METHOD DEVICE AND SYSTEM FOR MONITORING SUB-CLINICAL PROGRESSION AND REGRESSION OF HEART FAILURE

Abstract

A method including sensing local accelerations or changes in sensor position, including orientation and displacement, with a local acceleration sensor mounted on a chest or abdomen of a patient, and calculating energy of polyphasic motions, based on sensed information of the local acceleration sensor and classifying severity of cardiac decompensation by calculating an excessive energy index (EEi) that compares excessive energy that appears in the polyphasic motions to energy required for inspiration at a basic respiratory rate.

Claims

1. A device comprising: a local acceleration sensor mountable on a chest or upper abdomen of a patient for sensing local accelerations or changes in a position, including orientation or displacement, of the local acceleration sensor; a data acquisition system and data storage for data analysis and comparison with past data; and a processor and a non-transitory, computer-readable storage medium in communication with the processor, wherein a computer-readable program code is embodied in the storage medium, and wherein the computer-readable program code is configured to classify phases of a respiratory cycle, including inspiration and expiration phases, and polyphasic motions within said phases, to calculate energy of said polyphasic motions based on sensed information of said local acceleration sensor and to classify severity of cardiac decompensation by calculating an excessive energy index (EEi) that compares excessive energy that appears in said polyphasic motions to energy required for inspiration at a basic respiratory rate.

2. The device according to claim 1, wherein said excessive energy index (EEi) comprises a ratio between a sum of the energies of said polyphasic motions at high order harmonics of basic respiratory frequencies and all energy at a frequency higher than a basic respiratory frequency, to the energy at the basic breathing frequency, at a measurement site of said local acceleration sensor.

3. The device according to claim 1, wherein said excessive energy index (EEi) is calculated based on an excessive respiratory peak that is pathognomonic to decompensated heart failure and the EEi comprises a ratio between an amplitude or energy of said excessive respiratory peak and an amplitude or energy at an inspiratory wave, at a measurement site of said local acceleration chest sensor.

4. The device according to claim 1, wherein said EEi comprises an Activity Duty Cycle (Adc), which comprises a ratio of duration of a vigorous breath activity to an entire duration of breath cycle, the vigorous breath activity being defined as a breath activity during a portion of the breath cycle which is more vigorous than other breath activity in other portions of the breath cycle.

5. A method comprising: sensing local accelerations or changes in sensor position, including orientation and displacement, with a local acceleration sensor mounted on a chest or abdomen of a patient, and calculating energy of polyphasic motions, based on sensed information of said local acceleration sensor and classifying severity of cardiac decompensation by calculating an excessive energy index (EEi) that compares excessive energy that appears in said polyphasic motions to energy required for inspiration at a basic respiratory rate.

6. The method according to claim 5, wherein said excessive energy index (EEi) comprises a ratio between a sum of the energies of said polyphasic motions at high order harmonics of basic respiratory frequencies and all energy at a frequency higher than the basic respiratory frequency, to the energy at the basic breathing frequency, at a measurement site of said local acceleration sensor.

7. The method according to claim 5, wherein said excessive energy index (EEi) is calculated based on an excessive respiratory peak that is pathognomonic to decompensated heart failure and the EEi comprises a ratio between an amplitude or energy of said excessive respiratory peak and an amplitude or energy at an inspiratory wave, at a measurement site of said local acceleration sensor.

8. The method according to claim 5, wherein said EEi comprises an Activity Duty Cycle (Adc), which comprises a ratio of duration of a vigorous breath activity to an entire duration of breath cycle, the vigorous breath activity being defined as a breath activity during a portion of the breath cycle which is more vigorous than other breath activity in other portions of the breath cycle.

9. The method according to claim 5, comprising using local acceleration sensors, one placed on an upper sternum, close to a supra-sternal notch, another placed on a left side of the cheat at a region of cardiac point of maximal impact (PMI) and another placed on a upper abdomen (epigastrium) near a diaphragm of the patient, and determining different polyphasic activities at these different sites.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The present invention will be understood and appreciated more fully from the following detailed description, taken in conjunction with the drawings in which:

[0055] FIG. 1 is a simplified illustration of a device for monitoring and patients with heart failure, constructed and operative in accordance with an embodiment of the present invention. The system detects excessive work of accessory respiratory muscle at the neck, which are very important for fixing the chest cage (detected by the sensor at the supra sternal notch), or the abdominal accessory muscle (detected by the sensor at the epigastrium). It can also monitor changes in the impact of the heart and the heart sounds and vibration (utilizing the sensor at the PMIpoint of maximal impact). Consequently it may detect also the heart rate, as an adjuvant index.

[0056] FIG. 2 is an example of successful treatment of patient with decompensated heart failure that was electively admitted for hospitalization with a leading symptom of dyspnea. This patient had been treated with diuretics and his clinical condition was balanced, and he was readmitted only after 90 days.

[0057] The hemodynamic indices (BP=blood pressure, HR-Heart rate), the pulse oximetry (SpO.sub.2) (on the right hand side) together with the calculated respiratory motion (derived by the algorithm of the system) at admission (upper part) and discharge (the bottom). Note that there were no changes in these parameters, and they were normal already at admission. On admission the respiratory rate was slightly elevated (17 bpm), but what was more prominent is the continuous and restless breathing, without a moment of rest (quantified by the Adc index). Respiratory rate decreased from admission to discharge (from 17 to 12 bpm), but more significantly, the decrease in the rate is due to the appearance of almost normal and prolonged period of rest during the expiratory phase (normal and low Adc). The most significant phenomenon in this example was a polyphasic and active expiration that was observed at admission (arrows) which diminished towards discharge (was quantified by the EEi indices). The arrows show the additional polyphasic expiratory movement at admission. Interestingly, there are also significant changes in the relative duration of the inspiratory (green) and expiratory (gray) phases. At admission the patient has significant and abnormal prolongation of the inspiratory phase, due to the decrease in the lung compliance. At discharge the ratio between these two phases is practically normal.

[0058] FIG. 3. The figure shows signals of a patient to whom the treatment did not help; he was re-hospitalized due to HF decompensation only 11 days post discharge. This patient has been treated with diuretics and his clinical condition, according to the novel system recording, was not balanced. The system predicted that the treatment was unsuccessful, and that the patient was discharged with severe signs of excessive respiratory work.

[0059] There were no significant changes in the following parameters: hemodynamic indices (BP=blood pressure, HR-Heart rate), the pulse oximetry (SpO.sub.2) (on the right hand side) together with the calculated respiratory motion (derived by the algorithm of the system) at admission (upper part) and discharge (the bottom). On admission respiratory rate was elevated (20 bpm). The respiratory rate increases from admission to discharge (from 20 to 30 bpm), but more significantly, there was an increase in the intensity of the active motion with various polyphasic waves during the inspiration and expiration (An increase in the Adc index). The arrows depict the appearance of prominent active expiratory movements that were observed at admission (arrows) and discharge (was quantified by the EEi indices).

[0060] FIG. 4. Heart failure patients have a pathognomonic sign of excessive expiratory work. Real respiratory patterns on admission (upper part) in a state of HF decompensation on admission and discharge (lower part). On admission there is an additional wave during the expiration (the humped curve pointed by the arrow) which indicates an additional increase of extra respiratory effort. It is used as a novel index to monitor the severity of congestion of heart failure patients.

[0061] FIG. 5. The power spectral analyses on admission (the graph that continues to the right) and discharge (the graph with the peak at 12/min) from a patient that was successfully treated for decompensated heart failure. The power spectral analysis shows the energy distribution of patient respiration. Most of the energy is concentrated around the basic breathing rate (denoted by P.sub.B). However at admission there were additional significant peaks at higher frequencies (denoted by P.sub.E). They represent the additional respiratory work during the expiratory phase (FIG. 3) and the other polyphasic motions due to the various accessory respiratory muscle activities, also during the inspiratory phase. At discharge the excessive work at the higher frequencies disappeared and there was only one significant peak of energy, at the patient natural breathing rate. This patient was not re-hospitalized during the 30 days post hospital discharge.

[0062] FIG. 6. The power spectral analyses on admission (the graph that starts with a lower y value) and discharge (the other graph) from a patient that did not respond to the diuretic treatment according to our system that monitor the severity of dyspnea, and was re-hospitalized within the 30 days post hospital discharge period. The figure depicts the power spectral analyses on admission and discharge. The power spectral analysis shows the energy distribution of patient respiration. Most of the energy is concentrated around the basic breathing rate (denoted by P.sub.B). However at admission there were additional significant peaks at higher frequency (denoted by P.sub.E). This additional peak represents the additional respiratory work during the expiratory phase (FIG. 3) and the other polyphasic motions due to the various accessory respiratory muscle activities, also during the inspiratory phase. Interestingly, at discharge the excessive work at this higher frequency increased from about 30% of the amplitude of the peak at the basic respiratory rate (P.sub.B) to 45% of this energy, suggesting that was a significant aggravation in the excessive respiratory work.

[0063] FIG. 7. The figure depicts the utility of the Excessive Energy index (EEI) and the respiratory rate in differentiating between patients that be re-hospitalized within less than 30 days those that will gain a prolonged free of symptom period. The group of patients that were free from a HF re-hospitalization during the first thirty days post discharge (triangles) had EEI value below 0.4 at discharge. While all the patients that were readmitted within the first month post hospital discharge (squares) had EEI value above 0.4. Interestingly, the respiratory rate was inconsistent and varied substantially in a wide range of frequencies, and could not differentiate between the two groups.

DETAILED DESCRIPTION OF EMBODIMENTS

[0064] Reference is now made to FIG. 1, which illustrates a device for monitoring the respiratory dynamics for detecting and quantifying the development of signs for excessive respiratory work that is pathognomonic to HF. The system operates in accordance with a non-limiting embodiment of the present invention.

[0065] The device can continuously acquire and analyze the respiratory dynamics for a single sensor on the chest or the abdomen. In one embodiment, the sensor is an accelerometer. Other sensors, such as gyro-meters or gyroscope, which sense and quantify the motion of a single point in the space, can be used (all these types of sensors are referred to as a local acceleration sensor). Other accelerometers or alike may be used, such as on the abdomen. The sensors may be embodied in a patch, referred to as a patch and sensor unit.

[0066] The system includes [0067] the hardware for data amplification and filtration. [0068] the required data acquisition system and data storage for continuous data analysis and comparison with the past. [0069] The system includes a processor (central processing unit), which may include additional channels for ECG acquisition with appropriate gain and filtering.

[0070] Software for analyzing the pathognomonic signs of progressing HF (in other words, the processor uses a non-transitory, computer-readable storage medium, and computer-readable program code is embodied in the storage medium, wherein the computer-readable program code is configured for analyzing the pathognomonic signs of progressing HF). [0071] User-friendly interface and display of the new indices for the excessive respiratory energy.

[0072] The system may also include contact electrode(s) for ECG, which may be in each patch and sensor unit.

[0073] All the sensors are connected to processor. The processor analyzes the sensed local accelerations and\displacement and\or motion of each sensor and quantifies the severity of the above mentioned indices for the excessive respiratory work.

[0074] The following is noted:

(i) Measurement of the acceleration provides the following advantages: [0075] 1. Enables assessment of the displacement and acceleration of a single anatomic point of interest. [0076] 2. Provides the position of the sensor in space, utilizing the gravity acceleration. [0077] 3. Flat frequency response form DC to the high sound frequency. Therefore it can detect the subsonic respiratory dynamics. [0078] 4. Miniature sensor, with low energy consumption. [0079] 5. The measured acceleration is proportional to the forces that operate on the mass to which the sensor is attached, assuming that the mass is constant.
(ii) Frequent hospitalizations due to decompensated HF are due to the development of lung congestion. Although lung congestion is associated with changes in the lung compliance and lung viscosity, there is no prior art device to monitor the changes in lung mechanical characteristic with the development of progressing HF.
(iii) Patients with mild to moderate decompensated HF have normal saturation and adequate lung ventilation. To maintain the adequate lung ventilation the increase the respiratory work and utilize also the accessory respiratory muscles.
(iv) Increase in the respiratory work can develop in various normal (anxiety) and pathological (as airway obstruction). However, decompensated heart failure is associated with pathognomonic feature, since also the underlying changes in lung characteristics are quite unique. Decompensated heart failure is associated with a decrease in the lung compliance and an increase in lung viscosity. Upper airways obstruction is associated with a simple increase in the resistance to flow. COPD is associated with an increase in lung compliance (opposite to HF) and an increase in the resistance at the lower airways level and mainly during the expiratory phase. These differences in the affected components (resistance vs. compliance) and lung compartments (upper airways vs. lower airways) are associated with different transfer functions and observed phenomena. The above described features are pathognomonic for HF and thus different indices should be analyzes and monitored in decompensated HF.

Study Protocol

[0080] The study enrolled patients who arrived at the hospital with decompensated heart failure. On admission and discharge, the patients had measurements of blood pressure, pulse rate, oxygen saturation, and weight, and underwent a physical examination. Blood level of BNP was measured on admission. At hospital discharge the patients filled a Likert scale questionnaire, as a subjective indicator to the improvement in respiratory effort. Three motion sensors, which record the motion signals, were placed on the chest and over the epigastric area of the patient. The patients were monitored close to hospital admission and at the day of discharge, each time for a period of about thirty minutes. During the monitoring period patient's hemodynamic indices were taken.

Sensors Placement

[0081] The sensors were placed in 3 locations:

[0082] One sensor was placed on the upper sternum below the supra-sternal notch in order to detect excessive work of the muscles of the neck. This sensor is also very sensitive to breathing.

[0083] A second sensor was placed near the cardiac apex at the point of maximal impact (PMI) in order to detect changes in the cardiac impact and heart sounds. The sensor is sensitive to signals originated from the heart.

[0084] The third sensor placed at the epigastrium is sensitive to the motion of the diaphragm and the accessory abdominal respiratory muscle. It detects excessive work of the abdominal muscles.

Experimental Setup

[0085] The system we used comprises of three, miniature, two axial accelerometers, connected to a data acquisition unit.

Data Analysis

[0086] Data were acquired at a sampling rate of 5 KHz and analyzed offline with a specialized analysis program implemented in MATLAB (Mathworks) under the scope of the study. Results of all experiments were summarized for statistical analysis.

Signals Pre-Processing

[0087] The sampled data carried noise originated by patients' movement and talking. Since we were interested in the respiratory dynamics, the data was down-sampled to 500 Hz.

Signal Phase Estimation

[0088] Both the longitudinal and radial axis carry important information about the breathing signal, therefor it is important to combine the contributions from both axis.

Clinical Data

[0089] We compare the respiratory signals of a patient at admission and discharge. We found several prominent differences between the signal from admission and discharge:

[0090] In most of the patient the respiratory rate decreases on discharge compared to admission, as shown in FIGS. 2 and 5. In few patients there was a slight increase in the respiratory rate.

[0091] On admission the breathing is continuous and restless with no quiescent phases during respiration, as shown in FIG. 2. On the other hand, at discharge, there are phases of rest during the late expiratory phase in the majority of the cases. When the patients are discharged with prominent and continuous active breathing they are readmitted for hospitalization within less the 30 days (FIG. 3)

[0092] The most significant phenomenon is a polyphasic and active expiration seen at admission and diminishes towards discharge. We document for the first time that decompensated heart failure patients increase respiratory work during the expiration phase by recruiting abdominal respiratory muscles as shown in FIGS. 2 and 4.

Excessive Effort Index

[0093] In normal condition expiration should be passive. However, we found in decompensated state, the respiratory signal is characterized by active and polyphasic expiratory phase. This phenomenon is depicted as additional movement or harmonic to the basic breath cycle. Therefore, it will be manifested as additional energy concentration at higher frequency harmonics in the signal spectrum.

[0094] We use Welch's method to estimate the power spectral density. Welch's method is based on averaging the periodogram. The periodogram is a nonparametric estimate of the power spectral density (PSD) of a wide- sense stationary random process.

[0095] Welch's technique to reduce the variance of the periodogram breaks the time series into overlapped segments. Welch's method computes a periodogram for each segment and then averages these estimates to produce the estimate of the power spectral density. Because the process is wide-sense stationary and Welch's method uses PSD estimates of different segments of the time series, the periodograms represent approximately uncorrelated estimates of the true PSD and averaging reduces the variability.

Energy Calculation

[0096] First, the signal is segmented into overlapping windows. The PSD resolution depends on the window length by:

[00001]$R.Math.E.Math.S_{P.Math.S.Math.D}=\frac{F.Math.s^M}{2.Math.N_{\mathrm{window}}}\left[\mathrm{Hz}\right]$

Where F.sub.Sis the sampling frequency, M is the number of windows Welch's method breaks the signal to, as was explained in the previous section. Usually M=8. N is the length of the window in to what we segment the signal. In order to maintain a PSD resolution of 1/60 Hz (1 bpm) with F.sub.Sof 500 [Hz], we need to choose a window length of at least 270 [sec]. That means that the periodogram will be performed to 30 [sec] signals, with 50% overlap and we can assume that the signals are stationary. Then, for each window, the PSD is evaluated. The highest energy concentration is found around the natural breathing rate, we consider this to be the breathing frequency-F.sub.b. The breathing energy is defined as:

[00002]$E_{\mathrm{Br}}=\frac{\sqrt{2}}{f_b}.Math.\underset{f=\frac{1}{\sqrt{2}}.Math.f_b}{\overset{\sqrt{2}.Math.f_b}{.Math.}}.Math..Math.P_{\mathrm{xx}}\left(f\right)$

[0097] In the analysis presented here the excessive energy at high frequency is calculated as:

[00003]$E_{\mathrm{Ex}}=\frac{1}{1-\sqrt{2}.Math.f_b}.Math.\underset{f=\sqrt{2}.Math.f_b}{\overset{1}{.Math.}}.Math..Math.P_{\mathrm{xx}}\left(f\right)$

[0098] Where f.sub.b is the breathing frequency and P.sub.xx is the power spectral density values. The Excessive Energy index (EEI) is defined as the ratio between the excessive energy and the energy at the basic breathing frequency (respiratory rate):

EEI=E.sub.Ex/E.sub.Br

[0099] In healthy subjects the EEI should be zero and it increase with the increase in the appearance of excessive work.

Respiratory Rate

[0100] The respiratory rate estimation process consists of several steps. First, to remove all the unwanted information, which is above patient breathing frequency, the signal was filtered using a low pass filter with cutoff frequency of 1 [Hz]. To maintain stationarity, we segment the signal into thirty seconds windows with an overlap of fifty percent between them. For every window we calculate the Hilbert transform and find its peaks.

[0101] One of the properties of Hilbert transform is that it is an odd function. That means that it will cross zero on the x-axis every time that there is an inflection point in the original waveform. Similarly a crossing of the zero between consecutive positive and negative inflection points in the original waveform will be presented as a peak in its Hilbert transformed conjugate.

Results

[0102] Changes in EEI from Admission to Discharge

[0103] Readmission after a hospitalization for heart failure is an important target for quality improvement. There is a high thirty-day readmission rate (30%). Recognition of the importance of readmissions as a measure of quality is still very recent. Early readmission within the first thirty days post discharge is considered as treatment failure. Several strategies to lower hospital readmission rate for patients with heart failure had been developed (24).

[0104] The study group included 14 patients (Table 1). All had severe stage 3 or 4 heart failure. We divide the patients into 2 groups: responders and non-responders. The responders (n=10) had free of hospitalization period longer than 30 days, and the non-responders (n=4) were re-hospitalized within less than 30 days. There was no significant difference between the two groups in their age (71.49.8 vs. 73.88.3 years), cardiac ejection-fraction (24.49.8% vs. 25.27.1%) of BNP levels (18431461 vs. 1489121). Interestingly. There was also no significant difference between the two groups in the hemodynamic indices: heart rate (7410 vs. 7715 bpm), or systolic pressure (12114 vs. 12654 mmHg) of diastolic pressure (Table 1). Not surprisingly, the pulse oxymetry were normal and identical in the two groups, since they were normal also at admission. Thus the hemodynamic and the saturation can't differentiate between the two groups. The responder have a slightly larger decrease in the weight than the non-responder (3.751.65 vs. 2.51.3 Kg), however this difference was not statistically significant. The responder have also a slightly larger decrease in the respiratory rate than the non-responder (3.83.6 vs. 1.14.3 bpm), however this difference was also not statistically significant. Only the EEI could differentiate between the two groups (p<0.01). In the responder group there was a significant decrease of 45% in the EEI, while the in the non-responder there was even an increase of 23% in the EEI.

[0105] Moreover, when we plot the EEI index against the respiratory rate (FIG. 7) we can see that there is a clear separation between the responders (triangles) and the non-responders (rectangular). In the responder group the EEI at discharge was below 0.4.

[0106] While in the non-responders it was above 0.4. In both groups, the respiratory rate, was inconsistent and varied substantially in wide range of frequencies.

[0107] FIG. 7 also verifies that the respiratory rate can't assist in the differentiation between the two groups, since there was a clear overlap between the two groups in the distributions of the respiratory rate at discharge (FIG. 7).

Discussion

[0108] For the first time, we documented that HF decompensation is associated with increased respiratory work during the expiration phase. HF decompensation leads to decrease in lung compliance and increase in the viscous loads, that leads to increase of the respiratory work and was associated with appearance of dyspnea.

[0109] Chronic patients adapt to their physiological condition and may preserve the required cardiac output by elevating the cardiac filling pressures. However, the increase in the filling pressure and the associated increase in the lung blood pool, further decrease lung compliance, leading to an increase in the respiratory work and to a sensation of dyspnea.

[0110] We have established a novel index (EEI) that can be used for assessing the severity of the excessive respiratory work, and it also correlates with the severity of the dyspnea. We have turned the subjective symptom of dyspnea into a measureable objective and absolute marker, the EEI that is proportional to the severity of congestion.

[0111] In contradiction to what might be expected, the BNP levels in the responder group were higher than the BNP levels in the non-responder group. However, it is important to note that these BNP levels were collected at admission, and the difference was not statistically significant.

[0112] In both groups there was a significant reduction in weight and respiratory rate during the hospitalization. However, there was no difference between the two groups at discharge. These observations strengthen the fact that the decreases in the weight or respiratory rate have low predictive values. EEI was the only statistically significant parameter that enabled us to differentiate between the two groups of patients. Therefore we suggest that the EEI should be used for decision making before discharging the patients from the hospital.

TABLE-US-00001 TABLE 1 Summary of the accumulated results of the clinical study. Long free Early Readmission event period (<30 day) (n = 10) (n = 4) Age (years) 71.4 9.8 73.8 8.3 EF (%) 24.4 9.8 25.2 7.1 BNP (pg/ml) 1843 1461 1489.6 1213 Heart Rate (bpm) 74 10 77 15 Blood Pressure 121 14 126 65 Systolic (mm Hg) Blood Pressure 73 11 68 13 Diastolic (mm Hg) SpO.sub.2 (%) 97 2 97 1 Weight (kg) 3.75 1.65 2.5 1.3 Respiratory Rate 3.8 3.6 1.13 4.3 (bpm) EEI Index 0.45 0.32 +0.23 0.25 P < 0.01

[0113] Bibliography:

[0114] 1. Stewart S. Financial aspects of heart failure programs of care. Eur J Heart Fail. 2005; 7:423-428.

[0115] 2. Cleland J G, et al. The effect of cardiac resynchronization on morbidity and mortality in heart failure. N Engl J Med. 2005; 352:1539-1549.

[0116] 3. Pfeffer M A, et al. CHARM Investigators and Committees. Effects of candesartan on mortality and morbidity in patients with chronic heart failure: the CHARM-Overall programme. Lancet. 2003; 362:759-766.

[0117] 4. Vanderheyden M, et al. Continuous Monitoring of Intrathoracic Impedance and Right Ventricular Pressures in Patients With Heart Failure. Circ Heart Fail. 2010; 3:370-377.

[0118] 5. Roger V L, et al. Heart disease and stroke statistics2011 update: a report from the American Heart Association. Circulation. 2011; 123:e18-e209.

[0119] 6. Bui A L, et al. Home Monitoring for Heart Failure Management. Am Coll Cardiol. 2012; 59(2): 97-104.

[0120] 7. Jencks S F, et al. Rehospitalizations among patients in the Medicare fee-forservice program. N Engl J Med. 2009; 360:1418-1428.

[0121] 8. Lee W C, et al. Economic burden of heart failure: a summary of recent literature.

[0122] Heart Lung. 2004; 33:362-371.

[0123] 9. Louis A A, et al. A systematic review of telemonitoring for the management of heart failure. Eur J Heart Fail 2003; 5:583-590.

[0124] 10. Fonarow G C, et al. Impact of a comprehensive heart failure management program on hospital readmission and functional status of patients with advanced heart failure. J Am Coll Cardiol 1997; 30:725-32.

[0125] 11. Lewin J, et al. Clinical deterioration in established heart failure: what is the value of BNP and weight gain in aiding diagnosis? Eur J Heart Fail 2005; 7:953-7.

[0126] 12. Maisel A S, et al. Cardiac natriuretic peptides: a proteomic window to cardiac function and clinical management. Rev Cardiovasc Med. 2003; 4:S3-S12.

[0127] 13. Maisel A S, et al. Rapid measurement of B-type natriuretic peptide in the emergency diagnosis of HF. N Engl J Med 2002; 347:161-7.

[0128] 14. [9] Bayes-Genis A, et al. N-terminal probrain natriuretic peptide (NT-proBNP) in the emergency diagnosis and in-hospital monitoring of patients with dyspnoea and ventricular dysfunction. Eur J Heart Fail 2004; 6:313-8.

[0129] 15. Wu Alan H. B, et al. Biological variation of the natriuretic peptides and their role in monitoring patients with heart failure. Eur J Heart Fail 2004; 6:355-8.

[0130] 16. Chaudhry S I, et al. Telemonitoring in patients with heart failure. N Engl J Med. 2010; 363:2301-2309.

[0131] 17. Koehler F, et al, on behalf of the TIM-HF Investigators. Impact of remote telemedical management on mortality and hospitalizations in ambulatory patients with chronic heart failure: the telemedical interventional monitoring in heart failure study. Circulation. 2011; 123:1873-1880.

[0132] 18. Heist E K. Et al. for the DEFEAT-PE Study Investigators. Analysis of Different Device-Based Intrathoracic Impedance Vectors for Detection of Heart Failure Events (from the Detect Fluid Early from Intrathoracic Impedance Monitoring Study). Am J Cardiol 2014; 114:1249-1256.

[0133] 19. Yu C M, et al. Intrathoracic impedance monitoring in patients with heart failure: correlation with fluid status and feasibility of early warning preceding hospitalization Circulation 2005; 112:841-848.

[0134] 20. Vollmann D, et al. Clinical utility of intrathoracic impedance monitoring to alert patients with an implanted device of deteriorating chronic heart failure. Eur Heart J 2007; 28:1835-1840.

[0135] 21. Abraham W T, et al. On behalf of the FAST investigators. Superior performance of intrathoracic impedance-derived fluid index versus daily weight monitoring in heart failure patients: results of the Fluid Accumulation Status Trial (FAST). J Card Fail 2009; 15:813.

[0136] 22. Van Veldhuisen D J, et al. Intrathoracic Impedance Monitoring, Audible Patient Alerts, and Outcome in Patients With Heart Failure. Circulation. 2011; 124:1719-1726.

[0137] 23. Udelson J E, et al. T.M.I. (Too Much Information)? Circulation. 2011; 124:1697-1699.

[0138] 24. Bradley E H, et al. Hospital Strategies Associated With 30-Day Readmission Rates for Patients With Heart Failure. Circ Cardiovasc Qual Outcomes. 2013; 6:444-450.