Following high-risk cardiac surgery, optimization of fluid and hemodynamic status is fundamental during perioperative patient care. Historically, central venous pressure (CVP) and pulmonary capillary wedge pressure (PCWP) as determined from pulmonary artery catheters (PACs) were surrogates for ventricular volume to guide fluid resuscitation; however, in recent years, studies have shown that CVP and PCWP do not always have a predictable correlation to ventricular preload and do not adequately reflect the change in intravascular volume following fluid administration.1 The failure of pressure measurements to correlate with ventricular volume is secondary to changes in myocardial compliance resulting from cardiac surgery.1 The relationship between pressure, volume, and flow is further altered in the setting of mechanical circulatory support—including ventricular assist devices (VADs) and extracorporeal membrane oxygenation (ECMO). Although most decisions for straightforward patients can be made with readily available information, complex patients with acute situations may need more comprehensive investigations. Another modality of cardiac monitoring is the visualization of the cardiac silhouette through echocardiography.2 In the cardiac operating room, anesthesiologists make hemodynamic assessments and subsequent interventions using transesophageal echocardiography (TEE).3 Imaging obtained from TEE allows for an accurate assessment of cardiac function and volume status. However, while TEE is readily available in the operating room, obtaining one in the intensive care unit (ICU) requires additional time and personnel. In the ICU, TEE serves as the gold standard for determining the cause of hemodynamic instability,4 although the assembly of additional equipment and staff can delay the intervention. Meanwhile, if clinicians have a high suspicion that the TEE will reveal a possible indication for surgical intervention (eg, cardiac tamponade), the operating room is commonly placed on standby for emergent surgery while the clinician awaits the study.
To overcome the widespread lack of availability of conventional TEE, a miniaturized TEE probe (ImaCor Inc., Garden City, NY) with a diameter of 5.5 mm was developed. The hemodynamic TEE (hTEE) probe is inserted transorally and can transmit images of the cardiac silhouette from its resting place in the esophagus. The images allow for an immediate assessment of cardiac function and volume status, while, at the same time, intensivists may make additional decisions to improve a patient’s hemodynamics because the probe can remain indwelling for up to 72 hours. In this report, we describe our experience using the hTEE probe and review its ability to be used as a point-of-care device to guide management or make the appropriate diagnosis
From May 2011 through June 2012, a total of 490 patients were admitted to our ICU. The ImaCor hTEE system was used as a point-of-care device in 61 (12%) patients. Patients who had the hTEE probe used experienced a triggering event during which conventional methods of assessing hemodynamics (Swan-Ganz catheter, continuous Sv02 monitoring) were unsuccessful, and they failed to respond to initial therapy. hTEE probes were used in this subset of patients to provide the clinician with additional information to assist in making the diagnosis and were not routinely used in all ICU patients.
After attaining approval by the local institutional review board, patient demographics and hTEE images were retrospectively entered into a structured database. The majority of patients who underwent hTEE imaging had an arterial line and either CVP or a PAC monitoring system.
All 61 patients who required the hTEE probe were separated based on demographics into three groups: patients with ECMO, patient’s status post-VAD insertion, and other (Table 1). There were 25 ECMO patients with venoarterial or venovenous ECMO. Patients were cannulated according to our standard practice to either VV or VA, depending on the etiology of the decompensation, and the hTEE probe was not used for ECMO placement. Of the 25 patients, 21 were placed on VA and four patients on VV. All VA cannulation was either fem–fem (n = 19) or central (n = 2), while all VV cannulation was via right internal jugular vein using a BiCaval dual-lumen cannula (n = 4). The VAD (HeartMate II, Thoratec, Pleasanton, CA) group had six patients, and lastly, hTEE was used in 30 postoperative patients designated as “other.” This last group included patients after coronary artery bypass grafting (CABG, n = 8), heart transplant (n = 5), valve replacement/repair (aortic valve replacement n = 2, mitral valve replacement n = 1), thoracic surgery (decortication n = 1, lobectomy n = 1), aortic surgery (type A repair n = 5, root replacement n = 1), acute heart failure/myocardial infarction (n = 4), and other general surgery (n = 2).
The ImaCor miniaturized hTEE probe has a diameter of 5.5 mm with 15 cm of penetration at 6.67 MHz (B-mode). The probe was inserted transorally and is Food and Drug Administration (FDA) approved to remain indwelling for up to 72 hours.
Monoplane imaging with the hTEE system consists of two views: a midesophageal four-chamber view and a transgastric short-axis view (Figure 1). Insertion of the probe and image interpretation can be performed by any hTEE-trained intensivists and does not necessitate a cardiology or radiology consult. To attain hTEE competency, a specifically designed training course has been created for the hTEE system. After successfully obtaining images, the probe was removed unless there was a need for continuous monitoring of the patients’ hemodynamics. The echocardiography results were concurrently reviewed by our hTEE-trained intensivists at the bedside. The diagnosis was made while obtaining the images, and appropriate interventions were performed based on the imaging.
Our results are summarized in Table 2, and the details are discussed below.
Extracorporeal Membrane Oxygenation
The ECMO group consisted of 15 patients with hemodynamic instability and 10 patients who were weaned off ECMO using the hTEE probe. The patients with hemodynamic instability were further subdivided into three subgroups: myocardial dysfunction, pericardial dysfunction, and device/circuit dysfunction.
Myocardial dysfunction group (n = 5).
hTEE demonstrated stunned myocardium in three patients, and inotropes were initiated; two patients recovered cardiac function, while one did not and care was subsequently withdrawn. For the last two patients in this category, hTEE imaging revealed continued LV dysfunction with adequate right ventricular (RV) function; thus, therapy was adjusted and continued until LVAD insertion.
Pericardial dysfunction group (n = 2).
In the first patient, hTEE demonstrated a pericardial mass that was biopsied and revealed an adenomatous tumor compressing the heart; the second patient had a large pericardial clot that was evacuated in the operating room.
Device dysfunction group (n = 8).
In three patients hTEE imaging revealed grossly under filled RA/RV, treated with aggressive volume infusion under hTEE. Three patients had imaging showing an intra-atrial clot, and all had anticoagulation increased, while the clot was observed for resolution with hTEE imaging. The last two patients in this subgroup had imaging revealing malpositioning of the ECMO cannula and required adjustment under hTEE.
Weaning was performed under hTEE guidance to evaluate cardiac function and volume status as the flow rate of ECMO was reduced and inotropes were administered in 10 patients. The hTEE findings demonstrated complete recovery of biventricular function in seven patients, and they were subsequently decannulated without incident. In one patient, hTEE showed RV recovery but continued LV dysfunction; thus, the patient remained on the ECMO circuit until a permanent LVAD was placed. Two patients had imaging revealing continued right ventricular failure and failed the ECMO wean protocol. Both patients were not VAD or transplant candidates; thus, subsequently, care was withdrawn.
Ventricular Assist Device
Of the hemodynamically unstable VAD patients (n = 5), two patients had the probe placed suspecting an inflow obstruction of the device limiting VAD function and hemodynamics. In the first case, hTEE showed a small thrombus in the LV cavity, and the patient was treated with anticoagulation instead of a take-back of the patient to operating room for exploration. The other patient’s imaging did not reveal a thrombus, which later was confirmed by conventional TEE; the patient was treated conservatively without surgery. Two VAD patients had hTEE imaging showing RV failure, which was demonstrated by a dilated right ventricle with a shift of the interventricular septum on the four-chamber view. Treatment was initiated with inotropes and by decreasing the VAD flow rate, while hTEE was continuously used for confirmation. The fifth patient in this subgroup had hTEE visualization of vegetation on the ICD lead that was subsequently removed in the operating room, and the patient recovered. The last patient in the VAD group experienced multiple LVAD alarms on postoperative day 1 without any reasonable cause. The hTEE probe was placed suspecting inflow obstruction due to intracardiac thrombus; however, imaging failed to show any possible cause. Therefore, it was determined that the VAD had an internal error, the patient was placed on ECMO, and the malfunctioning VAD was removed.
The third group consisted of a mix of surgical patients. The majority of probe utilizations were in patients with profound hemodynamic instability (n = 12) and to rule out tamponade (n = 14). Two smaller groups, possible pulmonary embolism (PE) (n = 2) and intra-aortic balloon pump (IABP) wean (n = 2), conclude this category.
Of the 12 patients in the hemodynamic instability subgroup, hTEE diagnosed six patients with RV or LV dysfunction which resolved with inotropic support. In four patients, hTEE showed preserved RV/LV function and low volume status; therefore, fluid resuscitation was performed. One patient who presented with acute myocarditis and went into profound cardiogenic shock had hTEE imaging showing cardiac standstill necessitating emergent ECMO. In the last hemodynamically unstable patient, we suspected a perivalvular leak; however, the hTEE was able to rule out this diagnosis and avoid a surgical exploration.
Among the 14 patients in whom we suspected tamponade, hTEE diagnosed mediastinal tamponade in five and hemothorax in three. The five patients who were diagnosed with mediastinal tamponade were taken to the operating room for emergent evacuation of a hematoma, and the three patients with a hemothorax were treated at the bedside through aggressive chest tube management. In six additional patients, we suspected cardiac tamponade due to clinical manifestations; however, hTEE eliminated the diagnosis of tamponade, and operative intervention was avoided.
Before withdrawal of an IABP, two patients were assessed using the hTEE probe for cardiac recovery. One showed continued dysfunction, while the other had marked improvement, and the balloon pump was removed. There were two general surgery patients who both had respiratory distress with cardiac instability and warranted a PE work-up. Both patients had an allergy to contrast dye and were hemodynamically unstable for a ventilation–perfusion scan. The hTEE probe was inserted and the pulmonary artery/right ventricle analyzed to look for dilation and signs of increased pressure. One patient displayed dilation and stress of the right ventricle, and a dilated pulmonary artery therefore was started on anticoagulation, while the other patient did not and was managed successfully without it.
Transesophageal echocardiography has been used as a gold standard as a monitoring modality for patients undergoing open-heart surgery, especially in the operating room. However, in the setting of our ICU, the utilization of echocardiography has been limited by ease of accessibility and cost. TEE services are markedly limited by availability of certified personnel and equipment outside of routine daytime hours.4,5 Transthoracic echocardiography (TTE) has a more modest staff requirement, making it more accessible; however, the visualization of the ventricle can be limited by variability in a patient’s body habitus, chest tubes, edema, and/or technician’s ability.6 In addition, the imaging obtained by conventional TEE and TTE studies is limited, in that it represents only a single “snapshot” in time. It is impractical to ask the technician to remain bedside for a number of hours, while an intervention is performed to determine whether the therapy was a success or not. At the same time, repeating a study at a later time point would instill confidence in the clinician that the correct diagnosis and therapy had been rendered but would also add additional costs. To diagnose and treat complicated patients, these studies may need to be repeated three to five times at any hour and on an immediate basis; thus, ImaCor hTEE is a perfect fit for this kind of practice. While the hTEE probe provides a monopolar view versus the multiplane seen with conventional TEE, the probe allows for reliable and reproducible assessment of both left and right ventricular volume status and contractility with results comparable with conventional TEE.
The management of ECMO patients presents many unique challenges, a fundamental one being the lack of preexisting information about baseline cardiac function due to emergent placement of ECMO in life-threatening situation. This uncertainty of biventricular function is a clinical challenge for the intensivist in regard to exit strategies following neurologic and end-organ recovery while on ECMO therapy. Weaning of patients from ECMO therapy is most commonly performed without clear guidelines or algorithms and largely falls on the clinician to ascertain the patient’s cardiac function.7 We were able to develop an hTEE-based weaning protocol to make a real-time assessment of both LV and RV function over a 4- to 6-hour period using a staged approach of decreased ECMO support, volume support, and inotropes.8 This protocol allowed us to accurately predict successful weaning of ECMO to bridge to LVAD, medical therapy, or stagnant biventricular failure, leading to withdrawal of care. No patient predicted to have RV recovery had any right ventricular issues after LVAD insertion.
Any emergency hemodynamic instability on ECMO, especially low ECMO flow, would be disastrous, and these situations need to be addressed immediately to avoid delay of diagnosis that usually results in dismal outcomes. The availability of the disposable, point-of-care, hTEE probes allowed our intensivist to immediately access the etiology of hemodynamic instability in our ECMO patients. We were able to accurately identify cannula malposition or clots in the venous cannula by hTEE without any delay, and appropriate interventions were performed. Using this technology, not only were we able to diagnose the problem but also able to visualize its resolution. No other point-of-care device currently exists that can be used by the cardiac intensivist for a duration of time that can be accessed at any time of the day for accurate monitoring of the ECMO patient.
In our LVAD population, by identifying septal position and ventricular filling, we could make real-time adjustments to optimize device speed. Patients with distended left ventricles could be unloaded. In patients demonstrating device dysfunction, those demonstrating thrombus at the inflow conduit were appropriately treated with anticoagulation, while those with no inflow occlusion were managed by observation.
In other cardiothoracic surgery patients with hemodynamic instability, we found a benefit using hTEE to visualize ventricular function and to guide appropriate interventions. Since most of the valvular issues were corrected by surgery or were normal before surgery, the postoperative hemodynamic issues were most likely related to left and right ventricular filling and function, which is easily detected by hTEE.
Cardiac tamponade is an emergent situation and may need surgical intervention. At our institution, the diagnosis of tamponade is first a clinical diagnosis that frequently necessitates definitive confirmation through an imaging study (TEE or hTEE). Although patient hemodynamics were suspicious enough to make a presumptive diagnosis of tamponade in 14 patients, only 57% of the study patients had evidence of tamponade based on confirmatory hTEE imaging. The five patients who had pooled blood in the pericardial space underwent surgical intervention for decompression, while the three patients with a hemothorax were managed with aggressive chest tube drainage under continuous hTEE observation. Utilization of the hTEE technology not only allowed us to successfully diagnose tamponade but also served as a helpful tool during the management phase of the complication.9
During the duration of time for this study, our center performed >200 individual studies on the 61 patients. The probes were developed to remain indwelling for 72 hours and approved by the FDA for this duration of time. While the probe can remain indwelling for up to 72 hours, our team decided to remove the probe after each use only to replace it when/if needed during the 72-hour window. If and when this window expired, a new probe was used. In most circumstances, it is not practical to leave a conventional TEE probe indwelling in the esophagus to obtain continuous imaging for 72 hours. The hTEE probe provides an advantage in its ability to continuously provide hemodynamic data. While the hTEE technology was not specifically designed with the ECMO population in mind, through its application in our patient population, we found it invaluable to augment their complex management. Clinicians have the option to place the probe and leave it for 72 hours or remove and replace if/when needed during the 72-hour window. Our practice is not to use the probes continuously but to use the same probe multiple times during each 72-hour session.
During this study period, there were no complications resulting from placement of the probe or its use. The complication rate for conventional TEE in ICU patients is reported to be 1.6%.10 The risk of complications at insertion is of significance in the setting of an edematous patient common in the postoperative cardiac setting, where multiple insertions of the larger, standard TEE probe represent a higher risk of potential esophageal perforation. Contraindications for the use of the hTEE probe include, but are not limited to, patients with an oropharyngeal injury, esophageal injury or obstruction, recent esophageal surgery, and tracheoesophageal fistula.
Our study is a nonrandomized retrospective chart review and thus has with it many limitations. The patients who underwent hTEE imaging at our institution were selected by the individual intensivists and were not randomized. A randomized trial looking at patient outcomes comparing patients managed with hTEE versus conventional methods has yet to be done and would provide more objective data on the use of this technology. While it is the opinion of the authors of this article that this new technology has a cost saving associated with the use of hTEE in the ICU setting of hemodynamic instability, a detailed cost analysis still needs to be done comparing the use of hTEE and traditional TEE.
There was a learning curve to master the instrumentation of the hTEE; however, the simulation-based training was helpful to remove the staff–instrument barrier. Since the instrumentation and analysis were simple enough to perform, most residents were comfortable using the hTEE equipment even during the nighttime hours. House staffs found this new technology to be helpful in gathering data and to assist in their decision-making process.
The hTEE device has proven to be an invaluable adjunct in the ICU. The hTEE probe allows a clinician, when faced with a complicated patient who is not responding to traditional therapy, to obtain at a moment’s notice direct imaging of the heart. Our experience using the hTEE probe is not meant to present this device as a replacement for the conventional TTE or TEE; however, in certain situations, it can prove to be a life-saving adjunct to the armamentarium available to the ICU staff.