The Added Value of Bioimpedance Analysis to NT-proBNP in Predicting Short-term outcome in Acute Heart Failure Patients

Background: Acute heart failure (AHF) is a significant clinical problem, which has a high prevalence, mortality, and rehospitalization. Congestion is critical in the pathogenesis of AHF, which is also a predictor of mortality and rehospitalization in patients with AHF. Objective: Knowing the effect of adding a %Total body water (TBW) test detected by Bioimpedance Analysis to NT-proBNP as a short-term clinical outside predictor of patients with acute heart failure Method: This research is an analytical observational study using prospective cohort methods. The research was conducted at Dr. Saiful Anwar Malang Hospital in January 2018-July 2019 with research subjects taken consecutively against all AHF patients hospitalized at RSUD dr. Saiful Anwar Malang. The data were taken in the form of NT-proBNP value during mission and %TBW before the patient leaves the hospital detected with NICaS examination. 90 days follow-up was conducted via phone calls to the patients, to evaluate the patient’s follow-up for cardiovascular death and rehospitalization. Results: This study involved 65 subjects who were on average 61 years old and the majority male. Non-survivor patients had a higher NYHA class, NT-proBNP, and %TBW pre-discharge than the survivor group. Based on a statistical analysis we found that the ROC curve shows NT-proBNP is a good predictor of mortality (HR: AUC 0.74; 95%CI 0.59-0.90) and rehospitalization (HI: AUC 0.88; 95%CI 0.78-0.97). Similarly , %TBW pre-discharge shows good predictors of mortality (HI: AUC 0.72, 95%CI 0.56-0.87) and rehospitalization (HI: AUC 0.83, 95%CI 0.73-0.94). The addition of the %TBW pre-discharge parameter to NT-proBNP results in the best predictor numbers among the three for both mortality (HI: AUC 0.84; 95%CI 0.72-0.96) and rehospitalization (HI: AUC 0.92; 95%CI 0.85-1.00). Conclusion: The addition of pre-discharge %TBW examination detected by a bioimpedance analysis tool to NT-pro BNP at admission, increases the predicted value of short-term clinical outpatient in the form of mortality and rehospitalization of patients with acute heart failure.

Acute heart failure (AHF) is a serious clinical problem that has led the patient to seek acute care and requiring urgent diagnosis, facilitating a treatment decision that is appropriate to the particular AHF mechanism. 1 Furthermore, in-hospital mortality rates are higher after discharge, ranging from 4% to 8%, are estimated to be 8-15% by 3 months, and are followed by regular rehospitalizations (30-38 percent at 3 months). 2,3 The presence of signs and symptoms of systemic and pulmonary congestion is the product of AHF clinical features. 4 The main feature in the pathogenesis of AHF is congestion or fluid overload, and it is present in most patients admitted to this disease. Its severity was related to the prognosis of acute heart failure patients. 5 Moreover, congestion is both a cause and an effect of worsening cardiovascular function in heart failure. As a result of elevated ventricular wall tension, these pathophysiological alterations relate to clinical deterioration and brain natriuretic peptide (BNP) secretion. 6 For all these concerns, in order to avoid this progressive hemodynamic pathway, AHF therapy is undertaken and, as a result, decongestion is an important priority in the care of these patients. 7 Nearly half of hospitalized AHF patients, however, are discharged with chronic congestion signs and symptoms, and this is of considerable prognostic importance because the capacity to keep patient congestion-free will define a good survival population. 8 The objective of this study is to investigate the additional prognostic power of combining congestion status detected by bioimpedance analysis (BIA) to standardized NT-proBNP.

36
Original Article This is an observational analytic, with cohort prospective study. The sample of this study were patients with acute heart failure (AHF) presenting and admitted to Saiful Anwar General Hospital Malang. The inclusion criteria were: patients with age > 40 years old, who admitted with dyspnea due to AHF or symptoms linked to AHF, with hospitalization predicted. According to European Society of Cardiology standards, each patient got routine treatment and BIA values were blinded to the physicians. Exclusion criteria were pregnancy, acute coronary syndrome, congenital heart disease, the patient experienced cardiac arrest, patient with stenotic valve disease, or prosthetic valve, COPD, pneumothorax, patient with malignancy or terminally ill with life expectancy less than 6 months, and patients who were discharged without beta-blockers or ACE inhibitor. The final diagnosis of AHF was made according to the recent guideline. Upon arrival at the ED, a blood examination for NT-proBNP was carried out for every patient. After that, the sample was sent to the central laboratory and was analyzed with a sandwich-type electrochemiluminescence sensor. We used a bioelectrical impedance analyzer, to analyze the percentage of total body water (TBW) of the patients, at the time before discharge (pre-discharge). TBW was measured non-invasively using NICaS (Non-Invasive Cardiac System, Ni Medical Ltd., Singapore), bedside at supine position Without contact with metal and with abducted inferior limbs at 45 degrees and with abducted superior limbs at 30 degrees to prevent contact with the flesh. A phone call follow-up was conducted 90 days after hospital discharge to assess the follow-up of the patient for coronary mortality and rehospitalization.

Statistical Analysis
Data were analyzed with SPSS 22 Software. Univariate analysis was performed for baseline characteristics. A bivariate analysis differences in medical therapy during hospitalization between survivors and non-survivors. All patients were discharged without any sign of clinical congestion.

Discussion
Both BIA and BNP are presently used as valid methods for managing patients with AHF. 9,10 Two facets of the same pathophysiological pathway are reflected: BNP is released due to elevated 37  percentage TBW at discharge, and NT-proBNP admission are significantly associated with 90 days mortality and rehospitalization. At multiple regression analysis, only NT-proBNP admission and percentage TBW predischarge maintain statistical significance (Table. 3). ROC analysis shows that percentage TBW pre-discharge is a predictor of death and rehospitalization (respectively, AUC : 0.72, 95% CI 0.56 -0.87.5 p<0.001 and AUC: 0.83, 95% CI 0.726 -0.940 p<0.001).

Discussion
Both BIA and BNP are presently used as valid methods for managing patients with AHF. 9,10 Two facets of the same differences in medical therapy during hospitalization between survivors and non-survivors. All patients were discharged without any sign of clinical congestiton.
From the univariate analysis, NYHA, serum creatinine, rales,  pathophysiological pathway are reflected: BNP is released due to elevated ventricular filling pressure and wall tension, while BIA variables reflect changes in body hydration of the failing cardiovascular system. 11 BNPs alone, however, may not indicate pathophysiological congestion mechanisms, and other congestion indicators can be useful in determining congestion and risk stratification in patients with HF. BIA is a quick and easy tool for the evaluation of body peripheral hydration status among emerging strategies, offering prognostic information in HF patients. 12,13 A recent study has shown that bioimpedance analysis expressed in hydration status or total body water was related to NT-proBNP and CVP. 14 In a recent analysis of the literature, Oremus et al. indicate that NT-proBNP in both decompensated and chronic healthy HF populations is correlated with all-cause mortality and composite outcome. 6 There is, however, insufficient evidence in the literature that NT-proBNP contributes incremental importance to other prognostic variables in the short and long-term estimation of all-cause and cardiovascular mortality. 15 Instead, when used in conjunction with BIA, NT-proBNP was more predictive of death and rehospitalization in our study. As for BIA variables at discharge, our analysis shows that with a chronic state of congestion, certain HF patients are discharged. Their detection is important since these patients are those at increased risk of death and rehospitalization for 90 days. 16 The role of BIA in detecting congestion and latent fluid overload, Valle et.al showed that it has better diagnostic sensitivity compared to other parameters such as fluid balance, and It also demonstrates an emerging role in the prognostic significance. 17 A previous study by Santarelli et.al indicates that the elimination of congestion observed by BIA during hospitalization of AHF patients is correlated with an improvement in the outcome of patients. 14,18 Donner Alves et al., who identify incremental improvements in hydration status during diuretic treatment in hospitalized AHF patients, demonstrate the same outcome. 19 Instead, BIA prognostic significance before discharge was measured in this study. It was performed taking into account that clinical symptoms of congestion are frequently absent at this point as a result of diuretic treatment, and clinical judgment may not be accurate.
This study offers valuable knowledge from this point of view that BIA analysis is beneficial as it facilitates the detection of certain patients that have subclinical congestion, and this could aid doctors in determining the discharge of the patient. Moreover, our findings support the effectiveness of BNP in the risk stratification of patients who presented to ED with AHF, but the most crucial and newest element is obtained from the combined use of NT-proBNP and BIA.

39
The previous study shows NT-proBNP and BIA can be very helpful in the diagnosis, but also the clinical care of patients with AHF. This research expands these results, showing that they would have a poorer result when AHF patients are discharged with a chronic congestive state, even if clinically latent. 18 The limitations of this study are the sample size due to only a single center was involved in our study and the period of follow-up was quite short to rule-out biases.

Conclusion
Our study shows that in AHF patients, adding hydration status, measured as a percentage of total body water by bioimpedance analysis device before hospital discharge, to admission NT-proBNP increases short-term prediction for mortality and rehospitalization in acute heart failure patients.

Ethics Approval and Consent to participate
Patient has provided informed consent prior to involve in the study.

Consent for publication
Not applicable.

Availability of data and materials
Data used in our study were presented in the main text.

Competing interests
Not applicable.

Funding source
Not applicable.