Point-of-care lung ultrasound for the assessment of pneumonia: a narrative review in the COVID-19 era

We herein describe the ultrasound findings of pneumonia that were established before the COVID-19 era.

When the air content decreases and the lung density increases in the lung surface, multiple B-lines become obvious and erase A-lines [2] (Fig. 2). As this change progresses, coalescent B-lines, which appear similar to white curtains, are observed. Several experimental and clinical studies indicate that B-lines originate from the accumulation of fluid just below the visceral pleura [35, 36], thickened interlobular septa [37], or deflation of the lung [38]. However, the sonographic–histologic correlation in B-lines has not been completely elucidated [32].

Multiple B-lines, including coalescent B-lines, are observed in many lung diseases [2]. The distribution of B-lines can aid in the differentiation of diseases. In general, multiple B-lines are diffusely distributed in patients with cardiogenic pulmonary edema, acute respiratory distress syndrome (ARDS), and interstitial lung diseases [2]. In patients with pneumonia, multiple B-lines are usually observed focally, multifocally, or patchily in ground glass opacities (GGOs) or around the areas of consolidation shown on CT images [27, 39, 40]. Multiple B-lines are not a specific finding of pneumonia; thus, this finding should be interpreted according to the clinical context.

‘Sonographic consolidation’ is defined as a small subpleural hypoechoic region or large hypoechoic region with liver- or tissue-like echotexture (Figs. 3, 4, 5). Sonographic consolidations have various causes, including pneumonia, pulmonary embolism, lung carcinoma and metastasis, and atelectasis [2, 24, 41]. Sonographic consolidations in pneumonia are rarely rounded, as they would appear in patients with lung carcinoma or metastasis. The margins are usually irregular, serrated, and somewhat unclear [24]. In the presence of subpleural consolidation, the pleural line is not obvious, and lung sliding is reduced or absent [39].

The sonographic consolidations in pneumonia usually include high-echo spots and tree-like structures, which indicate air in the small bronchi. The tree-like structures seem to correspond to air bronchogram in radiology [24]. However, no study thus far has demonstrated the extent to which the tree-like structures on LUS correspond to air bronchogram on CT imaging [26]; thus, it is reasonable to use the term ‘sonographic air bronchogram’ at present in this context. Sonographic air bronchogram is not a specific sign of pneumonia [26]; however, movement with respiration is reported to be an important finding for differentiating pneumonia from obstructive atelectasis, in which the movement is not usually observed [42]. This phenomenon is called ‘dynamic air bronchogram’ [42].

LUS allows for follow-up care after the preliminary diagnosis of pneumonia and is capable of demonstrating whether the size of consolidations decreases after the initiation of treatment [25, 26]. When the dimensions of consolidations do not decrease or symptoms do not regress, follow-up LUS can help clinicians consider the possibility of malignancies or other diseases [26].

In addition to the B-mode image, color and power Doppler ultrasonography or contrast-enhanced ultrasonography is useful for evaluating the type of vascularization in consolidations for a differential diagnosis [5, 24]. However, with the exception of pulmonologists or radiologists, many clinicians are unfamiliar with these modalities in point-of-care LUS; thus, they are beyond the scope of this article.

The pleural line is recognized as a thin horizontal hyperechoic line in normal subjects. ‘Pleural line abnormality (irregularity)’ is defined as an irregularly thickened pleural line that can be observed in patients with pneumonia [23, 40], ARDS [2, 43], and interstitial lung diseases [44, 45]. Li et al. [44] reported on the application of patterns of pleural line abnormalities and their correspondence to CT findings for identifying specific pathologies. Pleural line abnormalities seem to be caused by several pathological changes. However, to the best of our knowledge, the sonographic–histologic correlation has not been investigated in pneumonia.

Several studies have shown that pleural effusion is detected by LUS in 30–46% of pneumonia patients [26‐28, 40]. In comparison with CT, LUS is an excellent modality for the detection of septations in pleural effusion [24].

Prospective studies on the application of LUS in the assessment of pneumonia (nonventilator associated) in adult patients were extensively carried out for more than 10 years before the COVID-19 era (Table 1) [9, 23, 25‐31, 39, 40, 46‐60]. These studies did not differentiate pneumonia based on the classification of pathogens or attempt to apply LUS in the differentiation of bacterial and viral pneumonia. The majority of the studies were performed in emergency departments (EDs). Other settings included intensive care units (ICUs), a stroke care unit, and an internal medicine ward. Many studies enrolled patients with suspected pneumonia/community-acquired pneumonia, acute dyspnea, or acute respiratory failure. In some studies that enrolled patients with acute dyspnea or acute respiratory failure, cardiac ultrasound and/or leg vein compression ultrasound was performed in addition to LUS to evaluate pneumonia and other pathologies causing these signs [9, 51, 53, 54, 57‐59]. The majority of the studies considered an LUS examination to be positive if signs of sonographic consolidation and/or focal multiple B-lines were observed. In some studies, positivity was defined based on the presence of sonographic consolidation alone or B-lines alone. Some study groups distinguished between small subpleural consolidations (or lesions) and larger consolidations with sonographic air bronchogram (air bronchogram) [23, 31]. The evaluation of dynamic air bronchogram in an area of consolidation was adopted in some studies [23, 30, 31, 39, 40, 49]. The majority of studies used the final diagnosis, while some studies use chest CT [23, 29, 31, 47, 55, 56] or chest CT/CXR [25, 39, 48, 60] as a reference standard for the primary analysis.

Orso et al. 16 reported a systematic review and meta-analysis (SR/MA) that included 17 studies  [23, 25, 27, 31, 39, 46, 48, 49, 51, 52, 54‐60], which evaluated the diagnostic accuracy of LUS for the diagnosis of pneumonia in ED patients of > 18 years of age. These studies provided a combined sample size of 5108 participants. The sensitivity ranged from 68% to 100%, and the specificity ranged from 25% to 100%. The pooled sensitivity and specificity were 92% (95% confidence interval [CI]: 86–95%) and 93% (95% CI 86–97%), respectively [16]. Other studies in ED settings were reported after the 17 studies. Amatya et al. [29] reported that the sensitivity and specificity were 91% and 61%, and Sezgin et al. [40] reported that the sensitivity and specificity were 98% and 96%, respectively. Several studies conducted in the ICU setting showed that the sensitivity ranged from 88% to 97%, and the specificity ranged from 50% to 94% [9, 47, 50, 53]. Some studies showed a small number of false-positive results in patients with neoplasia [26, 30, 31, 52], atelectasis [27, 31], pulmonary infarction [30, 52], abscess [27], congestive heart failure [52], chronic obstructive pulmonary disease [52], and fibrotic band [31]. As mentioned above, a significant difference in accuracy existed among these studies, and the misinterpretation of LUS findings may occur. Some studies compared the accuracy of LUS with CXR for the diagnosis of pneumonia (Table 2) [23, 27, 29‐31, 40, 49, 52]. The sensitivity of LUS was better than that of CXR when using either the final diagnosis as a reference standard or the chest CT findings as a reference standard. These results indicate that point-of-care LUS has the potential to be an initial imaging modality for the diagnosis of pneumonia.

Two studies showed that the accuracy of LUS improved under the guidance of physical examinations. Reissig et al. [25] reported that a combination of auscultation and LUS findings increased the positive likelihood ratio (LR) and decreased the negative LR. Nazerian et al. [31] found that sensitivity and specificity increased in patients complaining of pleuritic chest pain, which allowed the initial examination to be focused on a limited chest area.

D’Amato et al. [26] showed the significance of LUS monitoring in pneumonia patients and reported the change in size of the areas of consolidation as follows: 6.3 ± 3.4 cm at 0 days, 2.5 ± 1.8 cm at 4–6 days, and 0.9 ± 1.4 cm at 9–14 days. Out of the 12 patients with delayed lesion healing, 7 were found to have lung cancer.

Ventilator-associated pneumonia (VAP) is a nosocomial infection in the ICU that affects patients who receive mechanical ventilation for more than 48 h. In mechanically ventilated patients, pulmonary infiltration on CXR frequently develops and may be associated with multiple etiologies, including VAP and noninfectious processes, such as atelectasis. A combination of clinical criteria, such as the Clinical Pulmonary Infection Score (CPIS), has been proposed for the diagnosis of VAP [61]. However, a previous study reported that CPIS showed poor diagnostic performance [62].

The LUS diagnosis of VAP in ICUs is more challenging in comparison with the diagnosis of community-acquired pneumonia in EDs because the mechanically ventilated patients cannot lie in the lateral decubitus position by themselves. The presence of extensive thoracic dressings or drainage tubes limits assessment by LUS in postoperative or trauma patients on mechanical ventilation. Furthermore, the higher prevalence of atelectasis in mechanically ventilated patients may cause lower specificity in the diagnosis of VAP [63]. However, LUS has a number of advantages in the ICU setting. It can be used immediately in real time at the bedside. In addition, it can also be used for sequential monitoring to detect the emergence of lesions [64] or to semi-quantify lung aeration [65]. Several studies have demonstrated the utility of LUS, based on the finding of sonographic consolidation, in the assessment of VAP in the ICU setting [63‐69]. Some of the studies distinguished small subpleural consolidation (> 0.5 cm) and lobar/hemilobar consolidation with dynamic air bronchogram [63, 64, 68, 69].

Berlet et al. [67] reported that daily LUS examination with an abbreviated scanning protocol showed 100% sensitivity and 60% specificity in the diagnosis of VAP. The distinction between dynamic and static air bronchograms was of no significant additional benefit for the diagnosis in this study. They hypothesized that the lower specificity was explained by the high prevalence of atelectasis.

LUS alone cannot be used to accurately differentiate between inflammatory and noninflammatory consolidations. Zagli et al. [66] proposed a new score based on LUS, procalcitonin, culture of tracheal aspirate, purulence of tracheal secretion, temperature, and oxygenation to improve the diagnosis of VAP: the Chest Echography and Procalcitonin Pulmonary Infection Score (CEPPIS). A CEPPIS of > 5 was significantly better in predicting VAP (sensitivity, 81%; specificity, 85%) than LUS findings alone (sensitivity, 59%; specificity, 85%) or a CPIS of > 6 (sensitivity, 40%; specificity, 83%). Mongodi et al. [63] showed that the results were best with the combination of consolidations with dynamic air bronchogram or ≥ 2 areas with small subpleural consolidations and positive gram staining. Zhou et al. [68] demonstrated the superiority of the combination of LUS with procalcitonin to LUS alone in specificity. The sensitivity and specificity of LUS were 92% (95% CI 79–97%) and 63% (95% CI 51–74%), respectively, whereas the sensitivity and specificity of the combination were 81% (95% CI 67–91%) and 86% (95% CI 75–92%).

These studies excluded patients with ongoing pneumonia or other pulmonary diseases at ICU admission [63, 66‐68]. Staub et al. [64] did not exclude such patients. As a result, there was a high frequency of sonographic consolidations before 48 h of mechanical ventilation. Instead, the emergence of consolidations was determined by comparison with the LUS findings of the previous day. They reported that serial LUS examinations detected the emergence of early and specific signs of VAP, and that the presence of sonographic consolidations was more specific indicators of VAP when they emerged in anterior lung areas, given that noninfectious infiltrates, such as passive or resorptive atelectasis, rarely compromise the anterior lung areas.

Bouhemad et al. [65] demonstrated that sonographic consolidation was replaced by B-lines when reaeration was confirmed by CT in patients treated with antibiotics, and LUS was more appropriate than CXR for quantifying lung reaeration.

A single-center diagnostic randomized controlled trial investigated whether LUS improved patient care. In the control group, VAP was diagnosed using a combination of CXR and clinical findings. In the intervention group, VAP was diagnosed using a combination of LUS as a daily monitoring tool and clinical findings. This trial showed that the index of ventilator-free days was higher in the intervention group than in the control group (7.8 ± 9.7 days versus 3.7 ± 6.4 days, p = 0.044). The use of LUS monitoring for the diagnosis of VAP may improve patient outcomes in comparison with the standard diagnostic strategy [69]. However, larger studies are needed to confirm the potential benefit of LUS for these patients.

When considering the application of LUS in clinical practice, the CT features of patients with COVID-19 pneumonia provide us with useful information. An SR/MA of 13 studies on the anatomic distribution of the CT findings of patients with COVID-19 pneumonia found that the bilateral lungs (78%) and periphery (77%) were the most common sites of involvement. The incidence rates were higher in the right lower lobe (87%), left lower lobe (81%), and bilateral lower lobes (65%). The right upper lobe (65%), right middle lobe (55%), and left upper lobe (69%) were also commonly involved [70]. Based on these CT features, scanning of the entire region, including the posterior regions, is recommended when performing LUS for the evaluation of COVID-19 pneumonia. Lu et al. and Castelao et al. [71, 72] reported, based on an LUS study, that the lower lobes and posterior regions had a greater tendency to be involved.

The LUS findings in COVID-19 pneumonia are similar to those described in patients with pneumonia before the COVID-19 era. Lopes et al. [73] showed an association between multiple B-lines on LUS and GGOs on CT as well as between subpleural consolidations on LUS and consolidations on CT. Mohamed et al. [74] reported an SR/MA including seven studies with a total of 122 COVID-19 patients which examined the role of LUS. The pooled proportion of multiple B-lines (including focal, multifocal, and coalescent types) (Fig. 6) detected by LUS was 0.97 (95% CI 0.94–1.00), that of pleural line abnormalities was 0.70 (95% CI 0.13–1.00), that of small subpleural consolidation (Fig. 6) or large consolidation (Fig. 7) was 0.39 (95% CI 0.21–0.58), and that of pleural effusion was 0.14 (95% CI 0.00–0.37). Multiple B-lines, including focal, multifocal, and coalescent types, were the most common and consistent findings; however, other LUS findings had a high degree of heterogeneity. Large lobar or trans-lobar consolidations with air bronchograms are less common in the early phases of COVID-19 pneumonia [20, 75]. When larger consolidations are observed initially, bacterial pneumonia or superimposed bacterial infection should be suspected [20, 75, 76]. Larger consolidations can be present in more advanced phases of COVID-19 pneumonia [75].

Interestingly, a newly described finding has been observed in patients in the early phases of COVID-19 pneumonia. Volpicelli et al. [20] named the finding the ‘light beam.’ This is a band-like vertical artifact that often appears and disappears from the screen with respiration. The artifact corresponds to the early appearance of GGO on CT. It is assumed that the light beam was included in the category of coalescent B-lines in some studies. A multicenter study is currently investigating the specificity of this sign [20].

A few studies have demonstrated high agreement of the severity between LUS and chest CT in severe patients with proven COVID-19 pneumonia [71, 77]. Yang et al. and Lopes et al. [73, 78] demonstrated that LUS was more sensitive than CT in the detection of subtle changes of the lung surface due to its high resolution; however, the clinical significance has not been proven. Meanwhile, a study demonstrated the diagnostic value of LUS using a COVID-19-positive result in a reverse transcriptase polymerase chain reaction (RT-PCR) assay as a reference standard. Peyrony et al. [79] reported the diagnostic accuracy of LUS with a pocket-sized device in 84 patients tested with RT-PCR in an ED. The sensitivity and specificity of a positive LUS result (defined as the presence of bilateral B-lines) were 77% (95% CI 62–88%) and 89% (95% CI 75–97%), respectively, and LUS was superior to CXR.

Time course of chest CT and LUS findings was reported in patients with COVID-19 pneumonia. Pan et al. [80] showed the time course of CT findings in 21 patients recovering from COVID-19 pneumonia without severe respiratory distress. Initial lung findings were small subpleural GGOs that grew larger with crazy-paving pattern (GGO with superimposed inter- and intralobular septal thickening) and consolidation. Lu et al. [71] reported that serial LUS mainly found B-lines to be increased and coalescent, and sonographic consolidations to be enlarged and increased, which was in high agreement with the CT findings. Extension toward larger consolidations, especially in a gravitational position (Fig. 7), indicates the phase of respiratory failure that requires invasive ventilation [81]. In the recovery phases, regression of the abovementioned findings can be observed with the re-emergence of A-lines [82, 83]. As mentioned above, serial LUS at the bedside is a promising method for monitoring COVID-19 pneumonia. Semi-quantification of lung aeration with LUS may be useful for monitoring the severity. Several semi-quantitative methods have been proposed [18, 20, 71, 84]. For example, each hemithorax is divided into six regions using anterior and posterior axillary lines and an axial line, and the following grades are used in each region: 0 points (A-lines and < 3 B-lines); 1 point (≥ 3 B-lines); 2 points (coalescent B-lines); 3 points (consolidation). The score is calculated as the sum of all regional scores ranging from 0 to 36 points [71].

Several studies indicate that baseline LUS score correlates with the eventual need for ICU admission and invasive mechanical ventilation, and the score is a predictor of mortality [85, 86]. The anterior region, although usually more spared, may have prognostic value because its involvement strongly correlated with the risk of requiring noninvasive respiratory support [72]. In deteriorating patients, LUS pathology worsened mostly in the anterior region [86].

Many case reports and observational studies with a small number of patients with COVID-19 pneumonia have been published in a short period. That is, the COVID-19 pandemic has made many clinicians interested in the usage of LUS for the assessment of COVID-19 pneumonia. However, to date, there is no robust evidence on the diagnostic accuracy, the efficiency as a monitoring tool, or the contribution to patients and medical personnel.

Ultrasound had not been considered as a diagnostic tool for the evaluation of pneumonia. To date, the concept of POCUS has been widely accepted, and many clinical studies on LUS for the evaluation of pneumonia have been performed based on the idea of POCUS. Unlike chest CT, LUS cannot visualize the whole lung. However, several LUS findings obtained on or via the pleurae can contribute to the diagnosis of pneumonia and monitoring of the condition if they are properly interpreted in a clinical context.

Due to the superior sensitivity of LUS in comparison with CXR [23, 27, 29‐31, 40, 49, 52, 65, 79], LUS has the potential to be an initial imaging modality or to reduce the number of CXR examinations [87]. It is also reasonable to consider the complimentary use of LUS, CXR, and chest CT owing to the excellent mobility of LUS at the bedside. LUS is a radiation-free imaging modality, which makes it useful for daily monitoring and particularly suitable for pregnant patients [88‐90]. LUS with portable devices can also be performed in nursing homes [91] or patient homes [92] to reduce the need to transport the patient to a hospital for a diagnosis or monitoring with traditional imaging resources.

Theoretically, LUS may also be advantageous in infection control in the COVID-19 era. For example, stethoscope usage is strictly restricted to avoid COVID-19 virus transmission. It is difficult to create specific protective covers for stethoscopes or for medical personnel wearing personal protective equipment (PPE) to use a stethoscope [93]. The hazards of CT use include overuse of hospital resources, including PPE and other protective equipments that are required to safely perform CT, and disease transmission and exposure among staff engaged in patient transportation, technologists in imaging departments, and nonaffected patients who require CT examination in the same departments [22]. In contrast, portable ultrasound machines can easily be deployed at the bedside and covered with transparent plastic sheets for protection [94, 95]. Some handheld ultrasound devices with a wireless function are more easily covered with sheets or bags [93, 96]. However, it has to be kept in mind that the comprehensive LUS examination by experienced operators requires 5 to 15 min [26‐29, 84]. It has not been scientifically evident whether introduction of LUS examination reduces the risk of the COVID-19 virus transmission.

When interest in LUS for the assessment of pneumonia is increasing, there are several issues to be considered and solved. First, some concerns have been raised regarding the robustness of the obtained results, because of the lack of a well-standardized methodology in the involved studies [16]. The issues in conducting clinical studies include the choice of reference standards for the final diagnosis, the competency of LUS examiners [84], and standardization of LUS findings. Therefore, more methodologically rigorous studies are still needed [16]. Second, there were few evidences on utility of LUS to differentiate bacterial pneumonia based on the classification of pathogens. To the best of our knowledge, there is only one case report [97] and few small clinical studies [98, 99] on LUS focused on viral pneumonia other than COVID-19 pneumonia in adult patients. Bacterial pneumonia is suspected when large lobar or trans-lobar consolidations are observed initially; however, there had been no prospective studies on the diagnostic accuracy of LUS in the differentiation of bacterial and viral pneumonia in adult patients before the COVID-19 era. On top of them, it has not been proven whether the features of LUS findings in COVID-19 pneumonia can be useful markers for differentiating COVID-19 pneumonia from bacterial pneumonia or viral pneumonia caused by other pathogens. Further studies are still needed to reveal whether LUS supports us to differentiate pneumonia based on the classification of pathogens. The robust evidence that is expected to accumulate in the COVID era may provide us with clues for the differentiation, which would facilitate more appropriate management. Third, standardization of the scanning regions and protocol has not been established in the assessment of pneumonia. Scanning all intercostal space is clinically time-consuming; therefore, the simplified approaches are needed, especially in time-sensitive situations. When selecting the scanning regions, it has to be kept in mind that distribution of the lesions is affected by pathogens and clinical course. On top of that, each approach suitable for the diagnosis, monitoring, or prediction of outcome may have to be developed according to the previous studies.

Utility of point-of-care LUS for the assessment of pneumonia has been shown mainly in the field of emergency medicine and intensive care medicine. Collaboration between these specialties and pulmonology is indispensable for the further development of this modality.