Bedside echocardiography in the assessment of the ... Echo (Mar-27-08).pdfwith suboptimal transthoracic imaging, performance of a TEE will allow for a more precise evaluation of ventricular

Bedside echocardiography in the assessment of the critically ill

Yanick Beaulieu, MD, FRCPC

Echocardiography can nonin-vasively provide diagnostic in-formation regarding cardiacstructure and mechanical

function. The supplementary informationprovided by this technique can help de-termine the cause of hypotension refrac-tory to inotropic support or vasopressorinfusions (1). It can also help in the di-agnosis of a wide spectrum of other car-diovascular abnormalities and guide ther-apeutic management. An adequateunderstanding of the proper use of echo-cardiography is thus a prerequisite forthe intensivist. General indications forperformance of an echocardiographic ex-amination in the intensive care unit(ICU) are listed in Table 1.

TRANSTHORACIC VS.TRANSESOPHAGEALECHOCARDIOGRAPHY IN THECRITICALLY ILL PATIENT

Accurate and prompt diagnosis is cru-cial in the ICU. The easiest and least in-vasive way to image cardiac structures isechocardiography using the transtho-

racic approach (1). This noninvasive im-aging modality is of great value in thecritical care setting because of its porta-bility, widespread availability, and rapiddiagnostic capability. In the ICU, trans-thoracic echocardiography (TTE) may, incertain cases, fail to provide adequate im-age quality because of different factorsthat can potentially hinder the quality ofthe ultrasound signal, be it air, bone,calcium, a foreign body, or any other typeof interposed structure. The failure rate(partial or complete) of the transthoracicapproach in the ICU setting has beenreported to be between 30% and 40% (2,3). However, improvements have beenmade in transthoracic imaging (e.g., har-monics and contrast and digital technol-ogies), resulting in a lower failure rate ofTTE in the ICU (10–15% in our institu-tion; unreported data).

Transesophageal echocardiography(TEE) is particularly useful for evaluationof suspected aortic dissection, prostheticheart valves (especially in the mitral po-sition), source of cardiac emboli, valvularvegetations, possible intracardiac shunts,and unexplained hypotension. This mo-dality allows better visualization of theheart in general and especially the poste-rior structures, owing to the proximity ofthe probe and favorable acoustic trans-mission (4). As a result of the signifi-cantly improved technical quality of TTEimaging, the majority of ICU patients canbe satisfactorily studied with this modal-ity. In a recent study by Joseph et al. (5),

bedside TTE imaging identified the greatmajority of cardiac causes of shock in ageneral critical care population of pa-tients (excluding cardiac surgery pa-tients). TTE image quality was adequatein 99% of cases. The authors concludedthat TTE should be considered not onlythe initial but also the principal echocar-diographic test in the critical care envi-ronment. However, immediate TEE isstill preferable in certain specific clinicalsituations in which TTE is likely to fail orbe suboptimal (3). Even when TEE isnecessary, data from the TTE examina-tion is often essential for the final clinicalinterpretation. The major indications forprimary TEE in the ICU (6, 7) are listed inTable 2. The most common transthoracicacoustic windows used for performanceof a goal-directed cardiac ultrasound ex-amination are illustrated in Figure 1.

HEMODYNAMIC EVALUATION

Ventricular FunctionLeft Ventricular Systolic Function.

Accurate and timely assessment of sys-tolic function should be an integral partof the medical management of hemody-namically unstable critically ill patients.Global ventricular function will often bequalitatively assessed by visual inspectionalone. This method has been found to bevery reliable when used by experiencedclinicians (8). Real-time visualization ofthe kinetics and size of the cardiac cavi-ties by an experienced critical care inten-sivist with sufficient echocardiographic

From the Hôpital Sacré-Coeur de Montréal, Uni-versité de Montréal, Montréal, Québec, Canada.

The author has not disclosed any potential con-flicts of interest.

For information regarding this article, E-mail:[email protected].

Copyright © 2007 by the Society of Critical CareMedicine and Lippincott Williams & Wilkins

DOI: 10.1097/01.CCM.0000260673.66681.AF

Advances in ultrasound technology continue to enhance its diag-nostic applications in daily medical practice. Bedside echocardio-graphic examination has become useful to properly trained cardiol-ogists, anesthesiologists, intensivists, surgeons, and emergencyroom physicians. Cardiac ultrasound can permit rapid, accurate, andnoninvasive diagnosis of a broad range of acute cardiovascularpathologies. Although transesophageal echocardiography was oncethe principal diagnostic approach using ultrasound to evaluate in-tensive care unit patients, advances in ultrasound imaging, includingharmonic imaging, digital acquisition, and contrast for endocardialenhancement, has improved the diagnostic yield of transthoracic

echocardiography. Ultrasound devices continue to become moreportable, and hand-carried devices are now readily available forbedside applications. This article discusses the application of bed-side echocardiography in the intensive care unit. The emphasis is onechocardiography and cardiovascular diagnostics, specifically ongoal-directed bedside cardiac ultrasonography. (Crit Care Med 2007;35[Suppl.]:S235–S249)

KEY WORDS: bedside ultrasonography; hand-carried ultrasound;transthoracic echocardiography; transesophageal echocardiogra-phy; cardiac function

S235Crit Care Med 2007 Vol. 35, No. 5 (Suppl.)

background will allow an immediatefunctional diagnosis.

If the TTE examination is technicallydifficult and the endocardium is poorlyvisualized, harmonic imaging and possi-bly contrast, if needed, can dramaticallyimprove endocardial border visualizationand subsequent evaluation of global sys-tolic function (9–13). For the remainingminority of technically challenging caseswith suboptimal transthoracic imaging,performance of a TEE will allow for amore precise evaluation of ventricularfunction in most critically ill patients be-cause of the higher image quality thatcan be obtained with this echographicmodality.

Left Ventricular Failure in the ICU.Clinical examination and invasive hemo-dynamic monitoring often fail to providean adequate assessment of ventricularfunction in the ICU setting. Assessmentof biventricular function is thus one ofthe most important indications for per-formance of an echographic study in theICU. In a study by Bruch et al. (14), 115critically ill patients were studied by TEE.The most common indication for TEEstudy was hemodynamic instability (67%of patients). Of these hemodynamicallyunstable patients, 20 (26%) were found tohave significant left ventricular (LV) dys-function (LV ejection fraction [EF] of�30%). In a study by Vignon et al. (15),TTE allowed adequate evaluation ofglobal LV function in 77% of mechani-cally ventilated ICU patients. AlthoughTEE was needed for most other indica-tions, TTE was shown to be an excellentdiagnostic tool for assessment of LV func-tion in the ICU, even when positive end-expiratory pressure was present.

Several important points should be em-phasized: 1) significant LV dysfunction iscommon in critically ill patients; 2) ventric-ular function should be assessed in all pa-tients with unexplained hemodynamic in-

stability, as this information is particularlyimportant for guiding resuscitation and in-forming decisions management; 3) it isnow possible to obtain adequate informa-tion about ventricular function in most ICUpatients using TTE, but TEE provides bet-ter accuracy in patients with suboptimalimaging by TTE.

Sepsis-Related Cardiomyopathy. Clas-sically, septic shock has been considereda hyperdynamic state characterized bynormal or high cardiac output (CO). But

echocardiographic studies indicate thatventricular performance is often mark-edly impaired in patients with sepsis (16,17). Parker et al. (18) were the first todescribe LV hypokinesis in septic shock.They reported that survivors manifestedseverely depressed LVEF but that ade-quate LV stroke output was maintained asa result of acute LV dilation (19). LVEFmight not be a reliable index of LV sys-tolic function in patients with early septicshock, as this is a state characterized by

Table 1. General indications for performance ofan echocardiographic examination in the inten-sive care unit

Hemodynamic instabilityVentricular failureHypovolemiaPulmonary embolismAcute valvular dysfunctionCardiac tamponadeComplications after cardiothoracic surgery

Infective endocarditisAortic dissection and ruptureUnexplained hypoxemiaSource of embolus

Figure 1. Illustrated above are the most common transthoracic acoustic windows (and correspondingechocardiographic images) used for performance of goal-directed cardiac ultrasound examination.Parasternal long-axis (A) and short-axis (B) views; apical four-chamber view (C); subcostal four-chamber view (D). AV, aortic valve; Desc Ao, descending thoracic aorta; LA, left atrium; LV, leftventricle; RA, right atrium; RV, right ventricle.

Table 2. Major indications for performance of a primary transesophageal echocardiographic study inthe intensive care unit

Diagnosis of conditions in which a high image quality is vitalAortic dissectionAssessment of endocarditisIntracardiac thrombus

Imaging of structures that may be inadequately seen by transthoracic echocardiographyThoracic aortaLeft atrial appendageProsthetic valves

Echocardiographic examinations of patients with conditions that prevent image clarity withtransthoracic echocardiography

Severe obesityEmphysemaMechanical ventilation with high-level positive end-expiratory pressurePresence of surgical drains, surgical incisions, dressings

Acute perioperative hemodynamic derangements

S236 Crit Care Med 2007 Vol. 35, No. 5 (Suppl.)

low systemic vascular resistance that un-loads the left ventricle (16). Therefore,normal or supranormal EF in early sepsismight lead clinicians to make the wronginference about cardiac reserve becauseLVEF might decrease if afterload is in-creased by the administration of vaso-pressor agents.

In the septic patient, bedside echocar-diography is valuable for identification ofthe cause of hemodynamic instability(which may be of hypovolemic, cardio-genic, or distributive origin) and for thesubsequent optimization of therapy (i.e.,fluid administration, inotropic or vaso-constrictor agent infusion, or variouscombinations of the above) (20). The abil-ity to perform repeat bedside examinationis vital in assessing the adequacy andefficacy of therapy (20).

Cardiac Arrest

In patients presenting with cardiac ar-rest (either from in-hospital or out-of-hospital cardiac arrest), the advanced car-diac life support algorithm should alwaysbe rigorously followed, and assessment ofairway-breathing-circulation and huntfor defibrillation must be aggressivelypursued (an A-B-C-D sequence). Whenassessing for the presence or absence ofsigns of circulation in such patients, pe-ripheral pulses are usually taken. The ul-timate goal of pulse assessment is to de-tect the presence of an underlying cardiacactivity and the associated CO generated.But there are situations in which a pulseis absent, despite the presence of a car-diac rhythm on the monitor. These situ-ations are typically called pulseless elec-trical activity and are often equated withan electromechanical dissociation (EMD)condition. Not infrequently, when an ur-gent bedside echocardiographic study isperformed in patients who are thought tohave EMD, many of these cases are foundto have some degree of cardiac activityand thus present pseudo-EMD and notfull-blown EMD. Making the diagnosis ofpseudo-EMD in such acutely sick patientscan be of tremendous diagnostic andprognostic importance because patientsin cardiac arrest who are found to have aresidual cardiac function (varying fromsevere dysfunction as seen in cases ofacute myocardial infarction to hyperdy-namic cardiac activity as seen in cases ofextreme volume depletion), have a betterprognosis than patients who are in trueEMD (21). Echocardiography can also beused for confirmation of asystole and

even ventricular fibrillation in patients inwhom the cardiac monitor may seem un-reliable or difficult to assess.

The optimal resuscitation sequence tofollow in a code situation should thusbecome A (airway), B (breathing), C (cir-culation), D (defibrillation), and E (forgoal-directed bedside echocardiography).

Bedside echocardiography can thus beof tremendous help in the assessment of“circulation” in patients presenting withcardiac arrest (22). Despite the usefulnessof echocardiography in such acute situa-tions, there exists no clear recommenda-tions on how to use the information ob-tained from a goal-directed cardiacexamination during a code. It is not yetclear how, when, and which informationshould be used in such situation to con-tinue or terminate resuscitation maneu-vers.

LV Diastolic Function. In the ICU, di-astolic dysfunction should be suspectedwhen ventricular filling pressure (pulmo-nary artery occlusion pressure) is ele-vated and EF is normal or supranormal(4). The filling patterns related to theintrinsic diastolic properties of the myo-cardium are influenced by many differentfactors, particularly left atrial pressure,heart rate, ischemia, ventricular hyper-trophy, and valvular pathologies. Onlymodest correlation has been found be-tween Doppler indices of diastolic func-tion and variables measured using moreinvasive means (23). Interpretation of di-astolic function must be done with cau-tion when caring for critically ill patients,given the many different factors that canacutely influence flow patterns in thispopulation of patients (24).

Right Ventricular Function andVentricular Interaction

In the critical care setting, right ven-tricular (RV) function can be altered bymassive pulmonary embolism (PE) andacute respiratory distress syndrome, thetwo main causes of acute cor pulmonalein adults (25–28). Any other perturba-tions that increase RV afterload, such aspositive end-expiratory pressure or in-creased pulmonary vascular resistance(from vascular, cardiac, metabolic, orpulmonary causes), will also have a sig-nificant effect on RV function. DepressedRV systolic function is also often associ-ated with RV infarction, most commonlyin the setting of inferior myocardial in-farction. Acute sickle-cell crisis, air or fatembolism, myocardial contusion, and

sepsis are other causes of acute RV dys-function.

In unstable critically ill patients, spe-cifically those with massive PE and acuterespiratory distress syndrome, a diagno-sis of concomitant significant RV dys-function may alter therapy (e.g., fluidloading, use of vasopressors, use ofthrombolytics) and provide informationabout prognosis (28, 29). Echocardio-graphic examination of the right ventri-cle requires primarily an assessment ofthe size and kinetics of the cavity andseptum (30, 31). RV size and functiongenerally are evaluated by visual compar-ison with the left ventricle. RV diastolicdimensions can be obtained by measur-ing RV end-diastolic area from an apicalfour-chamber view, using either TTE orTEE. Because pericardial constraint nec-essarily results in LV restriction when theright ventricle acutely dilates (i.e., thereis ventricular interaction), one of the bestways to quantify RV dilation is to measurethe ratio between the RV and LV end-diastolic areas, an approach that cancelsout individual variations in cardiac size(30, 31). Moderate RV dilation corre-sponds to a diastolic ventricular ratio of0.6–1.0; severe RV dilation correspondsto a ratio �1 (30, 31). RV diastolic en-largement is usually associated with rightatrial dilation, inferior vena caval dila-tion, and tricuspid regurgitation. Whenpressure in the right atrium exceeds pres-sure in the left atrium, the foramen ovalemay open. Pressure and volume overloadof the right ventricle can lead to distor-tion of LV geometry and abnormal mo-tion of the interventricular septum. Withconditions of high strain imposed on theRV (volume or pressure overload), the in-terventricular septum flattens and the LVappears to have a D shape (30, 31). This“paradoxic” septum motion will also beseen at the interatrial level.

Pulmonary Embolism. Hemodynamicinstability from acute cor pulmonale as aconsequence of massive PE is a relativelycommon occurrence in critically ill pa-tients. Echocardiography is well suitedfor diagnosis of PE because it can be donewithin minutes at the bedside. The diag-nosis of acute cor pulmonale at the bed-side with TTE has good positive predic-tive value for the indirect diagnosis ofmassive PE (32, 33). The diagnosis is in-direct in the sense that, in most situa-tions, it is the acute RV dilation and dys-function resulting from a large PE that isvisualized and not the emboli itself (sel-dom seen). Thus, it is important to stress


that echocardiography may not be sensi-tive enough for smaller PEs and that in asituation in which the clinical suspicionof a PE is moderate to high, one must notexclude PE solely based on a normal RVsize and function on echocardiography.The finding of RV dilation and dysfunc-tion is not specific for PE, as these find-ings may be observed with a variety ofother conditions associated with increasedRV strain. In a study by McConnel et al.(34), patients with acute PE were foundto have a distinct regional pattern of RVdysfunction, with akinesia of the mid-freewall but normal motion at the apex byTTE. These findings contrasted withthose obtained in patients with primarypulmonary hypertension, who had abnor-mal wall motion in all regions. RegionalRV dysfunction had a sensitivity of 77%and a specificity of 94% for the diagnosisof acute PE; positive predictive value was71% and negative predictive value was96%. The presence of regional RV dys-function that spares the apex should raisethe level of clinical suspicion for the di-agnosis of acute PE.

Central pulmonary emboli are presentin half of patients with symptoms of PEand acute cor pulmonale on TTE (35).Emboli lodged in the proximal pulmo-nary arteries usually cannot be visualizedusing TTE (35). As other clinical condi-tions can produce acute cor pulmonale inthe ICU, better visualization of the pul-monary arteries is needed to achieve highaccuracy for the diagnosis of PE. Thisgoal can be achieved by using TEE. TEEhas a good sensitivity for detecting em-boli that are lodged in the main and rightpulmonary arteries but is limited for thedetection of more distal or left pulmonaryemboli (35–37). If an embolus is visual-ized, the diagnosis is made, but if thestudy is negative when the index of sus-picion for PE is high, then TEE must befollowed up by a more definitive test,such as angiography or helical computedtomography. Also, when there is highclinical suspicion for PE but no emboliare visualized using TEE, the potentialfor nonthrombotic causes of PE, such asair or fat emboli, must be kept in mind.

The demonstration of acute cor pul-monale with echocardiography has im-portant prognostic and therapeutic impli-cations (38 – 41). The presence of corpulmonale with massive PE is associatedwith increased mortality, whereas the ab-sence of RV dysfunction is associated witha better prognosis (29).

Assessment of CO

Measurement of CO remains a corner-stone in the hemodynamic assessment ofcritically ill patients. Several methods fordetermining CO have been described us-ing both two-dimensional and Dopplerechocardiography (42– 45). With thistechnique, stroke volume and CO can bedetermined directly by combining Dop-pler-derived measurements of instanta-neous blood flow velocity through a con-duit with the cross-sectional area of theconduit. Of these methods, the one usingthe left ventricular outflow tract and aor-tic valve as the conduit is probably themost reliable and most commonly used.There is excellent agreement with ther-modilution in most situations (45–49).

Another ultrasound-based technologyto noninvasively estimate CO in adultsuses a small transesophageal Dopplerprobe to measure blood flow velocitywaveforms in the descending aorta com-bined with a nomogram (based on height,weight, and age) for estimation of aorticcross-sectional area. This minimally inva-sive esophageal probe can be inserted eas-ily in sedated patients and left in placesafely for several days to provide contin-uous monitoring of cardiac function (50,51). However, several technical problemscan limit the accuracy of CO measure-ments by esophageal Doppler monitoring(50), and although initial results arepromising (52–54), more studies areneeded to make a decision regarding theaccuracy of this technique in critically illpatients.

Assessment of Filling Pressuresand Volume Status

Adequate determination of preloadand volume status is important for propermanagement of critically ill patients. In-vasive pressure measurements to assessLV filling are commonly used at the bed-side to make inferences regarding LV pre-load. These pressure measurements,however, only weakly correlate with LVvolume (55). Data from invasive monitor-ing using a pulmonary artery catheter(PAC) may be misleading because ven-tricular compliance is altered by numer-ous factors (56, 57). Differences in dia-stolic compliance among patients mayaccount for the weak correlation betweenpressure and volume and may limit theability to use pressure measurementsalone to derive information concerningLV preload (58). Echocardiography can

be of great help for adequately assessingpreload. Variables that can be measuredusing two-dimensional imaging are LVend-diastolic volume and LV end-dia-stolic area (EDA). Using Doppler interro-gation, additional information, mainlytransmitral diastolic filling pattern andpulmonary venous flow, can be obtained.

Two-Dimensional Imaging. Echocar-diography has been validated for LV vol-ume measurements (59). Subjective as-sessment of LV volume by estimating thesize of the LV cavity in the short- andlong-axis views is often adequate to guidefluid volume therapy at the extreme endsof cardiac filling and function, but moreprecise, quantitative values are desirableand can be obtained by tracing the innercontour of the endocardium of the LVcavity (endocardial border tracing).LVEDA measured in the left parasternalshort-axis view at the level of the mid-papillary muscle is commonly used toestimate volume status. Two-dimensionalTTE evaluation of ventricular dimensionshas been found to be useful in assessingpreload and in optimizing therapy of ICUpatients (16, 60). Nevertheless, imagequality may be suboptimal and precludeadequate visualization of the endocardialborder by TTE. This potential limitationof TTE has partly been circumvented inrecent years with the advent of harmonicimaging and contrast echocardiography,but in cases in which endocardial bordervisualization remains suboptimal, TEE isthe modality of choice. With TEE, LVvolume can be rapidly estimated by sub-jective assessment of the LV size. Quan-titatively, it is most often estimated bydetermining LV cross-sectional area atthe end of diastole, most commonly us-ing the transgastric short-axis view at thelevel of the mid-papillary muscle. Thissection is used because of the reproduc-ibility of the view and because changes inLV volume affect the short axis of theventricle to a greater degree than thelong axis (58). The EDA must be mea-sured consistently from the same refer-ence section. EDA measured with TEEcorrelates with LV volume determined byradionuclide studies (60).

Systolic obliteration (dynamic ob-struction) of the LV cavity accompaniesdecreased EDA and is considered to be asign of severe hypovolemia. Although asmall EDA generally indicates hypovole-mia, a large EDA does not necessarilyindicate adequate preload in patients withLV dysfunction. Also, when systemic vas-cular resistance is low, as in early sepsis,


LV emptying is improved because of thelowered afterload. In these situations, itmay be difficult to differentiate hypovole-mia from low systemic vascular resis-tance by echocardiography alone, as bothconditions are associated with decreasedEDA.

Knowledge of LV end-diastolic volumeor absolute preload does not necessarilyallow for accurate prediction of the hemo-dynamic response to alterations in preload(61). Tousignant et al. (62) investigated therelationship between LV stroke volume andLVEDA in a cohort of ICU patients andfound only a modest correlation (r � .60)between single-point estimates of LVEDAand responses to fluid loading. Based onthe assumption that changes in EDA occurbecause of changes in LV volume, the de-termination of this area and its subsequentdegree of variation after a fluid challengecould help better assess preload responsive-ness. Studies have demonstrated thatchanges in EDA measured by TEE usingendocardial border tracing are closely re-lated to changes in CO and are superior tomeasurements of pulmonary artery occlu-sion pressure for predicting the ventricularpreload associated with maximum CO (63).

Circulating volume status also can beassessed by two-dimensional echocardi-ography by indirectly estimating rightatrial pressure. This is often done by as-sessing the diameter and change in cali-ber with inspiration of the inferior venacava. This method has been shown todiscriminate reliably between right atrialpressures of �10 or �10 mm Hg (64). Adilated vena cava (diameter of �20 mm)without a normal inspiratory decrease incaliber (�50% with gentle sniffing) usu-ally indicates elevated right atrial pres-sure. In mechanically ventilated patients,this measure is less specific because of ahigh prevalence of inferior vena cava di-lation (65–67). A small vena cava reliablyexcludes the presence of elevated rightatrial pressure in these patients (65–67).

Variation of the diameter of the infe-rior vena cava with respiration (Fig. 2)has also recently been demonstrated to bea reliable guide to fluid therapy. Feissel etal. (68) studied 39 patients on mechanicalventilation with septic shock in whomthey assessed CO and change in inferiorvena cava diameter (by echocardiogra-phy) before and immediately after admin-istering a volume load (8 mL/kg 6% hy-droxyethylstarch over 20 mins). Theyfound that in patients who responded tovolume loading (increase in CO by�15%), the variation in the IVC diameter

before the fluid challenge was greaterthan in nonresponders. A 12% cutoffvalue in IVC diameter variation beforevolume loading identified those patientswho would respond to a fluid challenge,with positive and negative predictive val-ues of 93% and 92%, respectively.

Doppler Flow Patterns. Informationobtained by analysis of the Doppler signalat the level of the mitral valve and pul-monary vein offers additional informa-tion about preload (69, 70). These Dopp-ler profiles can be obtained by either TTEor TEE. Transmitral variables that havebeen studied include the relation of earlyto late transmitral diastolic filling (E/Aratio), isovolumetric relaxation time, andthe rate of deceleration of early diastolicinflow (deceleration time) (1).

Pulmonary venous flow can also beused to assess left atrial pressure (LAP).Both transmitral and pulmonary veinDoppler patterns are strongly dependenton intrinsic and external factors and arenot purely affected by the loading condi-tions of the left ventricle. It is thus ofutmost importance that interpretation ofDoppler variables be done in conjunctionwith a global analysis of cardiac function

and other available hemodynamic or an-atomic variables.

Positive pressure ventilation altersstroke volume by transiently increasingintrathoracic pressure and thereby de-creasing preload. This phasic variation instroke volume results in a cyclic fluctua-tion in arterial pressure (63, 71). Themagnitude of respiratory variation in aor-tic blood velocity (as recorded echocar-diographically by pulsed-wave Doppler atthe level of the aortic annulus) is a dy-namic variable that is superior to staticmeasurement of LVEDA (or of LV end-diastolic volume) to predict fluid respon-siveness in critically ill patients (61, 72).Feissel et al. (73) demonstrated thatwhen patients in septic shock experi-enced a magnitude of respiratory varia-tion of peak aortic velocity of 12%, infu-sion of 500 mL of fluid increased strokevolume and CO by 15%, while decreasingproportionately the magnitude of the re-spiratory variation of peak aortic veloci-ties. Although practical and reliable, useof this echocardiographic dynamic vari-able to assess volemic status can be ap-plied only to patients who are receivingmechanical ventilation and who are per-

Figure 2. Variation of the diameter of the inferior vena cava with respiration has recently beendemonstrated to be a reliable guide to assess fluid responsiveness in patients on mechanical ventila-tion. Top left, ultrasonographic longitudinal view of the intrahepatic segment of the inferior vena cava(IVC) as assessed in the subcostal area. In patients on positive pressure breathing (and synchronouswith the ventilator), the maximal diameter of the IVC will be obtained at the end of inspiration. Right,in such a patient, the minimal IVC diameter will be found at the end of expiration (complete collapseof the IVC is illustrated). Bottom left, precise measurement of the IVC diameter at the end-inspiratory(right arrow) and end-expiratory (left arrow) phases can be reliably obtained by using M-mode. Thiswill allow precise assessment of the IVC diameter variation with respiration before volume loading andhelp identify those patients who would respond to a fluid challenge.


fectly adapted (synchronous) to the ven-tilator and have no cardiac arrhythmia.

LV Dynamic Obstruction

In general, TTE has good sensitivity fordiagnosing the presence of a small, hyper-dynamic left ventricle, the most typicalfinding in severely hypovolemic patientswith underlying normal cardiac function.When dynamic LV obstruction is present,CO is low, and even in the presence ofmarked hypovolemia, pulmonary artery oc-clusion pressure is high. Paradoxic worsen-ing of hypotension after intravascular vol-ume loading may be the first clue todynamic LV obstruction in critically ill pa-tients. This entity should be recognizedearly because inadequate management ofthis condition can rapidly lead to worsen-ing of hemodynamic status and death. Bytwo-dimensional echocardiography, the leftventricle appears to be small and hyperdy-namic, and there is motion of the anteriorleaflet (or chordae) toward the septum insystole. With color Doppler, a “mosaic” pat-tern of flow is seen in the left ventricularoutflow tract due to the high velocity andturbulence. Variable degrees of asymmetricmitral regurgitation may also be present.Continuous-wave Doppler often demon-strates the presence of a significant gradi-ent in the left ventricular outflow tract. Asmall, hypertrophied left ventricle (typicallyseen in elderly patients with chronic hyper-tension), reduced afterload, and significantcatecholaminergic stimulation are factorsthat will predispose to the development orworsening of LV dynamic obstruction. Dy-namic LV obstruction also has been de-scribed in patients with acute myocardialinfarction, mostly in association with apicalinfarction (74–79).

Both TTE and TEE have been demon-strated to play a key role in making thediagnosis of hypovolemia and LV dynamicobstruction, leading to a dramatic effecton therapy (76–82).

Assessment of PulmonaryArtery Pressure

Pulmonary hypertension is commonin critically ill patients and is a manifes-tation of various pulmonary, cardiac, andsystemic processes. Pulmonary hyperten-sion is said to be present when systolicpulmonary pressure is �35 mm Hg, dia-stolic pulmonary pressure is �15 mmHg, and mean pulmonary pressure is�25 mm Hg (49). A number of echocar-diographic methods have been validated

for noninvasive estimation of pulmonaryartery pressure (49, 83). These methodscan be of great help in the ICU setting.Systolic and diastolic pulmonary arterypressures are determined from the tricus-pid and pulmonary regurgitation veloci-ties, respectively (some degree of regur-gitation is essential to be able to obtain aDoppler signal and subsequently deter-mine pulmonary artery pressure). Tricus-pid regurgitation is present in �75% ofthe normal adult population (59) and inapproximately 90% of critically ill pa-tients (84). Approximately 70% of criti-cally ill patients have an adequate Dopp-ler signal of pulmonic insufficiency forthis calculation (82). Tricuspid and pul-monary regurgitation are present at thesame time in �85% of subjects (85).

Assessment of ValvularFunction and Integrity

Attention has been drawn to the lim-itations of the physical examination forthe detection of cardiovascular abnormal-ities (86, 87). This problem is enhancedin acutely ill patients in the ICU, andmany cardiovascular abnormalities maybe concurrent with noncardiac illnesswithout being clinically suspected (88).Significant valvular abnormalities are agood example of such cardiovascular pa-thologies that can be present in the crit-ically ill patient without being clinicallyrecognized (88). Even in the presence ofinvasive monitoring, significant valvularpathologies may be missed. Precise eval-uation of the valvular apparatus may thusoften be warranted in the ICU. The mostcommon indications for bedside echocar-diography for evaluation of valvular appa-ratus in this population are for suspectedendocarditis (89, 90), acute aortic or mi-tral valve regurgitation (91, 92), or pros-thetic valve dysfunction (93). Echocardi-ography is uniquely suited to theevaluation of valvular heart disease be-cause of its ability to provide informationregarding the pathogenesis and severity

of valvular lesions. In the ICU, TTE canprovide valuable information concerningvalvular integrity and function (93) butmay be suboptimal and not sensitiveenough to detect endocarditis, a dysfunc-tional mitral valve, or prosthetic valvedysfunction. Thus, TEE is often war-ranted.

Evaluation of the PericardialSpace

In the ICU, the most common clinicalindication for assessment of the pericar-dial space is suspected tamponade. Thepericardium is a potential space that canbecome filled with fluid, blood, pus, oruncommonly, air. Presence of fluid inthis space is detected as an echo-freespace. Pericardial fluid is usually easilydetected with TTE. The parasternal long-and short-axis and the apical views usu-ally reveal the effusion. In many criticallyill patients with suboptimal TTE imagequality, the subcostal view is often theonly adequate window available to detectthe presence of a pericardial effusion. Inthese ICU patients with poor acousticwindows and in the postcardiac surgicalsetting, TEE may be needed to assess thepericardial space adequately.

In addition to assisting in the diagno-sis of pericardial effusion and tamponade,two-dimensional echocardiography canalso assist in its drainage, as pericardio-centesis can be performed safely undertwo-dimensional echocardiographicguidance (94, 95). By determining thedepth of the effusion and its distancefrom the site of puncture, it is possible tooptimize the needle placement. Echocar-diography also can be used to immedi-ately monitor the results of the pericar-diocentesis.

Cardiac Tamponade in the ICU

The most common causes of cardiactamponade in the ICU are listed in Table3. Echocardiographic two-dimensional

Table 3. Most common causes of cardiac tamponade in the intensive care unit

● Myocardial or coronary perforation secondary to catheter-based interventions (i.e., afterintravenous pacemaker lead insertion, central catheter placement, or percutaneous coronaryinterventions)

● Compressive hematoma after cardiac surgery● Proximal ascending aortic dissection● Blunt or penetrating chest trauma● Complication of myocardial infarction (e.g., ventricular rupture)● Uremic or infectious pericarditis● Pericardial involvement by metastatic disease or other systemic processes


signs of tamponade are a direct conse-quence of increased pericardial pressure,leading to diastolic collapse of one ormore cardiac chambers (usually on theright side first). Usually, collapse of theRV free wall is seen in early diastole andright atrial wall collapse is seen in latediastole (58). This latter sign is sensitivebut not specific for tamponade. It is, how-ever, specific for a hemodynamically sig-nificant effusion if the right atrial col-lapse lasts longer than one third of theR-R interval (58, 96). In the presence of amassive effusion, the heart may have a“swinging” motion in the pericardial cav-ity. This finding is not always present incardiac tamponade, as the amount offluid in the pericardial space may besmall but still cause a tamponade physi-ology, depending on the acuity withwhich the effusion accumulates and thecompliance of the pericardium. In post-sternotomy patients, tamponade may bemissed by TTE (even in cases in whichimaging quality seems adequate) becausehematomas causing selective cardiacchamber compression are often in theform of loculated clots, located in the farfield of the ultrasound beam in the pos-terior heart region (even when the ante-rior pericardium is left open) (97). Theright atrium and right ventricle may bespared in such cases secondary to postop-erative adhesions or tethering of the rightventricle to the chest wall anteriorly (97).

Another (indirect) sign of a hemody-namically significant pericardial effusionon two-dimensional imaging is plethoraof the inferior vena cava with bluntedrespiratory changes (1). The latter sign isless valuable in mechanically ventilatedpatients because they often have a stiffdilated inferior vena cava, even in theabsence of a pericardial effusion.

Doppler findings of cardiac tamponadeare based on characteristic changes inintrathoracic and intracardiac hemody-namics that occur with respiration. Incritically ill patients, however, mechani-cal ventilation, bronchospasm, signifi-cant pleural effusion, respiratory distress,and arrhythmias make the Doppler find-ings difficult to interpret. In some cir-cumstances, echocardiographic signs oftamponade may be very subtle or evenabsent so one must keep in mind that thediagnosis of tamponade remains a clinicalone and that the echocardiographic signsmust be analyzed in conjunction with theclinical findings.

Complications After CardiacSurgery

Bedside echocardiography has provedto be of particular value in the criticalcare management of patients with hemo-dynamic instability after cardiothoracicoperations (77, 89, 98–101). TTE is oftenseverely limited in this group of patients(35, 89) (Fig. 3). TEE is thus the modalityof choice in this setting because it pro-vides detailed information that can helpdetermine the cause of refractory hypo-tension. The most frequent echocardio-graphic diagnoses encountered in thispopulation of patients are LV or RV fail-ure, tamponade, hypovolemia, and valvu-lar dysfunction. Schmidlin et al. (102)studied 136 patients after cardiac surgeryand showed that a new diagnosis wasestablished or an important pathologywas excluded in 45% of patients under-going TEE. A therapeutic effect wasfound in 73% of cases. The main indica-tions for TEE in this study were controlof LV function (34%), unexplained hemo-dynamic deterioration (29%), suspicionof pericardial tamponade (14%), cardiacischemia (9%), and “other” (14%).Reichert et al. (99) performed TEE inhypotensive patients after cardiac sur-gery. LV failure was found in 27% ofpatients, hypovolemia in 23%, RV failurein 18%, biventricular failure in 13%, andtamponade in 10%. Comparison with he-modynamic variables showed agreementon diagnosis (hypovolemia vs. tamponadevs. cardiac failure) in only 50% of thecases. Echocardiography identified twopatients with tamponade and six with hy-povolemia that were not suspected based

on standard hemodynamic data. In fivepatients with hemodynamic findings sug-gestive of tamponade, unnecessary reop-eration was prevented as TEE ruled outthis diagnosis. Costachescu et al. (82)also demonstrated the superiority of TEE,compared with conventional monitoringwith a PAC, in diagnosing and excludingsignificant causes of hemodynamic insta-bility in postoperative cardiac surgical pa-tients. Descriptions of the echocardio-graphic findings of LV dysfunction,tamponade, hypovolemia, and valvulardysfunction have been described in ear-lier sections of this article.

INFECTIVE ENDOCARDITIS

Occurrence of infective endocarditisin patients hospitalized in an ICU is notan uncommon event. It is often in thedifferential diagnosis of febrile patients inthe ICU. Infective endocarditis was thesecond most common indication for per-formance of an echocardiogram amongcenters reporting their experience (35).Echocardiography is the test of choice forthe noninvasive diagnosis of endocarditis.The echocardiographic features typicalfor infective endocarditis are a) an oscil-lating intracardiac mass on a valve orsupporting structure or in the path of aregurgitant jet or an iatrogenic device, b)abscesses, c) new partial dehiscence of aprosthetic valve, or d) new valvular re-gurgitation (49, 103, 104). Sensitivity forthe echographic diagnosis of endocarditisis 58–62% for TTE and 88–98% for TEE(105, 106). TEE is particularly useful fordetecting small vegetations (107) and de-tecting vegetations on prosthetic valves.

Figure 3. In certain specific clinical situations in which transthoracic echocardiography is likely to failor be suboptimal, immediate transesophageal echocardiography will be preferable. In the aboveexample, a 65-yr-old patient presented with hemodynamic instability after having undergone cardiacsurgery. A transthoracic echocardiographic study was initially performed, but no adequate ultrasono-graphic window could be obtained (left, suboptimal parasternal short-axis view demonstrated), andcardiac function was impossible to assess with such an approach. A transesophageal echocardiographicstudy was then performed and a complete and reliable assessment of ventricular and valvular functionwas made (right, transesophageal short-axis view of the left and right ventricles from the transgastricplane demonstrated).


TEE has also been clearly shown to besuperior to TTE for diagnosing complica-tions of endocarditis, such as aortic rootabscess, fistulas, and ruptured chordaetendineae of the mitral valve (93). As con-cluded by Colreavy et al. (89), perfor-mance of TEE in the ICU for suspicion ofinfective endocarditis should be reserveda) for cases associated with a clinical like-lihood of endocarditis and a negative TTEexamination, b) for suspected prostheticvalve endocarditis, c) to assess complica-tions in known cases of endocarditis, andd) for cases of Staphylococcus aureusbacteremia when the source is unknownor blood cultures remain positive despiteantibiotic therapy.

ASSESSMENT OF THE AORTA

Suspected aortic pathologies can beencountered in different ICU settings.The aorta may need to be imaged to ruleout dissection, rupture, aneurysm, aorticdebris, or aortic abscess. TTE is a goodinitial imaging modality for evaluation ofthe proximal aorta (ascending aorta andarch) (49). The descending thoracicaorta, however, cannot be adequately as-sessed and visualized with this modality.Because of the close anatomic relation-ship between the thoracic aorta and theesophagus, TEE allows optimal visualiza-tion of the entire thoracic aorta.

Aortic Dissection and Rupture. Pa-tients presenting with suspected aortic dis-section need emergency diagnosis andtreatment. Different noninvasive tests havebeen advocated for evaluation of suspectedaortic dissection: TEE, computed tomogra-phy, and magnetic resonance imaging (35,108). Nienaber et al. (108) compared allthree modalities and found similar sensitiv-ities (98%). Magnetic resonance imaginghad higher specificity than TEE (98% vs.77%). A limitation of the study was thatsingle-plane TEE was used. With multi-plane TEE, specificity is improved to �90%(109). TEE was compared with computedtomography and aortography in the multi-center European Cooperative Study (110),and it was demonstrated that TEE was su-perior to both modalities for the diagnosisof aortic dissection (sensitivity, 99%).Other studies have confirmed the high ac-curacy of TEE (110–113) (Fig. 4). A nega-tive TEE for the diagnosis of aortic dissec-tion, even in a high-risk population, hashigh negative predictive value (114).

Additional very helpful features of TEEin the evaluation of aortic pathologies arethe ability to detect or assess: extensionof dissection into the proximal coronaryarteries; the presence of pericardial ormediastinal hematoma or effusion; thepresence, severity, and mechanism of as-sociated aortic valve regurgitation; thepoint of entry and exit between the trueand false lumens; the presence of throm-bus in the false lumen; and ventricularfunction (93).

Intraaortic Balloon Counterpulsation.Bedside TEE may be of help in differentaspects of intraaortic balloon counterpul-sation management. Before insertion, itcan rule out the presence of significantaortic regurgitation, which would repre-sent a contraindication to intraaortic bal-loon counterpulsation use. After inser-tion, TEE can confirm the position of theintraaortic catheter in the descendingthoracic aorta, ensure correct function-ing of the balloon (visualization of infla-tion and deflation), and rule out the pres-ence of important complications of aorticcatheter insertion like aortic dissection.TEE may also be used for monitoring ofthe ventricular function while separating

the patient from the intraaortic ballooncounterpulsation device.

ASSESSMENT FORINTRACARDIAC ANDINTRAPULMONARY SHUNTS

In critically ill patients, clinical suspi-cion for an intracardiac or intrapulmo-nary shunt will most often be raised inthe context of unexplained embolicstroke or refractory hypoxemia. In suchcases, the presence of a right-to-leftshunt needs to be excluded. Common or-igins of right-to-left shunt are atrial sep-tal defect or patent foramen ovale at thecardiac level (35) and arteriovenous fis-tula at the pulmonary level (35). To beable to detect the presence of such ashunt at the bedside, a contrast study isoften needed, as the shunt is usually notwell visualized with two-dimensionalechocardiography alone. Color-flow im-aging increases the detection rate of in-tracardiac shunt to some extent, but usu-ally only when the shunt is large.Accordingly, a contrast study should beperformed routinely as part of a TEE orTTE examination when evaluating a pa-

Figure 4. A 54-yr-old male patient with no medical history presented to the emergency room withsevere chest pain. His electrocardiogram was within normal limits and so was an initial urgent bedsidetransthoracic echocardiographic examination. A subsequent bedside transesophageal echocardio-graphic study revealed the presence of an important dissection of the descending thoracic aorta, as canbe seen from the two-dimensional images obtained in the short-axis (top left) and long-axis views (topright). A color-Doppler study performed in the same two views (bottom left and bottom right) showedthe presence of flow in a severely narrowed true lumen, with no flow detected in the relatively largefalse lumen. The patient was urgently taken to the operating room.


tient with unexplained embolic stroke orrefractory hypoxemia in the ICU. For thispurpose, agitated saline contrast is usu-ally used. Approximately 0.5 mL of air ismixed with 10 mL of normal saline and isthen vigorously agitated back and forthbetween two syringes connected to thepatient by a three-way stopcock. After anadequate echocardiographic view of theright and left atrial cavities has been ob-tained, the agitated saline is forcefullyinjected intravenously. After injection,the contrast is seen in the vena cava,right atrium, right ventricle, and the pul-monary artery. In the absence of a shunt,only a minimal amount of contrastshould be seen in the left-sided cavities,as most of the microbubbles from theagitated saline are not able to passthrough the pulmonary capillaries. If anintracardiac shunt is present, such as anatrial septal defect or patent foramenovale, left-sided contrast will be observedimmediately after right-sided opacifica-tion, and the contrast will be seen going

through the interatrial septum (Fig. 5).Performance of a Valsalva maneuver bythe patient during contrast injection in-creases the sensitivity of the bubble studyto detect right-to-left shunting. Right-to-left shunting can also be caused by thepresence of pulmonary arteriovenous fis-tulas. These are often associated withend-stage liver disease (hepatopulmonarysyndrome). With this type of shunt, con-trast is seen to appear in the left atriumfrom the pulmonary veins instead ofthrough the atrial septum; this finding isbest detected by TEE, which usually per-mits visualization of all four pulmonaryveins. The characteristic of intrapulmo-nary vs. intracardiac shunt is that there isa longer delay (3–5 cardiac cycles) be-tween the appearance of contrast fromthe right-sided to left-sided cavities in thepresence of an intrapulmonary shunt (5).Agitated saline is a simple and easy to usecontrast at the bedside. In critically illpatients, TEE is in general more usefulthan TTE for evaluation of patent fora-

men ovale, atrial septal defect, and pul-monary arteriovenous fistula (115) due tothe close proximity of the lesion to theultrasound transducer.

SOURCE OF EMBOLUS

In the setting of acute unexplainedstroke, echocardiography will often be re-quired to determine whether a potentialembolic source of cardiac origin ispresent. TEE is the modality of choice forthis purpose. Possible cardiac sources ofemboli to the arterial circulation includeleft atrial or appendicular thrombus, LVthrombus, thoracic atheromatosis, andright-sided clots (right atrium, right ven-tricle, vena cava) combined with a right-to-left intracardiac shunt (leading to aparadoxic embolus). Cardiac tumors andvegetations are other potential sources ofemboli of cardiac origin that need to beconsidered.

In the critically ill patient with atrialfibrillation or flutter in whom cardiover-sion is considered, performance of TEEwill be very helpful for evaluating the leftatrium and appendage for the presence ofthrombus. If no intracardiac clots aredocumented, cardioversion can then beperformed with minimal embolic risks.

COMPARISON BETWEENBEDSIDE ECHOCARDIOGRAPHYAND PULMONARY ARTERYCATHETER IN THE ICU

Since its introduction into clinicalpractice in 1970, the PAC has been thestandard hemodynamic monitoring tech-nique for critically ill patients in the ICU(116–118). The PAC provides clinicianswith indices of cardiovascular function toassist in therapeutic decision making. APAC can be a very useful diagnostic tool,aiding in the management of critically illpatients. Nevertheless, poor interpreta-tion of the data it provides can lead toexcessive morbidity and mortality (51,116, 119, 120). Conventional monitoringusing a PAC has been shown to often belimited in the evaluation of global ven-tricular function (80, 81), and echocar-diographic studies have established thatpulmonary artery occlusion pressure of-ten does not allow accurate assessment ofLV preload (17, 57, 121). The frequentchanges in ventricular compliance andloading conditions occurring in criticallyill patients can affect both systolic anddiastolic function. In such cases, conven-tional monitoring does not enable early

Figure 5. A 67-yr-old male patient had moderate to severe hypoxemia 2 days after a cardiac surgery.He was still on mechanical ventilation, with a high level of positive end-expiratory pressure. The chestradiograph did not show significant abnormalities. The possibility of having some degree of right-to-left shunt via a patent foramen ovale was entertained, and a transesophageal microbubble contraststudy was performed at the bedside. Top left, an adequate TEE imaging plane to view the left atrium(LA), right atrium (RA), interatrial septum (the membrane seen between the LA and RA), and distalpart of the superior vena cava (SVC) was obtained before injection of the microbubbles. Top right,microbubbles have just been injected and are seen completely opacifying the SVC and RA. Bottom left,within 2–3 heartbeats, the microbubbles are seen passing from the RA to the LA via a slit-like openingin the superior aspect of the interatrial septum, thus confirming the presence of a right-to-left shuntvia a patent foramen ovale. Bottom right, a few seconds later, more of the left atrium gets filled withthe bubble contrast. In such patient, the finding of a right-to-left shunt via a patent foramen ovale isof great importance, and an attempt will be made to lower the positive end-expiratory pressure andintrathoracic pressure as much as possible to decrease the degree of right-to-left shunting andassociated hypoxemia.


detection of acute changes in function,and it does not allow the clinician todiscern systolic from diastolic changes(80). In critically ill patients, echocardi-ography, particularly TEE, has the abilityto clarify diagnosis and define pathophys-iologic process more precisely than PAC.In a prospective study of limited-scope,goal-directed TEE, Benjamin et al. (81)found that TEE-derived data disagreedwith the PAC evaluation of intracardiacvolume in 55% of cases and with the PACassessment of myocardial function in39% of cases. These authors also demon-strated that the post-PAC therapeutic rec-ommendations were different from thepost-TEE therapeutic recommendationsin 58% of patients. In a retrospectiveanalysis of 108 critically ill patients whounderwent a TEE, Poelaert et al. (122)found that of 64% of patients with a PAC,44% underwent therapy changes afterTEE (41% in the cardiac and 54% in theseptic subgroup). Also, they found that in41% of patients without a PAC, TEE ledto a change in therapy. They concludedthat TEE produced a change in therapy inat least one third of their ICU patients,independent of the presence of a PAC(122). Another significant advantage ofechocardiography in the ICU is the speedwith which it can be performed relative toPAC. In the study by Benjamin et al. (81),TEE was performed in 12 � 7 mins vs.�30 mins for PAC insertion. In a study byKaul (123), the average time required toplace a PAC and record the data was 63 �45 mins vs. 19 � 7 mins to performbedside TEE. Reported complications ofPAC include pneumothorax, hemothorax,bacteremia, sepsis, cardiac arrhythmias,pulmonary artery rupture, cardiac perfo-ration, and valvular damage (81). Com-pared with PAC, bedside echocardiogra-phy has a better safety profile, as reportedin a previous section of this article.

A major advantage of the PAC vs. TEEexamination is that the catheter canmore easily serve as a continuous moni-toring technique to assess the response toa therapeutic intervention (81). However,this potential advantage may provide lit-tle benefit in patients in whom the infor-mation is misinterpreted or inadequate.In some ICUs, TEE has completely re-placed the PAC for assessment of circula-tory status of mechanically ventilated pa-tients (28). Despite having multiplelimitations, the PAC still has a role in theICU and remains a useful diagnostic toolwhen used by physicians who have exten-sive experience with it (122, 124). A com-

bination of invasive pressure monitoringand TEE imaging probably offers themost complete evaluation at the bedsideon morphology and intracardiac hemody-namics and provides a more precise pres-sure–volume evaluation of both LV andRV function and filling (82, 122).

EFFECT OF BEDSIDEECHOCARDIOGRAPHY IN THECRITICALLY ILL PATIENT ONDIAGNOSIS AND MANAGEMENT

Several studies have examined the effectof bedside echocardiography, particularlyTEE, on the management of critically illpatients. Published studies have reportedchanges in management after TEE in 30–60% of patients, (15, 122, 125, 126) leadingto surgical interventions in 7–30% (15, 92,126, 127). Effect varies depending on thetype of ICU population being studied. Sev-eral studies have reported the clinical effectof urgent TEE in hemodynamically unsta-ble patients (126, 128, 129). In a prospec-tive study of surgical ICU patients by Bruchet al. (14), echocardiography altered man-agement in 50 of 115 patients (43%). Alter-ations in medical management induced byTEE included administration of fluids andinitiation or discontinuation of inotropicagents, anticoagulants, or antibiotics.These findings are similar to those reportedin patients in medical or coronary careICUs (2, 127). In a retrospective study doneby Colreavy et al. (89) of a mixed medicaland surgical ICU population, TEE findingsled to a significant change in managementin 32% of all studies performed. In a pro-

spective study by Heidenreich et al. (130) of61 critically ill patients with unexplainedhypotension, new diagnoses were made in17 patients (28%), leading to surgical in-tervention in 12 (20%). Prospective ran-domized trials to study the ultimate effectof bedside echocardiography on mortalityand morbidity in the ICU are needed. Suchstudies will be difficult to perform, how-ever, given the growing use and impor-tance of this technology in the critical caresetting.

HAND-CARRIED ULTRASOUND

Hand-carried ultrasound (HCU) devicesare a new generation of portable ultrasoundmachines that are lightweight (6–10 lbs),battery powered, and less expensive($15,000–50,000) than the sophisticatedhigh-end machines. The tremendous po-tential of HCU to immediately provide di-agnostic information at the bedside not as-sessable by the physical examination alonehas been increasingly demonstrated andrecognized in the last few years (131–139).These devices may facilitate the full clinicalpotential of ultrasound imaging in the ICU,with true portability, ease of use, and lowercost (Fig. 6). They are especially powerfulwhen used as an adjunct to the physicalexamination (136, 137).

An examination using HCU is usuallydirected toward a specific clinical ques-tion and is in general significantlyshorter in duration (�6 mins in somestudies) than one using traditional echo-cardiography (133, 135, 140–144). Thedisadvantage of such directed examina-

Figure 6. Hand-carried ultrasound is a new generation of portable ultrasound machine that islightweight (6–10 lbs), battery powered, and much less expensive than the sophisticated high-endmachines. The small size is a tremendous advantage in the acute care environment because space isoften significantly limited, as shown in this picture of an intensive care unit patient receivingmechanical ventilation and continuous renal replacement therapy.


tions with hand-carried devices is thatthey are not as comprehensive and canpotentially miss some findings comparedwith traditional echocardiographic exam-inations. However, the HCU devicesshould not be compared with the yield orquality of the high-end machines. TheHCU should be viewed more as an exten-sion to the physical examination (133,136, 142–144). In general, accuracy ofimages created by these devices hasshown good agreement when comparedwith standard echocardiogram machineswith respect to two-dimensional findings(133, 142, 145). Studies have shown HCUsensitivity of two-dimensional imagingfor finding abnormal LV function torange from 76% to 96%, with lower sen-sitivity for color-Doppler assessment ofvalvular regurgitation (52% to 96%)(133, 142, 145). Most studies comparingHCU with standard echocardiographywere done in the inpatient ward or out-patient practice setting (133, 136). A re-cent study done by Gorcsan et al. (135)investigated the utility of HCU when spe-cifically used as an extension of the phys-ical examination on consultative cardiol-ogy rounds (n � 235). The HCUdemonstrated an excellent close overallagreement (92–100%, r � .91–.96) forestimation of EF, LV hypertrophy, re-gional wall motion abnormalities, andpericardial effusion (as assessed by two-dimensional imaging) when comparedwith an echocardiographic study using afull-size echocardiographic system. Thegoal-directed HCU study was performedin �10 mins and was focused on theabove-mentioned diagnoses. HCU datainfluenced treatment decisions in 149 pa-tients (63%); 50% had a change in med-ical therapy, and 22% had a change intheir diagnostic workup. In all, 12 pa-tients (5%) had an immediate change inthe decision for cardiac catheterization orpericardiocentesis. The authors of thisstudy concluded that use of “goal-directed HCU has the potential to influ-ence bedside patient treatment decisionsand expedite health care” (135). Concernshave been raised that HCU devices maycompare less favorably with standardechocardiography when performed incritically ill patients because of the morefrequent occurrence of a limited acousticwindow. In a study of 80 critically illpatients that compared HCU vs. standardechocardiography, 85% of clinical ques-tions could be addressed by the HCU de-vice (146). HCU failed to detect a clini-cally significant finding in 31% of

patients; however, the majority of thesemissed findings were Doppler-based diag-noses (e.g., valvular regurgitation). Amore recent study by Vignon et al. (138)compared the diagnostic capability ofHCU and of conventional TTE (used as agold standard) in a population of 106critically ill patients on mechanical ven-tilation. In this study, the HCU examswere performed by echocardiography-trained intensivists and the TTE examswere reviewed by a cardiologist experi-enced in echocardiography. They showedthat the number of acoustic windows wascomparable when using the HCU andconventional TTE in this population ofpatients. HCU had a lower overall diag-nostic capacity than TTE (199 of 251vs. 223 of 251 clinical questions solved,p � .005), mainly due to its lack of spec-tral Doppler capability. However, diag-nostic capacity based on two-dimensionalimaging was comparable for both ap-proaches (129 of 155 vs. 125 of 155 clin-ical questions solved, p � .4). Also, HCUand TTE had a similar therapeutic effect.Results from these studies suggest thatthe accuracy of HCU with respect to two-dimensional imaging remains very goodin the critically ill patient when com-pared with standard ultrasound machinesbut that information derived from HCUcolor-Doppler imaging should be inter-preted cautiously in this patient popu-lation.

It should be emphasized that thegoal of using HCU in the ICU shouldnot be to replace high-end machinesbut to provide diagnostic data not de-tected on physical examination. HCUshould allow critical care physicians todiagnose certain cardiopulmonary pa-thologies more rapidly than with stan-dard echocardiography (which is oftenperformed with a variable delay afterhaving been requested). Provided thatphysicians performing point-of-care ex-aminations with the HCU have ade-quate training (133, 143, 147), have re-alistic expectations, and understand thelimitations of the device, then the HCUhas the potential to create a tremen-dous advantage for bedside assessmentand treatment of the ICU patient.

PERFORMANCE OF BEDSIDEECHOCARDIOGRAPHY BY THEINTENSIVIST

It is usually not feasible to have a cardi-ologist or sonographer available on imme-diate call on a 24-hr basis to perform bed-

side ultrasonographic examinations in theICU. The value of immediate bedside echo-cardiography for aiding in diagnosis andmanagement of acute hemodynamic dis-turbances has been well demonstrated inboth the ICU and the emergency room (24,148, 149–151). It is recognized that ultra-sound technologies are not exclusive to theradiologist or cardiologist. Appropriatelytrained emergency room physicians, sur-geons, anesthesiologists, and intensive carespecialists have been performing echocar-diographic examinations with great suc-cess. Anesthesiologists were instrumentalin many of the pioneering studies of TEE inthe operating room and ICU (152–154).Successful performance of bedside echocar-diography by noncardiologist intensivistshas also been well demonstrated in the lit-erature (137, 138, 143). A recent study byManasia et al. (155) demonstrated that aftera brief (10 hrs) formal training in using ahandheld echocardiographic system, inten-sivists were able to successfully perform alimited TTE in 94% of patients and inter-preted their studies correctly in 84%. Lim-ited TTE provided new cardiac informationand changed management in 37% of pa-tients. This study supports the concept thatintensivist-performed goal-directed TTEcan be easily feasible and have significantclinical effect. However, adequate trainingis essential, and this must be individualizedand tailored to the specific needs and appli-cations of the user (156). With expertbackup, focused bedside ultrasonographyby intensivists is not only feasible but canalso be done safely and rapidly and yieldinformation pertinent to the managementof critically ill patients. However, inappro-priate interpretation or application of datagained by a poorly skilled user may haveadverse consequences (156). To avoid mis-using this technology, adequate training isessential.

The importance of adequate trainingand subsequent maintenance of compe-tence cannot be overemphasized, as inap-propriate use or misapplication could po-tentially temper the acceptance ofintensivist-performed bedside ultra-sound. Performance of emergency bed-side ultrasound should provide rapid an-swers to clinical questions that mayprofoundly affect medical and surgicalmanagement decisions. Training in goal-directed echocardiography and generalultrasonography should be incorporatedin the critical care fellowship as part ofthe training program of intensivists. Theera of a technology-extended physical ex-amination (136) seems to have arrived,


and there seems to be a role for user-specific, focused ultrasound examina-tions (156, 157).

REFERENCES

1. Poelaert J, Schmidt C, Colardyn F: Trans-esophageal echocardiography in the criti-cally ill. Anaesthesia 1998; 53:55–68

2. Hwang JJ, Shyu KG, Chen JJ, et al: Useful-ness of transesophageal echocardiographyin the treatment of critically ill patients.Chest 1993; 104:861–866

3. Cook CH, Praba AC, Beery PR, et al: Trans-thoracic echocardiography is not cost-effective in critically ill surgical patients.J Trauma 2002; 52:280–284

4. Stamos TD, Soble JS: The use of echocar-diography in the critical care setting. CritCare Clin 2001; 17:253–270

5. Joseph MX, Disney PJS, Da Costa R, et al:Transthoracic echocardiography to identifyor exclude cardiac cause of shock. Chest2004; 126:1592–1597

6. Foster E, Schiller NB: Introduction totransesophageal echocardiography (TEE)with a historical perspective. Cardiol Clin2000; 18:675–679

7. Townend JN, Hutton P: Transesophagealechocardiography in anaesthesia and inten-sive care. Br J Anaesth 1996; 77:137–139

8. Mueller X, Stauffer JC, Jaussi A, et al: Sub-jective visual echocardiographic estimate ofleft ventricular ejection fraction as an alter-native to conventional echocardiographicmethods: Comparison with contrast an-giography. Cardiol Clin 1991; 14:898–902

9. Lang RM, Mor-Avi V, Zoghbi WA, et al: Therole of contrast enhancement in echocar-diographic assessment of left ventricularfunction. Am J Cardiol 2002; 90(Suppl10A):28J–34J

10. Yong Y, Wu D, Fernandes V, et al: Diagnos-tic accuracy and cost-effectiveness of con-trast echocardiography on evaluation ofcardiac function in technically very difficultpatients in the intensive care unit.Am J Cardiol 2002; 89:711–718

11. Senior R, Soman P, Khattar RS, et al:Improved endocardial visualization withsecond harmonic imaging compared withfundamental two-dimensional echocardio-graphic imaging. Am Heart J 1999; 138:163–168

12. Becher H, Tiemann K, Schlosser T, et al:Improvement in endocardial border delin-eation using tissue harmonic imaging.Echocardiography 1998; 15:511–518

13. Nixdorff U, Matschke C, Winklmaier M, etal: Native tissue second harmonic imagingimproves endocardial and epicardial borderdefinition in dobutamine stress echocardi-ography. Eur J Echocardiogr 2001; 2:52–61

14. Bruch C, Comber M, Schmermund A, et al:Diagnostic usefulness and impact on man-agement of transesophageal echocardiogra-phy in surgical intensive care unit.Am J Cardiol 2003; 91:510–513

15. Vignon P, Mentec H, Terre S, et al: Diag-nostic accuracy and therapeutic impact oftransthoracic and transesophageal echocar-diography in mechanically ventilated pa-tients in the ICU. Chest 1994; 106:1829–1834

16. Jardin F, Fourme T, Page B, et al: Persistentpreload defect in severe sepsis despite fluidloading: A longitudinal echocardiographicstudy in patients with septic shock. Chest1999; 116:1354–1359

17. Jardin F, Valtier B, Beauchet A, et al: Inva-sive monitoring combined with two-dimen-sional echocardiographic study in septicshock. Intensive Care Med 1994; 20:550–554

18. Parker M, Shelhamer J, Barach S, et al:Profound but reversible myocardial depres-sion in patients with septic shock. Ann In-tern Med 1984; 100:483–490

19. Parillo J: Pathogenic mechanism of septicshock. N Engl J Med 1993; 328:1471–1477

20. Parker MM, Suffredini AF, Natanson C, etal: Responses of left ventricular function insurvivors and non-survivors of septic shock.J Crit Care 1989; 4:19–25

21. Textbook of Advanced Cardiac Life Support.Dallas, American Heart Association, 2004

22. Salen P, Melniker L, Chooljian C, et al: Doesthe presence or absence of sonographicallyidentified cardiac activity predict resuscita-tion outcomes of cardiac arrest patients?Am J Emerg Med 2005; 23:459–462

23. Stoddard MF, Pearson AC, Kern MJ, et al:Left ventricular diastolic function: Compar-ison of pulsed Doppler echocardiographicand hemodynamic indexes in subjects withand without coronary artery disease. J AmCardiol 1989; 13:327–336

24. Price S, Nicol E, Gibson DG, et al: Echocar-diography in the critically ill: Current andpotential roles. Intensive Care Med 2006;32:48–59

25. Enger EL, O’Toole MF: Noncardiogenicmechanisms of right heart dysfunction.J Cardiovasc Nurs 1991; 6:54–69

26. Bunnell E, Parillo JE: Cardiac dysfunctionduring septic shock. Clin Chest Med 1996;17:237–248

27. Jardin F, Gueret P, Dubourg O, et al: Two-dimensional echocardiographic evaluationof right ventricular size and contractility inacute respiratory failure. Crit Care Med1985; 13:952–956

28. Vieillard-Baron A, Schmitt JM, Augarde R,et al: Acute cor pulmonale in acute respira-tory distress syndrome submitted to protec-tive ventilation: Incidence, clinical implica-tions, and prognosis. Crit Care Med 2001;29:1551–1555

29. Vieillard-Baron A, Page B, Augarde R, et al:Acute cor pulmonale in massive pulmonaryembolism: Incidence, echocardiographicpattern clinical implications and recoveryrate. Intensive Care Med 2001; 27:1481–1486

30. Jardin F, Dubourg O, Bourdarias JP: Echo-

cardiographic pattern of acute cor pulmo-nale. Chest 1997; 111:209–217

31. Vieillard-Baron A, Prin S, Chergui K, et al:Echo-Doppler demonstration of acute corpulmonale at the bedside in the medicalintensive care unit. Am J Respir Crit CareMed 2002; 166:1310–1319

32. Kasper W, Meinertz T, Kersting F, et al:Echocardiography in assessing acute pul-monary hypertension due to pulmonaryembolism. Am J Cardiol 1980; 45:567–572

33. Jardin F, Dubourg O, Gueret P, et al: Quan-titative two-dimensional echocardiographyin massive pulmonary embolism: Emphasison ventricular interdependence and left-ward septal displacement. J Am Coll Cardiol1987; 10:1201–1206

34. McConnell MV, Solomon SD, Rayan ME, etal: Regional right ventricular dysfunctiondetected by echocardiography in acute pul-monary embolism. Am J Cardiol 1996; 78:469–473

35. Heidenreich PA: Transesophageal echocar-diography in the critical care patient. Car-diol Clin 2000; 18:789–805

36. Vieillard-Baron A, Qanadli SD, Antakly Y, etal: Transesophageal echocardiography forthe diagnosis of pulmonary embolism withacute cor pulmonale: A comparison withradiological procedures. Intensive Care Med1998; 24:429–433

37. Wittlich N, Erbel R, Eichler A, et al: Detec-tion of central pulmonary artery thrombo-emboli by transesophageal echocardiogra-phy in patients with severe pulmonaryembolism. J Am Soc Echocardiogr 1992;5:515–524

38. Goldhaber SZ: Echocardiography in themanagement of pulmonary embolism. AnnIntern Med 2002; 136:691–700

39. Grifoni S, Olivotto I, Cecchini P, et al:Short-term clinical outcome of patientswith acute pulmonary embolism, normalblood pressure, and echocardiographicright ventricular dysfunction. Circulation2000; 101:2817–2822

40. Kasper W, Konstantinides S, Geibel A, et al:Management strategies and determinants ofoutcome in acute major pulmonary embo-lism: Results of a multicenter registry. J AmColl Cardiol 1997; 30:1165–1171

41. Hamel E, Parcouret G, Vincentelli D, et al:Thrombolysis or heparin therapy in massivepulmonary embolism with right ventriculardilatation. Chest 2001; 120:120–125

42. Savino JS, Troianos CA, Aukburg S, et al:Measurement of pulmonary blood flow withtransesophageal two-dimensional andDoppler echocardiography. Anesthesiology1991; 75:445–451

43. Roewer N, Bednarz F, am Esch JS: Contin-uous measurement of intracardiac andpulmonary blood flow velocities with trans-esophageal pulsed Doppler echocardiogra-phy: Technique and initial clinical experi-ence. J Cardiothorac Anesth 1987; 1:418–428

44. Estagnasie P, Djedaini K, Mier L, et al: Mea-


surement of cardiac output by transesoph-ageal echocardiography in mechanicallyventilated patients: Comparison with ther-modilution. Intensive Care Med 1997; 23:753–759

45. Ihlen H, Amlie JP, Dale J, et al: Determina-tion of cardiac output by Doppler echocar-diography. Br Heart J 1984; 51:54–60

46. Katz WE, Gasior TA, Quinlan JJ, et al:Transgastric continuous-wave Doppler todetermine cardiac output. Am J Cardiol1993; 71:853–857

47. Darmon PL, Hillel Z, Mogtader A, et al:Cardiac output by transesophageal echocar-diography using continuous wave Doppleracross the aortic valve. Anesthesiology1994; 80:796–805

48. Feinberg MS, Hopkins WE, Davila-RomanVG, et al: Multiplane transesophageal echo-cardiography Doppler imaging accuratelydetermines cardiac output measurementsin critically ill patients. Chest 1995; 107:769–773

49. Oh JK: The Echo Manual. Second Edition.Philadelphia, Lippincott-Raven, 1999

50. Chaney JC, Derdak S: Minimally invasivehemodynamic monitoring for the intensiv-ist: Current and emerging technology. CritCare Med 2002; 30:2338–2345

51. Marik PE: Pulmonary artery catheterizationand esophageal Doppler monitoring in theICU. Chest 1999; 116:1085–1091

52. Cariou A, Monchi M, Joly LM, et al: Nonin-vasive cardiac output monitoring by aorticblood flow determination: Evaluation of theSometec Dynemo-3000 system. Crit CareMed 1998; 26:2066–2072

53. Valtier B, Cholley BP, Belot JP, et al: Non-invasive monitoring of cardiac output incritically ill patients using transesophagealDoppler. Am J Respir Crit Care Med 1998;158:77–83

54. Sinclair S, James S, Singer M: Intraopera-tive intravascular volume optimization andlength of hospital stay after repair of prox-imal femoral fracture: Randomized con-trolled trial. BMJ 1997; 315:909–912

55. Cheung AT, Savino JS, Weiss SJ, et al:Echocardiographic and hemodynamic in-dexes of left ventricular preload in patientswith normal and abnormal ventricularfunction. Anesthesiology 1994; 81:376–387

56. Douglas PS, Edmonds HL, Sutton MS, et al:Unreliability of hemodynamic indexes of leftventricular size during cardiac surgery. AnnThorac Surg 1987; 44:31–34

57. Hansen RM, Viquerat CE, Matthay MA, et al:Poor correlation between pulmonary arte-rial wedge pressure and left ventricular end-diastolic volume after coronary artery by-pass graft surgery. Anesthesiology 1986; 64:764–770

58. Troianos CA, Porembka DT: Assessment ofleft ventricular function and hemodynamicswith transesophageal echocardiography.Crit Care Clin 1996; 12:253–272

59. Schiller NB, Shah PM, Crawford M, et al:Recommendations for quantification of the

left ventricle by two-dimensional echocar-diography. J Am Soc Echocardiogr 1989;2:358–367

60. Clements FM, Harpole DH, Quill T, et al:Estimation of left ventricular volume andejection fraction by two-dimensional trans-esophageal echocardiography: Comparisonof short axis imaging and simultaneous ra-dionuclide angiography. Br J Anaesth 1990;64:331–336

61. Gunn SR, Pinsky MR: Implications of arte-rial pressure variation in patients in theintensive care unit. Curr Opin Crit Care2001; 7:212–217

62. Tousignant CP, Walsh F, Mazer CD: The useof transesophageal echocardiography forpreload assessment in critically ill patients.Anesth Analg 2000; 90:351–355

63. Swenson JD, Harkin C, Pace NL, et al:Transesophageal echocardiography: An ob-jective tool in defining maximum ventricu-lar response to intravenous fluid therapy.Anesth Analg 1996; 83:1149–1153

64. Kircher BJ, Himelman RB, Schiller NB:Non-invasive estimation of right atrial pres-sure from the inspiratory collapse of infe-rior vena cava. Am J Cardiol 1990; 66:493–496

65. Jue J, Chung W, Schiller NB: Does inferiorvena cava size predict right atrial pressurein patients receiving mechanical ventila-tion? J Am Soc Echocardiogr 1992;5:613–619

66. Nagueh SF, Kopelen HA, Zoghbi WA: Rela-tion of mean right atrial pressure to echo-cardiographic and Doppler parameters ofright atrial and right ventricular function.Circulation 1996; 93:1160–1169

67. Jardin F, Vieillard-Baron A: Ultrasono-graphic examination of the venae cavae. In-tensive Care Med 2006; 32:203–206

68. Feissel M, Michard F, Faller JP, et al: Therespiratory variation in inferior vena cavadiameter as a guide to fluid therapy. Inten-sive Care Med 2004; 30:1834–1837

69. Nishimura RA, Abel MD, Hatle LK, et al:Relation of pulmonary vein to mitral flowvelocities by transesophageal Doppler echo-cardiography: Effect of different loadingconditions. Circulation 1990; 82:1488–1497

70. Kuecherer HF, Muhiudeen IA, KusumotoFM, et al: Estimation of mean left atrialpressure from transesophageal pulsedDoppler echocardiography of pulmonary ve-nous flow. Circulation 1990; 82:1127–1139

71. Currie PJ, Seward BJ, Chan KL, et al: Con-tinuous wave Doppler determination ofright ventricular pressure: A simultaneousDoppler-catheterization study in 127 pa-tients. J Am Coll Cardiol 1985; 6:750–756

72. Hatle L, Angelsen BA, Tromsdal A: Non-invasive estimation of pulmonary arterysystolic pressure with Doppler ultrasound.Br Heart J 1981; 45:157–165

73. Feissel M, Michard F, Mangin I, et al: Re-spiratory changes in aortic blood velocity asan indicator of fluid responsiveness in ven-

tilated patients with septic shock. Chest2001; 119:867–873

74. Mintz GS, Kotler MN, Segal BL, et al: Sys-tolic anterior motion of the mitral valve inthe absence of asymmetric septal hypertro-phy. Circulation 1978; 57:256–263

75. Haley JH, Sinak LJ, Tajik AJ, et al: Dynamicleft ventricular outflow tract obstruction inacute coronary syndromes: An importantcause of new systolic murmur and cardio-genic shock. Mayo Clin Proc 1999; 74:901–906

76. Madu EC, Brown R, Geraci SA: Dynamic leftventricular outflow tract obstruction incritically ill patients: Role of transesopha-geal echocardiography in therapeutic deci-sion making. Cardiology 1997; 88:292–295

77. Joffe II, Jacobs LE, Lampert C, et al: Role ofechocardiography in perioperative manage-ment of patients undergoing open heartsurgery. Am Heart J 1996; 131:162–176

78. Blazer D, Kotler MN, Parry WR, et al: Non-invasive evaluation of mid-left ventricularobstruction by two-dimensional and Dopp-ler echocardiography and color flow Dopp-ler echocardiography. Am Heart J 1987;114:1162–1168

79. Chenzbraun A, Pinto FJ, Schnittger I:Transesophageal echocardiography in theintensive care unit: Impact on diagnosisand decision making. Clin Cardiol 1994;17:438–444

80. Fontes ML, Bellows W, Ngo L, et al: Assess-ment of ventricular function in criticallypatients: Limitations of pulmonary arterycatheterization. J Cardiothorac Vasc Anesth1999; 13:521–527

81. Benjamin E, Griffin K, Leibowitz AB, et al:Goal-directed transesophageal echocardiog-raphy performed by intensivists to assessleft ventricular function: Comparison withpulmonary artery catheterization. J Cardio-thorac Vasc Anesth 1998; 12:10–15

82. Costachescu T, Denault A, Guimond JG, etal: The hemodynamically unstable patientin the intensive care unit: Hemodynamic vs.transesophageal echocardiographic moni-toring. Crit Care Med 2002; 30:1214–1223

83. Stevenson JG: Comparison of several non-invasive methods for estimation of pulmo-nary artery pressure. J Am Soc Echocar-diogr 1989; 2:157–171

84. Balik M, Pachl J, Hendl J: Effect of thedegree of tricuspid regurgitation on cardiacoutput measurements by thermodilution.Intensive Care Med 2002; 28:1117–1121

85. Lee RT, Lord CP, Plappert T, et al: Prospec-tive Doppler echocardiographic evaluationof pulmonary artery diastolic pressure inthe medical intensive care unit. Am J Car-diol 1989; 64:1366–1370

86. Johnson JE, Carpenter JL: Medical housestaff performance in physical examination.Arch Intern Med 1986; 146:937–941

87. Mangione S, Nieman LZ: Cardiac ausculta-tory skills of internal medicine and familypractice trainees: A comparison of diagnos-tic proficiency. JAMA 1997; 278:717–722


88. Bossone E, DiGiovine B, Watts S, et al:Range and prevalence of cardiac abnormal-ities in patients hospitalized in a medicalICU. Chest 2002; 122:1121–1123

89. Colreavy FB, Donovan K, Lee KY, et al:Transesophageal echocardiography in criti-cally ill patients. Crit Care Med 2002; 30:989–996

90. McLean AS: Transesophageal echocardiog-raphy in the intensive care unit. AnaesthIntensive Care 1998; 26:22–25

91. Font Ve, Obarski TP, Klein AL, et al: Trans-esophageal echocardiography in the criticalcare unit. Cleve Clin J Med 1991; 58:315–322

92. Oh JK, Seward JB, Khandheria BK, et al:Transesophageal echocardiography in criti-cally ill patients. Am J Cardiol 1990; 66:1492–1495

93. Alam M: Transesophageal echocardiographyin critical care units: Henry Ford hospitalexperience and review of the literature.Prog Cardiovasc Dis 1996; 38:315–328

94. Callahan JA, Seward JB, Tajik AJ: Cardiactamponade: Pericardiocentesis directed bytwo-dimensional echocardiography. MayoClin Proc 1985; 60:344–347

95. Callahan JA, Seward JB: Pericardiocentesisguided by two-dimensional echocardiogra-phy. Echocardiography 1997; 14:497–504

96. Feigenbaum H: Pericardial disease. In:Echocardiography. Feigenbaum H (Ed).Philadelphia, Lea and Febiger, 1994

97. Russo AM, O’Connor WH, Waxman HL:Atypical presentations and echocardio-graphic findings in patients with cardiactamponade occurring early and late aftercardiac surgery. Chest 1993; 104:71–78

98. Al-Tabbaa A, Gonzalez RM, Lee D: The roleof state-of-the-art echocardiography in theassessment of myocardial injury during andfollowing cardiac surgery. Ann Thorac Surg2001; 72:S2214–S2219

99. Reichert CLA, Visser CA, Koolen JJ, et al:Transesophageal echocardiography in hypo-tensive patients after cardiac operations:comparison with hemodynamic parame-ters. J Cardiovasc Thorac Surg 1992; 104:321–326

100. Wake PJ, Ali M, Carroll J, et al: Clinical andechocardiographic diagnoses disagree in pa-tients with unexplained hemodynamic in-stability after cardiac surgery. Can J An-aesth 2001; 48:778–783

101. Poeze M, Ramsay G, Greve JW, et al: Pre-diction of postoperative cardiac surgicalmorbidity and organ failure within 4 hoursof intensive care unit admission usingesophageal Doppler ultrasonography. CritCare Med 1999; 27:1288–1294

102. Schmidlin D, Schuepbach R, Bernard E, etal: Indications and impact of postoperativetransesophageal echocardiography in car-diac surgical patients. Crit Care Med 2001;29:2143–2148

103. Braunwald E, Zipes DP, Libby P (Eds):Heart Disease: A Textbook of Cardiovascular

Medicine. Sixth Edition. New York, WBSaunders, 2001

104. ACC/AHA guidelines for the management ofpatients with valvular heart disease: A re-port of the American College of Cardiology/American Heart Association. Task Force onPractice Guidelines (Committee on Man-agement of Patients with Valvular HeartDisease). J Am Coll Cardiol 1998; 32:1486–1588

105. Shively BK, Gurule FT, Roldan CA, et al:Diagnostic value of transesophageal com-pared with transthoracic echocardiographyin infective endocarditis. J Am Coll Cardiol1991; 18:391–397

106. Karalis DG, Bansal RC, Hauck AJ: Trans-esophageal echocardiographic recognitionof subaortic complications in aortic valveendocarditis: Clinical and surgical implica-tions. Circulation 1992; 86:353–362

107. Erbel R, Rohmann S, Drexler M, et al: Im-proved diagnostic value of echocardiogra-phy in patients with infective endocarditisby transesophageal approach: A prospectivestudy. Eur Heart J 1988; 9:43–53

108. Nienaber CA, von Kodolitsch Y, Nicolas V,et al: The diagnosis of thoracic aortic dis-section by noninvasive imaging procedures.N Engl J Med 1993; 328:1–9

109. Keren A, Kim CB, Hu BS, et al: Accuracy ofbiplane and multiplane transesophagealechocardiography in diagnosis of typicalacute aortic dissection and intramural he-matoma. J Am Coll Cardiol 1996; 28:627–636

110. Erbel R, Engberding R, Daniel W, et al:Echocardiography in diagnosis of aortic dis-section: The European Cooperative StudyGroup for Echocardiography. Lancet 1989;1:457–461

111. Erbel R, Mohr-Kahaly S, Rennolet H, et al:Diagnosis of aortic dissection: The value oftransesophageal echocardiography. ThoracCardiovasc Surg 1987; 35:126–133

112. Ballal RS, Nanda NC, Gatewood R, et al:Usefulness of transesophageal echocardiog-raphy in assessment of aortic dissection.Circulation 1991; 84:1903–1914

113. Sommer T, Fehske W, Holzknecht N, et al:Aortic dissection: A comparative study ofdiagnosis with spiral CT, multiplanar trans-esophageal echocardiography, and MR im-aging. Radiology 1996; 199:347–352

114. Barbant SD, Eisenberg MJ, Schiller NB: Thediagnostic value of imaging techniques foraortic dissection. Am Heart J 1992; 124:541–543

115. Aller R, Moya JL, Moreira V, et al: Diagnosisof hepatopulmonary syndrome with con-trast transesophageal echocardiography:Advantages over contrast transthoracicechocardiography. Dig Dis Sci 1999; 44:1243–1248

116. Iberti TJ, Fischer EP, Leibowitz AB, et al: Amulticenter study od physicians’ knowledgeof the pulmonary artery catheter: Pulmo-nary Artery Catheter Study Group. JAMA1990; 264:2928–2932

117. Swan HJ, Ganz W: Complications with flow-directed balloon-tipped catheters. Ann In-tern Med 1979; 91:49–96

118. Gore JM, Goldberg RJ, Spodick DH, et al: Acommunity-wide assessment of the use ofpulmonary artery catheters in patients withacute myocardial infarction. Chest 1987;92:721–727

119. Connors AF, Speroff T, Dawson NV, et al:The effectiveness of right heart catheteriza-tion in the initial care of critically ill pa-tients. JAMA 1996; 276:889–897

120. Trottier SJ, Taylor RW: Physicians attitudestoward and knowledge of the pulmonaryartery catheter: Society of critical care med-icine membership survey. New Horiz 1997;5:201–206

121. Thys D, Hillel Z, Goldman M, et al: A com-parison of hemodynamic indices derived byinvasive monitoring and two-dimensionalechocardiography. Anesthesiology 1987; 67:630–634

122. Poelaert JI, Trouerbach J, De Buyzere M, etal: Evaluation of transesophageal echocar-diography as a diagnostic and therapeuticaid in a critical care setting. Chest 1995;107:774–779

123. Kaul S: Value of two-dimensional echocar-diography for determining the bases of he-modynamic compromise in critically ill pa-tients: A prospective study. J Am SocEchocardiogr 1994; 7:598–606

124. Eisenberg PR, Schuster DP: Clinical evalu-ation compared with invasive hemodynamicassessment. In: Oxygen Transport in theCritically Ill. Snyder JV, Pinsky MR (Eds).Chicago, Year Book Medical Publishers,1987, pp 199–204

125. Foster E, Schiller NB: Transesophagealechocardiography in the critical care pa-tient. Cardiol Clin 1993; 11:489–503

126. Khoury AF, Afridi I, Quinones MA, et al:Transesophageal echocardiography in criti-cally ill patients: Feasibility, safety, and im-pact on management. Am Heart J 1994;127:1363–1371

127. Pearson AC, Castello R, Labovitz AJ: Safetyand utility of transesophageal echocardiog-raphy in the critically ill patient. Am Heart J1990; 119:1083–1089

128. Cicek S, Demirkilic U, Kuralay E, et al:Transesophageal echocardiography in car-diac surgical emergencies. J Card Surg1995; 10:236–244

129. Sohn DW, Shin GJ, Oh JK, et al: Role oftransesophageal echocardiography in he-modynamically unstable patients. MayoClin Proc 1995; 70:925–931

130. Heidenreich PA, Stainback RF, Redberg RF,et al: Transesophageal echocardiographypredicts mortality in critically ill patientswith unexplained hypotension. J Am CollCardiol 1995; 26:152–158

131. Johnson JE, Carpenter JL: Medical housestaff performance in physical examination.Arch Intern Med 1986; 146:937–941

132. Mangione S, Nieman LZ: Cardiac ausculta-tory skills of internal medicine and family


practice trainees: A comparison of diagnos-tic proficiency. JAMA 1997; 278:717–722

133. Liang D, Schnittger I: Accuracy of hand-carried ultrasound. Echocardiography2003; 20:487–490

134. Bossone E, DiGiovine B, Watts S, et al:Range and prevalence of cardiac abnormal-ities in patients hospitalized in a medicalICU. Chest 2002; 122:1370–1376

135. Gorcsan J III, Pandey P, Sade LE: Influenceof hand-carried ultrasound on bedside pa-tient treatment decisions for consultativecardiology. J Am Soc Echocardiogr 2004;17:50–55

136. DeCara JM, Lang RM, Spencer KT: Thehand-carried echocardiographic device asan aid to the physical examination. Echo-cardiography 2003; 20:477–485

137. Spevack DM, Spevack DM, Tunick PA, et al:Hand carried echocardiography in the crit-ical care setting. Echocardiography 2003;20:455–461

138. Vignon P, Chastagner C, Francois B, et al:Diagnostic ability of hand-held echocardi-ography in ventilated critically ill patients.Crit Care 2003; 7:R84–R91

139. Fedson S, Neithardt G, Thomas P, et al:Unsuspected clinically important findingsdetected with a small portable ultrasounddevice in patients admitted to a generalmedicine service. J Am Soc Echocardiogr2003; 16:901–905

140. Bruce CJ, Zummach PL, Prince DJ: Per-sonal ultrasound imager: Utility in the car-diology outpatient setting. Abstr. Circula-tion 2000; 102(Suppl):II-364

141. Rugolotto M, Chang CP, Hu B, et al: Clini-cal use of cardiac ultrasound performedwith a hand-carried device in patients ad-

mitted for acute cardiac care. Am J Cardiol2002; 90:1040–1042

142. Spencer KT, Anderson AS, Bhargava A, et al:Physician-performed point-of-care echocar-diography using a laptop platform com-pared with physical examination in the car-diovascular patient. J Am Coll Cardiol 2001;37:2013–2018

143. Duvall WL, Croft LB, Goldman ME: Canhand-carried ultrasound devices be ex-tended for use by the noncardiology medi-cal community? Echocardiography 2003;20:471–476

144. Medicine in the 21st century: Hand-carriedultrasound. The visual stethoscope [mono-graph]. Today in Cardiology, 2003

145. Rugolotto M, Hu BS, Liang DH, et al: Rapidassessment of cardiac anatomy and functionwith a new hand-carried ultrasound device(OptiGo): A comparison with standardechocardiography. Eur J Echocardiogr2001; 2:262–269

146. Goodkin GM, Spevack DM, Tunick PA, et al:How useful is hand-carried bedside echo-cardiography in critically ill patients? J AmColl Cardiol 2001; 37:2019–2022

147. Kronzon I: The hand-carried ultrasoundrevolution. Echocardiography 2003; 20:453–454

148. Mandavia DP, Aragona J, Chan L, et al:Ultrasound training for emergency physi-cians: A prospective study. Acad Emerg Med2000; 7:1008–1014

149. Counselman FL, Sanders A, Slovis CM, et al:The status of bedside ultrasonographytraining in emergency medicine residencyprograms. Acad Emerg Med 2003; 10:37–42

150. Beaulieu Y, Marik P: Bedside ultrasonogra-

phy in the ICU: Part 1. Chest 2005; 128:881–895

151. Beaulieu Y, Marik P: Bedside ultrasonogra-phy in the ICU: Part 2. Chest 2005; 128:1766–1781

152. Liebson PR: Transesophageal echocardiog-raphy in critically ill patients: What is theintensivist’s role? Crit Care Med 2002; 30:1165–1166

153. Mathew JP, Fontes ML, Garwood S, et al:Transesophageal echocardiography inter-pretation: A comparative analysis betweencardiac anesthesiologists and primary echo-cardiographers. Anesth Analg 2002; 94:302–309

154. Denault AY, Couture P, McKenty S, et al:Perioperative use of transesophageal echo-cardiography by anesthesiologists: Impactin noncardiac surgery and in the intensivecare unit. Can J Anaesth 2002; 49:287–293

155. Manasia AR, Nagaraj HM, Kodali RB, et al:Feasibility and potential clinical utility ofgoal-directed transthoracic echocardiogra-phy performed by noncardiologist inten-sivists using a small hand-carried device(SonoHeart) in critically ill patients. J Car-diothorac Vasc Anesth 2005; 19:155–159

156. Seward JB, Douglas PS, Erbel R, et al:Hand-carried cardiac ultrasound (HCU) de-vice: Recommendations regarding newtechnology. A report from the Echocardiog-raphy Task Force on New Technology of theNomenclature and Standards Committee ofthe American Society of Echocardiography.J Am Soc Echocardiogr 2002; 15:369–373

157. Haar GT: Commentary: Safety of diagnos-tic ultrasound. Br J Radiol 1996; 69:1083–1085


Bedside echocardiography in the assessment of the ... Echo (Mar-27-08).pdfwith suboptimal transthoracic imaging, performance of a TEE will allow for a more precise evaluation of ventricular

Documents