Cardiac US. Transthoracic echocardiography enables a detailed assessment of cardiac structure and function, and a non-invasive estimation of both left ventricular filling pressures (LVFP) and pulmonary pressures [12]. The ratio of early transmitral flow velocity to early diastolic myocardial velocity (E/e’) is generally accepted as a proxy for LVFP and used to evaluate diastolic dysfunction [17, 18]. Assessment of left atrial size and function are also important markers that estimate LVFP and stratify cardiovascular risk [19, 20]. However, quantification of LVFP using echocardiography lacks universal agreement [10]. More recently, an equation has been proposed to estimate pulmonary capillary wedge pressure from standard echocardiographic variables [12]. Said equation was validated against the gold standard of RHC, which shows the potential utility of echocardiography in this setting, especially when an invasive evaluation is not necessary (e.g., in low-risk patients with stable HF), is unfeasible or carries excessive risk. On the other hand, measuring the peak tricuspid regurgitation velocity allows the estimation of the systolic pulmonary arterial pressures. Elevated pressures in the pulmonary circulation are associated with fluid extravasation in the lung interstitium and parenchyma [21].

Lung ultrasound (LUS). LUS is a valuable tool for detecting and quantifying extravascular lung fluid by evaluating B-lines. B-lines are vertical hyperechoic reverberation artefacts arising from the pleural line and extending to the bottom of the ultrasound screen and indicate the presence of extravascular lung water [14] (Fig. 2). In patients with HF, B-lines exhibit rapid changes in response to therapies that modify intravascular volume or pulmonary pressures. For instance, their number reduces during renal dialysis or after treatment with loop diuretics, while they might increase with exercise [21,22,23,24,25,26,27]. In patients presenting to the emergency room with acute dyspnoea, LUS has shown superior sensitivity and specificity in diagnosing interstitial pulmonary oedema compared to clinical examination or chest X-ray [28, 29]. However, B-lines are not exclusive of HF, as they can also appear in non-cardiogenic pulmonary oedema (such as acute respiratory distress syndrome) [30] and interstitial lung disease [30, 31], and might be low in patients who are obese [32].

Lung ultrasound (LUS). Left. LUS evaluation of a single zone where 2 B-lines can be seen. Right. LUS evaluation of a single zone where at least 5 B-lines can be seen (i.e., at least 50% of the sector is covered by B-lines), suggesting pulmonary congestion

From a technical point of view, LUS can be performed rapidly with the patient in the supine position, placing the phased array or linear transducer either perpendicularly or transversely to the ribs on the anterior thorax, preferably using an eight-zone protocol (four symmetrical zones per each hemithorax) [10]. Maintaining consistent patient positioning during serial exams is advisable, as the count of B-lines can vary with changes in position, typically increasing when the patient is supine [33]. There are several methods to evaluate B-lines, which include both scoring and counting approaches. In scoring methods, a specific intercostal space with a minimum number of B-lines (i.e., ≥ 3) is marked as positive, and the total number of positive zones is recorded [28, 34]. Alternatively, B-lines can be individually counted within a specific area of the chest or approximated by the proportion of the sector area they cover beneath the pleural line [35, 36]. The quantitative technique is more precise and reliable in terms of consistency within and between different observers [37, 38]. A cut-off value of ≥ 3 B-lines in at least two zones per hemithorax has been proposed to identify acute HF with high sensitivity and specificity [28]. In chronic HF, a higher number of B-lines is associated with an increased risk of subsequent HF admissions and mortality, even in patients with few or no symptoms [3, 23, 39].

Inferior vena cava (IVC). The diameter of the IVC and its respiratory changes are regarded as indicators of CVP [40]. From a pathophysiologic point of view, a high CVP leads to IVC distention and, eventually, to reduced collapsibility during inspiration [40]. Thus, US evaluation of the IVC is valuable for assessing systemic venous congestion, monitoring a patient’s response to diuretic treatment, and predicting outcomes [40]. Importantly, many ambulatory patients with chronic HF who are presumed to be clinically free from congestion are subsequently found to have a dilated and/or non-collapsing IVC, which is associated with elevated natriuretic peptides and an increased risk of poorer outcomes, regardless of LVEF [3, 39, 41]. Furthermore, in patients discharged after a hospitalisation due to acute decompensated HF, the presence of a persistently dilated IVC before discharge is associated with a high risk of rehospitalisation [42]. Surprisingly, invasive studies on patients undergoing RHC revealed only a modest correlation between RAP and the diameter of the IVC measured by echocardiography [40]; this correlation is even weaker in patients mechanically ventilated, as positive intrathoracic pressures generated by mechanical ventilation cause IVC dilation and reduce its collapsibility [43].

To perform an accurate ultrasound evaluation of the IVC, a phased array or curvilinear transducer should be used with the patient lying supine. The diameter of the IVC should be measured in the subcostal long-axis view, 1–2 cm proximal to its connection with the right atrium. The IVC diameter should be measured throughout the respiratory cycle at the same sagittal level, perpendicular to the long axis of the IVC (a 3D probe may assist in the measurement, thanks to the possibility of imaging long-axis and short-axis IVC simultaneously; Fig. 3A) [15]. It is recommended that an IVC diameter < 2.1 cm, collapsing by more than 50% with inspiration, suggests normal RAP, typically around 3 mmHg (range 0–5 mmHg). Conversely, an IVC diameter greater than 2.1 cm with inspiratory collapse < 50% indicates very high RAP (range 10–20 mmHg) [15]. As a potential limitation, not all patients can tolerate the assessment of IVC, and prolonged training and experience are required to identify and evaluate IVC in most patients [44].

A Ultrasound assessment of the inferior vena cava. B Ultrasound assessment of the internal jugular vein at rest and after Valsalva (Jugular Venous Distensibility ratio ~ 1.5, which suggests severe congestion)

Internal jugular vein (IJV). IJV distension is a proxy for CVP and represents an additional strategy to evaluate systemic venous congestion. Similar to the IVC, IJV distension is associated with a high risk of HF hospitalisation and disease progression [45]. A clinically distended IJV is a powerful marker of poor prognosis in patients with heart failure [45]. However, estimating JVP only by physical exam is often challenging, as the reliability of the evaluation depends on the clinician’s experience, examination time, body size, and intra-thoracic pressure fluctuations [46]. US-based IJV evaluation allows for precise assessment of the diameter of the IJV and its variations during the Valsalva manoeuvre [10] and identifies patients with higher RAP at greater risk [45, 47].

The right IJV lies close to the carotid artery beneath the sternocleidomastoid muscle, allowing easy access for ultrasound examination. For precise evaluation, the patient is studied in a semi-supine position, with the head and neck elevated at approximately 45 degrees, using a high-frequency linear transducer positioned below the jaw angle along the sternocleidomastoid muscle, moving downward toward the angle of Louis [10]. Evaluations performed with the patient in the supine or sitting position should be avoided as they are associated with overdistension and collapse of the IJV, respectively. During the US examination, the diameter of the IJV is measured both at rest and during the Valsalva manoeuvre, either with M-mode or 2D-US (Fig. 3B). The JVD (Jugular Venous Distensibility) ratio is also calculated as the ratio of the IJV diameter during Valsalva to the IJV diameter at rest; this parameter is consistent across different observers [10]. In euvolemic patients with HF, the resting IJV diameter is usually ~ 0.15 cm and increases 5–6 times (i.e., up to 1 cm) during a Valsalva manoeuvre [48, 49]. The maximal diameter achievable during Valsalva remains largely consistent across patients due to the limited compliance of the vessel. Therefore, fluid overload is associated with increased IJV diameter at rest and, in turn, a lower JVD ratio (< 4). [39, 48, 49]. A reduced JVD ratio correlates with more severe symptoms, elevated natriuretic peptides, right ventricular dysfunction, and tricuspid regurgitation [48], but not with LVEF. It also predicts an increased risk of HF-related hospitalisations and death independently of natriuretic peptides [39, 46, 49]. Furthermore, some studies have suggested that the cross-sectional area of the IJV, measured during the Valsalva manoeuvre, might be a more reliable index of RAP [50, 51]. Again, significant variations in the IJV area during Valsalva usually indicate normal CVP and are associated with better prognostic outcomes [50, 51].

One potential limitation of the US evaluation of the IJV is the difficulty in avoiding excess pressure with consequent IJV compression during the examination [52]. This issue may affect the reliability of the measurements and introduce interobserver and intraobserver variability, especially when estimating the volume status and fluid responsiveness in critically ill patients. Therefore, care must be taken to apply the minimum pressure required for the examination [52]. Furthermore, the Valsalva manoeuvre, crucial for effective measurement, can be particularly challenging for acutely unwell patients. For such cases, alternative techniques such as passive leg raising might be considered to estimate CVP, though their impact on the IJV diameter needs further investigation.

Liver US. Doppler evaluation of venous blood flow from the splanchnic district to the liver (i.e., portal venous flow, PVF) and from the liver to the heart (i.e., hepatic venous flow, HVF; Fig. 4) refines the characterisation of systemic congestion and its impact on abdominal organs in HF [16, 53]. Under normal conditions, due to the distensible nature of the hepatic veins, HVF is pulsatile, reflecting the dynamic changes in RAP with each cardiac cycle [16]. The typical waveform observed in the hepatic veins is triphasic, with S- and D-waves directed towards the heart (and away from the liver), indicating systolic and early diastolic right atrial filling; the S-wave is physiologically higher than the D-wave. During late diastole, in the absence of arrhythmias such as atrial fibrillation, the atrial contraction generates a brief A-wave directed towards the liver (and away from the heart). Since venous wall distensibility diminishes as the distance from the right atrium increases, venous flow becomes continuous in more peripheral systemic veins. In particular, the portal vein is at least partially insulated from the right atrium by the resistance of hepatic sinusoids; as a consequence, physiological PVF is continuous or only mildly pulsatile and displays a hepatopetal direction (i.e., it is directed towards the liver) [16, 53]. However, in the presence of worsening venous congestion, the compliance of the IVC reaches a plateau, and the amount of pressure transmitted upstream from the right atrium increases. This results in venous dilation and loss of the physiologic triphasic waveform in the hepatic veins, with a blunted or even reversed S-wave (especially if significant tricuspid regurgitation is present) and the appearance of a positive V-wave indicating right atrial overfilling with subsequent flow recoil towards the liver. Likewise, a shift from continuous to pulsatile PVF may appear, possibly with a brief flow reversal phase (i.e., hepatofugal PVF) during systole in most severe cases [54]. The severity of the alterations in flow dynamics in the portal vein can be quantitatively assessed using parameters like the pulsatility fraction, calculated as the difference between the maximal and minimal PVF velocities, divided by maximal PVF velocity and expressed as a percentage. While no agreed-upon cut-points exist for this parameter, an increased PVF pulsatility fraction has been associated with right ventricular dysfunction and increased CVP [16, 55, 56].

Right. Doppler evaluation of portal vein flow (PVF). A Healthy subject; PVF pulsatility fraction = 7%. B Outpatient with HF with mildly abnormal PVF (pulsatility fraction = 42%.). C Patient hospitalised due to acute HF showing systolic flow reversal (hepatofugal PVF). HF: heart failure. PVF: portal venous flow. Left. Doppler evaluation of hepatic venous flow (HVF). (1) Healthy subject with typical triphasic HVF: the S-wave is higher than the D-wave, with a brief A-wave generated by atrial contraction. (2) Patient with heart failure and atrial fibrillation, showing a blunted S-wave without the A-wave. (3) Patient with severe tricuspid regurgitation and atrial fibrillation, showing a reversed S-wave without the A-wave

The patient is typically supine or in left lateral decubitus for Doppler evaluation of the hepatic and portal veins. Liver vessels can be identified by positioning a curvilinear or phased array probe in the mid-to-posterior axillary line with the probe marker directed toward the patient’s head and color Doppler imaging with the flow scale adjusted to low-flow velocities (preferably ≤ 20 cm/s) [16]. Measurements should be performed during a breath-hold to mitigate any distortions caused by breathing [16]. Doppler evaluation of the HVF and PVF presents specific challenges. Even in patients with severe tricuspid regurgitation, HVF might not be impaired if the compliance of the right atrium is maintained; this indicates that the evaluation of HVF could potentially underestimate the degree of venous congestion [57]. On the other hand, pulsatile PVF can be observed in healthy, thin individuals and patients with arteriovenous malformations [57]. Also, patients with increased liver stiffness, e.g., liver cirrhosis or metabolic dysfunction-associated steatotic liver disease, may have a non-pulsatile PVF despite severe venous congestion due to a lack of pressure transmission from the right atrium through the liver sinusoid [57].

Renal venous flow (RVF). Doppler-derived evaluation of RVF is emerging as a valuable tool in managing HF, though its clinical usefulness is still under investigation [10]. In euvolemic patients, RVF is continuous, with negligible fluctuations during the cardiac cycle. As CVP increases and intravascular congestion worsens, these fluctuations increase in magnitude, resulting in discontinuous RVF, which can be further categorised qualitatively as pulsatile, biphasic, or even monophasic (Fig. 5). Quantitative metrics such as the venous impedance index (VII) and venous discontinuity index (VDI) further refine the characterisation of renal congestion. The VII is calculated as the difference between the maximum and minimum flow velocities, divided by the maximum flow velocity and expressed as a percentage [3, 10]. Thus, a higher VII indicates increased impedance, with a score of 0 denoting uniform flow and 1 indicating complete flow interruption during the cardiac cycle [3, 10]. Conversely, the VDI is defined as the proportion of the cardiac cycle where flow is absent, again expressed as a percentage [10]. Abnormal RVF patterns correlate with heightened CVP and poorer prognosis in both acute and chronic HF [58, 59]. This is not surprising, as in the presence of elevated CVP, pressure in the interstitium of the encapsulated kidneys is also raised [54], which may lead to further deregulated release of neurohormones, partial collapse of the nephrons, and renal ischaemia, increasing the risk of acute kidney injury [60, 61, 10]. Interestingly, changes in RVF may precede alterations in other echographic indices of congestion, highlighting RVF as a potential early indicator of fluid overload and diuretic response [10].