Right-to-left intrapulmonary and intracardiac shunts (RLS) are important causes of profound hypoxia as they permit passage of deoxygenated venous blood into the systemic circulation.1 A patent foramen ovale (PFO) can be found in up to 20–25% of healthy individuals, although the presence of an acute RLS is relatively rare.2,3,4 These RLSs are an underappreciated cause of significant hypoxia, necessitating an early and accurate diagnosis to optimize management. A clue in the suspicion and/or identification of an intracardiac shunt occurs when the patient’s changes in blood pressure and oxygen saturation are proportional (Fig. 1). In patients with intracardiac shunts, desaturation will occur when the right atrial pressure exceeds the left atrial pressure.1 Increasing the left atrial pressure with the use of a vasopressor such as noradrenaline or phenylephrine, as shown in Fig. 1, will reduce the RLS and proportionately increase the oxygen saturation.

Fig. 1
figure 1

Perioperative monitoring during percutaneous interatrial septal defect closure in a 37-yr-old woman. The changes in arterial pressure (AP) and pulse oximetry saturation (SpO2) before and after closure at 13:00 (white arrow) are illustrated. Note that before closure there were parallel changes in AP and SpO2 but once the septal defect was closed the relationship between AP and SpO2 was lost. Parallel changes in AP and in SpO2 should always raise the possibility of an intracardiac shunt. EtCO2 = end-tidal carbon dioxide; HR = heart rate.

Full size image

Venous thromboemboli can pass through to the systemic circulation by bypassing the lungs through either a PFO, atrial septal defect (ASD), or a pulmonary arterio-venous malformation (PAVMs).1,5,6,7 An intracardiac defect with RLS has been detected in about 10% of patients with ischemic stroke,8 although other studies have reported an incidence up to 46% in those with cryptogenic stroke.9,10 Left-to-right cardiopulmonary shunts (LRS), including ventricular septal defect and ASD, may lead to severe pulmonary hypertension secondary to right heart volume overload.1

Ultrasound-enhancing agents (UEAs) have become a useful tool for cardiovascular imaging, particularly for the diagnosis and assessment of cardiopulmonary shunts. Agitated saline (AS) is the agent used primarily and comprises large bubbles (approximately 30 μm) that cannot flow through the pulmonary circulation.1 In patients without RLSs, no bubbles are present in the left side of the heart during a bubble study.11 Because AS does not routinely pass to the left side of the heart, it cannot evaluate the left heart, but does allow RLSs to be visualized.1,12 If bubbles appear in the left heart on the ultrasound monitor, a cardiopulmonary shunt must be present. This is colloquially known as a “bubble study.”

Right-to-left pulmonary and cardiac shunts may be found in patients with various lung diseases that cause the veno-capillary tree to dilate. There are a variety of modalities available to evaluate these shunts, including transesophageal echocardiography (TEE) and transcranial Doppler ultrasonography (TCD).12 While TEE is the clinically accepted standard for diagnosing RLSs, it is an invasive procedure requiring sedation, a skilled operator, and specialized equipment.12,13 Furthermore, evidence suggests that contrast-enhanced transthoracic echocardiography (TTE) can detect the majority of RLSs with better sensitivity than TEE examinations (100% vs 85%),11,14,15 and with better specificity than contrast-enhanced TCD.16

Point-of-care ultrasound in the anesthesiologist’s or intensivist’s hands assists in the evaluation of clinically important RLSs, such as in severe refractory hypoxemia or paradoxical embolism.17 Early POCUS evaluation for patients with end-stage heart failure is important to assess for left ventricular assist device (LVAD) candidacy.18 The physiology associated with an LVAD includes reduced left heart filling pressures and may precipitate RLS and hypoxemia in patients with a PFO.19 Using POCUS for an early diagnosis allows patients with clinically significant RLSs to receive expedited therapy.

Article selection methods

In this review, the role of AS-enhanced POCUS for detecting RLSs with emphasis on the proper application of shunt-enhancing maneuvers is illustrated. Focus is placed on the most prevalent RLS types seen in the intensive care unit (ICU), including PFO, ASD, and PAVMs. The authors searched PubMed and Google Scholar for articles with the keywords “bubble study” OR “agitated saline contrast injection” OR “right to left shunts” AND “echocardiography”, AND “intensive care”. Authors included randomized-controlled trials, narrative reviews, systematic reviews and meta-analyses, book chapters, case reports and series, and clinical guidelines, as well as prospective and retrospective cohort studies. Only articles published in English were included in this review. The initial literature search revealed over 600 articles. Authors reviewed all the relevant articles and decided which studies to include in the review by consensus. A total of 105 resources were selected for inclusion in this narrative review.

Principles of AS echocardiography

As part of the piezoelectric effect, an ultrasound device’s mechanical energy deforms the crystals at the end of the transducer and emits inaudible sound waves (1–5 MHz). These sound waves propagate and reflect through tissues of varying acoustic impedance (resistance to sound waves), then these signals return to the transducer. Once again, the crystals are deformed, and this mechanical energy is converted to electrical energy, which is then amplified to produce a pixelated image on the ultrasound display screen

As sound waves travel from one tissue to another, the change in acoustic impedance at the interface also reflects sound waves. The ability of a sound wave to propagate from one tissue to another depends on the difference between their respective impedance.20 The larger the difference in acoustic impedance, the more the ultrasound waves reflect, and the more echogenic the surface will appear.20

Red blood cells, despite generally appearing anechoic, actually produce a very weak echogenic signal because of the differing densities of red blood cells and the surrounding blood plasma.21 This can present as “spontaneous echo contrast”, an echogenic swirling pattern seen in settings of low-velocity blood flow or stasis.22 Gas is approximately 100,000 times less dense than blood.1 Thus, the gas bubbles of AS enhance this difference in acoustic impedance at the blood-saline-air interface to generate a greater reflection of sound waves compared with between red blood cells and plasma alone.23

It is important to differentiate between AS and other UEAs. Because of their broad applicability and advantageous safety profile, many different types of UEAs have been developed, and guidelines on their use have been recently updated.24 They consist of varying sized bubbles, shell material, and gas cores. Initial ultrasound enhancement employed non-encapsulated nitrogen and oxygen from ambient air. These bubbles are incapable of crossing the pulmonary vascular capillary bed and were primarily used for right heart evaluation. Nevertheless, in the case of RLS, these bubbles were noted in the left heart. The first-generation UEAs are manufactured air bubbles that are encapsulated or attached to one of a variety of microparticles.11,25 These agents are injected intravenously and appear in the left heart.25 Second-generation agents (i.e., activated perfluorocarbon and stabilized sulfur hexafluoride microbubbles surrounded by a phospholipid shell) replace air with a fluorocarbon gas that increases stability as well as the duration of its ultrasonic effect.25,26 More recent agents include a polymer shell, leading to more consistent enhancement.11,26 Regardless, all UEAs have a long track record of safety, even in patients with known cardiac shunts.27,28 Nevertheless, AS is the preferred method for evaluating RLSs using POCUS in the acute care setting because of its ease of availability and low cost, although several variations of UEAs have the potential to be used as well (Table 1).

Table 1 Common ultrasound-enhancing agents and recommended dosing
Full size table

Equipment, instrumentation, and technique

Vascular access and contrast preparation

Performing and optimizing a bubble study relies on consistent preparation and injection of AS, which is best performed with an assistant performing the injection while the sonographer obtains the images.11 To begin with, the bubbles require thorough agitation immediately prior to injection.11 When using AS for RLS detection, the technique relies on utilizing two 10-mL syringes connected to a three-way stopcock to agitate the saline (Fig. 2).11 As AS separates into its component parts quickly, the saline should be re-agitated before each injection.11 It is recommended to use a Luer Lock syringe to avoid accidentally spraying saline during agitation.11 It is recommended that the AS be composed of 9.0–9.5 mL of normal saline and 0.5–1.0 mL of room air.11 The American Society of Echocardiography guidelines recommend a mixture of 8 mL of normal saline, 1 mL of the patient’s own blood, and 0.5–1.0 mL of room air as an alternative.4 This may increase the intensity of the microbubbles visualized by echocardiography.29 Additionally, some authors have reported using AS with added benzyl alcohol (9 mg·mL−1 as a preservative in saline) as opposed to AS alone.30 The alcohol acts as a surfactant, leading to smaller bubbles and better overall intravenous contrast effect.31 This saline-air mixture should be forcefully injected between two syringes until the bubbles appear uniform, without visibly large air bubbles.11,32 When this is achieved, the entire syringe’s contents should be injected, as the AS mixture quickly separates into its composite parts.11 This AS injection should be followed by a second syringe of “flush” solution to ensure that the AS reaches its destination faster and in a more complete manner.

Fig. 2
figure 2

Materials necessary for point-of-care bubble study: 18-G peripheral intravenous catheter, 9.5 mL of normal saline with 0.5 mL ambient air, two 10 mL Luer Lock syringes, and a three-way stopcock with extension tubing. Note: positioning the stopcock “off” to the patient allows saline agitation between the two syringes.

Full size image

Injection of UEAs through a small-bore catheter destructs the thin-walled external capsule; however, this is not the case for unencapsulated AS, which may be used with any size of intravenous catheter.33 Nevertheless, AS should be administered through a 20-G or larger cannula, most commonly in the antecubital veins, although using femoral and dorsal hand veins has been reported.34,35 As the patient may be placed in left lateral decubitus to enhance imaging windows, venous access should be preferentially gained on the patient’s right side.

Compared with an antecubital vein, utilization of a femoral vein for AS injection increases the sensitivity for PFO detection during bubble studies.34,36,37,38 This is because inferior vena cava blood flow is directed towards the atrial septum, while superior vena caval flow is pointed towards the tricuspid valve (Fig. 3).39 There are some practical disadvantages to using the femoral vein, notably an increased rate of vascular complications, including inadvertent arterial puncture, thrombosis, bleeding, and, in rare cases, arterio-venous fistula formation.40 Nevertheless, femoral vein access remains a valuable option in cases of incomplete right atrial opacification or if there is a high index of suspicion for an RLS despite a negative bubble study using an upper limb vein.

Fig. 3
figure 3

A. Venous flow patterns in the superior vena cava and inferior vena cava. The inferior vena cava flow is directed towards the atrial septum, while superior vena cava flow is pointed towards the tricuspid valve. This allows the agitated saline to preferentially pass through an intracardiac right-to-left shunt. B. Mechanism of right-to-left shunt augmented by Valsalva. An arrow facing upward indicates increase in pressure and facing downward indicates decrease in pressure. LA = left atrium; LV = left ventricle; PFO = patent foramen ovale; PLA = left atrial pressure; PRA = right atrial pressure; RA = right atrium; RV = right ventricle.

Full size image

Imaging technique

Before administration of AS, an apical four-chamber view of the heart should be acquired. Alternatively, one may obtain a subcostal view or a parasternal short axis view in which the atrial septum is visualized (Fig. 4).35,41 The parasternal long axis view does not adequately evaluate the right atrium next to the atrial septum.1 When possible, the patient should lie in the left lateral decubitus position, which brings the heart anterior and lateral, improving image acquisition.1 Nevertheless, this may not be feasible in the critically-ill perioperative patient, in which case supine positioning is appropriate. If necessary, the patient’s ribcage may be expanded by placing their left arm behind their head, thereby increasing space (acoustic window) for the phased array transducer in between the ribs.1 The equipment settings should be configured to acquire a 20-sec duration clip. The ultrasound’s depth and gain functions should be set to achieve the highest image resolution, with the screen’s focal zone near the base of the heart. Retrospective capture is preferred and allows the operator to record the previous 20-sec loop when saving a clip. Prospective capture, conversely, records the upcoming 20-sec loop and may not allow adequate recording of the RLS. In particular, this may occur in those with very low cardiac output who require more time for AS to fully opacify the right atrium. This may be mitigated by starting prospective capture at, or just before, injection of the AS.

Fig. 4
figure 4

The focused cardiac ultrasound study (FOCUS) exam. This includes these important transthoracic echocardiographic views. Scanning should be performed in a systematic, clockwise fashion, from three main areas: 1) parasternal (A,D), 2) apical (E,F), and 3) subcostal (G,H). Ao = aorta; LA = left atrium; LV = left ventricle; RA = right atrium; RV = right ventricle. (Reproduced by permission of Taylor and Francis Group, LLC, a division of Informa plc. from Denault et al.).105

Full size image

Shunt-enhancing maneuvers and their associated physiology

Although right atrial pressure is typically lower than the left atrial pressure, this is not always the case. During the patient’s normal respiration, transient reversal of the interatrial pressure gradient (right atrial pressure > left atrial pressure) may be sufficient to allow a short-lived period of right-to-left shunting, particularly if a large defect is present. Nevertheless, this finding is not clinically reliable, and RLSs are difficult to assess on echocardiography without the aid of shunt-enhancing maneuvers such as Valsalva or lung recruitment maneuvers with subsequent rapid release.42 In the critically-ill perioperative patient unable to participate in a Valsalva maneuver, a lung recruitment maneuver can mimic the physiologic effects of Valsalva.43 Lung recruitment maneuvers aim to quickly apply increased airway pressures for a short period of time, which increases intrathoracic pressure and decrease venous return. Release of this maneuver (and release of Valsalva) increases venous return to the right heart and augments right atrial filling, precipitating a transient right-to-left atrial pressure gradient, exposing an RLS. Nevertheless, successful use of the Valsalva maneuver or the release of a lung recruitment maneuver requires precise timing and adequate technique to accomplish.1 It is often beneficial to practice these maneuvers prior to injection of AS.

For a proper Valsalva maneuver, the patient must pause their breathing after inhalation and bear down for approximately ten seconds.1 Many patients cannot produce an adequate Valsalva even with coaching. In these patients, pressing on the abdomen with one hand and asking the patient to push out against the sonographer’s hand, and then “relax”, may be more successful.

These augmentation maneuvers change the position of the heart upon release. Upon inspiration or upon application of the recruitment maneuver, the heart moves caudad towards the abdomen, whereas upon release of the Valsalva or lung recruitment, it moves cephalad. The sonographer must adjust the imaging plane and continue to visualize the heart after release of these maneuvers as it returns to its original position.1 Additionally, the septum is normally convex toward the right atrium, as the pressure in the right atrium is less than that of the left atrium.39 During these shunt-enhancing maneuvers, the atrial septum shifts toward the left (Fig. 5). With release, this pressure differential transiently changes and allows the septal curvature to alter, as the atrial septum shifts towards the right. Release of a Valsalva or lung recruitment maneuver is the optimal time to have a fully opacified right atrium, providing the highest sensitivity for detection of RLS.13

Fig. 5
figure 5

A. Apical four-chamber view during Valsalva. B. Note the atrial septum shifting toward the left atrium (red arrow) as well as the full opacification of the right atrium. LA = left atrium; LV = left ventricle; RA = right atrium; RV = right ventricle.

Full size image

During the Valsalva maneuver, the saline mixture is agitated as described above and injected just prior to release of the Valsalva breath.1 Clip acquisition should begin at the time of AS injection. Immediately after AS administration, the patient should begin to breathe normally.1 At this time, AS should completely opacify the right heart. Agitated saline bubbles are incapable of passing the pulmonary circulation.11 Therefore, any evidence of bubble effect in the left-sided circulation is likely due to the presence of a cardiopulmonary shunt.11,23,44

Colour flow Doppler is used frequently in sonography to semi quantitate general velocity and direction of blood flow within the heart, which may aid in identifying intracardiac shunts. Nevertheless, the interatrial pressure gradient is small without the aid of shunt-enhancing maneuvers, which leads to low velocities on colour Doppler and poor visualization of RLS.35 Furthermore, the PFO jet is typically found in the far field of the ultrasound, impairing image resolution.35 These both contribute to the low sensitivity (22%) of colour flow imaging to detect a PFO.35,45 Colour Doppler echocardiography is useful for visualizing high-velocity flow signals present in LRS but performs poorly when evaluating for low-velocity flow signals in transient RLSs.12,46

Furthermore, any decrease in left heart diastolic compliance or elevation of left atrial pressure will make an RLS more difficult to evaluate.47 This may be seen in a variety of conditions, including hypertension, restrictive or infiltrative cardiomyopathies, and valvular disease.47 This has been borne out in the literature, including one study showing that 5% of patients with evidence of left heart disease have an RLS, compared with 29% in patients with no evidence of left heart disease.48,49 Alternatively, patients with reduced right ventricular compliance and those with tricuspid valve disorders have increased right atrial pressure, aiding in the evaluation of RLSs, as well as increasing the potential for clinical hypoxia.47 Theoretically, any conditions elevating right heart pressures can be associated with an RLS.47

Maneuvers that increase venous return to the right heart are often required to accurately diagnose RLSs (Fig. 3).12,50,51,52 Techniques that directly alter venous return, such as abdominal compression or placing the patient in the Trendelenburg position may also be employed for this purpose, but they are not nearly as sensitive as techniques that indirectly increase venous return such as a well-executed Valsalva or release of lung recruitment maneuver.53,54 The presence of positive end-expiratory pressure increases the intrathoracic pressure, right ventricular afterload, and right atrial pressure while decreasing the left ventricular preload; together, these improve RLS detection.55,56,57 When determining the efficacy of a shunt-enhancing maneuver on echocardiography, one should see opacification of the right atrium when the atrial septum bows toward the left heart (Fig. 5).35,58,59 The maneuvers are deemed inadequate if there is no evidence of septal bowing, which should prompt a repeat attempt.35 Similarly, a repeat bubble study is necessary if a minimal amount of AS appears in the left heart, which may indicate an inadequate shunt-enhancing maneuver.

Echocardiography for the detection of anatomic cardiopulmonary shunts

Intracardiac shunts

There are many etiologies of an intracardiac RLS, including a variety of ASD subtypes and coronary sinus defects. Nevertheless, this review is limited to the two most common intracardiac shunt pathways—PFOs and secundum ASDs. Compared with ASDs, a PFO is a natural part of cardiac embryology.1 Commonly, the foramen ovale closes before adulthood; however, up to 25% of the population has evidence of a PFO.1,3 On conventional echocardiography without AS, a PFO typically appears as piece of myocardial tissue adjacent to the foramen ovale attached to the atrial septum (Fig. 6).1 The PFO opens intermittently whenever the pressure in the right atrium exceeds the left atrial pressure, for example, during normal respiratory-phasic variation.60 Nevertheless, whenever the pressure in the left atrium exceeds the right atrial pressure, the PFO remains closed, with a small LRS visible on colour flow Doppler (Fig. 6). An ASD may also be associated with an RLS, most commonly via a secundum ASD.1 A secundum ASD is most commonly located in the middle of the interatrial septum. Secundum ASDs appear as an aperture on echocardiography, which are typically difficult to visualize without the aid of AS.1

Fig. 6
figure 6

Patent foramen ovale (PFO). (A,B) A PFO is shown by colour Doppler (Nyquist 44 cm·sec−1) in a mid-esophageal (ME) bicaval view. (C,D) ME right ventricular inflow/outflow view shows opacification of the right-sided cardiac chambers during intravenous injection of agitated normal saline. At the release phase of a Valsalva maneuver, microbubbles are seen crossing from the right atrium (RA) to the left atrium (LA) through a PFO. AoV = aortic valve; IVC = inferior vena cava; RV = right ventricle; SVC = superior vena cava. (Reproduced by permission of Taylor and Francis Group, LLC, a division of Informa plc. from Denault et al.).105

Full size image

Compared with PFOs, ASDs confer an increased risk of right heart volume overload.1 A persistent interatrial communication is more likely to develop an LRS compared with the PFO intermittently communicating flap, as the PFO flap is typically pushed against the septum secundum because of elevated left atrial pressure.61,62 When evaluating for the presence of an RLS, the AS injection must be precisely timed to the release of the shunt-enhancing maneuver.1,63 Nevertheless, it should be noted that it may not be possible with POCUS alone to differentiate between a secundum ASD (small) and a large PFO. Because of atrial dilatation, severe stretching of the PFO may occur, making it indistinguishable from an ASD.

When using POCUS, an apical four-chamber view is the preferred image to detect AS in the right atrium.49 After opacification of the right atrium, the septum is visualized to determine appropriate bowing, and cardiac cycles are counted.49 Any appearance of AS in the left atrium or ventricle is diagnostic of an RLS, sometimes referred to as a “positive contrast effect”.51 The definition of a positive bubble study result, based on either the timing of AS appearance or number of bubbles visualized, remains under debate.12 The absolute number of AS bubbles necessary to define a positive test result varies; most commonly between one, three, or five bubbles visualized in the left heart.6,64,65,66 Nevertheless, a generally accepted definition of an RLS is any evidence of AS passing from the right to left heart after complete right atrial opacification.11 Some have postulated that shunt size may be approximated by counting the number of bubbles visualized in the left heart in the three cardiac beats after right atrial septal opacification.64,67 Right-to-left pulmonary and cardiac shunt size is categorized as small (< ten bubbles), moderate (> ten bubbles), and large (full opacification of the left atrium).1,51,68 The amount of AS passing through the RLS is dependent on the atrial pressure gradient, which is augmented by shunt-enhancing maneuvers. The largest number of bubbles seen at any one time after shunt-enhancing maneuvers determines the size of the shunt.51

Left-to-right cardiopulmonary shunts, most commonly ASDs, can cause irreversible pulmonary hypertension due to right heart volume overload.1 Any patient with echocardiographic evidence of right heart volume overload may warrant further investigation for an ASD and accompanying LRS.1 Nevertheless, AS bubble studies should not be used to evaluate for patients with small ASDs.1 Septal defects are primarily LRS because the left heart pressure is typically higher than the right heart pressure.1 These ASDs may be best visualized with colour flow Doppler as a colour jet on the right side of the heart when left atrial pressure is higher than right atrial pressure.1 If this jet is not visualized, and there is no evidence of increased right heart pressure, then an ASD is not likely to be present.1

Intrapulmonary shunts

Intrapulmonary RLS pathways include both macroscopic and microscopic PAVMs as well as intrapulmonary arteriovenous anastomoses. Recent literature suggests that intrapulmonary RLSs are present in up to 20% of the general population.69,70,71 Pulmonary arterio-venous malformations represent a direct pathway between the pulmonary arteries and veins.71,72 These are most commonly secondary to Osler-Weber-Rendu syndrome and are a well-described RLS pathway of thrombotic particles, occurring in up to one half of untreated patients.12,73,74,75 Intrapulmonary shunts may be suggested by the presence of AS in the left atrium or ventricle 3–6 beats after atrial opacification, termed the “3–6 beat rule”.12,76 This is due to the increased “transit time” through the lungs as opposed to immediately through an intracardiac shunt.

Differentiating between intrapulmonary and intracardiac shunts

Clinicians attempt to differentiate between extracardiac and intracardiac RLSs by examining the time it takes for AS to appear in the left heart.76 This is commonly referred to as the “3 beat rule”.76 If an RLS is present within 3–6 cardiac cycles after AS injection, the culprit lesion is likely to be intracardiac. Nevertheless, the clinical utility of this rule has recently been called into question.45,77,78 Microbubbles appearing earlier than within 3–6 beats may be caused by a combination of intracardiac and extracardiac RLSs.4,12,79 In the presence of a PFO, intrapulmonary shunts cannot be reliably detected because the AS may cross the interatrial septum while simultaneously crossing an extracardiac shunt.1 Furthermore, poor atrial septal shifting may lead to late-appearing bubbles from a PFO, as has been reported numerous times in the literature.79,80,81 Until more data are available to determine the clinical significance of the “3 beat rule”, the clinician must be vigilant when employing AS to evaluate intracardiac and extracardiac RLSs.

Pitfalls in bubble study interpretation

False positives

There are multiple false-positive and false-negative interpretations when evaluating for an RLS with POCUS (Table 2). For instance, a prolonged Valsalva maneuver causes blood to stagnate in the pulmonary veins. These red blood cells aggregate into a rouleaux formation and may mimic AS when Valsalva is released.82,83,84,85 Nevertheless, it should be noted that the only reported cases have occurred during TEE. Nonetheless, if this is suspected, the shunt-enhancing maneuver should be performed again without using AS. Appearance of bubbles during the attempt without AS would indicate that pseudo-bubbles are present.49

Table 2 Common pitfalls of interpreting echocardiography bubble study results, along with associated underlying mechanisms and potential solutions
Full size table

When performing a bubble study, rapid appearance of AS in the left atrium or ventricle should prompt consideration for the presence of an ASD.49 Investigation of the interatrial septum with colour flow Doppler may identify an ASD. An interatrial LRS may present with a “negative contrast effect” during a bubble study.12,86 This is seen as “a sharply delineated washout phenomenon appearing on the right atrial side of the interatrial septum in continuity with the contrast-free left atrium” (Fig. 7).12,86 While the “negative contrast effect” may indicate the presence of an ASD, sensitivity ranges between 37% and 58% in the published literature.87,88,89 Other findings indicative of an ASD and LRS include the presence of AS bubbles in the inferior vena cava after right atrial opacification, as well as characteristic flow patterns on colour Doppler imaging.61,88 Finally, the presence of an enlarged right heart should prompt consideration for an ASD and LRS.

Fig. 7
figure 7

(A,B) Negative contrast effect suggestive of an atrial septal defect. Note the absence of agitated saline on the right side of the heart immediately adjacent to the inter-arterial septum (red arrow) and no evidence of bubbles in the left atrium. LA = left atrium; LV = left ventricle; RA = right atrium; RV = right ventricle.

Full size image

False negatives

One cause of a false-negative study is a large eustachian valve, an embryologic remnant of the valve of the inferior vena cava”.90 During fetal circulation, the eustachian valve preferentially facilitates blood flow toward the foramen ovale, left heart, and systemic circulation.49 While the eustachian valve reabsorbs after birth in most people, there are multiple reports of persistent eustachian valves in adults.49,90,91 The eustachian valve is a part of the Chiari network, which is described as a “fenestrated, net-like embryonic remnants of valves of sinus venosus, lying closely in relation to the inferior vena cava and coronary sinus”.92 The eustachian valve may be mistaken for the right atrial septum on ultrasound, potentially causing a false positive if injected AS is seen on either side of the valve.1 Agitated saline delivered via the antecubital vein may stream directly along this valve into the right ventricle, resulting in incomplete opacification of the right atrium and a false-negative result.93,94 In contrast, AS delivered via the femoral vein is preferentially directed towards the atrial septum in patients with a persistent eustachian valve.94

Chronically elevated left heart pressures predispose to a false-negative bubble result. This is because right atrial pressures may never exceed left atrial pressures, even in the presence of Valsalva augmentation.48,95 Increased inferior vena cava flow may disrupt right atrial opacification, resulting in a false-negative bubble result.35 In this scenario, blood flow through the inferior vena cava may be diminished by pressing firmly on the liver.35 This maneuver may assist in differentiating between a “negative contrast effect” from an ASD and a bubble study disrupted by increased blood flow through the inferior vena cava.35,59

Normal variation in right heart pressures during respiration play an important role in false-negative bubble study results.49 It is known that respiratory-phasic increases in right atrial pressure during end-inspiration may reveal a transient RLS across a PFO.49 Nevertheless, this RLS may occur more than 3–6 beats after injection of AS, resulting in misclassification of an intracardiac shunt as an intrapulmonary shunt due to the timing of the phases of ventilation.49 This pitfall can be overcome by evaluating the patient’s respirations during the AS bubble study, or preferably, interpreting AS injection with Valsalva.49

An atrial septal aneurysm is a congenital deformity consisting of redundant and mobile interatrial septal tissue, which bulges into the right or left atrium.96 If a patient has an atrial septal aneurysm, intracardiac shunts should be suspected, as these have strong associations with both PFOs and ASDs.49,97,98 The atrial septal aneurysm can be used to evaluate the effectiveness of a patient’s Valsalva, as the septum will deviate towards the side of the heart with the lowest pressure at all times.49 When an effective Valsalva maneuver is released, the septum should transiently deviate towards the left atrium, the lower pressure chamber.49

Other alternative methods in the diagnosis of right-to-left pulmonary and cardiac shunts

The use of pulsed-wave Doppler on any arterial vessel is a simple way to diagnose the presence of air in the arterial circulation, particularly if no adequate cardiac windows are available. Transcranial Doppler monitoring modalities used in the operating room or in the ICU can detect and even measure the number of emboli that can occur (Fig. 8). Using this technique to obtain TCD signals using two-dimensional ultrasonography of the brain has been reported.99 These embolic signals are called “high-intensity transient signals” (HITS) and are commonly observed during extracorporeal membrane oxygenation.100 A simple saline injection can be associated with the appearance of those HITS even without contrast stressing the importance of bubble filters in those patients. When a standard TTE probe is used to perform brain ultrasound with a transcranial profile, high-intensity linear hair-like artefacts can also be seen in any artery such as the cerebral, hepatic, or renal arteries (Fig. 9). Techniques to obtain those signals have been reported previously.99,101,102,103,104 Indeed, TCD is currently used as a diagnostic modality to exclude RLS in patients with unexplained stroke resulting from a PFO and performs as well as TEE.16

Fig. 8
figure 8

A. Transcranial Doppler monitoring in a patient under extra-corporeal membrane oxygenation (ECMO). B. Note the appearance of several high-intensity transient signals (HITS) during 10 mL saline injection through a central venous catheter. Air-trapping devices are not as efficacious as cardiopulmonary bypass with the use of ECMO. EDV = end-diastolic velocity; MV = mean velocity; PI = pulsatility index; PSV = peak systolic velocity.

Full size image
Fig. 9
figure 9

Two patients undergoing cardiac surgery with simultaneous linear hair-like artefacts in the hepatic artery (A) and with several high-intensity transient signals (HITS) in the brain. In another patient, the artefacts are seen simultaneously in the renal artery (C) and in the brain (D). CPB = cardiopulmonary bypass; EDV = end-diastolic velocity; HITS = high-intensity transient signals; MV = mean velocity; PI = pulsatility index; PSV = peak systolic velocity. (Fig. 9A with permission of Denault et al.).104

Full size image

Conclusions

Point-of-care ultrasound echocardiography with AS is a quick, efficient method for investigating cardiopulmonary shunts and a useful clinical tool for the intensivist. This imaging modality is helpful for cases of acute RLS presenting with severe, refractory hypoxemia.2 The paradigm for interpreting bubble study results is built on the tenet that AS does not cross into the left heart in patients without a RLS. Although bubble studies are the clinical standard for RLS identification, POCUS practitioners should be knowledgeable of the potential pitfalls leading to both false-positive and false-negative results. Pulsed-wave Doppler interrogation of any arterial signal can also be used to detect abnormal RLS but requires further investigation to determine the exact etiology and mechanism.