Infective endocarditis (IE) is an important pathology to detect in the Emergency Department (ED), but the diagnosis is too often delayed or missed1. IE ticks many boxes as a diagnosis we can ‘own’ in Emergency Medicine (EM):
1. It is time-critical with worsening morbidity/mortality as the disease progresses1-4
2. Patients are often critically unwell
3. Along with recent dental work and congenital heart disease, risk factors include uncontrolled diabetes and intravenous drug use5, and patients in the latter two groups often have poor access to healthcare and poor compliance with follow-up instructions. Their ED visit may well be the best opportunity we have to make a life-saving diagnosis.
The Modified Duke Criteria combines clinical, lab and imaging criteria for diagnosing IE, with the gold standard imaging modality being transoesophageal echocardiography (TOE). However access to TOE is extremely limited in the developing world for the majority of patients, and even in the developed world there are often geographic, expertise or resource constraints to accessing this study. In addition TOE requires some preparation - multiple staff, an empty stomach, sedation, topical anaesthesia - all followed by meticulous probe sterilisation and storage.
In contrast, transthoracic echocardiography (TTE) is a widely available modality which can be performed immediately by clinicians at the bedside, and is steadily becoming cheaper and more accessible in emergency departments.
As with so many things point-of-care ultrasound (POCUS), when dealing with valvular vegetations we are blessed with this rule of thumb:
BIGGER = WORSE = MORE URGENT = EASIER TO FIND6-10.
A 2017 meta-analysis published in JASE11 summarised the accuracy of comprehensive TTE for diagnosing IE. The statistics don’t apply directly to the ED setting as we are not performing comprehensive echocardiography - our scans are limited or focused and often performed with lower-quality machines. However TTE for native valve endocarditis has an impressive positive likelihood ratio (PLR) of 14.6, and a sensitivity of 66%. (PLR and sensitivity are the numbers we are most interested in, as we are attempting to rule-in, NOT rule-out this pathology.)
What have some of our societies and guidelines had to say about TTE for IE?
a) The Australasian Society of Ultrasound in Medicine (ASUM) Rapid Cardiac Echocardiography (RCE) unit includes assessment of the tricuspid, mitral and aortic valve appearance12.
b) The ASUM Critical Care Diploma of Diagnostic Ultrasound (DDU) - targeted at specialists including emergency physicians - includes assessment of valvular lesions13.
c) The 2016 American Society of Echocardiography (ASE) guideline for Appropriate Use Criteria for Evaluation of Cardiac Sources of Emboli states that suspected IE, new murmur, stroke and peripheral emboli are all valid reasons to perform transthoracic echocardiography14.
d) In 2016 a group including emergency physician Dr Mike Blaivas published Guidelines for the Appropriate Use of Bedside General and Cardiac Ultrasonography in the Evaluation of Critically Ill Patients15. They gave a grade 2c recommendation for screening with Basic Cardiac Ultrasound (BCU) in suspected IE. They state: “The intensivist with basic-level training may be able to recognise obvious vegetations. In low-risk patients, BCU could lead the physician to pursue alternative diagnoses, and in high-risk patients, it could help to identify large lesions easily.”
I identified eight published cases where IE was diagnosed by an emergency physician using bedside echo16-23.
The following is a further series of three patients who presented to us at Janus General Emergency Department24 with IE, whose bedside echo in ED expedited their diagnostic and management pathway. The cases are all males aged 20-50 - it is worth noting that the incidence of IE is higher in men (1.5:1 - 3:1 M:F)10.
A male intravenous drug user in his 20s presented with acute left-sided weakness. A week prior he had presented with cough, fever and murmur, and was discharged on oral antibiotics. Subsequently his blood cultures had grown streptococcus mitis. On examination he had 0/5 power in his left upper and lower limbs, fever, a systolic blood pressure of 90, tachycardia, inspiratory crackles in bilateral lung fields, and a loud diastolic murmur throughout the precordium.
Bedside echo demonstrated anterior and posterior mitral valve leaflets (AMVL/PMVL), thickened by heavy vegetations. His PMVL was prolapsing, with severe mitral regurgitation (MR) causing acute pulmonary oedema.
He was urgently transferred to a cardiothoracic unit and stabilised. His comprehensive echo showed severe MR and moderate-sized mitral vegetations with a possible impact lesion. He eventually underwent mitral valve replacement and thankfully on discharge months later had minimal residual neurological deficit.
2. A male in his 40s presented septic, drowsy, with vague complaints of pain in his R flank and down his R thigh. Examination did not reveal any septic focus relevant to the location of the pain, his urine was clear, and he had no septic focus in his abdominal, vertebrae or soft tissues. He did however have an obvious diastolic murmur, and it was unknown whether this was acute or chronic.
Bedside echo showed severe aortic regurgitation (AR). The LV was not yet significantly dilated, suggesting that the AR was acute. My imaging was sub-optimal, and I did not identify an obvious aortic valve (AV) vegetation - his scan was ‘indeterminate’.
I admitted the patient with heightened suspicion for IE as the primary septic focus. His comprehensive echocardiogram the following day confirmed an AV non-coronary cusp vegetation with acute severe AR.
3. Another male in his 40s presented with a week of fever and heart failure symptoms - orthopnoea and dyspnoea on exertion. He had no known IE risk factors. He was overweight, however otherwise looked reasonably well from the end of the bed. On examination he had bilateral lung crackles, mild peripheral oedema, and a new murmur.
Bedside echo revealed an enormous 22mm aortic valve lesion prolapsing into the LVOT. Considering the size of the lesion he was compensating very well and didn’t appear septic, but was transferred semi-urgently to a tertiary hospital. His comprehensive echocardiogram showed a congenital bicuspid AV (previously undiagnosed) with mild stenosis, and a large mobile echodensity >2cm attached to the the base of the LVOT aspect of the anterior leaflet. There was severe eccentric AR with a pressure half-time (PHT) of 93 ms, and pan-diastolic flow reversal in the aorta.
Some learning points:
1. IE can have a widely variable, non-specific presentation - pain, dyspnoea, sepsis, septic emboli. Bedside echo is a tool which should be used liberally for sick patients, and also when you are uncertain about the cause of a patient’s fever or murmur.
2. In the ED, we don’t necessarily have to visualise a vegetation to make a provisional diagnosis of IE. If you find secondary signs including regurgitation and pericardial effusion, you might label the study as ‘indeterminate’, with a heightened suspicion for IE.
3. We all know TTE is not highly sensitive for IE. Take this one step further - even a TOE cannot exclude infectious endocarditis! “Vegetations not identified on this study” is as definitive as you’ll see on a TOE report.
4. Next time you’re assigned the painful task of checking your ED’s un-reviewed lab results, keep a close eye out for unexplained staphylococcal, streptococcus viridans (as in case 1), enterococcal and HACEK organisms in blood cultures.
5. Speaking of blood cultures, take those 3 sets over the course of an hour BEFORE giving antibiotics25. If you skip this step in the rush to meet your departmental sepsis guideline, your cardiologists may end up having to use the difficult label of Blood Culture-Negative Infective Endocarditis, and the choice of antibiotic to use long-term may not be optimal.
6. We’ve only discussed native valve endocarditis here - echocardiography for prosthetic valve endocarditis is a massive further leap of complexity! Safest to say that prosthetic valves are a risk factor for IE, and if IE is on the differential your patient needs comprehensive imaging as soon as possible.
To conclude, bedside echo in the emergency department to diagnose IE is not yet considered standard practice. However the POCUS Perfect Storm of decreasing cost, widespread training and technological imaging advances means emergency physicians today have the potential to diagnose and manage this life-threatening condition more rapidly than ever before.
Dr Jonathan Henry
Blogs are written by our EMUGs Team from across Australasia.