Point-of-care ultrasound: aortic, lung and cardiac (the RUSH protocol)

Point-of-care ultrasound (POCUS) of the aorta, the lung and the heart is the bedside skill that lets the emergency physician localise the cause of shock, breathlessness and chest pain within minutes, at the point of care, without moving the patient. Three focused examinations are woven together: the aortic sweep for the abdominal aortic aneurysm, the lung survey for the artefact patterns that separate oedema from pneumothorax and from the normal lung, and the cardiac windows for the effusion, the left-ventricular function and the right ventricle. The RUSH protocol — Rapid Ultrasound in Shock — integrates them into a single resuscitation algorithm that maps the pump, the tank and the pipes, so that the sonographic findings drive the next drug, the next fluid bolus or the next call to theatre.[1][8]

Definition and the principle

The principle is that each examination answers a single, focused question at the bedside, and the answers combine to discriminate the four shock states. The aorta is measured in its longitudinal and transverse planes from the diaphragm to the bifurcation, and an outer-to-outer anteroposterior diameter over 3 centimetres defines the abdominal aortic aneurysm.[6][10] The lung is interrogated through the intercostal spaces for the pleural line and its artefacts: the horizontal A-lines (reverberation artefacts of a dry, air-filled lung, normal or COPD), the vertical B-lines (comet-tail artefacts arising from the pleural line that erase the A-lines and move with sliding, indicating a water-logged interlobular septum and pulmonary oedema), and the absent lung sliding with a lung point that confirms pneumothorax.[2][3][4] The heart is viewed through the parasternal, the apical and the subxiphoid windows to find the pericardial effusion, to eyeball the left-ventricular systolic function, and to detect the right-ventricular dilatation of massive pulmonary embolism.[7][8]

Indications

The aortic, lung and cardiac surveys are indicated whenever a sonographic answer would change the next decision in a critically ill patient, and the Fellowship candidate should reach for the probe early rather than late. The undifferentiated hypotension or shock is the prime indication — the RUSH exam is performed within the first minutes of the resuscitation to separate the hypovolaemic from the cardiogenic, the obstructive and the distributive causes.[1] The acute dyspnoea or respiratory failure triggers the lung and the cardiac survey together: bilateral B-lines flag cardiogenic pulmonary oedema, the A-line pattern with absent sliding flags pneumothorax, and the A-line lung with a dilated right ventricle flags pulmonary embolism — the BLUE protocol's discriminant profiles.[5] The syncopal, the chest-pain or the back-pain patient of the aneurysmal age gets the aortic sweep, as does any pulsatile abdominal mass or the older male smoker with new abdominal pain. The unexplained tachycardia, the peri-arrest patient and the cardiac arrest all warrant the cardiac and the lung survey to find the reversible causes.

Contraindications — when not to delay for the scan

There is no absolute contraindication to the examination itself, which is non-invasive and adds no radiation. The contraindication is to delaying a life-saving intervention for the sake of the scan. The haemodynamically shattered patient with a pulsatile abdominal mass and the classic aneurysm triad goes to theatre without a scan; the patient in refractory ventricular fibrillation gets the defibrillation and the CPR, with the cardiac windows grabbed during the rhythm check; the tension pneumothorax is decompressed on clinical grounds before any probe is placed. A technically limited scan — the morbidly obese, the patient with extensive subcutaneous emphysema, the agitated uncooperative patient, the heavily bandaged chest — is documented as indeterminate rather than falsely read as negative, and the decision is made on the clinical picture and the other windows that do succeed. [1]

Differential diagnosis — the shock states and the sonographic pattern

The Fellowship candidate must hold the four shock states and their sonographic fingerprints at once, because the RUSH exam exists to discriminate them at the bedside. Each row below is a complete sonographic profile — the pump, the tank and the pipes read together. [1]

The discriminating power of the protocol comes from reading the patterns together rather than in isolation: a single dilated right ventricle is meaningless without the A-line lung and the plethoric IVC that complete the pulmonary-embolism profile, just as a single effusion is harmless until the right-ventricular free wall buckles. [1]

The aortic window — the abdominal aortic aneurysm

The abdominal aorta is scanned with the curvilinear probe in the longitudinal and the transverse planes, swept from the xiphisternum to the bifurcation, because an aneurysm may lie at any point and a single measurement misses it. The probe is placed in the epigastrium, the depth set to see the full width of the aorta and the vertebra behind it, and the gain adjusted so the vessel lumen is dark. The anteroposterior diameter is measured outer-to-outer (the leading edge of the anterior wall to the leading edge of the posterior wall) at the widest point, in the transverse plane, with the measurement perpendicular to the long axis of the vessel — a tangential cut overestimates the size and overcalls the aneurysm.[10] The sweep captures the suprarenal and the infrarenal aorta and continues onto the common iliac arteries; the proximal extent matters because it determines the surgical and the endovascular options. An aortic diameter over 3 centimetres is an aneurysm, and the diameter that predicts rupture is the large one — the rupture risk rises steeply above 5.5 centimetres in men, the threshold for elective repair established by the screening trials.[6]

In the shocked patient with back or abdominal pain, the aortic sweep is read alongside the FAST: an aneurysm plus a retroperitoneal fluid stripe or free intraperitoneal fluid is rupture until proven otherwise, and the patient goes to theatre or to endovascular repair without a CT. The echoconfirms whether tamponade complicates a dissection. The classic error is the tangential measurement that overcalls a normal aorta, and the second is accepting a single transverse slice that misses a saccular aneurysm at an unsampled level. [1]

The lung windows — A-lines, B-lines and pneumothorax

The lung is examined with a high-frequency linear probe (or the phased-array probe) at eight zones — the upper and lower anterior and lateral chest on each side — and the Fellowship candidate should run the survey to a fixed sequence so no zone is missed. The pleural line lies between the two rib shadows (the bat-wing sign), and its movement and its artefacts are the whole of the examination.[2]

Lung sliding is the to-and-fro shimmer of the visceral pleura against the parietal pleura with respiration; on M-mode it produces the seashore sign — a grainy sandy pattern below the pleural line over horizontal wave-like lines above. Its absence abolishes the shimmer and converts M-mode to the barcode or stratosphere sign of parallel horizontal lines throughout. Absent sliding is pneumothorax, but also apnoea, a main-stem intubation, the post-pleurodesis lung, severe bullous disease and ARDS — so sliding is read alongside the artefacts.[2]

The A-line pattern — horizontal reverberation lines deep to the pleural line at regular intervals — is the normal dry lung, also seen in COPD and asthma where the lung is air-filled and the septa dry. The B-lines (also called comet-tail artefacts) are vertical, laser-like, erasing the A-lines, arising from the pleural line and moving with sliding; three or more in a zone indicate the water-logged interlobular septum of pulmonary oedema, and bilateral diffuse B-lines with a hyperdynamic or failing heart on the cardiac view confirm cardiogenic oedema against the A-line alternative.[3] The BLUE protocol harnesses these profiles: bilateral A-lines with a normal left ventricle point to asthma or COPD; bilateral B-lines with a low ejection fraction point to pulmonary oedema; and the A-line lung without B-lines, with a dilated right ventricle, points to pulmonary embolism — the dry-lung, large-heart profile.[5]

Pneumothorax is sought over the anterior chest in the supine patient (where the air rises). The signs, in order of specificity, are the absent lung sliding, the absent B-lines, and the lung point — the boundary on the chest wall where the sliding lung meets the non-sliding pneumothorax, seen as a flickering transition on two-dimensional and a transition on M-mode. The lung point is the only sign near-diagnostic of pneumothorax (specificity approaching 100 per cent), but it is insensitive because it is seen only when the pneumothorax is partial and the probe happens to sit on the margin; a completely collapsed lung has no lung point.[4]

The cardiac windows — effusion, function and the right ventricle

The focused cardiac exam uses four windows — the parasternal long-axis, the parasternal short-axis, the apical four-chamber and the subxiphoid — and the candidate should be able to obtain at least two in any patient. The phased-array probe is used, set to the cardiac preset, with the depth at 15 to 18 centimetres in the adult and the ECG gated where available.[8]

The pericardial effusion appears as an anechoic stripe separating the visceral from the parietal pericardium, circumferential or loculated. The Fellowship candidate must distinguish the effusion that threatens tamponade from the one that does not: right-ventricular free-wall collapse in early diastole and right-atrial collapse in late diastole are the echocardiographic signs of tamponade, and they correlate with the clinical picture of the Beck triad and the pulsus paradoxus. A large effusion that has not yet produced chamber collapse is not tamponade; a small effusion under pressure in the trauma patient may be. Tamponade is always a clinical diagnosis the scan confirms, never a scan diagnosis alone. [1]

Left-ventricular systolic function is estimated by eye — the eyeball ejection fraction — by watching the inward movement of the walls and the change in cavity size through the cycle. Hyperkinetic (the walls nearly kissing), the normal-to-hyperdynamic pump with a small cavity points to hypovolaemia; the poorly contracting dilated ventricle points to cardiogenic failure and explains the B-lines on the lung survey; the regional wall-motion abnormality — a segment that bulges outward in systole — points to the acute coronary syndrome. The eyeball method is less precise than formal quantification but is accurate enough to drive the resuscitation decision.[8]

The right ventricle is examined for the dilatation and the hypokinesis of acute pulmonary embolism. The normal right ventricle is smaller than the left and contracts briskly; in massive embolism it dilates to equal or exceed the left, the interventricular septum bows into the left ventricle in systole (the D-shaped left ventricle in the short-axis view), and the contraction becomes hypokinetic — except, classically, at the apex, which is the McConnell sign of regional right-ventricular dysfunction.[7] A dilated right ventricle with A-line lungs and a plethoric fixed IVC completes the obstructive profile; the candidate reads the whole RUSH pattern, not the single chamber.

The inferior vena cava — the volume status

The inferior vena cava is imaged in the subxiphoid long-axis view, a few centimetres from the right atrium where the vessel is straight, with the diameter measured through the respiratory cycle. A small (under 1.5 centimetre) IVC that collapses by more than half with a sniff implies a low volume state and predicts fluid responsiveness; a plethoric, fixed IVC (over 2.5 centimetres, little or no collapse) implies right-heart strain, tamponade, tension pneumothorax or volume overload. The IVC gives a dynamic, repeatable read on the volume status that complements the static vital signs, and the collapsibility index (the change in diameter with respiration) is the bedside surrogate for fluid responsiveness, though its accuracy is modest and it is read alongside the pump and the lungs rather than in isolation.[9]

Equipment and the machine setup

The three examinations are run from a single cart with a curvilinear probe (the low-frequency 2 to 5 megahertz transducer) for the aorta and the deep cardiac views, a phased-array probe for the intercostal and the subxiphoid cardiac windows, and a high-frequency linear probe (5 to 10 megahertz) for the pleural line and the lung sliding. The depth is set to 15 to 18 centimetres for the aorta and the heart so the target fills the screen, and to 4 to 6 centimetres for the lung so the pleural line sits in the near field. The gain is set so the anechoic structures (the aortic lumen, the effusion) are genuinely black and the parenchyma is mid-grey; too much gain fills free fluid with echoes and hides small collections. The cardiac preset is selected for the heart and the abdominal preset for the aorta; M-mode is added for the lung sliding. A warm gel, a clean probe and a probe cover complete the setup, and the probe marker convention is fixed so the images are reproducible. [1]

Patient preparation and consent

The patient is examined supine, fully exposed, with the head of the bed flat for the aortic and the subxiphoid views and slightly raised for the apical window if tolerated; the left lateral decubitus position brings the heart forward for the apical and the parasternal windows in the harder patient. The probe is placed directly on the skin with coupling gel, and the windows are swept to the optimal image. Consent is implicit in the emergency setting — the scan is non-invasive and adds no radiation — and the team proceeds under the doctrine of necessity, explaining the procedure to the conscious patient as it runs. The scan is integrated into the resuscitation: it is performed by one team member while the airway, the breathing, the access and the haemorrhage control proceed in parallel, and it is interrupted and resumed as the patient demands. [1]

Stepwise technique — the RUSH protocol

The RUSH protocol (Rapid Ultrasound in Shock) runs to a fixed sequence, structured as the pump, the tank and the pipes, so that no window is missed under the pressure of the resuscitation.[1]

The structural error is to read each window in isolation — the single effusion called tamponade, the single B-line called oedema — instead of synthesising the pump, the tank and the pipes into the shock profile. The protocol is an algorithm, not a checklist, and the synthesis is the mark. [1]

Drug doses and the resuscitation decisions

The scan does not itself deliver a drug, but every finding drives a defined therapy, and the Fellowship candidate must state the doses the protocol triggers. The flat, collapsing IVC of the hypovolaemic is challenged with a 500 millilitre crystalloid bolus (or 20 millilitres per kilogram in the child) and rescanned; a responsive IVC and a rising pressure confirm volume responsiveness, and the bleeding patient is crossmatched and transfused to a permissive target. The B-line lung of cardiogenic oedema is treated with glyceryl trinitrate (sublingual 300 to 600 micrograms, then an infusion of 10 to 200 micrograms per minute) or with furosemide 40 to 80 milligrams intravenously, and supported with non-invasive ventilation. The dilated right ventricle of massive pulmonary embolism is treated with systemic thrombolysis — alteplase 50 milligrams intravenously over 2 minutes in the arrest, or 100 milligrams over 2 hours in the haemodynamically unstable — or with embolectomy where lytic is contraindicated. The aneurysm with rupture triggers the massive transfusion protocol and the tranexamic acid 1 gram over 10 minutes, on the way to theatre. The tension pneumothorax is decompressed before any scan, and the tamponade is drained. [1]

Complications — the interpretive errors

The examination is non-invasive and carries no procedural complication beyond the theoretical probe pressure. The real complications are interpretive — the wrong disposition driven by a misread scan. The tangential aortic measurement overcalls an aneurysm and sends a stable patient to an unnecessary CT; the single unsampled level misses a saccular aneurysm at an unsampled point. The B-line called pneumonia or the A-line called normal when the lung is dry but the patient has PE loses the diagnosis. The effusion called tamponade without chamber collapse, or the small effusion called harmless when it is under pressure in the trauma patient, drives the wrong drainage decision. The dilated right ventricle of cor pulmonale misread as acute embolism sends the patient to needless thrombolysis. Each is prevented by honest reporting, by correlation with the clinical picture and the other windows, and by repeating the scan as the patient changes. [1]

Pitfalls and practical tips

The pitfalls are the inverse of the discipline. Accepting a single window as the answer — the effusion without the chamber, the B-line without the heart, the aorta without the bifurcation. The tangential aortic cut that overestimates the diameter. Forgetting the lung point is the only specific pneumothorax sign — its absence never excludes one, so a high-clinical-suspicion pneumothorax is decompressed on the clinical picture even with a sliding lung. Confusing the IVC with the aorta in the subxiphoid view — the IVC is to the patient's right, enters the right atrium, and collapses with respiration; the aorta is to the left, is pulsatile and does not. Over-reading the eyeball ejection fraction as precise — it is a gross estimate, and the regional wall-motion abnormality, not the number, drives the ischaemia decision. Missing the apical window in the obese or the bandaged patient and declaring the heart unassessable when the subxiphoid or the parasternal window would have answered. The practical tips are the opposite: scan all the windows every time, measure the aorta outer-to-outer at multiple levels, read the pump-tank-pipes together, label indeterminate honestly, decompress the tension pneumothorax clinically, and treat the patient — not the scan. [1]

Post-procedure care and disposition

The disposition hinges on the synthesised shock profile and the response to the triggered therapy. The ruptured aneurysm goes to theatre or to endovascular repair, the massive transfusion running and the vascular surgeon waiting. The tamponade goes to pericardiocentesis or to theatre. The massive pulmonary embolism goes to thrombolysis or to embolectomy, with the critical-care bed ready. The cardiogenic pulmonary oedema responds to the diuretic and the non-invasive ventilation, and is admitted to the monitored bed for the workup of the cause. The hypovolaemic patient who stabilises with fluid is admitted for the source — the bleeding, the sepsis, the loss. The stable patient with an incidental small aneurysm is referred for surveillance, the interval dictated by the diameter. The scan is repeated serially — at any change in physiology, and at the milestones of the resuscitation — because the sonographic picture is dynamic and the single snapshot is never the whole story. [1]

Special populations

The pregnant patient is scanned with the gravid uterus displacing the viscera and the relative hypervolaemia masking the shock until late; the team tilts the uterus left, and the cardiac and the lung windows remain reliable when the aortic view is distorted. The paediatric patient is examined with the same windows at a shallower depth and a smaller probe, and the IVC collapsibility and the lung patterns hold; the volume bolus is weight-based (20 millilitres per kilogram). The elderly patient carries the comorbidity, the anticoagulation and the diminished reserve that lower the threshold to scan, and in whom the B-lines of heart failure are distinguished from the effusion and consolidation of pneumonia by the bilateral profile and the cardiac function. The obese patient is the technically limited scan, honestly labelled indeterminate where the windows fail, and the chronic lung disease patient is the one in whom the A-lines of COPD and asthma are not mistaken for the normal lung, and in whom the B-lines of cor pulmonale are distinguished from the acute oedema by the chronic right-heart profile. [1]

Evidence and regional guidelines

The diagnostic framework is built on the RUSH protocol of Perera and colleagues, which formalised the pump-tank-pipes approach to the undifferentiated shock patient and showed that clinician-performed bedside ultrasound changed the working diagnosis and the management in a substantial proportion of the critically ill.[1] The lung ultrasound evidence rests on the international consensus of Volpicelli and the International Liaison Committee, which standardised the technique, the terminology (A-lines, B-lines, the lung point) and the eight-zone survey, and on the BLUE protocol of Lichtenstein, which demonstrated that the artefact profiles discriminated pulmonary oedema, pneumothorax, COPD or asthma, and pulmonary embolism with high accuracy in acute respiratory failure.[2][5] The B-line to pulmonary artery occlusion pressure correlation of Lichtenstein showed that the diffuse bilateral B-line profile corresponds to the wedge pressure above 18 millimetres of mercury, and the lung point of Lichtenstein was the sign that made the sonographic diagnosis of pneumothorax specific.[3][4] The abdominal aortic aneurysm diameter that defines the disease and the rupture risk was established by the Multicentre Aneurysm Screening Study (MASS), and the emergency-department aortic measurement dates to the early sonography series of Shuman.[6][10] The right-ventricular regional dysfunction of acute pulmonary embolism — the McConnell sign — was defined by McConnell and colleagues, and the clinician-performed focused echo was validated in the emergency department by Mandavia.[7][8] The IVC collapsibility as a fluid-responsiveness surrogate was synthesised in the meta-analysis of Orso and the International Liaison Committee on Lung Ultrasound.[9]

Exam pearls

• The pump, the tank and the pipes — the three-line summary of the RUSH protocol, and the structure the examiner wants for any "undifferentiated shock" station.
• AAA is an outer-to-outer diameter over 3 cm — measure at the widest point, in the transverse plane, perpendicular to the long axis; a tangential cut overestimates.
• A-lines are normal or COPD; B-lines are pulmonary oedema; absent sliding with a lung point is pneumothorax — the three-line lung summary.
• The lung point is the only specific pneumothorax sign — its absence never excludes one; its presence confirms it.
• Bilateral B-lines with a failing LV is pulmonary oedema; A-line lungs with a dilated RV is pulmonary embolism — the BLUE protocol discriminator.
• A pericardial effusion with RV free-wall collapse in diastole is tamponade — a clinical diagnosis the scan confirms.
• A dilated hypokinetic RV with septal bowing, sparing the apex, is the McConnell sign of massive PE — read it with the A-line lung and the plethoric IVC.
• A flat, collapsing IVC (under 1.5 cm, more than half collapse on sniff) is the empty tank; a plethoric fixed IVC (over 2.5 cm) is the obstructed or overloaded tank.
• Synthesise the three, do not read a single window — the single effusion is not tamponade, the single B-line is not oedema, the single chamber is not PE. [1]

Exam practice

Red flags