Initial Clinical Trial of Substernal EpicardialEchocardiography: SEEing a New Window to thePostoperative HeartAnthony P. Furnary, MD, Carmelindo Siqueira Jr, MD, Robert I. Lowe, CCP,Timothy Thigpen, BS, YingXing Wu, MD, and H. Storm Floten, MDProvidence St. Vincent Medical Center, and Division of Cardiothoracic Surgery, Oregon Health Sciences University, Portland,Oregon

Background. Postoperative echocardiography windowsare often of poor quality because of intervening airspaces around the heart and great vessels. We assessedthe utility of a new commercially available adult chestdrainage tube that has been modified with the additionof a sterile sleeve to accommodate the introduction of anonsterile pediatric transesophageal echocardiography(TEE) probe. The TEE probe lies in a substernal epicar-dial position on the postoperative heart allowing one toperform substernal epicardial echocardiography (SEE).

Methods. Informed consent was obtained from 21 con-secutive adult cardiac operation patients. At the comple-tion of the operation the SEE drainage tube was insertedthrough the rectus muscle and into the pericardium.After chest closure, all patients underwent a full echoexamination using an Acuson pediatric biplane probe inthe SEE position. Views obtained and ease of insertionwere judged on a 1 to 10 (worst to best) scale.

Results. Full SEE examinations were completed in anaverage of 12 minutes. Ease of probe entry and manipu-lation was excellent (ratings of 9.3 and 9.6, respectively).The quality of the anatomic images was also excellent.Substernal epicardial echocardiography tube positioningwas integral to the orientation of the images obtained.There were no complications related to the placement ofthe SEE tubes or TEE probes. In 4 of 21 patients (19%) theSEE methodology was used serially in the intensive careunit to accurately assess ventricular function and fillingduring weaning of an intraaortic balloon and inotropicagents.

Conclusions. Substernal epicardial echocardiographyis a safe and highly effective methodology for the serialechocardiographic assessment of the postoperative heart.

(Ann Thorac Surg 2001;72:S1077–82)© 2001 by The Society of Thoracic Surgeons

Serial postoperative assessment of the cardiac surgicalpatient’s ventricular function is of paramount im-

portance to the assurance of a successful outcome [1].Current modalities for postoperative cardiac assessmentinclude catheters placed in the right atrium, pulmonaryartery, or left atrium. These catheters can be used todirectly monitor intracardiac pressures—central venouspressure, pulmonary artery pressure, pulmonary capil-lary wedge pressure, and left atrial pressure—fromwhich ventricular and valvular function can be second-arily deduced. In addition, mixed venous oxygen satura-tion (MVO2) can be directly measured from blood drawnfrom a pulmonary artery catheter. Pulmonary arterythermodilution catheters (Swann-Ganz) are used fre-quently in the postoperative period to obtain secondaryestimates of cardiac output and index.

These primary pressure measurements, combinedwith the secondary estimates of cardiac output or MVO2,allow the physician to further deduce tertiary estimates

of ventricular function and mitral valve competence, leftventricular end diastolic pressure, and the probability oftamponade. These tertiary deductions are dependent onseveral assumptions related to the cardiovascular systemincluding the absence of intracardiac, pulmono-systemic,and peripheral shunts; the absence of tricuspid regurgi-tation; and the presence (in the case of MVO2) of aeu-metabolic state [2].

With the advent of modern echocardiographic (echo)techniques, physicians no longer need to rely on physi-ologic assumptions and derive tertiary estimates of ven-tricular and valvular function. Transthoracic and trans-esophageal echo (TEE) techniques have long given thecardiologist a direct view of ventricular and valvularfunction as well as a view of the pericardial space. Thus,in the nonoperated chest, a primary (direct) analysis ofcardiac function can be made and not estimated orassumed.

However, current postoperative cardiac echo imagingis not optimal, primarily because air trapped within thepostoperative mediastinum creates poor acoustic win-Presented at the Seventh Annual Cardiothoracic Techniques and Tech-

nologies Meeting 2001, New Orleans, LA, Jan 24–27, 2001.

Address reprint requests to Dr Furnary, Division of CardiothoracicSurgery, Oregon Health Sciences University, Mailcode L353, 3181 SW SamJackson Park Rd, Portland, OR 97201; e-mail: tfurnary@starrwood.com.

Drs Furnary and Lowe disclose that they have afinancial relationship with Medtronic, Inc.

© 2001 by The Society of Thoracic Surgeons 0003-4975/01/$20.00Published by Elsevier Science Inc PII S0003-4975(01)02952-6

dows through which ultrasound waves cannot pass. Bothtransthoracic and TEE images are severely degraded.Postoperative TEE images are better than transthoracicimages, but the implementation of TEE is comparativelycomplex, requiring sedation and a trained physicianechocardiographer. Still, postoperative TEE images of theright heart, ascending aorta, arch, and pericardium aresuboptimal.

If an ultrasound transducer could be placed on theepicardium of the postoperative heart, the intervening airinterface problems would be obviated. This has beenaccomplished in the animal laboratory where ultrasonicpiezoelectric crystals are sewn to the epicardium forpostoperative monitoring [3]. However, these piezoelec-tric crystals cannot be repositioned or manipulated tooptimize the view. Another solution attempted by aJapanese surgeon was to sterilize a TEE probe and insertit through a subcostal incision [4]. This technique wascostly, manipulation was cumbersome, and sterility re-mained an issue, but the images were excellent.

An alternative, presented here, is to modify a chestdrainage tube to accommodate a TEE probe. The TEEprobe would lie within a closed “sock” in an epicardialposition on the underside of the chest drain. The epicar-dial positioning would eliminate air interface distortionand the chest tube sock would allow rotational andvertical manipulation of the TEE probe. This new echomodality has been termed substernal epicardial echocardi-ography (SEE) [5]. This initial clinical trial of theMedtronic (Minneapolis, MN) SEE imaging technologydevice (IT) drainage system was designed to evaluate andelucidate the quality and utility of SEE images obtainedusing a pediatric TEE probe. Special attention was paid topositioning of the SEE IT in relation to the quality of theimages obtained.

Material and Methods

A modified IT drainage system was designed and createdto enable the performance of SEE with a pediatric TEEultrasound probe. The Medtronic SEE cannula model19611 (SEE IT) is a mediastinal drain coupled with aseparate, collapsible, closed-ended ultrasound trans-ducer sleeve (Fig 1). The product was intended for use indrainage of the pericardial or thoracic cavity while simul-taneously creating an insertion port or pathway for anultrasonic transducer probe (Fig 2). The cannula is con-structed entirely of silicone to provide a soft flexibleprofile and to minimize any attenuation of the acoustic

Fig 1. Features of the Medtronic sub-sternal epicardial echocardiographyimaging technology device (SEE IT).Clockwise from upper left: Multiori-fice 36F flat silicone drainage tube;distal SEE IT end showing integral;underlying echo sock; suture tabs; andinjection port for introduction of inter-face coupling media into distal echosock.

Fig 2. (A) Close-up of the substernal epicardial echocardiographyimaging technology device (SEE IT) with an Acuson transesopha-geal pediatric biplane inserted into the distal tip of the Echo sock.(B) Overview of the SEE CT with the Seimens Acuson pediatric bi-plane in place.

S1078 CTT SUPPLEMENT FURNARY ET AL Ann Thorac SurgSUBSTERNAL EPICARDIAL ECHOCARDIOGRAPHY 2001;72:S1077–82

signal. Suture tabs on the drain are used to anchor thecannula, and depth marks are provided to indicate posi-tion and any tube migration. The transducer sleeve issized for a pediatric TEE probe and is coated with ahydrophilic coating to minimize friction. The female Luerconnector permits the introduction of sterile water orsaline into the distal end of the transducer sleeve toactivate the hydrophilic coating and to provide an echo-coupling medium. Representative samples of the SEEcannula were qualified through rigorous tests to ensuresafety and efficacy. These preclinical tests included sim-ulated usage under typical environmental conditions andoverstress exposures to ensure hydrophilic coating dura-bility, functional longevity, and product structuralintegrity.

Twenty-one consecutive adult cardiac surgery patientsoperated on by a single surgeon were enrolled in thestudy after obtaining informed consent. After the cardiacportion of the operation was completed, and before theplacement of sternal wires, the SEE IT was insertedthrough a 2.5- to 3-cm subcostal incision, through therectus abdominus and into the pericardial space. Therectus fascia was dilated with a Kelly clamp beforeinsertion of the SEE IT to facilitate passage of the pedi-atric TEE probe through the abdominal wall. Immedi-ately before insertion of the pediatric TEE probe into theSEE sock, 20 mL of sterile water was injected through theultrasonic coupling media port to activate the hydrophiliccoating.

The location (left, right, midline) and direction (hori-zontal, oblique, vertical) of the incision was recorded, aswas the location of the distal tip (right ventricle, rightventricular outflow tract (RVOT), pulmonary artery,aorta, right atrium) of the SEE IT. After sternal and skinclosure, echocardiographic examinations were per-formed. In those patients in whom adult multiplane TEEprobes (Aligent adult omniplane probe model #21364-A,powered by Aligent’s model 5500 [Aligent Corporation,Palo Alto, CA]) were placed intraoperatively for cardiacmonitoring, a postoperative closed chest TEE examina-tion was performed for comparison with SEE. A full SEEechocardiographic examination was attempted throughthe SEE CT using a pediatric biplane TEE probe (Sei-mens, Inc; Acuson model #V-705B) powered by an Acu-son Aspen cardiac ultrasound unit. The SEE and TEEexaminations were performed by an experienced echo-cardiographer who then graded the quality and clarity ofthe specific anatomic images obtained (on a scale of 1 to10, from worst to best). The length of time required toperform the SEE examination, and the ease of SEE probeinsertion and manipulation were graded. Data werecollected on scannable forms and imported directly into adatabase. Statistical analyses (mean 6 SEM, paired-samples t tests and independent samples t tests) wereperformed using SPSS 10.0 (SPSS, Chicago, IL).

Results

The 21 consecutive adult patients enrolled in the studywere operated on between December 11, 2000, and Jan-

uary 17, 2001. There were 18 coronary artery bypassoperations, 3 aortic valve operations, and 1 mitral valverepair. Two of these 21 patients had an intraaortic balloonpump (IABP) placed preoperatively for cardiogenicshock. Nine patients had concomitant adult multiplaneTEE probes placed intraoperatively for comparison withpediatric biplane SEE. There were no perioperative com-plications (infection, tamponade, bleeding) related to theplacement of the SEE IT. The hydrophilic coating on theSEE IT resulted in an excellent mediastinal drainagetube, which did not allow clotting to form on the walls ofthe tube (thus manual stripping was not required).

Performance of Substernal EpicardialEchocardiographyThe SEE IT provided rapid and facile access to thesubsternal epicardial position required for performanceof SEE. Ease of passage of the ultrasound transducer“into” the mediastinal space (ease of insertion) wasgraded as excellent (9.3 6 0.2). Ease of probe manipula-tion within the hydrophilic-coated sock was also superb(9.7 6 0.1). The average length of time for performance ofa full echocardiographic examination by SEE was 12 6 2minutes. No arrhythmias were observed during the ma-nipulation of the TEE probe within the SEE sock. Serialpostoperative SEE studies in the intensive care unit wereperformed in 4 of 21 (19%) patients. These additional SEEexaminations were clinically indicated to definitivelyassess ventricular filling and contractility in critically illpatients being weaned from inotropic or IABP support.Three of these 4 patients were awake and extubatedduring their postoperative serial SEE examinations.

Substernal Epicardial Echocardiography ImagesThe results of the SEE image scores are presented inTable 1. In general the images obtained from SEE wereexcellent, with an average score of 8.8 6 0.2. Left rectuspositioning of the insertion site was found to be superiorto right rectus positioning (average scores 9.1 6 0.17versus 7.1 6 0.71, p 5 0.001). Higher-quality images of thegreat vessels were also obtained when the distal end ofthe SEE IT was positioned over the distal ascending aortaor innominate vein. Thus, the optimal positioning of theSEE IT to produce the highest-quality images with abiplane pediatric probe is from a left rectus insertion siteinto the pericardial space along the right side of theventricular septum (parallel to the left anterior descend-ing artery) over the RVOT, between the aorta and pul-monary artery, with the tip anterior to the innominatevein.

Comparison With Adult Multiplane TransesophagealEchocardiographySix patients who received left rectus positioning of theSEE IT also had postoperative TEE examinations avail-able for direct comparison (Table 2). This comparisonrevealed that extreme posterior structures (left atriumand pulmonary veins) were visualized better with theadult multiplane TEE. However, pediatric SEE images ofthe majority of intracardiac structures were better than or

S1079Ann Thorac Surg CTT SUPPLEMENT FURNARY ET AL2001;72:S1077–82 SUBSTERNAL EPICARDIAL ECHOCARDIOGRAPHY

equal to adult TEE. In particular, visualization of thepostoperative pericardial space, right ventricle, rightatrium, tricuspid valve, anterior wall of the left ventricle,and intraventricular septum were superior with pediatricSEE as compared with adult TEE.

Comment

The echocardiographic evaluation of the postoperativeheart is technically difficult because of ultrasonographicinterference by intervening air in the pericardial space.Because of this interference, postoperative transthoracicecho assessment of the heart is virtually impossible.Transesophageal echocardiography is somewhat betterin quality but remains an invasive procedure requiringsedation, intubation, a trained echocardiographer, and asterile TEE probe. Substernal epicardial echocardiogra-phy may provide a cost-effective solution to these issueswhile at the same time resulting in higher-quality echoimages.

This initial clinical trial of the Medtronic SEE IT hasshown that the device and technique are safe and effec-tive. The SEE IT device allows for the rapid deploymentof a nonsterile TEE probe into the substernal epicardialposition. The anatomic detail obtained from these SEEimages was excellent. There were no SEE IT-associatedcomplications during the performance of SEE or after theremoval of the SEE IT.

The technique was also cost effective. The TEE probeused for SEE does not come in contact with either thepatient’s skin or body fluids, it only touches the sterileinner walls of the SEE IT sock. Therefore, the probe usedfor SEE need not be sterilized, thus decreasing prepara-

tion costs as compared with TEE. In addition, a singleTEE probe can be used serially in multiple patients. Theprobe may be simply wiped down with an isopropylalcohol-impregnated cloth between uses as an addedaseptic precaution.

The SEE IT was found to be an extremely effectivechest drainage tube. Blood did not coagulate along itsinner lumen and the tube drained freely without theneed for manual stripping. The tubes were judged by thecritical care nurses to be less painful to patients thanpolyvinyl chloride drainage tubes or even Blake drains.Removal was also easily accomplished with less painbecause of the slippery nature of the hydrophilic coating.

Images obtained through pediatric SEE were mostlyequal to or superior to those obtained through adult TEE.These favorable comparisons occurred despite the factthat the quality of the images was technically biased infavor of the adult TEE multiplane probes. These adultTEE probes have a longer built-in focal length than thepediatric probes. There are twice as many crystals in theadult TEE multiplane array as there are in the pediatricbiplane. Finally, the utility of the multiplane in obtainingdesired images cannot be overstated. Despite these sig-nificant biases, pediatric biplane SEE compared favor-ably with adult multiplane TEE. In general, anteriorstructures were visualized better with SEE, whereasposterior structures were visualized better with TEE. Inaddition, routine cardiac views of the heart were morerapidly achieved with SEE than with TEE. A full cardiacexamination can be completed easily in 8 to 12 minutes.

Table 1. Comparison of SEE Image Quality by Incision Site

Anatomy

Mean Score (SEM)

Left Right

Anterior LV wall 9.8 (0.10) 7.0 (1.73)Posterior LV wall 9.2 (0.23) 7.3 (1.76)Septum 9.9 (0.06) 7.7 (0.88)Lateral LV wall 9.7 (0.11) 7.3 (1.67)Mitral chordae 8.4 (0.28) 5.0 (1.00)Mitral leaflets 8.9 (0.27) 5.0 (1.00)Left atrium 8.5 (0.20) 6.0 (1.15)Pulmonary veins 6.6 (0.52) 4.3 (0.88)Aortic valve 9.3 (0.18) 8.0 (1.00)Ascending aorta 9.2 (0.39) 9.0 (1.00)Aortic arch 8.4 (0.63) 4.0 (3.00)Pulmonary artery 8.8 (0.47) 9.7 (0.33)RV free wall 9.7 (0.14) 9.0 (0.58)Tricuspid valve 9.6 (0.15) 9.0 (0.58)Right atrium 9.4 (0.16) 9.7 (0.33)Pericardial space 9.2 (0.25) 6.3 (2.33)Long axis 9.1 (0.26) 5.7 (0.88)Short axis 9.7 (0.11) 8.7 (0.88)

Image quality was rated on a scale of 1 to 10 (worst to best).

LV 5 left ventricular; RV 5 right ventricular; SEE 5 substernalepicardial echocardiography.

Table 2. Comparison of SEE With TEE Image QualityScores

Anatomy

Mean Scores(Standard Error)

p(2-tailed)SEE TEE

Anterior LV wall 9.9 (.14) 8.7 (.18) 0.005Posterior LV wall 9.0 (.31) 9.1 (.26) 0.356Septum 9.9 (.14) 8.4 (.37) 0.025Lateral LV wall 9.6 (.20) 9.1 (.34) 0.078Mitral chordae 8.1 (.59) 8.6 (.37) 0.289Mitral leaflet 8.9 (.40) 9.3 (.28) 0.078Left atrium 8.1 (.40) 9.1 (.26) 0.004Pulmonary veins 6.0 (.72) 8.4 (.36) 0.015Aortic valve 9.1 (.34) 8.9 (.26) 0.457Ascending aorta 8.4 (.95) 6.6 (.84) 0.297Aortic arch 7.3 (1.51) 5.4 (1.57) 0.515Pulmonary artery 8.0 (1.07) 5.7 (1.25) 0.291RV free wall 9.7 (.18) 6.9 (.46) 0.002Tricuspid valve 9.7 (.18) 7.6 (.48) 0.011Right atrium 9.4 (.30) 7.6 (.48) 0.032Pericardial space 9.0 (.44) 7.1 (.51) 0.039Long axis 9.2 (.48) 9.3 (.21) 0.695Short axis 9.5 (.22) 8.3 (.42) 0.110

Image quality was rated on a scale of 1 to 10 (worst to best). Pairedsamples t test, n 5 7.

LV 5 left ventricular; RV 5 right ventricular; SEE 5 substernalepicardial echocardiography; TEE 5 transesophageal echocardiography.

S1080 CTT SUPPLEMENT FURNARY ET AL Ann Thorac SurgSUBSTERNAL EPICARDIAL ECHOCARDIOGRAPHY 2001;72:S1077–82

Positioning of the SEE IT was found to be critical inobtaining optimal views with the biplane probe becauseof the fixed axes of the biplanar piezoelectric crystals inrelation to the SEE IT sock. Several positions were triedbefore finding the optimal course for the SEE IT. Theseincluded:

• Left rectus . parallel to septum . aorta• Right rectus . RVOT . pulmonary artery• Right rectus . RV . aorta• Right rectus . right atrium . right side of aorta.

Through trial and error it was found that an obliqueincision over the left rectus placed the SEE IT at or nearthe ventricular apex as the tube entered the pericardium.The optimal positioning was completed by a courserunning along the right side of the intraventricular sep-tum, across the RVOT, over the pulmonary valve, ante-rior to the ascending aorta or pulmonary artery with thetip of the sock lying just inferior to the innominate vein(Fig 3). The course along the anterior intraventricularseptum is important because it diminishes movement ofthe probe during the cardiac cycle. The addition of asingle pericardial reapproximation suture inferiorly alsohelped to improve motion artifact and increased visual-ization of the apical ventricle. One SEE IT position notincluded in this trial was inferior to the heart along thediaphragmatic surface. We thought that this positioningwould require the creation of a premolded, right-angledSEE IT for safe insertion and manipulation of the echoprobe within the sock. However, if implemented, thistype of positioning would give SEE views similar totransgastric TEE views.

Depending on the type of operation and the anatomicarea of postoperative interest, the SEE IT could bepositioned to look at specific structures. However, withthe release of a SEE IT that will accommodate a multi-plane adult TEE probe, positioning will become muchless critical. The advantages of the adult TEE probe in theSEE position will be wide current availability of probes,multiplane adjustments of cardiac imaging, deeper focaldepths, and even clearer images because of the increasednumber of crystals in the adult array. This will likelyeliminate the posterior advantage of TEE in visualizingthe left atrium and pulmonary veins. Disadvantages ofthe adult version of the SEE IT include a larger incisionand the possibility of more discomfort during the intro-duction of the probe into an awake patient.

The SEE IT was used to serially follow postoperativeventricular dysfunction in 19% (4 of 21) of the patients inthis study. In 3 of these patients serial examinations weredone while the patient was awake, extubated, and watch-ing the echo monitor. They each described a dull pres-sure-type pain as the probe head passed through therectus sheath. Once the probe head was in the medias-tinum, the pain ceased and the patients could comfort-ably watch their own examination. These patients wereon significant doses of inotropic agents and IABP support

Fig 3. Placement of the substernal epicardial echocardiography im-aging technology device for optimal postoperative cardiac views witha biplane echo probe.

Fig 4. First postoperative day substernal epicardial echocardiographysystolic images of the left ventricle with (top) and without (bottom)intraaortic balloon pump (IABP) support demonstrating depressedventricular systolic function with the IABP off. This patient was notweaned from IABP support at this time.

S1081Ann Thorac Surg CTT SUPPLEMENT FURNARY ET AL2001;72:S1077–82 SUBSTERNAL EPICARDIAL ECHOCARDIOGRAPHY

after urgent revascularization with poor preoperativeventricular function. Using daily SEE examinations wewere able to definitively wean patients off IABP andinotropic support with real-time visual assessment ofventricular function. Simply placing the IABP on standbyand visualizing changes in ventricular contractility (Fig 4)was an expeditious and effective method to evaluateunderlying native ventricular dysfunction. There is nowno longer a need to wean support and then wait for 1 to2 hours to check the cardiac index. Rather, this techniqueis more akin to placing a temporary pacemaker onstandby while assessing the underlying native rhythm onan electrocardiogram monitor.

During these serial postoperative SEE examinationswe found that the quality of the SEE images improveddramatically over the first 24 hours. This improvementwas thought to arise from the gradual diminution ofintrapericardial air. At 24 hours postoperatively the car-diac images were spectacular. All structures could beeasily visualized using a 7-MHz ultrasound transducer.

Like the placement of postoperative temporary pacingwires, the placement of a SEE IT provides an extrameasure of security in the management of the postoper-ative cardiac patient. Directed questions regarding ven-tricular function and filling can be answered easily anddefinitively. Pericardial tamponade can be ruled in or outquickly and accurately. Valvular function may also beassessed rapidly. As the placement of pacing wires doesnot require the initiation of cardiac pacing, so too theplacement of a SEE IT does not mandate utilization of thesock for an SEE examination unless it is indicated. TheSEE IT can be used simply as an effective mediastinaldrainage system, on standby for SEE ultrasound probeinsertion if needed.

This initial trial of the Medtronic SEE IT has shownthat the device is safe and effective for the echocardio-

graphic evaluation of the postoperative heart. The SEE ITaligns clinical incentives for use in the postoperativeperiod. For the surgeon it provides an expeditious “sur-gical” view of the heart. The device can be implementedeasily and allows for direct assessment of ventricularfunction and filling. This further enhances definitiveweaning of inotropic and IABP support. For the cardiol-ogist the SEE IT provides a new and accurate echoapproach. For the hospital administrator it is a cost-effective implementation of a common echocardio-graphic tool, which can now be used to efficiently weanventricular support. For the patient it has great potentialto provide better postoperative care.

We express our sincere appreciation to David Sahn, MD,Medtronic, Inc, and Acuson, a Siemens Company, for theirinsight and support of this study. We also thank Eric Johnson forhis invaluable assistance as research coordinator for St. VincentMedical Center.

References

1. Kyo S, Takamoto S, Matsumura M, et al. Immediate and earlypost operative evaluation of results of cardiac surgery bytransesophageal two-dimensional Doppler echocardiogra-phy. Circulation 1987;76:V113–21.

2. Headley JM. Invasive hemodynamic monitoring: applyingadvanced technologies. Crit Care Nurs Q 1998;21(3):73–84.

3. Lee KF, Dignan RJ, Parmar JM, et al. Effects of dynamiccardiomyopathy on left ventricular performance and myocar-dial mechanics in dilated cardiomyopathy. J Thorac Cardio-vasc Surg 1991;102(1):124–31.

4. Yozu R, Mori A, Shimizu H, et al. Development and clinicalapplication of a next generation implantable echocardiogra-phy probe for monitoring cardiac function under assistedcirculation after open heart surgery. ASAIO J 1994;40:m482–5.

5. Hanlon JT, Lowe RI, Furnary AP. Substernal epicardial echo-cardiography: a new ultrasonic window to the postoperativeheart. J Am Soc Echocardiogr 2000;13:35–8.

S1082 CTT SUPPLEMENT FURNARY ET AL Ann Thorac SurgSUBSTERNAL EPICARDIAL ECHOCARDIOGRAPHY 2001;72:S1077–82