Three years’ experience of focused cardiovascular ultrasound in the peri-operative period

Authors


Dr Brian Cowie
Email: brian.cowie@svhm.org.au

Summary

Ultrasound applications in peri-operative medicine have become common place in modern anaesthesia practice. Anaesthetists have performed transoesophageal echocardiography in cardiac and selected non-cardiac surgery for over two decades. We aimed to assess the indications, impact on clinical management and accuracy of focused cardiovascular ultrasound performed by anaesthetists in the peri-operative period. One hundred and seventy patients over a 3-year period had a focused transthoracic echocardiogram. Adequate images to answer the clinical question were obtained in 167 out of 170 patients (98%). The undifferentiated systolic murmur was the commonest indication (98 out of 170, 58%). Some degree of aortic stenosis was present in 47 out of 170 (26%) of patients; mitral valve disease (30 out of 170 (18%)) and pulmonary hypertension (25 out of 170 (14%)) were also common. Changes in peri-operative management occurred in 140 out of 170 (82%) patients and major findings correlated with a formal cardiology transthoracic echocardiogram in 52 out of 57 (92%) patients. Focused cardiovascular ultrasound performed by anaesthetists in the peri-operative period accurately detects major cardiac pathology and significantly alters peri-operative management.

Small portable ultrasound machines with transthoracic echocardiographic capability are now readily available in most operating rooms. Anaesthetists have performed transoesophageal echocardiography (TOE) in cardiac surgery for two decades and do so increasingly in non-cardiac surgery [1–3]. However, TOE is invasive and not without risk, with recent data suggesting a 1:1000 incidence of gastroesophageal injury and a 1:5000 incidence of death [4]. The use of focused or limited transthoracic echocardiography by anaesthetists in the peri-operative period has recently been described [5–7]. Anaesthetists are able to obtain diagnostic quality images in the majority of patients and focused echocardiography is feasible and alters patient management [5]. Traditionally, transthoracic echocardiography services have been provided by a formal cardiology based echocardiography unit, but many hospitals are unable to access a formal transthoracic echocardiogram at short notice, with waiting times as long as 10 days for inpatients [8] and over 20 weeks for outpatients [9]. After a preliminary pilot study to assess the feasibility of an anaesthesia run focused transthoracic echocardiography service [5], we aimed to assess the indications, impact on clinical management and accuracy of such a service in the peri-operative period.

Methods

Over a 3-year period commencing from May 2007, the Department of Anaesthesia at St. Vincent’s Hospital in Melbourne established an anaesthesia run peri-operative transthoracic echocardiography service. This project was approved by the St. Vincent’s Hospital Human Research Ethics Committee. Including the initial 50 patients in the pilot study [5], data were collected on 170 patients. Clinical indications for a focused transthoracic echocardiogram included, but were not limited to:

  • 1 Haemodynamic instability
  • 2 Undifferentiated murmur/valve disease
  • 3 Ventricular function assessment
  • 4 Dyspnoea/hypoxaemia
  • 5 Poor functional capacity
  • 6 Other (e.g. to estimate pulmonary artery pressure, investigate arrhythmia).

Patients were referred for a focused transthoracic echocardiogram by the primary anaesthetist involved with the case as necessary, based on the above indications. This echocardiogram occurred either pre-operatively in the pre-anaesthesia room, in the operating room itself or in the postanaesthesia care unit.

A focused study was performed by experienced cardiac anaesthetists with formal TOE training and qualifications, with additional training in transthoracic echocardiography. On each day, a separate anaesthetist was scheduled for the echocardiography service. All studies were performed in patients having non-cardiac surgery as intra-operative TOE is routine in our institution in cardiac surgery cases.

The focused echocardiography views and the qualitative and quantitative measurements obtained were at the discretion of the anaesthetist performing the examination, depending on the exact indication and the windows available in each patient. In general, some estimation of left and right ventricular size, volume and function (systolic and diastolic) were made. All four cardiac valves were examined and relevant measurements such as ejection fraction, aortic valve velocity/gradient, right ventricular systolic pressure, inferior vena cava size and collapsibility were recorded on a structured data sheet that formed part of the clinical report.

Management decisions that were influenced by the focused study were recorded. These included fluid and vasoactive drug administration, invasive monitoring, changes in anaesthesia technique, cancellation of the case or referral for a formal transthoracic echocardiogram. If a formal cardiology echocardiogram was performed, this result was compared with the focused study to assess the accuracy of major findings and any missed diagnoses.

Focused studies were performed on four different ultrasound machines, depending on their availability: the Acuson Cypress (Siemens Healthcare, Mountain View, CA, USA); the Vivid I (GE Medical Systems, Milwaukee, WI, USA); the M-Turbo (Sonosite, Bothell, WA, USA); and the iE33 (Philips Medical Systems, Andover, MA, USA).

Results

During the 3-year study period, approximately 30 000 cases were performed in the operating rooms and 170 patients underwent a focused transthoracic echocardiogram in the peri-operative period (approximately one study per 175 patients). Images of adequate quality to answer the clinical question were obtained in 167 out of 170 (98%) patients. In the three patients where image quality was non-diagnostic, morbid obesity and inability to position the patient in the left lateral position because of pain and fractured neck of femur were contributory. The median (range) age was 70 (16–96) years, with 86 men and 84 women.

Most focused studies were performed pre-operatively in the pre-anaesthesia room (126 out of 170 (74%)) with the remainder in the operating room (22 out of 170 (13%)), the postanaesthesia care room (20 out of 170 (12%)) and the ICU (2 out of 170 (1%)). The most common indication was the undifferentiated systolic murmur (Table 1); 47 out of 98 (48%) of these patients had at least some degree of aortic stenosis (Table 2). In addition, nine patients had aortic sclerosis with calcification and mild restriction of aortic leaflets, but no significant gradient across the aortic valve. Some element of mitral valve disease was present in 30 out of 170 (18%) patients. Evaluation of hemodynamic instability and ventricular function were the other most common indications (Table 1). A normal study was reported in 30 out of 170 patients (18%); this finding was most common after the development of atrial fibrillation or supraventricular tachycardia in the peri-operative period (10 out of 30, 33%).

Table 1.   Indications for focused transthoracic echocardiogram in 170 patients. Values are number (proportion).
Murmur detected/valve disease suspected98 (58%)
Haemodynamic instability37 (22%)
Ventricular function assessment17 (10%)
Dyspnoea/hypoxaemia6 (4%)
Other5 (3%)
Pericardium4 (2%)
Poor functional capacity3 (2%)
Table 2.   Echocardiographic diagnosis of patients with murmurs; 98 patients had a total of 112 lesions detected. Values are number of patients with detected pathology (proportion of total having focused transthoracic echocardiogram, n = 170). Grade of aortic stenosis as defined by the American Heart Association/American College of Cardiology [24]. Values are number (proportion).
Aortic stenosis (mild)28 (16%)
Aortic stenosis (moderate/severe)19 (11%)
Mitral regurgitation24 (14%)
Tricuspid/pulmonary valve disease13 (8%)
Aortic sclerosis10 (6%)
Aortic regurgitation8 (5%)
Mitral stenosis6 (4%)
Left ventricular outflow obstruction4 (2%)

Some change in patient management resulting from the focused study occurred in 140 out of 170 (82%) patients (Table 3). Most commonly, this change was a decision on the use of invasive blood pressure monitoring. All patients with moderate to severe aortic stenosis had insertion of a pre-operative arterial line when it would not have otherwise been clinically indicated. Arterial lines were also inserted in patients with unexpected mitral stenosis and severe pulmonary hypertension. In contrast, in patients with only mild stenosis, aortic sclerosis, aortic regurgitation, flow murmurs or isolated mitral or tricuspid regurgitation, invasive monitoring was avoided. Changes in fluid management were also common, particularly fluid boluses to patients found to be unexpectedly hypovolaemic; furthermore, in a number of patients, extra fluid was avoided due to apparent volume overload and ventricular dysfunction.

Table 3.   Changes in management following peri-operative focused transthoracic echocardiogram in 170 patients. Values are number (proportion).
Invasive monitoring avoided45 (24%)
Referred for formal cardiology echocardiogram34 (20%)
None30 (18%)
Fluid bolus administered25 (15%)
Invasive monitoring placed22 (13%)
Change in anaesthesia technique20 (12%)
Vasoactive drug administered12 (7%)
Postoperative recovery location altered12 (7%)
Procedure cancelled7 (4%)
Fluid restriction prescribed3 (2%)
Vasoactive drugs ceased3 (2%)

Changes in anaesthesia technique included the use of local anaesthesia and sedation in patients found to have previously undiagnosed severe aortic stenosis or pulmonary hypertension. If general anaesthesia was necessary, drugs known to be more haemodynamically stable were selected. Changes in postoperative management of patients occurred in 12 out of 170 (7%) patients, such as overnight admission of patients with previously unknown cardiac pathology initially scheduled day case surgery, or admission to ICU or the cardiac care unit after more major surgery.

Cancellation of the planned surgical case occurred in six patients with previously undiagnosed severe aortic stenosis, including one with combined aortic and mitral stenosis, and one patient with severe pulmonary hypertension with a right ventricular systolic pressure of 75 mmHg. All were deemed to be high-risk patients having non-essential surgery, and a decision not to proceed was reached after discussion with the surgeons and informed decision-making by the patients.

In patients who had a formal transthoracic echocardiogram, either by direct referral from a focused study or for other reasons, the formal echocardiography report and major findings were compared with the peri-operative focused study. Fifty-seven patients (34%) had a formal transthoracic echocardiogram and this finding confirmed the major clinical findings of the focused study in 52 out of 57 (91%) of these patients (Table 4). No patients with significant aortic stenosis, severe ventricular dysfunction (ejection fraction < 30%) or significant pericardial effusion were missed comparing formal cardiology with focused echocardiograms. Pulmonary hypertension was diagnosed in four patients on formal echocardiography that was not detected in the focused study. Each of these four patients had additional cardiac pathologies including aortic and mitral valve disease and severe left ventricular dysfunction. One patient was diagnosed with aortic stenosis in the focused study and formal echocardiogram additionally identified subaortic stenosis due to a subaortic membrane in the left ventricular outflow tract. Another patient was diagnosed with aortic stenosis in the focused study but had systolic anterior motion of the mitral valve with mitral regurgitation and left ventricular outflow obstruction on the formal echocardiogram. In three patients, the formal echocardiogram missed findings detected on the focused study (atrial septal aneurysm, atrial septal defect and pulmonary hypertension). In a further 10 patients, a focused study was requested specifically when the diagnosis had already been made, to assist guiding catheter placement in closed chest cardiac procedures. These patients have not been included in the data analysis, given the specific indication for the study. This indication included pigtail catheter placement in patients with pericardial effusion and pericardial tamponade, and also placement of an inferior vena caval cannula in a patient requiring extracorporeal membrane oxygenation.

Table 4.   Major diagnosis of formal transthoracic echocardiography compared with focused peri-operative study in 57 patients. The patient with SAM/LVOT obstruction and mitral regurgitation was diagnosed with aortic stenosis in the focused study. The patient with the subaortic membrane had aortic stenosis in addition to the membrane. Values are number (proportion).
 Formal transthoracic echocardiographyFocused study
  1. SAM/LVOT, systolic anterior motion of mitral valve with left ventricular outflow obstruction.

Aortic stenosis17 (30%)18 (32%)
Aortic sclerosis4 (7%)4 (7%)
Aortic regurgitation2 (4%)2 (4%)
Flow murmur2 (4%)2 (4%)
Mitral or tricuspid regurgitation2 (4%)2 (4%)
Pulmonary valve1 (2%)1 (2%)
SAM/LVOT obstruction1 (2%)0
Sub-aortic membrane1 (2%)0
Left ventricular dysfunction11 (19%)11 (19%)
Right ventricular dysfunction2 (4%)3 (5%)
Pleural/pericardial effusion2 (4%)2 (4%)
Pulmonary hypertension5 (9%)3 (5%)
Normal3 (5%)4 (7%)
Left atrial enlargement/atrial septal defect2 (4%)3 (5%)
Left ventricular hypertrophy2 (4%)2 (4%)

Discussion

This study supports the utility of focused transthoracic echocardiography by anaesthetists in the peri-operative period, with a significant number of patients having some alteration in their management as a result of information obtained from the study. It also supports the diagnostic accuracy of limited transthoracic echocardiography, with good correlation reported when compared with formal studies performed by the cardiology service.

Like many institutions, we are unable to organise a reliable, formal transthoracic echocardiogram at short notice for patients who present with undifferentiated murmurs or previously unrecognised cardiac symptoms and signs. With increasing numbers of patients being admitted on the day of surgery or presenting for emergency surgery, this situation will continue to be a challenge for anaesthetists.

In this series, the most common differential diagnoses of the undifferentiated systolic murmur were aortic stenosis, aortic sclerosis, aortic regurgitation or mitral regurgitation. Clinical differentiation of these pathologies is notoriously unreliable [10], with many patients being asymptomatic even with severe aortic valve disease [11]. Severe aortic stenosis is a major clinical predictor of peri-operative morbidity and mortality in non-cardiac surgery [12]. Previous data have shown that knowledge of severity of disease influences the choice of anaesthesia technique, invasive monitoring and use of vasoactive drugs [5, 13]. With aortic regurgitation, classic teaching suggests that these patients should have a predominantly diastolic murmur. However, all of our patients had a systolic murmur and this problem is increasingly recognised, presumably because of increased stroke volume during systole; a diastolic murmur may in fact be uncommon [14].

When faced with the dilemma of the undifferentiated systolic murmur, echocardiography is the definitive diagnostic modality. Without an echocardiogram, the anaesthetist may cancel or delay elective surgery, or proceed employing unnecessary invasive monitoring and changes in anaesthesia technique. A focused transthoracic echocardiogram, however, should not replace a thorough pre-operative assessment and examination. It may, however, confirm examination findings and provide rapid, point-of-care diagnostic information in the peri-operative period.

Unexplained peri-operative hypotension and cardiovascular collapse is a class one indication for TOE during non-cardiac surgery [1]. However, the patient can be similarly assessed non-invasively by using transthoracic echocardiography, with images obtainable in well over 90% of critically ill patients in the ICU and the operating room [15]. In this emergency situation, there is rarely time to call for a formal transthoracic echocardiogram and information is urgently required regarding volume status, filling pressures, ventricular function and response to therapy. A focused study is our first-line investigation in this situation and has replaced TOE.

This study has also demonstrated that the focused study, generally performed in around 10 min [15], correlates with major echocardiographic findings on formal echocardiography in over 90% of cases. This finding is consistent with our early data [5] and those in other limited transthoracic echocardiography series [16–18]. Estimation of the right ventricular systolic pressure is now a routine part of our focused studies, after our preliminary experience showed this systolic pressure could be missed if not specifically looked for, particularly when associated with other cardiac pathology. A dilated and dysfunctional right ventricle, septal shift and significant tricuspid regurgitation are not always present and even experienced cardiologists can significantly under or overestimate right ventricular systolic pressure and pulmonary artery pressure [19]. The presence of a subaortic membrane in combination with aortic stenosis, and with a similar gradient, is a subtle sign and had no additional implications for anaesthesia compared with aortic stenosis alone. Left ventricular outflow tract obstruction with systolic anterior motion of the mitral valve and coexistent mitral regurgitation can be confused with aortic stenosis, particularly if aortic valve views are suboptimal and continuous wave Doppler evaluation catches part of the mitral regurgitant jet, overestimating the aortic velocity and gradient. Importantly, both of these patients were referred by the anaesthetists for formal transthoracic echocardiography and this procedure is essential in situations where imaging is suboptimal, if pathology or echocardiographic signs are unusual, when the peri-operative echocardiographer is inexperienced, and when further postoperative follow-up is required. Echocardiography has some element of subjectivity with inter- and intra-observer variability. When in doubt, a second opinion or formal echocardiography is always indicated.

A focused study or limited echocardiography does not replace a formal transthoracic echocardiogram. However, in the peri-operative setting, where rapid diagnostic data are required to answer a particular clinical question, detailed analysis of a number of parameters (for example venous flow in all four pulmonary veins) is likely to be time consuming and is unlikely to add significantly to immediate clinical management. How focused a limited echocardiogram should be in the peri-operative period will depend on the specific indication for the study and the available acoustic windows. In our view, this focus should include qualitative estimates of right and left ventricular function and an estimate of aortic valve velocity/gradient, right ventricular systolic pressure and intravascular volume status as minimum requirements in all patients.

Despite clearly altering patient management, it is unclear whether these changes actually influenced patient outcome. Just as it seemed intuitive that monitoring patient physiology with a pulmonary artery catheter and altering management based on these variables would be beneficial for patients, monitoring is unlikely to improve outcome unless it is directly linked to a proven therapy. There are no data that support invasive arterial blood pressure monitoring, fluids or vasoactive drugs in peri-operative hypotension or a particular anaesthesia technique in patients with aortic stenosis, yet these decisions are fundamental in our daily practice. There is data that TOE alters patient management in cardiac surgery [20], but there have been no randomly allocated controlled trials or studies with control groups and there are no data that it convincingly improves patient outcomes [21]. It is unlikely these studies will ever be done, now that technology exists to evaluate cardiac structure and function and valvular integrity, and few centres would undertake valvular repair without it. Transthoracic echocardiography is able to acquire almost all of the information obtainable with a pulmonary artery catheter or TOE, but in a non-invasive way with no known risks.

Outcome data on emerging technology in anaesthesia are sadly lacking and to evaluate transthoracic echocardiography, specific management based on echocardiographic findings would be required. One example would be to look at the use of transthoracic echocardiography in patients with fractured neck of femur, most of whom are elderly with multiple comorbidities. Almost 50% of unselected elderly patients have ejection systolic murmurs and 26% have some element of aortic stenosis [22]. These patients could be managed with and without pre-operative and peri-operative transthoracic echocardiography. Strict protocols on invasive monitoring, anaesthesia technique, fluid management and use of vasoactive drugs would be required, and outcome variables would need to be clearly defined. This study would be challenging, but not impossible, and has been performed in high-risk surgical patients with the pulmonary artery catheter [23].

Focused transthoracic echocardiography performed by anaesthetists in the peri-operative period allows diagnosis of undifferentiated systolic murmurs, unexplained hypotension and cardiovascular collapse. It allows stratification of risk with alterations in invasive monitoring, anaesthesia technique and location of postoperative monitoring. It assists rational use and response to fluid and vasoactive drugs and guides interventional percutaneous cardiac procedures. Major clinical findings correlate with those obtained in a formal echocardiogram in over 90% of cases. When indicated, focused transthoracic echocardiography alters peri-operative management of most patients.

Competing interests

No external funding or competing interests declared.

Ancillary