Advice

Evidence review

Clinical and technical evidence

Three clinical studies were identified in which the hTEE system was used as an intervention for people who are haemodynamically unstable. These were a prospective observational study (Vieillard-Baron et al. 2013), a case series (Maltais et al. 2013) and a case report (Mykytenko et al. 2012).

The prospective observational study was a multicentre investigation of 94 patients with haemodynamic instability of any cause, across 4 intensive care units during a 4‑month period (Vieillard-Baron et al. 2013). The primary outcomes were feasibility of insertion, complications, image quality and influence on management. No failure of probe insertion was observed, however, in 17% of cases the nasogastric tube had to be removed to enhance probe insertion. Two patients had minor, self‑limited gastric bleeding that was reported to have negligible clinical consequence. In 1 of these patients, endoscopy revealed oesophageal ulceration. An additional 2 patients developed a mechanical ulceration of the superior lip, caused by prolonged contact with the probe. Image quality was judged as 'adequate or optimal' in 91/94 (97%) of cases in the superior vena cava view, 89/94 (95%) of cases in the 4 chamber view, and 86/94 (91%) of cases in the short axis view. However, all superior vena cava, mid‑oesophageal 4 chamber and transgastric short axis views were optimal in only 45% of patients. Haemodynamic assessment was not possible in 15% of patients, which the authors attributed to the technical limitations of the single‑patient use, miniature probe. The authors expressed the opinion that the ClariTEE probe had lower overall imaging quality than that usually obtained with a conventional transoesophageal echocardiography probe, although no direct comparisons were made. The mean duration for which the probe remained in place was 32 ± 23 hours, allowing for a mean of 2.8 ± 1.6 haemodynamic evaluations per patient (which led to a mean of 1.4 ± 1.5 therapeutic changes per patient). Of the 263 haemodynamic assessments made, 132 (50%) had a direct therapeutic effect in 62 patients (66%). A summary of this study is presented in table 1.

Table 1 Summary of the Vieillard-Baron et al. (2013) prospective observational study

Study component

Description

Objectives/hypotheses

To evaluate the haemodynamic monitoring capability and safety of the ImaCor single-use miniaturised TEE probe, when left in place for up to 72 hours in mechanically ventilated ICU patients being assessed for a cardiorespiratory compromise.

Study design

Prospective observational study.

Setting

Multicentre study during a 4-month period with 4 participating ICUs.

Inclusion/exclusion criteria

Inclusion criteria: all patients requiring mechanical ventilation for circulatory failure or acute lung injury/acute respiratory distress syndrome and who needed TEE for diagnosis and management. Circulatory failure was defined as the presence of sustained hypotension (systolic blood pressure less than 90 mmHg or a mean arterial pressure less than 65 mmHg) associated with clinical evidence of tissue hypoperfusion which needed the use of vasopressors or inotropes.

Exclusion criteria: at least 1 of the following: <18 years old, pregnant, contraindication for a TEE study.

Primary outcomes

Feasibility of probe insertion

Complications

Image quality

Influence on patient management.

Statistical methods

Continuous variables expressed as mean ± standard deviation.

Participants

n=94 ventilated, critically ill patients.

Of these: n=57 in septic shock, n=7 in cardiogenic shock, n=13 with post-resuscitation syndrome, n=4 in haemorrhagic shock, n=13 on ventilation for acute lung injury/acute respiratory distress syndrome.

Results

Feasibility of probe insertion

No failure of probe insertion was observed. It was inserted at the first attempt in 78/94 patients (83%) and the nasogastric tube had to be removed in the remaining 17 % of cases. No additional sedation or paralysis was necessary.

Complications

Two patients had minor, self-limited gastric bleeding without relevant clinical consequence. In 1 of these, upper endoscopy revealed an oesophageal ulceration. An additional 2 patients developed a mechanical ulceration of the superior lip, due to prolonged contact of the probe. Technical dysfunction of the probe precluded further imaging in 2 patients who had already been monitored for 48 hours.

Image quality

Image quality was judged as adequate or optimal in 91/94 (97%) of patients in the superior vena cava transverse view, 89/94 (95%) in the mid-oesophageal 4 chamber view, and 86/94 (91%) in the transgastric short axis view. However, all views were optimal in only 45% of patients and haemodynamic assessment was not possible in 15% of patients.

Influence on patient management

The mean duration of probe insertion was 32 ± 23 hours, allowing for a mean of 2.8 ± 1.6 haemodynamic evaluations per patient (which led to a mean of 1.4 ± 1.5 therapeutic changes per patient). Among the 263 haemodynamic assessments, 132 (50%) had a direct therapeutic effect in 62 patients (66%).

Conclusions

In expert hands, the probe provided relevant information for the haemodynamic monitoring of ventilated ICU patients with cardiorespiratory compromise and had a therapeutic effect in most patients.

Abbreviations: ICU, intensive care unit; n, number of patients; TEE, transoesophageal echocardiography.

The case series by Maltais et al. was conducted in a single-institution university hospital setting during a 9‑month period in 21 postoperative, cardiac surgery adult patients who were haemodynamically unstable (Maltais et al. 2013). The primary outcomes were fluid volume responsiveness, and agreement between echocardiographic and standard haemodynamic assessments. Two patients needed reoperation for bleeding and tamponade physiology. Right ventricular dysfunction was diagnosed by episodic transoesophageal echocardiography monitoring in 7 patients (33%), and hypovolaemia was documented in 12 patients (57%). Volume responsiveness was documented in 11 patients (52%). Discordance between haemodynamic monitoring and episodic transoesophageal echocardiography was qualitatively observed in 14 patients (66%), meaning that there was a change in treatment as a result of echocardiographic findings. A summary of this study is presented in table 2.

Table 2 Summary of the Maltais et al. (2013) case series

Study component

Description

Objectives/hypotheses

Objective: to assess whether episodic monoplane TEE with a limited examination would help guide the postoperative management of high-risk cardiac surgery patients.

Hypothesis: that episodic monoplane TEE guides assessment of intravascular/myocardial volume, inotrope need, vasopressor use, and assessment of pericardial effusions in critically ill, postoperative, cardiac surgery patients.

Study design

Prospective, consecutively enrolled, non-blinded, descriptive case series.

Setting

Single institution in a university hospital setting during a 9-month period (June 2010 to February 2011).

Inclusion/exclusion criteria

Inclusion criteria: haemodynamically unstable at any time in the ICU (defined as persistent systolic BP < 100 mmHg, cardiac index < 2.2 l/min/m2, SvO2 < 60%, suspected pericardial effusion with tamponade physiology, base deficit >8 mEq/l, or lactate >5 mg/dl despite persistent inotropic, vasopressor, and/or volume resuscitation, and concern for or known right ventricular failure).

Exclusion criteria: none defined.

Primary outcomes

Fluid volume responsiveness

Concordance/discordance between echocardiographic data and standard haemodynamic data, with discordance defined as a change in direction of management as a result of echocardiographic findings.

Statistical methods

Descriptive statistics for categorical variables were reported as frequency and percentage, and continuous variables were reported as mean (standard deviation) or median (range) as appropriate.

Participants

n=20 unstable postoperative cardiac surgery patients and n=1 in septic shock (mitral valve endocarditis) for surgical evaluation.

Results

Fluid volume responsiveness

Volume responsiveness was documented by echocardiography in 11 patients (52%), with no response in 10 patients (48%).

Concordance/discordance between echocardiographic data and standard haemodynamic data

Discordance between standard haemodynamic monitoring and episodic transoesophageal echocardiography was qualitatively observed in 14 patients (66%), meaning that there was a change in direction of management of these patients as a result of echocardiographic findings.

Additional results reported

Mean number of imaging sessions was 3.28, median 3 per patient.

Two patients (10%) needed reoperation for bleeding and tamponade physiology.

Right ventricular dysfunction was diagnosed by episodic TEE monitoring in 7 patients (33%), and hypovolaemia was documented in 12 patients (57%).

Conclusions

The authors demonstrated the ImaCor hTEE system's ability to change the clinical management of unstable, postoperative, cardiac surgery patients. Patients considered to be fluid responsive by echocardiography were more likely to be appropriately resuscitated 6 hours after initiation of imaging. Haemodynamic monoplane TEE assessment was judged a useful adjunct, extending the haemodynamic assessment capabilities of TEE from the operating room to the ICU.

Additionally, the authors reported that they have used more than 200 probes in the cardiovascular ICU and more than 50 elsewhere with no complications to date.

Abbreviations: hTEE, haemodynamic transoesophageal echocardiography; ICU, intensive care unit; n, number of patients; SvO2 , mixed venous oxygen saturation.

The retrospective case report Mykytenko et al. describes a single patient who developed septic shock complicated by right ventricular dysfunction and underwent hTEE to guide fluid resuscitation, vasopressor, and inotropic therapy (Mykytenko et al. 2012). Invasive haemodynamic monitoring with an arterial line and central venous line was started, as the patient was given a 5 litre fluid resuscitation and multiple vasopressors. Subsequently, hTEE demonstrated a range of clinically abnormal cardiac anatomical and functional features. A summary of this study is presented in table 3.

Table 3 Summary of the Mykytenko et al. (2012) case report

Study component

Description

Objectives/hypotheses

To review a single patient with septic shock complicated by right ventricular dysfunction, in which hTEE guided fluid resuscitation, vasopressor, and inotropic therapy.

Study design

Retrospective descriptive case report.

Setting

Postoperative ICU.

Inclusion/exclusion criteria

None defined.

Primary outcomes

Response to fluid resuscitation, vasopressor, and inotropic therapy.

Statistical methods

None.

Participants

n= 1; a 48-year old man who had an uneventful off-pump coronary artery bypass graft and 2 days later developed septic shock from pneumonia complicated by right ventricular dysfunction.

Results

Invasive haemodynamic monitoring with an arterial line and central venous line was initiated as the patient was administered with a substantial 5 litre fluid resuscitation and multiple vasopressors. Subsequently, hTEE demonstrated a range of clinically abnormal cardiac anatomical and functional features, with resultant images reproduced as 3 figures in the paper: shortly after the onset of septic shock, 12 hours later and 2 days later.

hTEE demonstrated new RV dysfunction with apical sharing, grossly decreased thickening of the RV free wall, paradoxical septal motion with grossly normal left ventricular function.

Low-dose epinephrine at 2 microgram/min was chosen in the setting of hypotension for inotropy and RV support with improvement of RV function over time. Subsequent to the initiation of epinephrine and initial improvement in RV function, a random cortisol level suggested relative adrenal insufficiency, so stress dose hydrocortisone was also added.

Over the next 24 hours, the patient's haemodynamics improved dramatically, and they were weaned off all vasopressors except for low-dose epinephrine, which was stopped a few hours later based on resolution of RV dysfunction.

The patient's ventilator was weaned over the next several days as they had diuresis and continued clinical improvement.

Conclusions

This study highlights the rapid interval development of RV dysfunction in a previously normal heart over the course of 48 hours secondary to septic shock and the utility of direct assessment of cardiac function with hTEE.

Abbreviations: hTEE, haemodynamic transoesophageal echocardiography; ICU, intensive care unit; n, number of patients; RV, right ventricle.

Two appropriate ongoing or in-development trials on the ImaCor hTEE system for haemodynamically unstable patients were identified in the preparation of this briefing.

The EVENT retrospective, matched-pair, case-control study was completed in January 2014, although results have not yet been published. Its primary hypothesis was patient outcomes and secondary measures were hospital expenses.

The ImaCor II randomised controlled trial commenced in January 2014 and aims to recruit 500 participants. It is scheduled to complete in September 2016. This trial, in collaboration with the system manufacturer, is designed to measure patient outcomes, thereby demonstrating the efficacy of hTEE monitoring compared with standard monitoring.

Costs and resource consequences

The manufacturer stated that 45 ClariTEE single-use, disposable probes have been sold to NHS customers, for use with 3 Zura imaging systems on loan. No Zura EVO imaging systems have yet been sold to the NHS (figures correct at May 2014).

In 2012/13, Hospital Episode Statistics Online recorded 33,405 transoesophageal echocardiograms in admitted NHS patients treated in England (code U20.2), of which 31,655 were adults. In 63.5% of all patients, transoesophageal echocardiography was recorded as a supplementary procedure during an admission, which would include its use for episodic haemodynamic monitoring in critical care. Health and Social Care Information Centre also recorded 48,577 adult critical care episodes in 2012/13 that were associated with cardiac surgery and primary cardiac conditions. However, in the UK most transoesophageal echocardiography is performed in the UK in the cardiology inpatient population for diagnosis and in cardiac surgery theatres for diagnosis and monitoring by cardiac anaesthetists. Therefore, it is difficult to determine the proportion of scans conducted in the specific population of people who are haemodynamically unstable and for episodic monitoring of cardiac function.

The ImaCor hTEE system is promoted as an adjunct to continuous haemodynamic monitoring, as an alternative to conventional transoesophageal echocardiography examination in the critical care setting. The Zura EVO imaging system, which is used in conjunction with the disposable ClariTEE probe, is kept in the critical care setting, rather than being in the echocardiography or radiology department of the hospital. Because of this, adoption of the hTEE system requires users within the intensive care unit to be trained to insert the ClariTEE probe and interpret the haemodynamic assessment images. The ImaCor website describes a tailored training programme that provides a wide range of educational materials. In current UK practice, the conventional transoesophageal echocardiogram would normally be conducted in the cardiac intensive care unit by an anaesthetist or intensivist, with training and accreditation available through a joint venture between the Association of Cardiothoracic Anaesthetists and the British Society of Echocardiography.

Cioccari et al. (2013) published a study on the feasibility of haemodynamic monitoring using the hTEE system in critically ill patients after a brief operator training period of 6 hours. They assessed 14 intensive care unit staff specialists over 148 hTEE examinations in 55 patients. They concluded that echocardiographic examinations after brief bedside training were feasible and of sufficient quality in most patients, with substantial inter-rater reliability between hTEE operators and an expert cardiologist. Results on relevant patient outcomes were not reported.

Training needs specific to the hTEE system will therefore have a measureable cost and resource impact, but no published evidence on resource consequences was identified.

Strengths and limitations of the evidence

The 3 studies are observational pilot studies of the hTEE system. In general, qualitative methods were used to report primary outcome measures, although the case series by Maltais et al. (2013) did capture some semi‑quantitative data on the assessment of left ventricular end‑diastolic area. The critical care settings of all 3 studies match the intended use of the hTEE system.

Although the study designs of all 3 identified papers are regarded as low quality in the conventional hierarchy of evidence, the study by Vieillard-Baron et al. (2013) was high quality as a prospective multicentre observational study with 94 patients across 4 centres. Its main strength was that the authors had previously validated their haemodynamic assessment test protocol against a more comprehensive and quantitative method and all 3 views of the heart were explored and assessed. However, the reliability of the analysis remains uncertain because blinding was not addressed. Another limitation is that inter-observer reproducibility was not tested.

The case series by Maltais et al. (2013) involved fewer participants. One of the key strengths of this study was the methodology where patients received intraoperative conventional transoesophageal echocardiography, and were continuously haemodynamically monitored postoperatively via a pulmonary artery catheter. When patients became haemodynamically unstable, the hTEE system was used as an adjunct to the pulmonary artery catheter to elucidate the pathology. This highlights the intended use for hTEE in clinical practice, and demonstrates its utility within the patient pathway. Incidental findings were also reported, such as average length of stay in the intensive care unit and observed in-hospital or 30-day mortality. However, a notable limitation of the study is that the authors obtained information from only the mid-oesophageal 4 chamber and transgastric short axis views; they did not address the absence of the superior vena cava view. It is important to note that this was an observational non-blinded case series, which leaves the results open to observer bias.

The third study was a retrospective case report on a single patient with right ventricular dysfunction (Mykytenko et al. 2012). Although this paper reports some interesting clinical observations in a single patient, the case report cannot be considered representative of the patient population in general.