Transthoracic lung ultrasound (LUS) has been used in critical care settings for well over fifty years with Dr. Joyner reporting on the evaluation of pleural effusions in 1967 (Joyner effusion). Initially, its use was thought to be of low utility in evaluating the lung because air in the lung interferes with the reflection of ultrasound beams. Dr. Lichtenstein detailed the use of LUS in the critical care setting in 1983, noting that the medical findings are in actually in the artifacts (Lichtenstein Lung) Since then LUS has found many uses in the medical setting and has been explored more extensively in adult populations. However, over the last decade, it has gained more popularity in pediatric patients. In 2012, International Guidelines further supported the use of LUS in pediatrics providing strong recommendations for its use in a variety of applications and notably remarking that adult lung findings are also found in the pediatric population (Volpicelli International). In 2013, Marin and Lewiss published a policy statement that provided a framework for point of care ultrasound training and integration into pediatric care (Marin Policy). Since then, The American Academy of Pediatrics has released a policy statement regarding training pathways and also provided framework for its integration into practice(AAP Point of Care). LUS can be completed at the bedside with real-time results, does not provide the added risk of ionizing radiation, and is generally well tolerated by patients. Further, more recent studies have shown that pediatric emergency physicians can be trained to acquire and interpret images as well as apply the information to their clinical practice(Shah, Deswani, Jones).
This article will cover respiratory pathology within the pediatric emergency setting focusing on image acquisition and current research regarding the use of LUS to diagnose and evaluate pleural, alveolar, and interstitial syndromes.
It is imperative to understand basic anatomy in order to acquire and evaluate the images. The pediatric chest is relatively small allowing for nearly any probe to be used to interrogate the superficial chest structures. Because of the inherently smaller pediatric thorax and the superficial location of the ribs and pleura, the linear probe provides enough penetration and resolution for evaluation of the pleura and superficial structures. The phased array and curvilinear probes can be used to evaluate the deeper structures including the costophrenic angles for effusions and consolidations. Once the probe is chosen, the focus should be on image acquisition. Lichtenstein has eloquently discussed the displacement and mixing of air and fluid in the lungs (Lichtenstein BLUE). With this in mind, the probe should be placed on the most anterior surface of the chest when evaluating for a pneumothorax in the supine patient (image anterior lung/apex) because the less dense air will move to the least dependent area of the chest. There is more water content in a pleural effusion, therefore, the phased array or curvilinear probe should be used at the costophrenic angle to initially evaluate for effusions (image probe costophrenic angle). Consolidations and interstitial syndromes tend to be a mix of fluid and water; therefore, the chest should be divided into areas and a zone approach should be followed where the lung is also interrogated in the sagittal and transverse plane (Volpicelli blus/Copetti/Milliner).
Ultrasound waves do not penetrate the calicified boney ribs; therefore, a dark anechoic shadow (the rib) immediately interrupts the field. Younger ribs are more cartilaginous, so there will be less rib shadowing. The pleural line is then appreciated as a white hyperechoic line. Immediately deep to the pleural line is the lung tissue(image bat sign). Normally, well aerated lung is seen as a series of repeating hyperechoic lines. These lines are termed A-lines and are simply equidistant reflections of the pleural line. A-lines indicate air and will be found in cases of a normal lung, pneumothorax, apnea, or mainstem intubation (image A-lines). Occasionally the pleural line will be interrupted by short jet like vertical projections that do not interrupt the A lines. These lines are generally inconsequential and fade with lung sliding; these are a type of comet tail artifact termed Z-lines(image comet tail artifact). Comet tail artifacts that interrupt the A lines and extend the entire length of the screen are termed B-lines. These lines arise from the visceral pleura and are a result of fluid within the interaveolar septum. These lines do not fade with pleural movement. Few B-lines are also generally inconsequential; however, repeating or more numerous B-lines are pathologic representing an increasing fluid balance within the alveoli as seen in interstitial pneumonia, bronchiolitis, and pulmonary edema (image nonpathological B-line). These artifacts provide the groundwork for obtaining and understanding the images acquired when performing a lung ultrasound.
A pneumothorax occurs when air is within the pleural space separating the parietal and visceral pleura. The less dense air does not allow the ultrasound beams to penetrate into the deeper tissues. This only allows the superficial structures to be visualized. Given this, the linear high frequency probe is superior in evaluating a pneumothorax particularly in the pediatric chest. There are several sonographic findings associated with a pneumothorax. The signs include the absence of lung sliding, the absence of B-lines, the presence of a lung point, and the absence of a lung pulse(Volpicelli PTX). When evaluating for a pneumothorax in the supine pediatric patient, the linear probe should be placed anteriorly on the chest near the midclavicular line along the second through fourth intercostal spaces scanning caudally in a longitudinal fashion. In the upright pediatric patient, air will accumulate at the apices (Husain LF, Hagopian L, Wayman D, Baker WE, Carmody KA. Sonographic diagnosis of pneumothorax. J Emerg Trauma Shock 2012;5:76-81)
The field should then include the most superior rib, the parietal pleura, and the inferior rib (image). Once this view is obtained, the sonographer should assess for pleural sliding. It is oftentimes difficult to appreciate the subtle sliding of the pleura; therefore, the M-mode and doppler function can be used to further determine if sliding is present. When using the M-mode function, the ultrasound caliper should be placed between the superior and inferior rib along the middle of the pleural line. In the normal lung, the “seashore sign” will be present. The more static soft tissue lacks movement and will appear as the “waves” while the to and fro movement of the pleural line during the respiration will appear as “sand.” (image) (video) When a possible pneumothorax is present, the sonographer will appreciate a “barcode” or “stratosphere” sign. This finding indicates that air is between the parietal and visceral. The ultrasound beams strike the air and return to the transducer. (image) (video) If the doppler function is used, the pleural line will appear to have color; this color will be absent if a pneumothorax is present. (image) (video) The presence of B-lines excludes the diagnosis of a pneumothorax. B-lines arise from the visceral pleura and move with respiration. These lines are not always present in normal patients; however, the presence of B-lines rules out a pneumothorax because they arise from the visceral pleura and move with lung sliding. There are also instances of false positive findings. Mimickers include mainstem intubation, adhesions, massive atelectasis or consolidations, and apnea. The identification of a “lung pulse” and “lung point” proves useful in these instances. The lung pulse can exclude a pneumothorax. It occurs when the heart beats causing the lung to subtly move vertically; using M-mode this will be appreciated by an alternating seashore and stratosphere sign. The lung point is evaluated by moving the probe laterally and posteriorly along the chest wall to identify an area where lung sliding or artifacts occur. This area is termed the lung point and effectively diagnoses a pneumothorax. This point might be absent in a complete pneumothorax as the lung is nearly completely retracted. Overall, the presence of lung sliding, B-lines, and a lung pulse rule out a pneumothorax. If the diagnosis remains unclear in the absence of lung sliding and B-lines, then additional findings should be sought including the lung point and a lung pulse.
At present a bulk of the literature regarding lung ultrasound usage in the diagnosis of a pneumothorax (excluding growing neonatal literature) is based on adult literature that provides strong evidence for using LUS to evaluate for pneumothoraces. However, pneumothoraces both traumatic and nontraumatic occur in the pediatric population as well. The literature suggests that physicians can be trained to quickly evaluate for a pneumothorax in the clinical setting; furthermore, ultrasound tends to be more accurate than CXR in rulingout the diagnosis of a pneumothorax.
In 2005, Lichtenstein et al detailed the use of ultrasound to diagnose occult pneumothoraces. This study compared the ultrasound to supine chest x-ray(CXR) and computed tomography of the chest (CT). His study found that the absence of lung sliding had a sensitivity of 100% and specificity of 78%. The absence of lung sliding in addition to the A line profile (absence of B-lines) showed a sensitivity of 94% and specificity of 95%. When the lung point was identified along with the absence of lung sliding and the A line profile, the specificity increased to 100% with a sensitivity of 79%. This study indicates that a pneumothorax can be ruled out in the setting of lung sliding as well as when B-lines are identified. A lung point will rule in a pneumothorax. Of note, the evaluation of the thoraces for a pneumothorax using sonography required 2 minutes to complete (Lichtenstein 2005). A study published in 2013 designed to assess the accuracy of ultrasound in diagnosing post-traumatic pneumothorax using a two-step process found that emergency physicians were able to accurately diagnose a pneumothorax in trauma patients who did not have obvious clinical signs present. The physicians were experienced in focus assessment of trauma but completed a two hour training session including lectures and hands on training prior to the study. The patient populations were evaluated by bedside ultrasound followed by CXR and a gold standard chest CT. Results showed the ultrasound had a sensitivity of 86.4%, specificity of 100%, positive predictive value of 100%, and negative predictive value of 85.1%. CXR showed a sensitivity of 48.64%, specificity of 100%, positive predictive value of 100%, and negative predictive value of 85.1%. The mean time to complete the ultrasound evaluation of pneumothorax was two minutes while the mean time to complete the CXR was 12 minutes. The study also considered the physicians’ accuracy using sonography after a subset of exams. The sensitivity improved from 60% after five exams to 89.47% after 20 exams with negative predictive value increasing from 89.5% to 95.6%. The study concluded that ultrasound evaluations completed by emergency physicians after brief training is an accurate modality for the diagnosis of pneumothorax (Abbasi PTX). This study is comparable in results to previously published results (Zhang PTX) . Several meta-analysis findings comparing ultrasound to CXR also report higher pooled sensitivities and similar specificities for clinician and radiologist performed ultrasounds. (Ding et al, Alrajhi) The 2012 International Lung Guidelines provide strong recommendations that lung ultrasound should be used in evaluating for a pneumothorax and further report that lung ultrasound is more sensitive than a supine CXR in ruling out the diagnosis of a pneumothorax. (Volpicelli International) These findings indicate that physicians can be trained to quickly evaluate the thorax for a pneumothorax and accurately rule out the diagnosis; further, the ultrasound evaluation is more sensitive in ruling out a pneumothorax than supine CXR.
Pleural effusions are appreciated when fluid accumulates between the parietal and visceral pleura. Fluid will most commonly occur in the most dependent portions of the chest cavity. With the aid of gravity, this dark, anechoic or hypoechoic area can be appreciated along the posterior costophrenic angle. Because fluid acts as a great acoustic window, the remaining lung inflated lung will appear brighter or “hyperechoic.” If the effusion is large enough to compress the lung, the lung will appear as a floating structure within the darker pleural effusion. Pleural effusions can be appreciated by the “quad sign.” This is a rectangular area with bilateral rib shadow boundaries, a superior parietal pleural border, and an inferior lung line or visceral pleural border; the center is an anechoic or hypoechoic structure(image). The area can be further defined by using M-mode. When M-mode is applied, the lung line or visceral pleura will move towards the parietal pleura during inspiration creating a “sinusoid sign” (image) (Lichtenstein Lung). Because fluid will most commonly accumulate at the most dependent portions, a lower frequency probe should be used to interrogate the costophrenic angles along the posterior axillary line (Volpicelli International). This angle is generally bordered by the diaphragm inferiorly imposed on the liver or spleen while the lung is appreciated superior to the diaphragm. Because of the proximity of the liver and spleen, color doppler can assist in the evaluation.
Research has shown that LUS is more sensitive than CXR in evaluating pleural effusions. Further, physicians can be trained to use LUS to identify and measure pleural effusions which would likely decrease the complication rate with performing procedures including thoracenteses. Lung ultrasound has been compared to CXR using chest CT as the gold standard in several studies. Lichtenstein et al assessed whether LUS could serve as an alternative in detecting the presence and extent of a several pathological processes including pleural effusions; the study further compared the diagnostic performance of auscultation, CXR, and LUS in patients with ARDS. 26% of the study patients were found to have pleural effusions. LUS proved to have higher sensitivity, specificity, and accuracy at 92%, 93%, and 93% versus 42%, 90%, 61% for auscultation and 39%, 85%, and 47% for CXR. Similarly, Xirouchaki et al evaluated ventilated patients in the ICU comparing the performance of LUS with CXR using CT as a gold standard. 42 patients were found to have pleural effusions. LUS had a sensitivity, specificity, and accuracy of 100%. CXR had a sensitivity of 65%, specificity of 81%, and accuracy of 69%. There are additional systemic review articles showing similar results; Grimberg evaluated four articles finding LUS to have a mean sensitivity of 93% and a mean specificity of 96% to detect pleural effusions (Grimberg). Yousefifard et al conducted a systematic review of 12 articles to determine the value of LUS and CXR in detecting pleural effusions. LUS showed a pooled sensitivity of 0.94, pooled sepcficity of 0.98. CXR showed a pooled sensitivity of 0.51 and a pooled specificity of 0.91 (yousefifard). International Guidelines provide strong recommendations that LUS is more accurate than supine CXR and as accurate as CT for detecting pleural effusions (Volpicelli Guidelines). While training is required to understand the basics of identifying and quantifying pleural effusions, studies have shown that with brief training, novice residents have been able to identify and quantify pleural effusions. Begot et al evaluated LUS-novice resident physicians’ ability to identify and quantify unloculated effusions in ICU patients. The resident physicians received three hours of training including didactics, case illustrations, and hands on teaching. When compared to physicians who were considered experts, the kappa coefficients showed good to excellent agreement for diagnosis pleural effusions and measuring interpleural distances. The study concluded that novice resident physicians can identify and quantify unloculated effusions using LUS with good agreement with physicians considered to be experts (begot). Beyond this a recent systematic review and meta-analysis regarding the rate of pneumothoraces following thoracenteses found that the overall rate of post procedural pneumothoraces of 6% decreased with an odds ratio of 0.3% with LUS use. The study also showed that a pneumothorax associated with two or more needle passes, following therapeutic thoracentesis in conjunction with periprocedural symptoms and nonsignifitantly with mechanical ventilation and less experience (Gordon PE complication). Cao et al evaluated the efficiency of ultrasound-giuided thoracentesis catheter drainage for pleural effusions using a control group who received a pleural puncture in comparison to an interventional group who received an ultrasound guided thoracentesis. The study showed an 84% success rate in the control group versus a 100% success rate in the intervention group. Beyond factors related to the type of tube used, the authors also noted that the complication rate was lower in the intervention group (Cao). The current data suggests that LUS is more accurate and sensitive than CXR at evaluating pleural effusions; further, physicians can be trained to identify and quantify pleural effusions. Further, ultrasound aids in procedural guidance and decreases complications.
Lung ultrasound has also proven useful in evaluating the lung parenchyma for pathological processes beyond the pleura. A more commonly encountered pediatric parenchymal pathological process includes pneumonia which is represented centrally by consolidations possibly accompanied by a combination of additional findings. In pediatric patients, most of these consolidations directly contact the visceral pleura causing a series of findings as air and water mix. Normally well-aerated lung shows A-lines with the possibility of no more than three B-lines per field. As the water content increases, the ultrasound wave transmission becomes easier generating images reflective of consolidated lung. Consolidations can have several appearances on ultrasound; they are mostly of mixed echogenicity or hypoechoic. Notable findings include the “shred sign,” and “the tissue-like sign” or “hepatization.” These areas of consolidation can also be associated with effusions, have B-lines in or around them, or have “air bronchograms.” The shred sign appears as an irregular pleural border between the consolidated and well aerated lung (Lichtenstein lung). This causes the pleural line to have a shredded appearance (image). At other times, the lung will take on a tissue-like appearance. This is termed the tissue-like sign or hepatization (image). If the consolidation has a mix of air within it, it is considered to have air bronchograms (image). When considering pneumonia, the air bronchograms will generally move with respiration unless there is a large enough consolidation that does not allow the pleura to move with respiration and appear as a branching pattern within the consolidation. The literature documents several scanning protocols for identifying consolidations (Copetti; Chavez; Milliner; Nazerian; Caiulo). Because the pediatric thorax is smaller and most findings are likely to abut the pleural line, the linear probe is most commonly used to evaluate the structures. However, a phased array or curvilinear probe can also be used especially when the costophrenic angles require interrogation. The anterior, axillary, and posterior lung fields can be evaluated in both longitudinal and transverse planes; the costophrenic angles can be interrogated by placing a lower frequency probe along the posterior axillary line such that the lung superior to the diaphragm overlying the liver or spleen can be evaluated (image). To date, there are more studies suggesting that utilizing sonography to evaluate for consolidations in pediatric patients. Moreover, studies have shown that ultrasound can be as accurate if not better than chest xray at detecting lung consolidations. These sonograms can be efficiently obtained at the bedside by trained pediatric emergency physicians. In fact, lung ultrasounds have been found to be more time and resource efficient by decreasing patient costs and by decreasing total emergency department length of stay. Over the last decade, dedicated pediatric studies have shown increased accuracy of lung ultrasonography in diagnosing pneumonia in children. Reali et al compared lung ultrasound to chest xray in diagnosing pneumonia in children 16 years of age and under. The lung ultrasounds were completed by either a pulmonologist or one of two resident physicians who had performed at least 100 prior scans. The chest xray results were determined by a pediatric radiologist. The final diagnosis of pneumonia was then based on a committee using recommendations from the British Thoracic Society. The study showed ultrasound, when compared to chest xray, had a sensitivity of 94% and specificity of 96% in diagnosis pneumonia in children. Esposito et al similarly evaluated the performance of lung ultrasound in children with pneumonia when compared to chest xray. The study enrolled patients who were 14 years old or younger who presented with fever and suspicion of pneumonia. Each participant received an xray that was interpreted by a radiologist followed by a lung ultrasound completed by a pediatric resident. The resident’s ultrasound training consisted three hours of lecture followed by four hours of hands on training prior to the study. The study showed lung ultrasound had a sensitivity and specificity of 97.9% and 94.5% respectively when compared to the standard of chest xray in diagnosing pneumonia. Additionally, a meta-analysis of eight articles showed a pooled sensitivity of 96% and pooled specificity of 93%. This meta-analysis indicates good accuracy in diagnosis and excluding pneumonia among sonographers of various levels of experience.
Shah et al performed a prospective observational study aimed at determining the accuracy of point of care ultrasound for the diagnosis in children and young adults. Fifteen pediatric emergency medicine physicians of various ultrasound experience levels completed 1 hour of lecture and hands on training prior to enrolling patients. The pediatric emergency medicine interpreted ultrasound findings were compared to the criterion, chest xray findings, as interpreted by a pediatric radiologist. On average, the point of care ultrasound took seven minutes to complete. Results showed a sensitivity of 86% and specificity of 89% for all pneumonias; when the size of consolidation increased above 1 cm, the specificity increased to 97%. The results suggest that after minimal training, point of care ultrasound is very good at ruling in pneumonia despite operator experience.
Similarly, Daswani et al showed a sensitivity of 87% and specificity of 94% with a positive likelihood ratio of 14.6 and negative likelihood ratio 0.14 when using ultrasonography to diagnose acute chest syndrome (defined as ultrasound consolidations with air bronchograms) in sickle cell patients who have fever comparing ultrasound to the criterion of chest xray. Fifteen pediatric emergency medicine physicians received 1 hour of training prior to the study. The study sonographers were blinded to the clinical examination and ultrasound was completed prior to the chest xray. The ultrasound study required an average of 9 minutes to complete. The study suggests that lung ultrasound is both sensitive and specific in diagnosing acute chest syndrome and may be an accurate imaging modality that reduces the need for chest xray.
The prior studies indicate that pediatric emergency medicine physicians can be adequately trained to identify sonographic consolidations completing exams under 10 minutes. Jones et al completed a randomized control trail comparing lung ultrasonography to chest xray in 191 children and adults up to the age of 21. The study aimed to determine the feasibility and safety of substituting lung ultrasound for chest xray in patients suspected of having pneumonia. The study consisted of an investigative arm where patients received a lung ultrasound performed by enrolling pediatric emergency medicine physicians. A chest xray was completed in the investigative arm only in instances of physician uncertainty, admit or primary physician or family request. The control arm received a chest xray followed by a lung ultrasound performed by pediatric emergency medicine physicians. The pediatric emergency medicine physicians were provided with 1 hour of training prior to the study; each ultrasound required an average of 7 minutes to complete. The patients were subsequently followed for rates of missed pneumonia or adverse events. The study ultimately showed a 38.8% reduction in chest xray use in the investigative arm. Further, there were no changes in management in patients within the investigative arm if a chest xray was requested after the ultrasound was completed, suggesting that the ultrasound findings were accurate. Beyond this, the investigative arm resulted in a $9200 overall reduction of costs by foregoing the chest xray suggesting there might be a cost benefit to performing the lung ultrasound. ED length of stay was also, on average, 27 minutes less for the ultrasound only investigative arm compared to the control arm. While the studies required 7 minutes of dedicated bedside care, the department length of stay was decreased by performing a bedside ultrasound and there was a cost reduction to the patient. A nonstatistically significant difference found in the rate of antibiotic use in the investigation arm which the authors attributed to likely treatment of subcentimeter consolidations. Similar to prior studies, this study suggests that pediatric emergency medicine physicians can be trained to accurately identify consolidations. It also shows that lung ultrasound might be a feasible and safe substitute for chest xray.
Lung ultrasound additionally proves useful in evaluating interstitial pathology as well including pulmonary contusions. There is limited data regarding the role of ultrasound in pediatric lung contusions. However, the more robust adult research has served to define ultrasongraphical appearances of lung contusions as well as provide data regarding ultrasound accuracy in diagnosing contusions. Lung contusions have been defined by Rocco et al as hypoechoic blurred lesions with indistinct margins whose dimensions do not change with respiration; these lesions might have air bronchograms. Lung contusions are also defined by multiple B-lines (Rocco). Soldata et al define lung contusions as involving multiple B-lines arising from the pleura in patients without clinical suspicion of cardogenic pulmonary edema indicative of alveolointerstitial syndrome or by peripheral parenchymal lesions represented by C lines, hepatization, or parenchymal changes with a localized effusion (Soldata).
Evaluation of pulmonary contusions can be completed with a linear probe as well as a phased array or curvilinear probe in fashion similar to the technique used to evaluate for pneumonia.
In 2009, Stone reported a case of bedside ultrasound being used to assist in the evaluation of a pulmonary contusion in a 10 year old male who presented after a pedestrian versus automobile event. The patient was noted to have CXR findings concerning for a right lower lobe consolidation; he used the curvilinear probe to evaluate the area of tenderness and CXR concern noting bilateral lung sliding with A-lines in the bilateral anterior and left lateral lung fields; however, a focality of B-lines were observed over the area of chest wall tenderness with CT confirming an area of lung contusion. Adult literature identifies lung ultrasound as more accurate and a better correlate to CT than CXR. Soldati et al compared LUS to CXR and a gold standard CT in patients presenting to the emergency department with chest trauma. A retrospective evaluation was obtained by reviewing chart data and images in 76 patients and prospectively with two sonographers in 12 patients. The sonographers used a convex probe to evaluate the patients within fifteen minutes of arrival scanning for both pneumothorax and lung contusing using a previously detailed approach. Supine AP CXR was then performed and analyzed by a radiologist blinded to the LUS results. A chest CT was performed within 60 minutes of arrival and was used as the gold standard. There were 10 positive CXR findings providing a sensitivity of 27%. LUS was positive 37 times. LUS proved a sensitivity of 94.6%, specificity of 96.1%, positive predictive value of 94.6%, negative predictive value of 96.1%, and accuracy of 95.4%. There were two false positive findings on LUS both with findings consistent with alveolointerstitial syndrome. LUS showed peripheral parenchymal lesions in 7 patients who also had alveolointerstitial findings. The data for peripheral parenchymal lesions sensitivity decreased to 18.9% while specificity increased to 100%. The study concludes that LUS can accurately detect lung contusions in trauma patients (Soldati). In 2008, Rocco et al showed similar results in a study aimed at evaluating the role of lung ultrasound compared to CXR using chest CT as a gold standard in trauma patients with respiratory failure admitted to an intensive care unit. The study found that CXR proved to be a poor predictor of the extent of lung injuries when compared to ultrasound. It also found that CXR was less accurate and less sensitive than LUS completed at presentation and 48 hours after the initial evaluation. The study included 22 patients who received an initial CXR, LUS, and CT scan within an hour between studies. 15 patients then required repeated imaging 48 hours later using same 1 hour timeframe. The lung ultrasound was completed using a convex and long probe in order to determine the presence of lung contusions or pleural effusions by operators who each had one year of lung ultrasound experience. The sonographers were blinded to the CXR and chest CT findings. The CXR was completed in a supine fashion and read by a blinded radiologist. The chest CT was also read by a radiologist blinded to the CXR and LUS results. Lung ultrasound was proven to have an initial sensitivity and specificity of 0.89 and 0.89 with an accuracy of 0.9 compared to CXR showing a sensitivity and specificity of 0.39 and 0.89 with an accuracy of 0.67. LUS also showed better correlation with chest CT than CXR with a correlation coefficient of 0.86; CXR had a correlation coefficient of 0.55. The study concludes that LUC is a reliable method for assessing acute respiratory failure in intensive care unit patients with chest trauma (Rocco). Helmy et al found similar results in a 2015 study investigating the applicability of LUS to diagnose lung contusions in comparison to CXR and a gold standard chest CT. The study included 50 patients who presented with blunt chest trauma (helmey 2015). LUS was completed using a convex probe in a previously described scanning protocol. (Lichtenstein AIS 2009) Chest CT was positive for lung contusions in 40 patients. When compared to the gold standard, LUS had a sensitivity of 97.5%, specificity of 90%, positive predictive value of 97.5%, negative predictive value of 90% and accuracy of 96%. With respect to peripheral parenchymal lesions, LUS showed a sensitivity of 92.5%, specificity of 100%, positive predictive value of 100%, negative predictive value of 76.92%, and accuracy of 94%. CXR was positive in 34% of patients. With regards to CT, CXR had a sensitivity of 40%, specificity of 90%, positive predictive value of 94.12%, negative predictive value of 27.27% and accuracy of 50%. The authors conclude that LUS is reliable study that might be of value in the diagnosis of lung contusions in blunt trauma patients (helmey).
The diagnosis of bronchiolitis has traditionally been made based on historical and clinical factors alone. However, there is growing research involving the role of sonography in both the diagnosis and management of bronchiolitis. Chest radiography rarely aids in the initial diagnosis but is used occasionally as an adjunct when the diagnostic picture is unclear. Studies have shown increasing B-lines and subpleural consolidations among patients with a final diagnosis of bronchiolitis. B-lines are considered a classic finding in interstitial syndromes. The lung evaluation should involve using Studies have shown various findings when evaluating bronchiolitis including subpleural consolidations, focal B-lines, compact B-lines, and pleural line abnormalities. Because the pediatric thorax is smaller and most findings are likely to abut the pleural line, the linear probe is most commonly used to evaluate the structures. However, a phased array or curvilinear probe can also be used especially when the costophrenic angles require interrogation. The anterior, axillary, and posterior lung fields can be evaluated in both longitudinal and transverse planes; the costophrenic angles can be interrogated by placing a lower frequency probe along the posterior axillary line such that the lung superior to the diaphragm overlying the liver or spleen can be evaluated (image). Caiulo et al compared the accuracy of ultrasound with chest radiography in children presenting with bronchiolitis; additionally, they further evaluated clinical and sonographical findings. The study involved children who received chest radiography and a clinical diagnosis of bronchiolitis. The clinical severity was classified according to the Downes’ score. Ultrasound was completed by a sonographer who was blinded to chest radiography findings. The control group then consisted of infants who were admitted with suspected GERD. Subpleural consolidations, pleural line abnormalities, multiple or compact B-lines, radiographic occult pleural effusions, and one radiographic occult pneumothorax were found in the infants with bronchiolitis. The infants were followed through their clinical course with lung ultrasound findings consistent with their clinical changes. Further, all patients showed resolution of lung ultrasound findings with clinical improvement. The study ultimately showed positive ultrasound findings in 47/52 patients and positive chest ray findings in 38/52 patients. The study suggests that ultrasound is a reliable tool for diagnosing and following the clinical course of patients with bronchiolitis. Varshney et al evaluated point of care lung ultrasound in children with respiratory tract infections and wheeze. The ultrasound was completed by a pediatric emergency medicine fellow who completed a two day course and five proctored ultrasound scans. The scans were reviewed by an expert sonographer who determined the final ultrasound read. Chest radiographs were ordered at the discretion of the treating physician who also made the final diagnosis independent of the ultrasound findings. The lung ultrasound was considered positive if three or more B-lines, consolidation, or pleural abnormalities were found. The study enrolled 72 patients with a final diagnosis of bronchiolitis. 33/72 patients had positive ultrasound findings. With regards to the final clinical diagnosis, ultrasound had a sensitivity of 45.8%, specificity of 72.9%, positive likelihood ratio of 1.7, and negative likelihood ratio of 0.7. Reliability between the expert and novice sonographer as a whole was noted to be good with a kappa agreement of 0.68.
Basile et al study that aimed to determine a correlation agreement between lung ultrasound findings and a clinical score in patients with bronchiolitis. Additionally, study evaluates the potential for lung ultrasound to predict the need for supplemental oxygen support. The study evaluated 106 infants who were admitted for bronchiolitis. The control arm consisted of 25 infants who were admitted for routine kidney and hip ultrasounds. A clinical physician who was blinded to the ultrasound results determined the severity of bronchiolitis based on a modified clinical score. A lung ultrasound severity score was based on a series of sonographic findings that included lung sliding with B-lines, confluent B-lines, and subpleural consolidations. A pediatric sonographer and radiological sonographer both blinded to the clinical score completed the scans and reviewed each scan individually. Agreement between the clinical physician and pediatric sonographer was 90.6. The agreement between the two sonographers was 89.6%. Further, ultrasound was able to predict the need for oxygenation with good accuracy showing a specificity of 98.6%, sensitivity of 96.6%, positive predictive value of 96.6 and negative predictive value of 98.7. The authors concluded that lung ultrasound can be considered as an extension of the clinical picture and perhaps incorporated into algorithms to assist with decision making.
Allinovi et al evaluated fluid overload by comparing the number of B-lines to weight changes in children receiving dialysis for acute kidney injury and end stage renal disease. Clinical fluid status was assessed on a scale from -1 to 3 indicating dehydration versus severe fluid overload. Unclothed weight was obtained prior to and after dialysis. Lung ultrasound was completed by a nephrology resident and a pediatric nephrologist consultant who received specific training and completed fifty sonograms. B-lines were quantified from 0-10 in specific positions along the anterior and lateral chest within fifteen minutes of initiating and completing dialysis. Each scan required 3-8 minutes to complete. 23 patients were included in the study for a total of 142 assessments. The observers found that the B-line score directly correlated with weight reduction in patients with acute kidney injury. Further increasing B-lines with a minimum of 12 were found in patients who were scored to have moderate to severe overload. A linear correlation between fluid overload and B-line score was found in both end stage renal disease (r =0.61) and acute kidney injury (r=0.83). The authors found lung ultrasound to be a real-time and practical method in quantifying fluid overload in dialysis patients. Further, the sonographic exam required 8 minutes maximum to complete and was well tolerated by the patients.
Aggarwal et al blinechf Adult studies have shown B-lines to correlate with the severity of fluid overload in patients with acute decompensated heart failure. One study showed that with regard to the diagnosis of acute decompensated heart failure, B-line findings on lung ultrasound showed a sensitivity of 91.9%, specificity of 100%, positive predictive value of 100%, negative predictive value of 62.9% and accuracy of 92%. The study further reports that if there is an absence of multiple B-lines or low lung comet scores, then pulmonary edema can be ruled out. A meta-analysis by Martindale et al used to determine the operating characteristics of the diagnostic elements used by the emergency phyisician to diagnose heart failure has shown that bedside lung ultrasound has a sensivitiy of 85.3%, specificity of 92.7%, positive likelihood ratio of 7.4, and a negative likelihood of 0.16. The study suggests that the presence of three or more B-lines in two or more fields is specific in ruling in heart failure, and the absence is sensitive in ruling out acute heart failure. Additionally Lichtenstein’s BLUE protocol showed a sensitivity of 97%, specificity of 95%, positive likelihoold ratio of 87, negative likelihood ratio of 99 in lung ultrasound diagnosis of pulmonary edema. There are additionally more adult studies related to lung ultrasound in sepsis showing that lung ultrasound might be usesful in predicting the severity of disease (santos) as well as lung ultrasonography’s role in the differentiation of undifferentiated hypotension (Volpicelli)
Pediatric studies are at present more limited, but research is currently pending including an anticipated overservational study to evaluate physiological changes after fluid bolus administration.