With the changing epidemiology of heart failure, characterized in particular by growing prevalence of the syndrome, it has become increasingly important that management strategies are directed at maintaining well-being in the community. In so doing, the need for hospitalization will be reduced, the major driver of the economic cost of this condition.
The risk of hospitalization for patients with heart failure is well described. All patients are at risk, but the most vulnerable period are the weeks following decompensation.1 Data from our own unit has underlined that patients in the first 3 months following discharge from hospital have approximately a 30% chance of demonstrating features of clinical deterioration requiring medical attention, and beyond this period there is a continuing 10% to 15% occurrence per quarter (Figure). The majority of such deterioration occurs over days rather than over a period of hours, thereby providing opportunity to intervene and prevent further decompensation to a level requiring admission.
Such early intervention requires effective warning systems of emerging fluid overload, which is the central problem in the majority of cases. Ideally, such systems should alert the patient and health care provider before symptoms and signs of decompensation appear. This would allow early intervention, predominantly with diuretic therapy. Once treatment has been initiated another critical aspect of care is how much extra diuretic use is required and for what duration. Excess diuretic use can lead to renal dysfunction and electrolyte imbalance, which can prolong length of stay or in the outpatient setting even precipitate admission. An accurate objective indicator of fluid status is essential to help minimize such problems and allow for more precise administration of the diuretic.
Outside of the setting of diagnosing and managing clinical deterioration, accurate assessment of fluid status is also required to determine optimal time for prescription of β-blockade in recovery from acute decompensated heart failure (ADHF) and less critically for angiotensin-converting enzyme (ACE) inhibition. Timing of β-blockade introduction requires precise knowledge of fluid status; if the patient is too dry or too wet, introduction of β-blockade could produce significant hypotension or clinical deterioration, which aside from the immediate concerns could inappropriately result in the patient being labeled β-blocker–intolerant in the long term. Therefore, it is critical to be as certain as possible regarding fluid status prior to initiation of this important therapy to ensure optimal response. Significant hypotension can also occur if ACE inhibition is initiated in patients who are hypovolemic following excess use of diuretic.
Therefore, it is clear that optimal assessment of fluid status is critically important in the management of ADHF. A critical foundation to the early and accurate diagnosis of emerging fluid retention is patient awareness of the importance of immediate contact with health care providers when features of decompensation are suspected. However, when patients present, how effective are our present strategies in defining fluid overload? This article will review the present state of knowledge of the various standard strategies available to assess fluid status in heart failure and review what is known about potential role of impedance measurement in this area.
Established Parameters for Assessment of Fluid Status in Heart Failure
While compromised by subjectivity, both with the patient and the physician, clinical history and physical examination remain the cornerstone of assessment of fluid status in the heart failure patient. The development of worsening dyspnea is a typical but nonspecific presentation in emerging clinical deterioration. When a patient complains of the more specific symptoms of orthopnea and paroxysmal nocturnal dyspnea, fluid retention is likely well-progressed.
Physical examination in emerging decompensation suffers from a lack of sensitivity to fluid overload, outside of the presentation of acute pulmonary edema where physical examination more often demonstrates typical signs of fluid overload such as lung crackles and a gallop rhythm.2 Body weight measurement, routinely advocated as an important self-care strategy for early detection of emerging deterioration, lacks sensitivity, but the recommended increase of 2 kg over 2 to 3 days has been shown to be specific for clinical deterioration.3,4 However, the consensus guideline stated above does underline the lack of precision of this parameter in this context, given that the change in weight deemed relevant is not adjusted for the patient’s dry weight and assumes, despite evidence to the contrary, that the emerging fluid retention occurs over a short period. It would seem reasonable to propose that a smaller weight change may be relevant especially in individuals with lower dry weight. Furthermore, aside from the lack of individuality, it is also possible that smaller increments of weight increase over a longer period may also be meaningful. In this regard, work from Chaudhry and colleagues5 observed weight gain several weeks before hospitalization. A further problem with daily weight is that dry weight may change with time and needs to be readjusted. Without doing so, weight changes may become even less accurate in helping to define deterioration at an early stage.
In patients with established heart failure and reduced ejection fraction, Stevenson and Perloff2 demonstrated that physical signs of fluid overload, even when sought by experienced cardiologists, were often absent in patients with markedly elevated pulmonary capillary wedge pressures. This may be explained by augmented lymphatic drainage of the alveolar-capillary interstitial space in the lung. Similarly, elevated jugular venous pressure, while present in 75% of patients with increased right atrial pressure, was not an acceptable surrogate of increased left-sided filling pressures.2
Efforts to guide diuretic dosing based on clinical assessment is also compromised by lack of sensitivity. Symptomatic improvement does indicate removal of excess fluid and movement toward euvolemia, but unfortunately it does not accurately define volume status. While excess diuresis can be suspected from a low jugular venous pressure, this feature does not confirm or aid in the estimation of pulmonary venous pressure.
In summary, while clinical assessment is of importance in the setting of patients with possible clinical deterioration, it must be stressed that the absence of physical signs of fluid overload does not adequately rule out the diagnosis. In particular, it needs to be emphasized, especially to patients, that in evolving clinical deterioration weight gain may occur in smaller increments over longer periods and that in some circumstances may not register at 2 kg over 48 hours, which forms the basis of advice given to the patient. This underlines the need for complimentary investigations to help secure or rule out the diagnosis.
While chest radiography remains a frequently performed test in the assessment of emerging clinical deterioration, it has been well-demonstrated that normal findings on chest radiograph do not exclude fluid overload.6 The reasons for this are likely multiple, including technical issues relating to radiographic penetrance, body habitus, and also the likelihood that significant fluid overload is required for the standard chest radiographic signs of fluid overload to appear.
Therefore, its value in the investigation of possible decompensation in heart failure is to rule out alternative diagnoses such as respiratory infection, but in regard to fluid status it lacks sensitivity and is not an effective guide to use of diuretic therapy.
Natriuretic peptide (NP) assessment is becoming increasingly used in the management of heart failure.7 The main role of this biomarker at present is in the initial diagnosis of heart failure, in defining prognosis in patients with established heart failure, and potentially as a guide to therapy. Whether NP may be helpful in the assessment of clinical deterioration in patients with an established heart failure is less clear. In heart failure patients presenting to the emergency department with significant dyspnea, elevated NP may be helpful in confirming the diagnosis. B-type natriuretic peptide levels in excess of 500 pg/mL are very suggestive of a heart failure diagnosis, while values <100 pg/mL provide a confident rule out value.8 However, certain caveats need to be taken into consideration when making this assessment. In patients with established heart failure, it is important not to interpret an NP value in isolation, but to compare it to the patients most recent euvolemic value if available. There are 2 important reasons for this comparison: first, a patient’s euvolemic value may remain significantly elevated, and therefore assessment of an isolated value on presentation to the emergency department may lead to incorrect interpretation. Second, the reference change value for NP is quite large, of the order of 50% for B-type natriuretic peptide (BNP) and modestly less for N-terminal proBNP, indicating that such a change from a stable value is required to be certain that it is clinically significant.9 Therefore, substantial change from a clinically stable value, especially when the current value at the time of potential deterioration is >500 pg/mL, indicates evolving fluid retention.
However, the role of NP in the community setting to detect more subtle presentations earlier in the natural history of emerging clinical deterioration is less certain. Work from Lewin and colleagues4 in our unit demonstrated that significant changes in BNP were specific for clinical deterioration, but similar to weight change, early clinical deterioration diagnosed by an experienced clinician occurred without significant change in NP. These data also demonstrated little correlation between weight increase and NP change.
These observations suggest that NP is a useful aid in diagnosing emerging fluid retention but that a lack of sensitivity, especially in the early stages of clinical deterioration, underlines the need for supportive diagnostic strategies.
Several stu-dies have assessed the ability of Doppler echocardiography to noninvasively estimate left ventricular filling pressures. Multiple parameters have been used, including mitral inflow pattern and pulmonary vein velocities, but these parameters are load-dependent, making their use less effective in heart failure. More recently, several authors have demonstrated that the ratio between the mitral E velocity and the tissue E velocity measured at the septal or lateral wall at the base of the left ventricle is more accurate in determining left-sided filling pressures.10 However, this technique remains operator-dependent and demands higher-level sonography skills. In addition, while based on Doppler signal, it still demands reasonable echocardiographic windows. Furthermore, not all data concur that the mitral E/E prime ratio accurately reflects left ventricular filling pressure. A recent paper by Mullens and colleagues11 demonstrated a relatively poor sensitivity for the averaged septal and lateral wall E/E prime in detecting a pulmonary capillary wedge pressure >18 mm Hg, interestingly worse in those with cardiac resynchronization therapy devices in situ. They also failed to observe a close relationship between change in pulmonary capillary wedge pressure and change in the E/E prime ratio. These data would at least raise a little concern in attempting to use this parameter as an indicator of fluid status or to guide diuretic or vasodilator therapy.
Invasive measurements remain the gold standard for assessing left-sided filling pressures. However, either the temporary placement of a Swan-Ganz catheter or the use of a chronic in dwelling pulmonary catheter is complicated by all the hazards of invasive procedure as well as costs.
The Role of Bioimpedance Measurement
It is clear from the above that while there are many approaches to the assessment and diagnosis of emerging fluid retention in patients with heart failure, all suffer from a lack of sensitivity and are compromised in their usefulness in guiding therapy. Therefore, further strategies are required. Given that the major concern is emerging fluid retention, attention has turned to the potential role of bioelectrical impedance.12,13 Impedance is a measure of resistance to flow of an electrical current. It can be measured by multiple methods and has been shown to accurately reflect hydration status in patients with renal, cardiac, and liver disease.12,13 Impedance is derived from 2 measurements, electrical resistance and reactance. As patients become fluid-overloaded, both of these parameters decrease, reflecting improved conductivity of electrical current.
Impedance values have been obtained from indwelling device therapies used in heart failure management such as defibrillators and cardiac resynchronization devices. Several reports have now demonstrated significant decreases in impedance values measured between the right ventricular lead and the device can, often observed up to weeks before the typical features of clinical deterioration present.14,15 This early warning would clearly be of benefit in our attempts to intervene at the earliest possible time to abort a threatened decompensation. False-positive readings can occur as a result of other thoracic pathologies such as pneumonia and hemothorax, which could reduce impedance values. However, while useful when available, it is likely not practical or cost-effective to place invasive devices for the sole purpose of measuring fluid status in heart failure patients.
Fortunately, seve-ral noninvasive techniques are also available to measure total body or regional impedance. A band electrode method has provided data on thoracic impedance in patients with heart failure and has demonstrated reduced values with impending congestion.16
Kataoka17 reported on the role of a novel weighing scales incorporating measurement of bioelectrical impedance and analysis of total body fat. The author demonstrated that combining weight change with alteration in impedance or body fat percentage proved more accurate in determining clinical deterioration than body weight change alone.
Tetrapolar impedance plethysmography, with electrodes placed on the periphery of ipsilateral upper and lower limbs for whole body impedance and proximally on the limbs for segmental thoraco-abdominal impedance, has also been assessed in patients with heart failure. Parrinello and colleagues18 used this strategy to assess fluid status in patients presenting to the emergency department with acute dyspnea of unclear cause. They observed that impedance and reactance measurements, either whole-body or segmental, accurately differentiated dyspnea from ADHF from non–heart failure–related dyspnea and normal controls. In addition, improvement in heart failure status during admission was mirrored by an increase in resistance and reactance and an overall increase in bioelectrical impedance. Furthermore, the authors demonstrated a strong correlation between impedance and natriuretic peptide levels. Finally, receiver operator curve analysis underlined the diagnostic accuracy of this technique with area under the curve values of 0.93 to 0.96 for total body and regional values. Patterna and colleagues assessed the value of impedance measurement in patients with refractory heart failure treated with furosemide with or without hypertonic saline.19 Similar to Parrinello and colleagues, they demonstrated that impedance is reduced in ADHF, increases with effective therapy in heart failure, and does so in a manner similar to BNP.
Since impedance specifically measures fluid status, it may more accurately define volume status than other techniques outlined above. Potential pitfalls with noninvasive impedance measurement include the influence of pleural and abdominal cavity fluid on measurements. Furthermore, skin conditions can significantly alter readings.
The above data indicate that monitoring bioelectrical impedance provides important incremental information on the fluid status of patients with heart failure. The information from invasive and noninvasive strategies indicates that this parameter may provide an earlier signal of evolving problems than presently available surveillance strategies such as weight monitoring. By identifying reduction in impedance weeks before clinical presentation, this approach underlines that significant changes are evolving remote from the time of clinical presentation, potentially allowing for earlier and more effective intervention. In addition, impedance values appear to track clinical response to intervention and therefore may allow for more precision in adjusting diuretic therapy. Development of a simple noninvasive system would further expand the role for this approach and potentially allow for patient home monitoring of this parameter, thus facilitating even earlier notification of impending problems, especially in high-risk patients.
Disclosure: Dr McDonald received an honorarium funded by an unrestricted educational grant from Abbott Laboratories and Otsuka America Pharmaceuticals for time and expertise spent in preparation of this article. He is a member of the Advisory Board of EFG Diagnostics Ltd, Belfast, Ireland.