)			.147
BMI? (kg/m2)			.037		Mean +- SD	32.7 +- 26.5	49.4 +- 36.7
Mean +- SD	24.9 +- 2.0	34.3 +- 0.56			Median (IQR)	26.6 (11.4-38.2)	40.3 (24.6-58.0)
Median (IQR)	24.6 (23.7-26.6)	34.3 (33.9-34.7)			Troponin			.015
Duration of hospitalization			.088		Mean +- SD	0.09 +- 0.20	0.14 +- 0.14
Mean +- SD	5.0 +- 4.5	6.8 +- 4.3			Median (IQR)	0.02 (0.0-0.12)	0.09 (0.05-0.19)
Median (IQR)	4.5 (2.0-6.2)	6.0 (3.0-8.0)			Procalcitonin (ng/mL)			.669
Re-hospitalization for CHF	8 (36.4%)	5 (45.5%)	.714		Mean +- SD	0.46 +- 1.32	0.14 +- 0.11
Ejection fraction b50% (%)	11 (52.4%)	4 (44.4%)	.690		Median (IQR)	0.10 (0.10-0.10)	0.10 (0.10-0.14)
Left ventricular end			.924		Creatinine (mg/dL)			.027
diastolic volume?? (ml)					Mean +- SD	1.40 +- 0.60	2.0 +- 0.77
Mean +- SD	73.4 +- 49.5	123.8 +- 123.7			Median (IQR)	1.20 (1.0-1.84)	2.0 (1.21-2.50)
Median (IQR)	70.0 (34.0-88.0)	76.0 (31.5-240.0)			BUN (mg/dL)			.012
Left ventricular end			.533		Mean +- SD	32.9 +- 19.2	55.4 +- 28.0
systolic volume??? (ml)					Median (IQR)	27.5 (19.2-42.2)	49.0 (32.0-76.0)
Mean +- SD	152.2 +- 45.0	182.9 +- 114.4			Hemoglobin (g/dL)			.566
Median (IQR)	147.0 (112.5-194.0)	141.0 (114.0-216.0)			Mean +- SD	11.5 +- 1.8	11.4 +- 1.5
Heart rate (per min)			.528		Median (IQR)	11.8 (10.2-12.8)	11.1 (10.1-13.1)
Mean +- SD	75.0 +- 11.3	73.8 +- 15.7			RDW			.14
Median (IQR)	75.0 (68.0-80.0)	72.0 (62.0-80.0)			Mean +- SD	15.7 +- 2.9	16.5 +- 1.5
Systolic blood pressure			.047		Median (IQR)	15.2 (13.9-16.3)	16.5 (15.2-17.7)
Mean +- SD	131.7 +- 24.6	115.2 +- 17.7			BNP (pg/mL)			.001
Median (IQR)	127.0 (111.7-148.0)	108.0 (102.0-127.0)			Mean +- SD	843.3 +- 646.9	1989.3 +- 861.8
Diastolic blood pressure			.491		Median (IQR)	629.0 (309.9-1159.9)	2031.4 (1173.4-2707.2)
Mean +- SD	64.9 +- 14.6	59.4 +- 13.8
Median (IQR)	63.5 (55.2-77.0)	61.0 (48.0-71.0)

?? Data for left ventricular end diastolic volume was available in 24 (72.7%) patients.

QRS duration N 120 (ms)	1 (4.5%)	2 (18.2%)	.252
Mechanical ventilation (%)	0 (0%)	2 (18.2%)	.104
Cardiopulmonary	0 (0%)	2 (18.2%)	.104
resuscitation (%) Defibrillation (%) 0 (0%)		0 (0%)	–
Cardiac catheterization (%) 1 (4.5%)		0 (0%)	1.0
Dialysis 0 (0%)		1 (9.1%)	.333
Prior history of CHF (%) 12 (54.5%)		10 (90.9%)	.040
Ischemic heart disease (%) 17 (77.3%)		8 (72.7%)	1.00
Hypertension 18 (81.8%)		7 (63.6%)	.391
Diabetes mellitus (%) 9 (40.9%)		4 (36.4%)	1.00
Peripheral vascular disease 3 (13.6%)		2 (18.2%)	1.00
Chronic lung disease (%) 2 (9.1%)		1 (9.1%)	1.0
Hyperlipidemia (%) 14 (63.6%)		3 (27.3%)	.071
Atrial fibrillation (%) 8 (36.4%)		10 (90.9%)	.004
Loop diuretics (%) 18 (81.8%)		11 (100%)	.276
?-Blockers (%) 18 (81.8%)		9 (81.8%)	1.0
Spirinolactone (%) 4 (18.2%)		2 (18.2%)	1.0
ACE-inhibitors/ARBs (%) 20 (90.9%)		8 (72.7%)	.304
Digoxin (%) 11 (52.4%)		4 (44.4%)	.606
* Data for BMI was available in 10 (30.3%) patient.

??? Data for left ventricular end systolic volume was available in 16 (48.5%) patients.

cause of death and therefore we were not able to assess the association with cause specific mortality. Nevertheless, the results of this study deserve consideration, and may serve for hypothesis generation for larger future studies.

The numbers of events in the first month were very small; therefore we could not assess the association with shorter term

BNP ?992.6 pg/ml

BNP >992.6 pg/ml

Log-rank P value = .007

(Table 1A). In addition, the duration of hospitalization was not associated with all-cause mortality in our patients. Our findings may suggest that persistent Neurohormonal activation that lags behind clinical improvement may better reflects ADHF severity and predict worse prognosis despite clinical improvement.

In summary, in patients hospitalized with ADHF, BNP level at discharge is essential in the risk stratification and decision making processes, as patients with higher levels have increased 6-month mortality risk when compared to patients with lower BNP levels. Therefore, we recommend incorporating BNP level at the time of discharge in all future models being developed to predict post- discharge mortality in patients hospitalized due to ADHF.

The main limitation of this study is the small sample size of participants in each group. In addition we didn’t have data on the

Fig. 1. Kaplan-Meier curves for 6-month mortality according to BNP levels, among patients with decompensated heart failure.

Table 3

Univariate and multivariate Cox proportional hazard analysis for 6-month mortality among patients with decompensated heart failure

Variables Univariate? Multivariate??

	HR (95% CI)	P		HR (95% CI)	P
BNP <=992.6 pg/mL	Reference		Reference
N 992.6 pg/mL	6.23 (1.34-28.9)	.019	9.58 (2.0-45.89)		.005
Albumin <=3.2 g/dL	6.46 (1.39-30.0)		–		–
N 3.2 g/dL	Reference	.017

BUN (per 10 mg/	1.24 (1.04-1.48)	.018	– –
dL increase)
Creatinine (per 1 mg/	2.76 (1.20-6.36)	.017	– –

dL increase)

Prior history of CHF	5.85(0.75-45.78)	.092	–	–
Atrial fibrillation	10.54(1.34-82.66)	.025	16.15 (2.0-130.31)	.009

* Only variables with P <= .1 are shown. Variables considered for univariate analysis includes; age, gender, smoking, prior history of CHF, comorbidities, duration of hospitalization, re-hospitalization for CHF, blood tests at time of discharge (sodium, albumin, IL-6, procalcitonin, BNP, troponin, hemoglobin, RDW, BUN, and creatinine), ejection fraction, cardiac volumes assessed by echocardiography, QRS duration, heart rate, systolic and diastolic blood pressure, and medications (?-blockers, ACE-inhibitors/ ARBs, spirinolactone, digoxin, and loop diuretics).

?? Variable with P <= .1 in univariate analysis were included in the multivariate model.

Only variable with P b.05 are shown.

mortality. In addition, for the purpose of this study we used biochemical biomarker levels that were tested at the time of discharge instead of levels tested at a predefined time from admission. However, this selection may even sound more logical because it represents the actual condition of the patient at the time of discharge as evidenced by the higher performance BNP at the time of discharge compared to the time of admission (Figs. 1 and 2). Moreover; for the purpose of this study we used serum creatinine as a proxy of renal function instead of the more accurate estimated Glomerular Filtration Rate.

In conclusion, BNP level at the time of discharge is an independent predictor of 6-month mortality in patients with acute decompensated

heart failure. Compared with BNP level in admission, BNP level at discharge better predicted all 6-month all-cause mortality.

References

1. Oh J, Kang SM, Hong N, et al. Relation between red cell distribution width with echocardiographic parameters in patients with acute heart failure. J Card Fail 2009;15:517-22.
2. Gheorghiade M, Pang PS. Acute heart failure syndromes. J Am Coll Cardiol 2009;53:557-73.
3. Hirota H, Izumi M, Hamaguchi T, et al. Circulating interleukin-6 family cytokines and their receptors in patients with congestive heart failure. Heart Vessels 2004;19:237-41.
4. Brunkhorst FM, Clark AL, Forycki ZF, Anker SD. Pyrexia, procalcitonin, immune activation and survival in cardiogenic shock: the potential importance of bacterial translocation. Int J Cardiol 1999;72:3-10.
5. Maisel A, Hollander JE, Guss D, et al. Primary results of the Rapid Emergency Department Heart Failure Outpatient Trial (REDHOT). A multicenter study of B- type natriuretic peptide levels, emergency department decision making, and outcomes in patients presenting with shortness of breath. J Am Coll Cardiol 2004;44:1328-33.
6. Baggish AL, van Kimmenade RR, Pinto Y, et al. New York Heart Association class versus amino-terminal pro-B type natriuretic peptide for acute heart failure prognosis. Biomarkers 2010;15:307-14.
7. Januzzi Jr JL, Maisel AS, Silver M, Xue Y, DeFilippi C. Natriuretic peptide testing for predicting adverse events following heart failure hospitalization. Congest Heart Fail 2012;18(Suppl 1):S9-S13.
8. Maisel AS, Peacock WF, McMullin N, et al. Timing of immunoreactive B-type natriuretic peptide levels and treatment delay in acute decompensated heart failure: an ADHERE (Acute Decompensated Heart Failure National Registry) analysis. J Am Coll Cardiol 2008;52:534-40.
9. Logeart D, Thabut G, Jourdain P, et al. Predischarge B-type natriuretic peptide assay for identifying patients at high risk of re-admission after decompensated heart failure. J Am Coll Cardiol 2004;43:635-41.
10. Dickstein K, Cohen-Solal A, Filippatos G, et al. ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure 2008: the Task Force for the Diagnosis and Treatment of Acute and Chronic Heart Failure 2008 of the European Society of Cardiology. Developed in collaboration with the Heart Failure Association of the ESC (HFA) and endorsed by the European Society of Intensive Care Medicine (ESICM). Eur Heart J 2008;29:2388-442.
11. Yancy CW, Jessup M, Bozkurt B, et al. 2013 ACCF/AHA Guideline for the Management of Heart Failure: A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2013;62:1495-539.

    Fig. 2. Receiver operating characteristic curves for the performance of BNP at the time of admission and at the time of discharge in patients with decompensated heart failure as primary discharge diagnosis. The numbers on the curves show the corresponding BNP cut point level (pg/mL). The area under the curve was 0.810 (P = .004) for BNP level at admission and .864 (P = .001) for BNP level at discharge.

    Di Somma S, Magrini L, Tabacco F, et al. Brain natriuretic peptide and N-terminal pro-B-type natriuretic peptide show a different profile in response to acute decompensated heart failure treatment. Congest Heart Fail 2008;14:245-50.

12. Di Somma S, Magrini L, Pittoni V, et al. In-hospital percentage BNP reduction is highly predictive for adverse events in patients admitted for acute heart failure: the Italian RED Study. Crit Care 2010;14:R116.
13. Maisel A, Barnard D, Jaski B, et al. Primary results of the HABIT Trial (heart failure assessment with BNP in the home). J Am Coll Cardiol 2013;61:1726-35.
14. Chen CY, Yoshida A, Asakura M, et al. Serum blood urea nitrogen and plasma brain natriuretic peptide and low diastolic blood pressure predict cardiovascular

    morbidity and mortality following discharge in acute decompensated heart failure patients. Circ J 2012;76:2372-9.

    Kociol RD, Horton JR, Fonarow GC, et al. Admission, discharge, or change in B-type natriuretic peptide and long-term outcomes: data from Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure (OPTIMIZE-HF) linked to Medicare claims. Circ Heart Fail 2011;4: 628-36.

15. Cheng V, Kazanagra R, Garcia A, et al. A rapid bedside test for B-type peptide predicts Treatment outcomes in patients admitted for decompensated heart failure: a pilot study. J Am Coll Cardiol 2001;37:386-91.