Critical care ultrasound (POCUS) in ICU

Overview and rationale

Critical care ultrasound (CCUS) is the bedside use of ultrasound by the intensivist to answer focused, time-critical clinical questions at the point of care. Unlike a departmental study — performed by a sonographer, reported by a radiologist or cardiologist, and delivered hours later — CCUS is performed and interpreted in real time by the clinician responsible for the patient, and the result is immediately folded into the differential diagnosis and the next therapeutic step. For this reason CCUS is best understood not as an imaging modality but as an extension of the clinical examination — the modern equivalent of the stethoscope, used to confirm or refute a hypothesis formed at the bedside.[1][5]

The value of CCUS in the ICU is that it is rapid (a focused exam takes minutes), non-invasive, repeatable (serial exams track change in real time), free of ionising radiation, and transportable to the patient who is too unstable to move. Its limitations are equally important to hold in mind: it is operator-dependent, its accuracy is bounded by image quality (obesity, surgical emphysema, bowel gas, bandages all degrade it), and it is a focused study — it does not substitute for formal echocardiography, computed tomography, or comprehensive departmental ultrasound when those are indicated and the patient is stable enough to obtain them.[1][4]

[1]

The scope of critical care ultrasound

CCUS is organised into five anatomical domains, each answering a defined set of questions, and several integrating protocols that combine domains to work through an undifferentiated presentation (shock, respiratory failure, trauma). The boundaries between a focused study (what the intensivist does at the bedside) and a comprehensive study (what the cardiology or radiology department does) are defined by international consensus and are a frequent exam point.[1][4][7]

[1]

Ultrasound principles — probe selection, frequency and depth

Effective CCUS begins with choosing the right transducer and setting the machine correctly for the question. The wrong probe or the wrong depth makes a straightforward finding invisible.[5]

Probe (transducer) selection

Three transducers cover almost every CCUS application. The choice is governed by the trade-off between penetration (reaching deep structures) and resolution (resolving superficial detail), which in turn is governed by frequency — higher frequency gives better resolution but loses penetration because it is absorbed more rapidly by tissue.[5]

[1]

A practical rule: curvilinear for the abdomen and lung, linear for vessels and superficial structures, phased array for the heart. Many modern machines combine these into a single broadband or switchable transducer, but the principle of matching frequency to depth is unchanged.[5]

Frequency, depth and gain

• Frequency: select the highest frequency that still penetrates to the target. For a carotid or femoral vein 1 cm deep, use the linear probe at its highest frequency for maximal resolution. For an AAA at 12 cm depth, drop to the lowest frequency of the curvilinear probe to penetrate.[5]
• Depth: set the region of interest to fill roughly the middle two-thirds of the screen. Too shallow and you miss the target; too deep and the target is a tiny structure in a sea of tissue with the focal zone in the wrong place.
• Gain (overall brightness): adjust so that the image is not too dark (under-gained, structures missed) or too bright/saturated (over-gained, noise obscures contrast). Use time-gain compensation to even out brightness with depth, since deeper structures return weaker echoes.
• Focal zone: where present, set the focal zone at the depth of the structure of interest, because lateral resolution is best there.

Artefact recognition — what is not real

Ultrasound images are built from assumptions about how sound travels in tissue. Whenever those assumptions are violated (sound bends, reflects repeatedly, or attenuates unevenly), the machine draws structures that are not anatomically there — artefacts. Recognising artefacts is essential both to avoid misdiagnosis (calling an artefact a lesion) and because several artefacts are themselves diagnostic — B-lines, the lung point and the lung pulse are all artefacts that signify disease.[3][12]

[1]

The cardinal rule of artefact interpretation: artefacts are reproducible with probe position, do not have distinct boundaries, and often disappear or change character when the angle or depth is changed — unlike real structures. The exceptions are the lung artefacts (A-lines, B-lines, lung point, lung pulse), which are reproducible precisely because they arise from the physics of the air–tissue interface and are the basis of lung ultrasound diagnosis.[3][12]

Cardiac ultrasound — focused (FoCUS) versus comprehensive

The distinction that defines scope

A focused cardiac ultrasound (FoCUS) is a limited, goal-directed bedside study performed by the non-cardiologist intensivist or emergency physician to answer a small number of yes/no or categorical questions: Is there a pericardial effusion? Is there tamponade physiology? Is global LV systolic function normal, moderately reduced, or severely reduced? Is the RV dilated or hypokinetic? Is the heart small and hyperkinetic (suggesting hypovolaemia) or full? Is there gross valvular pathology (e.g. a flail leaflet, a vegetation, a severely calcified or immobile valve)? FoCUS does not quantify (no ejection fraction by Simpson's biplane, no formal valve gradients, no detailed haemodynamics) and does not replace a comprehensive transthoracic or transoesophageal echocardiogram.[4][7]

A comprehensive echocardiogram is performed by a trained sonographer and reported by a cardiologist; it quantifies chamber sizes and function, measures valve gradients and areas, assesses diastolic function, estimates pulmonary artery pressure, and integrates Doppler haemodynamics. The intensivist must know the boundary: FoCUS is for the unstable patient who needs a decision now; comprehensive echo is for the stable patient who needs a definitive anatomical and haemodynamic answer.[4][7]

[1]

The core FoCUS views

Four views, remembered as the FATE / mnemonic windows, are sufficient for most focused questions. They are acquired in whatever order gives the best image; in the intubated patient the subcostal and apical views are often the most achievable.[4]

1. Parasternal long-axis (PLAX): probe in the 3rd–4th left intercostal space, parasternal line, indicator toward the right shoulder. Shows the RV outflow, LV, mitral and aortic valves, aortic root and descending aorta in cross-section. Best for LV size and function, aortic and mitral valve motion, and a posterior pericardial effusion.
2. Parasternal short-axis (PSAX): rotate 90 degrees from PLAX at the same window. Cross-sections of the LV at mitral, papillary and apical levels; RV crescent wraps around the LV. Septal flattening (D-shaped LV) here = RV pressure or volume overload. Best regional wall motion assessment.
3. Apical four-chamber (A4C): probe at the point of maximal impulse, indicator toward the patient's left. All four chambers in one frame; the natural plane to compare RV and LV size (the RV is normally two-thirds of the LV; if the RV approaches the LV size it is dilated) and to detect an apical effusion.
4. Subcostal / subxiphoid: probe just below the xiphisternum, indicator toward the patient's left, beam aimed at the heart under the liver. The most reliable view for a pericardial effusion (fluid tracks circumferentially) and for assessing the IVC as it enters the right atrium; often the only usable view in the obese, bandaged, or arrested patient. [1]

What FoCUS looks for in shock (the RUSH 'pump')

In the undifferentiated hypotensive patient, the cardiac FoCUS answers questions that map directly onto the type of shock.[4]

• Hyperkinetic, small, kissing LV cavity → hypovolaemic or distributive (septic) shock — the empty, hypercontractile heart.
• Poorly contracting, dilated LV with global hypokinesis → cardiogenic shock.
• Dilated RV with septal shift (D-sign), tricuspid regurgitation, RV hypokinesis (McConnell's sign) → obstructive shock from massive PE or severe pulmonary hypertension.
• Pericardial effusion with RV free-wall diastolic collapse, a distended IVC, and a small, underfilled LV → obstructive shock from cardiac tamponade.
• A dilated, poorly contracting LV with a pericardial effusion but no tamponade physiology may be myocarditis or end-stage cardiomyopathy. [1]

Pericardial effusion and tamponade

FoCUS is highly sensitive for a pericardial effusion, which appears as an anechoic (black) stripe between the visceral and parietal pericardium circumferentially around the heart. The key exam point is that an effusion is not the same as tamponade: tamponade is a clinical and haemodynamic diagnosis, of which the echocardiographic hallmarks are right atrial systolic collapse (early, sensitive) and right ventricular free-wall diastolic collapse (later, more specific), accompanied by a distended, non-collapsing IVC and, on Doppler, exaggerated respiratory variation in inflow. The intensivist's job at the bedside is to integrate the ultrasound with the clinical picture — tachycardia, hypotension, raised JVP, muffled heart sounds (Beck's triad) — and to act: a patient with clinical tamponade and an effusion with RV collapse needs emergency pericardiocentesis (often ultrasound-guided), not a wait for formal echo.[4][7]

Valve assessment — the limit of focused ultrasound

FoCUS can screen for gross valve pathology with 2D and, where available, colour Doppler: a flail or prolapsing mitral leaflet, a vegetation (especially on the aortic or mitral valve in a septic patient), a fixed/stuck leaflet suggesting severe calcific stenosis, or an obviously regurgitant valve on colour. It cannot grade valve disease reliably — stenosis severity requires continuous-wave Doppler to measure gradients and the continuity equation for area, and regurgitation severity requires quantitative Doppler. A vegetation seen on FoCUS is highly suggestive of infective endocarditis and mandates a formal study (and often TOE), but a normal FoCUS does not exclude endocarditis.[4][7]

Lung ultrasound — the BLUE protocol

Lung ultrasound is the domain in which POCUS most dramatically outperforms the bedside chest X-ray and approaches CT. Because air and water interact with ultrasound in predictable, opposite ways, the aerated lung, the oedematous lung, the consolidated lung, the pleural effusion and the pneumothorax each produce a characteristic sonographic signature — and these signatures are the basis of the BLUE protocol (Bedside Lung Ultrasound in Emergency), a decision tree that identifies the cause of acute respiratory failure in minutes.[2][3]

The eight BLUE points

Each hemithorax is examined at standardised points: the upper blue point and lower blue point on the anterior chest (lined up with the hands of BLUE — the operator's two hands placed side by side from the sternum), and the posterolateral alveolar and/or pleural syndrome (PLAPS) point on the lateral/posterior chest. Both lungs are examined, giving a profile that drives the diagnostic algorithm.[2]

The five elemental lung findings

1. Lung sliding (the 'seashore' sign): the to-and-fro movement of the visceral against the parietal pleura with respiration, visible in 2D as a shimmering of the pleural line and in M-mode as a granular 'sandy' pattern below a stratified 'wave' pattern (the seashore sign). Present sliding = the pleural layers are apposed = no pneumothorax at that point.[3]
2. A-lines: horizontal reverberation artefacts below the pleural line at multiples of the skin-to-pleura distance. A-lines + sliding = normal aerated lung (the 'A' profile).[2]
3. B-lines (comet-tail artefacts): vertical, laser-like, hyperechoic artefacts arising from the pleural line, erasing A-lines, spreading to the bottom of the screen without fading, and moving with sliding. Three or more B-lines in a single interspace = interstitial syndrome. Bilateral diffuse B-lines = pulmonary oedema, ARDS, or interstitial fibrosis; focal B-lines = pneumonia, atelectasis, contusion, or infarction.[2][12]
4. Consolidation (alveolar syndrome / C-profile): when alveoli fill with fluid or pus, the lung becomes tissue-like and transmits sound — appearing as a hypoechoic, liver- or spleen-like ('hepatatised') area with echogenic punctate or branching air bronchograms within it. A 'shred sign' marks the irregular deep border of a superficial consolidation.[3]
5. Pleural effusion (PLAPS): an anechoic (or echoic/complex if exudative or haemorrhagic) collection above the diaphragm, with the lung floating in it; the volume can be estimated from the interpleural distance at end-expiration.[3]

The pneumothorax sequence

Pneumothorax is air in the pleural space, so the visceral pleura is separated from the parietal pleura by air — air that reflects all ultrasound and hides the lung beneath. The diagnosis is therefore made by the absence of normal findings and confirmed by one specific positive finding.[9]

The BLUE decision tree

The BLUE protocol combines the lung profile with a venous (DVT) assessment to predict the cause of acute respiratory failure with reported accuracy around 90 per cent.[2]

[1]

The FALLS protocol (Fluid Administration Limited by Lung Sonography) extends this idea to shock: the intensivist gives fluid boluses while watching the lung for the appearance of B-lines, using lung ultrasound as a safety check against iatrogenic pulmonary oedema — the moment B-lines appear, the lung has reached its threshold for interstitial oedema and further fluid is likely harmful.[2]

Abdominal ultrasound — FAST, gallbladder, IVC, bladder, aorta

FAST and eFAST

The Focused Assessment with Sonography in Trauma (FAST) examines four abdominal windows for free (anechoic) intraperitoneal fluid, which in the trauma context is presumed to be blood until proven otherwise:[5]

• Hepatorenal recess (Morison's pouch) — the most dependent supramesocolic space in the supine patient; the most sensitive single view.
• Splenorenal recess — the left upper quadrant mirror of Morison's pouch.
• Suprapubic / pouch of Douglas — the most dependent space in the pelvis; views the bladder (which should be full, acting as an acoustic window) and the rectovesical/rectouterine space behind it.
• Subxiphoid / pericardial — the cardiac view for haemopericardium (an effusion in trauma is tamponade until proven otherwise). [1]

The extended FAST (eFAST) adds lung views to detect a pneumothorax (often missed on supine chest X-ray) and, in some protocols, a more detailed aortic assessment. A positive FAST in a haemodynamically unstable trauma patient is an indication for immediate operative intervention; a negative FAST in an unstable patient does not exclude injury and warrants further assessment (CT if stable enough, or repeat/dynamic FAST). Outside trauma, the same FAST views detect free fluid from ruptured ectopic pregnancy, ruptured abdominal aortic aneurysm, haemorrhagic pancreatitis, ascites, and bowel perforation.[5]

The gallbladder

Bedside assessment of the gallbladder identifies stones (echogenic foci that cast clean acoustic shadows and move with position), wall thickening (greater than 3 mm, measured anterior wall), pericholecystic fluid, and a sonographic Murphy's sign (maximal tenderness as the probe presses directly over the sonographically localised gallbladder). These are the components of acute calculous cholecystitis and are especially relevant in the critically ill patient at risk of acalculous cholecystitis (prolonged fasting, TPN, sepsis, burns), where the finding is a distended, thick-walled gallbladder with sludge and pericholecystic fluid in the absence of stones.[5]

The inferior vena cava for volume and fluid responsiveness

The IVC is examined in the subcostal longitudinal plane as it enters the right atrium, in M-mode just below the hepatic vein inflow. Its absolute diameter and its respiratory variation (collapsibility) are used as surrogates for right atrial pressure and, with important caveats, for fluid responsiveness.[8]

[1]

The IVC is a better bedside guide than the central venous pressure, but inferior to a passive leg raise (which is itself a reversible, endogenous fluid challenge) and to dynamic indices (stroke volume variation, pulse pressure variation) for predicting fluid responsiveness. It is best used as one data point, not the sole determinant of fluid therapy.[8]

Bladder volume

Bladder volume is estimated by measuring the bladder in two perpendicular transverse and sagittal planes and applying the prolate ellipsoid formula (volume = length × width × height × 0.523), or by the machine's automated calculation. It is used to measure post-void residual (a residual over 100–150 mL suggests urinary retention and may warrant catheterisation) and to confirm a full bladder before suprapubic catheterisation or a suprapubic FAST window.[5]

Abdominal aortic aneurysm screening

The abdominal aorta is measured in the transverse plane from the outer wall to outer wall at the suprarenal (diaphragmatic hiatus), renal, and infrarenal levels. A diameter greater than 3 cm defines an aneurysm; greater than 5.5 cm marks the threshold at which elective repair is usually offered because rupture risk exceeds operative risk. In the unstable patient with abdominal/back pain and a pulsatile mass, finding an aortic diameter greater than 3 cm (especially with retroperitoneal fluid) supports the diagnosis of a ruptured AAA and mandates immediate vascular surgery involvement.[5]

Vascular ultrasound — DVT compression and CVC guidance

Compression sonography for deep vein thrombosis

Compression sonography exploits the fact that a normal vein is fully compressible under light probe pressure (its walls coapt and it disappears), whereas a vein containing thrombus is non-compressible. The test is therefore one of compressibility, not of directly seeing clot.[10][11]

[1]

A non-compressible common femoral or popliteal vein confirms proximal DVT with specificity around 96 per cent and warrants anticoagulation. In the ICU, where DVT prevalence is high and compression is rapid, a positive compression study in the right clinical context is sufficient to start treatment without waiting for a formal departmental study.[10][11]

The DVT assessment is also the final step of the BLUE protocol: an A-profile (normal lung) combined with a positive compression study strongly suggests pulmonary embolism as the cause of respiratory failure — a haemodynamically significant PE is, in effect, diagnosed by finding its source clot in the leg.[2]

Ultrasound-guided central venous catheterisation

Ultrasound guidance for central venous catheter (CVC) insertion is now standard of care for the internal jugular, subclavian, and femoral routes. It is used in two ways: static (pre-procedural mapping — locate the vein, confirm patency, mark the site, then proceed 'blindly') and dynamic (real-time visualisation of the needle entering the vein throughout the procedure). Real-time dynamic guidance is preferred.[6]

The evidence is consistent: ultrasound guidance reduces mechanical complications (arterial puncture, haematoma, pneumothorax, haemothorax), increases first-pass and overall success, and reduces the number of attempts, with the benefit largest for the internal jugular vein. The Cochrane review by Brass and colleagues confirmed this benefit for subclavian and femoral routes as well.[6]

Competency, training and certification

Because CCUS is operator-dependent, international consensus statements define what a competent practitioner must be able to do, and how that competency is acquired and assessed. The defining documents are the American College of Chest Physicians / Société de Réanimation de Langue Française (ACCP/SRLF) competence statement, the international evidence-based recommendations for point-of-care lung ultrasound (Volpicelli et al, WINFOCUS-affiliated, Intensive Care Medicine 2012), the international evidence-based recommendations for focused cardiac ultrasound (Via et al, 2014), and the European Association of Cardiovascular Imaging focus cardiac ultrasound core curriculum and syllabus (Neskovic et al, 2018).[1][3][4][7]

These consensus documents converge on a tiered competency model: a basic level (recognising the common life-threatening findings — effusion/tamponade, gross LV dysfunction, RV dilatation, B-lines, absent lung sliding, free intraperitoneal fluid, a non-compressible vein, ultrasound-guided vascular access) that all intensivists and emergency physicians should achieve, and an advanced level (comprehensive echocardiography, advanced lung and abdominal protocols) reserved for those with dedicated training. The boundary is emphasised repeatedly: basic competency is not a licence to call oneself an echocardiographer.[1][7]

Named certifications and pathways include: [1]

• FATE (Focused Assessed Transthoracic Echocardiography) — the Scandinavian protocol (Jensen and colleagues), widely taught and certified internationally; emphasises the four core views plus IVC and lung, for haemodynamic assessment.
• FoCCUS / CEUS (Comprehensive Focused Cardiac Ultrasound via the Intercollegiate Committee for Critical Care Echocardiography, UK) — the UK ICS/FICM pathway to focused echo competency and accreditation.
• FUSIC (Focused Ultrasound in Intensive Care) — the UK Intensive Care Society's one-stop, multi-domain POCUS accreditation (heart, lung, abdomen, vascular, procedural).
• ACCP / CHEST critical care ultrasonography — the North American College of Chest Physicians competency framework and certification.
• WINFOCUS (World Interactive Network Focused on Critical Ultrasound) — the international society that has driven the evidence-based recommendations and runs training and consensus globally.
• ACEP emergency ultrasound — the American College of Emergency Physicians guidelines and credentialing for point-of-care ultrasound in the emergency department. [1]

Competency is acquired through a combination of structured courses, supervised scanning, logbook accumulation, and assessment (often image-acquisition and interpretation reviews). The consensus statements emphasise that competency requires both the technical skill to acquire diagnostic images and the cognitive skill to interpret them and integrate them into clinical decisions — and that both must be maintained through continued practice and quality assurance.[1][7]

POCUS versus formal imaging

A recurring exam theme is when POCUS is sufficient and when it must defer to formal imaging. The principle is that POCUS answers a focused question in real time to guide immediate therapy; formal imaging answers the full question definitively. POCUS does not replace a comprehensive echocardiogram in the stable patient with suspected valve disease, nor a CT pulmonary angiogram in the haemodynamically stable patient with suspected PE, nor a departmental vascular or abdominal study when the question is anatomical and not time-critical.[1][4]

[1]

The correct relationship is sequential and complementary: POCUS makes the immediate decision (start fluids? pericardiocentesis? chest tube? anticoagulate? intubate?); formal imaging, when the patient is stabilised, makes the definitive diagnosis and plans definitive treatment. [1]

[1] [1]

Clinical pearls

Red flags

Key trials and evidence

[1] [1] [1] [1] [1] [1] [1] [1] [1] [1] [1] [1]

Limitations and pitfalls

CCUS is powerful precisely because it is fast and repeatable, but its limitations are the mirror image of its strengths and must be held in mind whenever the probe is picked up.[1][5]

• Operator dependence. Every CCUS finding depends on the operator's ability to acquire and interpret the image. Inadequate training degrades both sensitivity and specificity, and a confident but wrong reading can be more dangerous than no reading at all.
• Image-quality barriers. Obesity, surgical emphysema (air reflects all ultrasound), bandages and dressings, poor acoustic windows, bowel gas (abdomen), and patient inability to cooperate (agitation, position) all reduce image quality. A non-diagnostic study must be reported as non-diagnostic.
• Focused, not comprehensive. CCUS answers the question asked; it does not exclude pathology outside that question. A negative FoCUS does not exclude endocarditis; a normal lung scan does not exclude a parenchymal nodule; a negative FAST does not exclude hollow-viscus or retroperitoneal injury.
• Artefact misinterpretation. A-lines, mirror artefact, side-lobe artefact, and shadowing can all masquerade as pathology. The operator must recognise the common artefacts and, where in doubt, change probe angle/depth to test whether the finding persists.
• The hard questions stay with formal imaging. Valve gradients and areas, ejection fraction quantification, pulmonary artery pressure, complex anatomy (CT), and tissue characterisation (CT/MRI) remain the province of formal studies in the stable patient.
• Documentation and quality assurance. Findings should be recorded (images saved, a focused report written) and reviewed, both for clinical communication and for the maintenance and improvement of competence over time.[1][7]

Putting it together — CCUS in the undifferentiated ICU patient

In practice, the intensivist does not perform 'a lung scan' or 'a cardiac scan' in isolation but integrates domains to work through the presentation. In the undifferentiated hypotensive patient, RUSH sequences the cardiac (pump), IVC (tank volume), aorta and venous (tank integrity), and lung/pleura (source and effect) assessments. In the undifferentiated breathless patient, BLUE sequences the lung profile with a venous assessment. In the trauma patient, eFAST adds lung to the abdominal and cardiac windows. The unifying habit is to form a focused question, choose the right probe and domain, acquire and interpret the image at the bedside, integrate it with the clinical picture, act, and then repeat to track the response — and to know when the question has outgrown POCUS and must be handed to formal imaging.[1][2][4][5]

[1]

CCUS is, in the end, a clinical skill: its value lies not in the image on the screen but in the decision it enables at the bedside. Performed within its scope, by a competent operator, and integrated with the clinical examination, it shortens the time to the right therapy in the ICU's most unstable patients — and, just as importantly, it tells the intensivist when to stop and obtain the definitive study. [1]

References

1. [1]Mayo PH, Beaulieu Y, Doelken P, et al. American College of Chest Physicians/La Société de Réanimation de Langue Française statement on competence in critical care ultrasonography Chest, 2009.PMID 19188546
2. [2]Lichtenstein DA, Mezière GA Relevance of lung ultrasound in the diagnosis of acute respiratory failure: the BLUE protocol Chest, 2008.PMID 18403664
3. [3]Volpicelli G, Elbarbary M, Blaivas M, et al. International evidence-based recommendations for point-of-care lung ultrasound Intensive Care Med, 2012.PMID 22392031
4. [4]Via G, Hussain A, Wells M, et al. International evidence-based recommendations for focused cardiac ultrasound J Am Soc Echocardiogr, 2014.PMID 24951446
5. [5]Moore CL, Copel JA Point-of-care ultrasonography N Engl J Med, 2011.PMID 21345104
6. [6]Brass P, Hellmich M, Kolodziej L, et al. Ultrasound guidance versus anatomical landmarks for subclavian or femoral vein catheterization Cochrane Database Syst Rev, 2015.PMID 25575245
7. [7]Neskovic AN, Skinner H, Price S, et al. Focus cardiac ultrasound core curriculum and core syllabus of the European Association of Cardiovascular Imaging Eur Heart J Cardiovasc Imaging, 2018.PMID 29529170
8. [8]Barbier C, Loubières Y, Schmit C, et al. Respiratory changes in inferior vena cava diameter are helpful in predicting fluid responsiveness in ventilated septic patients Intensive Care Med, 2004.PMID 15034650
9. [9]Volpicelli G, Boero E, Sverzellati N, et al. Semi-quantification of pneumothorax volume by lung ultrasound Intensive Care Med, 2014.PMID 25056671
10. [10]Heijboer H, Büller HR, Lensing AW, et al. A comparison of real-time compression ultrasonography with impedance plethysmography for the diagnosis of deep-vein thrombosis in symptomatic outpatients N Engl J Med, 1993.PMID 8413431
11. [11]Cogo A, Lensing AW, Koopman MM, et al. Compression ultrasonography for diagnostic management of patients with clinically suspected deep vein thrombosis: prospective cohort study BMJ, 1998.PMID 9451260
12. [12]Lichtenstein DA Lung ultrasound in the critically ill Ann Intensive Care, 2014.PMID 24401163