Article Text
Abstract
The development of hand-carried, battery-powered ultrasound devices has created a new practice in ultrasound diagnostic imaging, called ‘point-of-care’ ultrasound (POCUS). Capitalising on device portability, POCUS is marked by brief and limited ultrasound imaging performed by the physician at the bedside to increase diagnostic accuracy and expediency. The natural evolution of POCUS techniques in general medicine, particularly with pocket-sized devices, may be in the development of a basic ultrasound examination similar to the use of the binaural stethoscope. This paper will specifically review how POCUS improves the limited sensitivity of the current practice of traditional cardiac physical examination by both cardiologists and non-cardiologists. Signs of left ventricular systolic dysfunction, left atrial enlargement, lung congestion and elevated central venous pressures are often missed by physical techniques but can be easily detected by POCUS and have prognostic and treatment implications. Creating a general set of repetitive imaging skills for these entities for application on all patients during routine examination will standardise and reduce heterogeneity in cardiac bedside ultrasound applications, simplify teaching curricula, enhance learning and recollection, and unify competency thresholds and practice. The addition of POCUS to standard physical examination techniques in cardiovascular medicine will result in an ultrasound-augmented cardiac physical examination that reaffirms the value of bedside diagnosis.
- Cardiac imaging and diagnostics
- Echocardiography
- Heart Disease
- Heart failure
- Educational goals and objectives
Statistics from Altmetric.com
- Cardiac imaging and diagnostics
- Echocardiography
- Heart Disease
- Heart failure
- Educational goals and objectives
Introduction
The patient’s bedside has been the epicentre of medical care for centuries. It is from this site, the ‘point of care,’ that the physician obtains the patient’s history and performs physical examination, anchoring initial diagnostic impressions and setting a pathway for subsequent triage, testing, treatment and referral. In today’s acutely ill patient, limited applications of ultrasound at the point of care are easily performed and have higher sensitivities for the same pathological processes sought by physical techniques.1 2 The detection of disease by taking a quick look for ultrasound ‘signs’ that augment traditional physical methods bridges the use of ultrasound to the venerable practice of physical examination3 (table 1). The following paper will specifically review the use of point-of-care ultrasound (POCUS) to improve the cardiac physical examination.
Value of ultrasound at the ‘point of care’
Miniaturisation and simplification of devices that perform ultrasound imaging has resulted in the development of a novel methodology and non-traditional user groups.4 The economic, quality and medico-legal implications of using ultrasound as a part of the physical examination or as a limited diagnostic test have resulted in controversy and affect the inclusion and dissemination of POCUS in medicine (table 1).
Despite the ongoing ambiguities of POCUS privileging and reimbursement, there is little debate over the value of improving a physician’s own physical examination. Perhaps more than any other organ system, physical examination of the heart is routinely performed regardless of the patient’s chief complaint, as it provides the status of the cardiovascular system under the current physiological stressors. Finding signs representing common abnormalities that have prognostic value or treatment implications should be the goal of bedside examination by any method and result, time permitting, in confirmation by more comprehensive evaluation. Published studies of the diagnostic accuracies of traditional physical examination versus POCUS examination demonstrate higher sensitivity of POCUS, whether compared as absolute accuracies (table 2) or when measured as incremental improvement in accuracy by the same observer on the same patient (table 3). The diminishing sensitivity of cardiac physical examination has caused much concern over recent years and has many potential explanations including a reduction in training and skills of the current generation of physicians, the increased number of patient care environments where physical examination techniques cannot be adequately practiced and reduced prevalence of physical findings in contemporary patient populations and disease presentations. Few studies have evaluated a truly ‘blended’ examination as often performed in clinical practice, where the incremental effect of adding a limited ultrasound examination to traditional bedside evaluation addresses the specific biases of user confidence in their own accuracies, particularly when attempting to reconcile discordant bedside data.
In addition to the capability of detecting disease at an earlier stage, inherent advantages exist in using ultrasound for bedside examination, particularly in modern acute care settings where ambient noise levels, frequent alarms and patient immobility can mask soft auscultatory findings or limit manoeuvring, accounting for some of the reduced sensitivities reported for physical findings in emergency settings.5 Unlike the original validation studies of physical techniques that primarily involved only expert practitioners and patients with a single, symptomatic lesion, initial studies using ultrasound have noted incremental value for cardiac diagnosis by medical students,6–9 residents,10–13 non-cardiologists14 and cardiologists,15–18 often bolstered by the detection of multiple findings, some incidental and asymptomatic, in each patient. In bedside practice, skills with a stethoscope or ultrasound probe change over time, as they are variably acquired, honed or lost, as dictated by the physician and by the patients encountered. A recent study of well-trained medical residents who had stopped ultrasound imaging has found loss of proficiency after only 2 years.19 Indeed similar to specific traditional physical examination techniques that are retained, maintenance of POCUS skills in practice may require frequent, repetitive use and simple, robust findings. Over time, the incremental value of adding less frequently employed ultrasound techniques to bedside examination, such as M-mode, colour and spectral Doppler, may differ between cardiologists and non-cardiologists largely due to prior echocardiography training and the prevalence of cardiac patients seen. The individual physician’s realisation of personal improvement in aspects of his/her own bedside accuracies directly attributable to the use of ultrasound may be the strongest ethical argument for self-determined use of POCUS by non-cardiologists (figure 1) and will require self-directed maintenance of the appropriate skill set. However, regardless of individual variability in bedside skills, there exists an overall need to structure a fundamental POCUS examination using traditional, time-tested targets to coordinate standards for patient care and medical education in ultrasound.
Accuracy of ultrasound and physical techniques for traditional targets of bedside examination
Detection of LV systolic dysfunction
A reduction of LV systolic function, even when asymptomatic, has diagnostic, treatment and prognostic implications and has long been a target for bedside techniques. Few data exist in screening by non-cardiologists for the asymptomatic patient with LV dysfunction20 or for the patient with incidental LV dysfunction who is admitted for another disease. When examined by cardiologists, the detection of an S3 was 68% sensitive and 73% specific in 52 patients with chronic LV systolic dysfunction of whom 37 had PCWP >18 mm Hg.21 In a more contemporary study of 1376 patients with symptomatic heart failure and LVEF <35% examined by site investigators, an S3 was observed in only 15% of patients.22 Similarly, a sustained apical impulse, although considered ‘very helpful’ to make the diagnosis in a systematic review,23 is frequently absent even when sought by experts.2 These data suggest that physical findings may relate to more advanced, decompensated stages of heart failure. As the suspicion of LV dysfunction by a primary or emergency room physician often generates testing and referral, studies of non-cardiologist ultrasound use have importantly verified a >90% accuracy of limited imaging protocols for LV systolic dysfunction, mainly employing subjective estimations of contractility or E-point septal separation6–10 12–18 24 25 (figure 2A), across all stages of heart failure. Multiple studies7–10 12 14 have evaluated the incremental effect of adding a limited ultrasound examination to traditional bedside evaluation specifically for LV systolic dysfunction by non-cardiologists and primarily confirm improved sensitivity with ultrasound use.
Detection of elevated left heart filling pressures
Perhaps more important than the screening for any numerical threshold of ejection fraction is the detection of elevated filling LV pressures as a marker of heart failure. As previously noted, the presence of an S3 or S4 is evidence of elevated filling pressures at the bedside, with studies showing poor sensitivity (11%–78%) but increased specificity (80%–99%) for the S3, and minimal diagnostic value, if any, for the S4 (2,5). As the presence of an S3 gallop, elevated filling pressures, or brain natriuretic peptide (BNP) levels could be transient, the detection of left atrial enlargement (LAE) may be the best indicator of sustained elevation of LA filling pressures.26 Studies27 28 using either M-mode or two-dimensional measures of LA size have shown prognostic implications to LAE, in regard to mortality and stroke, and in patients with cardiomyopathy and mitral disease. The size of the left atrium could be helpful in determining a cardiac source of dyspnoea, signify the onset of symptoms in LV dysfunction or aortic stenosis, predict chronicity and stroke risk in a patient with newly-recognised atrial fibrillation or confirm the significance of chronic left-sided valvular disease. Clinically significant LAE cannot be detected by physical examination techniques but can be recognised with 80% accuracy by briefly trained medical residents with a quick-look ultrasound sign that compares the LA anterior-posterior diameter to a 4 cm reference standard29 or to the overlying aorta in a single parasternal view24 30 (figure 2A). It is conceivable that mitral inflow by Doppler could provide additional useful information by POCUS imaging; however, it requires more extensive knowledge. As a structural adaptation, LAE importantly signifies a predisposition for acute decompensation of heart failure, where a further unaccommodated rise in filling pressures results in pulmonary congestion.
Determination of pulmonary congestion
Excessive extravascular lung water can manifest as interstitial oedema and pleural effusion. Rales or crackles felt to be generated by oedematous alveoli, which open on inspiration, have a poor sensitivity (19%–64%) and high specificity (82%–94%) for elevated filling pressures,2 relate to the presence of pulmonary oedema and severity of symptoms, and are confounded by their evanescence with treatment, presence of atelectasis and gravitational forces in the supine or intubated position. The ultrasound ‘equivalent’ of rales, B lines or ‘comet-tail’ artefacts are ring-down phenomena proposed to be due to thickening of peripheral pleural or interstitial septae due to oedema or fibrosis31 (figure 2B,C). In oedema, B-lines have similar gravitational predilections as rales and are also thought to resolve quickly, having less specificity for heart failure when found in the lung bases and more when found in the apices.32 Despite the lack of a validating tissue model and potential interdevice variation, the presence of B-lines are a more sensitive marker of decompensated heart failure than clinical assessment,33 may precede the appearance of rales and have significant prognostic impact whether tallied in a 28-zone lung imaging protocol, 8-zone protocol or simply detected as 3 or more in the lung apices.34 In a recent study of 195 ambulatory patients with chronic heart failure,35 the subgroup with 3-or-more B-lines (n=59) had a four-fold higher risk of heart failure hospitalisation or all-cause mortality and yet a relatively low 19% prevalence of rales on review of the cardiologists’ clinic notes. The number of B-lines has been shown to be related to ejection fraction, degree of diastolic dysfunction and New York Health Association class.36 Similar to rales, the presence of B-lines due to infectious, autoimmune, drug-related or fibrosing interstitial abnormalities can confound attempts at diagnosis of superimposed acute oedema.
The detection of small, bilateral pleural effusions, often indicative of decompensated heart failure and a worse prognosis, is difficult by physical techniques but much improved by using ultrasound imaging of the costophrenic angles (figure 2D,E). Physical findings such as dullness to percussion have reported sensitivity and specificity of >80% in studies that used chest radiography as a gold standard37 and therefore likely apply to effusions of >200 ccs. Accordingly, these studies overestimate the sensitivity of physical techniques when applied to small effusions, especially compared with ultrasound studies that used CT as a gold standard. In a study of decompensated heart failure using 60 patients and 22 controls, expert examination by two cardiologists and CT as the gold standard, pulmonary auscultation had a 55% sensitivity and 91% specificity for heart failure compared with the 90% sensitivity and 95% specificity of lung ultrasound.38 Pleural effusions were detectable in 90% of heart failure admissions. In a small study of 32 subjects with adult respiratory distress syndrome and 10 normal volunteers, lung auscultation by a senior investigator had a sensitivity and specificity for pleural effusion of 42% and 90%, and for alveolar-interstitial syndrome of 34% and 90%, as compared with ultrasound use that had a sensitivity and specificity of 92% and 93% for effusions, and 98% and 88% for alveolar-interstitial syndrome.39 Using CT as a gold standard, ultrasound also outperformed chest radiography in this ICU population. Additionally, lung POCUS has been used to quantify the amount of effusion and detect exudative complications such as particulate matter or early septation. Lung ultrasound has been touted to have 92% sensitivity and 92% specificity in the detection of pneumonia40 and 79% sensitivity and 98% specificity in pneumothorax,41 both entities encountered by cardiologists and poorly detected by auscultation. Conversely, some disorders, such as airway narrowing causing stridor or wheezing, are better found by auscultation. Detection of lung congestion through ultrasound signs of B-lines and effusions provides physiological insight that complements echocardiography or BNP data and can suggest a need for more aggressive treatment to perhaps avoid hospitalisation or readmission. Although elevated LA pressures can predispose to the accumulation extravascular lung water, lymphatic drainage of the lung may also determine the temporal resolution of lung congestion through a dependence on reduction in central venous pressures.
Determination of right heart function and central venous pressures
Enlargement of the right ventricle can be detected on physical examination by a sustained parasternal heave or right-sided S3, both with minimal validation of accuracy. Although most likely signifying chronic pulmonary hypertension, right ventricular enlargement can also occur with RV infarction, acute pulmonary embolism, isolated severe tricuspid regurgitation, atrial septal defect, acute ventricular septal rupture and RV dysplastic syndromes. In POCUS, the utility of the RV/LV ratio >0.6 as obtained in the apical four-chamber view can be used to diagnose submassive pulmonary embolism but is very subjective, has low sensitivity (55%) and specificity (69%) even in expert hands,42 and can be falsely increased by foreshortening of the LV during imaging, underestimate RV enlargement in the setting of LV enlargement and be confounded by pre-existing cor pulmonale.
The determination of central venous pressures by jugular venous pressure (JVP) has been considered a quintessential skill of physical examination. The accuracy of JVP assessment has an estimated 78%–95% sensitivity and 89%–93% specificity2 for a central venous pressures >12 cm of H20, with details of x and y descents providing additional diagnostic information in tamponade and constriction. However, physical techniques are diminished in more obese patients with thicker necks, and limited in supine intubated patients. Ultrasound can be used to find the height of the jugular venous column, although both visual and ultrasound observations may underestimate true central venous pressures.43 Ultrasound can also estimate central venous pressure by the respiratory dynamics of the inferior vena cava (IVC) in longitudinal view44 (figure 2F), a method also used to determine ‘fluid responsiveness’ in the critically ill.45 When comparing physical examination of JVP versus IVC ultrasound techniques performed by four medical residents on 40 consecutive patients after right heart catheterisation, sensitivity was 14% versus 82% for right atrial pressure >10 mm Hg.11 In addition to variability in technique, the question of accuracy may be in the actual difference between these two vessels in their behaviour to central venous pressures. The JVP estimate, as a manometer, relies on an approximation of the true phlebostatic axis2 and stable cerebral venous return. The IVC size is confounded by diaphragmatic motion, obesity and intra-abdominal pressures. Nonetheless, IVC findings obtained at the bedside by non-cardiologists specifically in decompensated heart failure have shown prognostic value in 75 patients for predicting readmission46 and in 80 patients for predicting in-hospital mortality.47
Detection of significant valvular disease
As echocardiography is essential in modern assessment of valve disease, POCUS facilitates care by identifying which patients should be referred for echocardiography. The common echo referral generated by auscultation of a systolic murmur frequently demonstrates benign or ‘functional’ morphology but can also detect aortic valve sclerosis, early aortic valve stenosis or mitral valve prolapse.48 Supporting the sensitivity of auscultation, in a study of 100 patients referred for ‘systolic murmur’, only 21% were free of abnormal findings on echocardiography;49 however, the ability of two cardiologists to identify specific cardiac lesions by physical examination was limited especially when more than one lesion was present. Using POCUS, limited imaging protocols have demonstrated the capability to detect significant calcific aortic stenosis50 with 84% sensitivity and 89% specificity and screen for mitral prolapse (figure 3A) or rheumatic heart disease. In the evaluation of murmur, the use of POCUS to localise the abnormal valve morphology improves specificity and the addition of colour Doppler imaging further increases sensitivity, detecting as little as mild regurgitation51 (figure 3B). However, in the setting of an asymptomatic patient with normal left atrial size and LV systolic function by POCUS, it is unclear whether the discovery of a murmur or small colour flow jets have clinical value that justify subsequent costs of formal echocardiography referral. In suspected endocarditis, POCUS could screen for pre-existing valve disease or significant regurgitation but cannot be expected to find small vegetations, inheriting at least the limitations of standard transthoracic echocardiography. Although POCUS detection of severe valvular disease is feasible by non-cardiologists, the use of POCUS to diagnose and categorise lesser valvular pathology is unknown and would require expert familiarity with specific valve lesions and the use of Doppler.
Expanded cardiac POCUS applications where the physical examination is known to be inadequate
Many diseases that pertain to the care of the cardiac patient manifest few or no physical findings and can now be diagnosed using POCUS (table 4). In the outpatient, the detection of early, non-stenotic subclinical atherosclerosis in the carotid (figure 3C) or femoral artery bifurcation portends a prognosis of a coronary heart disease equivalent and can affect lipid management. Similarly, the presence of an echogenic or fatty liver (figure 3D) can be a marker for the metabolic syndrome, liver disease or occult alcohol use. The detection of left ventricular hypertrophy may affect treatment decisions in hypertension or suggest hypertrophic cardiomyopathy in preparticipation athletic screening. Non-cardiac findings of the urinary tract in cardiorenal syndromes (figure 3E) and the spleen in subacute bacterial endocarditis or unexplained thrombocytopaenia may be helpful. For the interventionalist, a brief vascular ultrasound survey for peripheral arterial disease or deep vein thrombosis can be used to guide or deter femoral arterial and venous cannulation. During cardiac arrest, in the least, ultrasound can note the presence of cardiac motion, assess the clearance of blood pool stasis (figure 3F) and update prognostic signs during pulse checks.
The future of an ultrasound-augmented physical
An evidence-based, multifaceted POCUS cardiac examination that detects the above-mentioned ultrasound ‘signs’ of traditional pathologies (figure 2) has already been described3 and successfully integrated into an internal medicine residency curriculum52 and can be easily expanded for comprehensive medical evaluation or subspecialty use (figure 4). The prototypical quick-look imaging protocol is constructed to synthesise diagnoses from findings and minimise technique errors due to transducer misplacement. Device design in the future may reduce cost and further leverage technological advancements in wireless transmission, potentially enabling remote, expert interpretation or telementoring of new users. Earlier, more accurate detection of disease as afforded by POCUS technologies will likely result in the saving of downstream costs,16 17 testing of clinical outcomes and changes in current clinical pathways by the next generation of ultrasound-enabled, cost-conscious physicians.
Conclusion
The patient’s bedside has become a critical site for the future of healthcare. Despite the promise of telemedicine, laboratory biomarker assessment, proliferation of comprehensive radiological imaging and the team approach in medical care, the physician’s physical presence at the patient’s bedside to perform a skilful examination develops patient trust, endorses physician responsibility and remains a cornerstone in patient-centred care. Evidence suggests that ultrasound increases the sensitivity of physical examination by detecting disease at an earlier, asymptomatic stage or by improving inadequacies in current practice for both cardiologists and non-cardiologists. The recent use of ultrasound at the point of care strengthens a traditional medical care model that is well worth preserving—at a location well worth recognising.
Acknowledgments
Lee Luniewski and Sondhaya McGowan, MLIS, AHIP, Scripps Mercy medical librarians for assistance in review of the literature.
References
Footnotes
Competing interests None declared.
Provenance and peer review Commissioned; externally peer reviewed.