key: cord-0910469-rt49jo2h
authors: Fox, William C.; Krishnamoorthy, Vijay; Hashmi, Nazish; Bronshteyn, Yuriy S.
title: Pneumonia: Hiding in Plain (Film) Sight
date: 2020-07-25
journal: J Cardiothorac Vasc Anesth
DOI: 10.1053/j.jvca.2020.07.062
sha: 57cb1c0262c1741b452b7fea09b7d19f37506bc4
doc_id: 910469
cord_uid: rt49jo2h

We present a case where point-of-care lung ultrasound was used to identify the source of progressive multi-organ failure when a chest X-ray (CXR) and other routine tests failed to provide a conclusive answer. Post-case discussion focuses on the following: (1) the relative strengths and weaknesses of CXR versus lung ultrasound in screening for lung disease; and (2) suggestions of how lung ultrasound practice can be standardized within the field of anesthesiology.

Although cardiac intensivists have traditionally relied upon portable chest X-rays (CXR) as the imaging modality of choice to rapidly narrow the differential diagnosis of respiratory failure, point-of-care lung ultrasound can also aid in this scenario. 1 This case describes the use of lung ultrasound to identify the source of progressive multi-organ failure when a CXR and other routine tests failed to provide a conclusive answer in a patient recovering from cardiac surgery.

On post-operative day (POD) 2 after an uneventful coronary artery bypass surgery, a 68-yearold male manifested mild, but subtly worsening multi-organ failure. Specifically, he developed increasing leukocytosis, persistent oxygen requirement, and hypotension (Supplementary Table   1 ). A transthoracic echocardiogram (TTE) identified the following: mildly reduced left ventricular (LV) systolic function (LV ejection fraction = 50%); mildly reduced right ventricular (RV) systolic function; no pericardial effusion; and no other significant abnormalities. These findings were grossly unchanged from the patient's pre-operative TTE. That day's portable chest X-ray (CXR) was officially read as showing, "bilateral heterogenous lung opacities" (Figure 1 ). Given the data available at the time, there did not appear to be an obvious unifying diagnosis to explain the patient's three mild abnormalities. First, given the patient's normal temperature, lack of sputum production, and the nearness of surgery, the CXR findings were more consistent with atelectasis than pneumonia. Second, the patient's leukocytosis was within the range seen from post-surgical neutrophil demargination. Third, the mild hypotension was consistent with some combination of routine hypovolemia, mild systolic dysfunction, and/or post-operative vasoplegia.

Thus, over the subsequent 24 hours, the patient's treating providers tried multiple maneuvers to address the suspected causes of hypotension. First, the patient was given intravenous fluids (750 ml 5% albumin in divided doses) and then his epicardial pacing rate was increased, but neither of these maneuvers resolved the hypotension. Finally, the patient was started on lowdose dopamine (3 mcg/kg/min), which did restore mean arterial pressure to 65 mm Hg.

Despite all these interventions, the patient's multi-organ failure generally progressed over the approximately 24 hours spanning from POD2 to POD3 (Supplementary Table 1 ). Though his oxygen requirement decreased somewhat, his leukocytosis and vasopressor requirement persisted and he developed new acute kidney injury (AKI). The POD3 portable CXR ( Figure 3) was officially read as showing right-sided opacities that "may represent atelectasis, aspiration, or infection." The CXR findings seemed unable to explain the patient's multi-organ failure, since the findings were highly subtle, the patient's oxygen requirement was actually decreasing, and the patient continued to have neither fever nor sputum production.

But whereas on POD2 the patient's mild symptoms seemed to fall within the scope of a routine post-operative course, on POD3 the patient's development of AKI challenged this assumption.

Since any unexplained organ failure necessitates inclusion of sepsis in the differential diagnosis 2 and since the CXR findings in this case seemed ambiguous, the ICU team decided to clarify those findings by performing point-of-care lung ultrasound. Specifically, the intensivist used a low-frequency, small footprint probe to examine the anterior and posterior-lateral lung fields (Supplementary Video 1). The findings were generally benign throughout except for a focal consolidation and small effusion in the right lung base (Figure 2 ; Supplementary Video 2).

Normally, ultrasound permits only visualization of visceral pleura, with the air content of the lung preventing visualization of lung parenchyma (Supplementary Video 3). However, consolidated lung can be seen on ultrasound as tissue-like ("sonographic hepatization") because this lung is "de-aired" in the setting of atelectasis or pneumonia.

1) When one encounters lung consolidation on ultrasound, what findings support a diagnosis of atelectasis versus an infiltrative process like pneumonia?

2) In what circumstances is a portable CXR superior to lung ultrasound and vice versa?

In this case, the sonographic consolidation contained preserved lung volume and dynamic air bronchograms (Supplementary Video 2), the latter of which is pathognomonic for pneumonia. 1, 3 The ultrasound diagnosis compelled the initiation of antibiotics and, after a few days of therapy, the patient fully recovered. In contrast to this image, ultrasound of atelectasis would show volume loss and absence of dynamic air bronchograms (Supplementary Video 4; Figure 4 ). 1, 3

A portable (aka anterior-posterior) CXR uses ionizing radiation to provide a static 2-dimensional compression of a 3-dimensional space. Based on these principles, a portable CXR is better suited than ultrasound to provide certain kinds of information. First, a portable CXR allows visualization of the entire chest in a single image, whereas ultrasound requires multiple views and will miss focal pathology in any interspace where ultrasound was not performed. Second, a portable CXR can detect pathology that lies deep to aerated lung, subcutaneous emphysema, or surgical dressings/chest tubes -any of which form an impenetrable barrier past which ultrasound waves cannot travel. Third, a portable CXR is less operator-dependent than lung ultrasound: whereas in the United States CXRs are obtained in a relatively standardized way and interpreted by board-certified Radiologists with standardized training, lung ultrasound is generally performed by providers with varying skill levels using variable image acquisition protocols.

Conversely, ultrasound has its own distinct advantages. First, an ultrasound probe's unlimited degrees of freedom on the chest wall allows clinicians to interrogate the three-dimensional nature of the chest cavity. This, in turn, allows ultrasound to clarify whether vague "opacities" seen on portable CXR are likely to represent intra-parenchymal edema, fully consolidated lung, extra-parenchymal effusion, or some combination of all of the above. 1, 4 Second, ultrasound may be more sensitive than supine CXR for detection of a small pneumothorax. 1 Third, as demonstrated in this case, ultrasound can help determine whether a lung consolidation is due to atelectasis or an infiltrative process like pneumonia. 1, 3 Fourth, in the pandemic era, handheld ultrasound probes can be easily disinfected and permanently localized within SARS-Cov-2 wards whereas X-ray machines must travel throughout the hospital and may be harder to fully disinfect given their larger surface area.

Since lung ultrasound is highly operator-dependent, standardizing practice is likely to enhance the clinical value of this modality. Standardization of lung ultrasound practice can, in turn, be accomplished by standardizing training and utilization protocols. Regarding training, this topic was recently evaluated by an Ad Hoc Committee on Point-of-Care Ultrasound (POCUS) of the American Society of Anesthesiologists (ASA). 5 The Ad Hoc Committee concluded that, based on the available literature, competence in lung ultrasound is likely to occur after: (i) completing a didactic curriculum that covers ultrasound physics and lung sonoanatomy/pathology; (ii) performing and interpreting 30 supervised lung ultrasounds; and (iii) interpreting an additional 20 supervised lung ultrasounds that need not be personally performed but cover a wide spectrum of lung ultrasound pathologies. Toward this goal, the ASA is currently developing a Certificate of Completion in Diagnostic POCUS that will include lung ultrasound as a modality. as well as posterior-lateral fields (e.g., pneumonia, atelectasis, and pleural effusions/hemothorax). 1 Notably, this approach is likely to miss focal pathology hidden behind aerated lung, such as a medially-located consolidation or a posterior, loculated pneumothorax.

In terms of equipment, we typically use a low-frequency (1-5 MHz), small footprint probe for the entire exam: low-frequency to visualize deep (i.e., parenchymal) pathology and small footprint to minimize acoustic shadowing from ribs (Supplementary Video 1). Conveniently, this probe is also the standard one used for transthoracic cardiac ultrasound. Notably, two other commonlyavailable probes can also be used for lung ultrasound, but with some limitations. A highfrequency linear probe (typically used for vascular access) can be used if the primary goal is to screen for pneumothorax, since one does not need deep penetration to sonographically detect air between the visceral and parietal pleura. Further, a low-frequency curvilinear probe (typically used for ultrasound-guided deep tissue nerve blocks and abdominal organ imaging) can also be used for lung ultrasound but at the expense of greater rib shadowing in the image.

Separate from variability in image acquisition technique, perioperative centers and ICUs may also differ in locally-appropriate indications for performing lung ultrasound based on local levels of experience with this modality and institution-specific concerns about the infectious risk of portable X-ray machines. For example, in the SARS-Cov-2 era, the indications for performing lung ultrasound may expand as hospitals look to minimize cross-contamination of portable X-ray equipment between patients. Currently within our department's critical care division, lung ultrasound is typically performed if a patient develops new/worsening respiratory insufficiency and CXR is either not rapidly available or a CXR has already been obtained but shows ambiguous findings.

In this case, the ambiguous findings on portable CXR combined with worsening, unexplained multi-organ failure prompted the ICU team to perform lung ultrasound. The ultrasound corroborated something that was already established by the CXR: the patient did not have a large pneumothorax. And more importantly, the ultrasound also clarified the "pulmonary opacities" seen on portable CXR as a pneumonia, rather than atelectasis or pulmonary edema.

Although a posterior-anterior/lateral (PA/L) CXR or computed tomography (CT) scan of the chest could have also helped to clarify the findings seen on portable CXR, these two modalities are an imperfect solution for this scenario for at least two reasons. 6 First, both tests would have required transport of a critically-ill, post-cardiac surgery patient out of the ICU for an off-site test, a transfer that risks dislodgement of critical hardware (e.g., epicardial pacing wires) and complicates management of the patient's hemodynamic instability. Second, because a PA/L CXR and CT scan are both static tests, they sometimes fail to resolve ambiguity about the nature of a consolidation. In contrast to these limitations, lung ultrasound is a portable test that is brought to the critically ill patient's bedside and allows visualization of the lung throughout the respiratory cycle. This increased temporal resolution allows ultrasound to capture dynamic findingssuch as dynamic air bronchograms pathognomonic for pneumoniathat static tests simply cannot detect (e.g., Supplementary Video 2).

In summary, lung ultrasound can detect many kinds of pathology visible on a portable CXR and even some conditions that are radio-occult. 1, 4 This could be useful to anesthesiologists not only in critical care, but also in the pre-operative evaluation of pulmonary symptoms to differentiate lower versus upper respiratory infections, especially in patients who may be uniquely harmed by ionizing radiation, such as children or parturients.

YB has performed paid consulting for Teleflex/Arrow in 2020 unrelated to diagnostic ultrasound. 

International evidence-based recommendations for point-of-care lung ultrasound

Sepsis-3)

The dynamic air bronchogram. A lung ultrasound sign of alveolar consolidation ruling out atelectasis

Lung ultrasound for the diagnosis of pneumonia in children: a meta-analysis

Ad Hoc Committee on Point-of-Care Ultrasound

Reading chest radiographs in the critically ill (Part II): Radiography of lung pathologies common in the ICU patient